Quarterly Bulletin - 2016 Q1

Quarterly Bulletin2016 Q1 | Volume 56 No. 1

Quarterly Bulletin2016 Q1 | Volume 56 No. 1

Topical articles

How could a shock to growth in China affect growth in the United Kingdom? 4

Wages, productivity and the changing composition of the UK workforce 12Box The productivity puzzle and compositional effects 19

Bank of England notes: the switch to polymer 23Box UK coin and Scottish and Northern Ireland banknotes 25Box Banknote characters 30

Capturing the City: Photography at the Bank of England 33Box The Archive cataloguing project 39

The small bank failures of the early 1990s: another story of boom and bust 41

Recent economic and financial developments

Markets and operations 54

Appendices

Contents of recent Quarterly Bulletins 64

Bank of England publications 66

Contents

The contents page, with links to the articles in PDF, is available atwww.bankofengland.co.uk/publications/Pages/quarterlybulletin/default.aspx

Author of articles can be contacted [email protected]

Except where otherwise stated, the source of the data used in charts and tables is the Bank of England or the Office for National Statistics (ONS). All data, apart from financialmarkets data, are seasonally adjusted.

Research work published by the Bank is intended to contribute to debate, and does notnecessarily reflect the views of the Bank or members of the MPC, FPC or the PRA Board.

www.linkedin.com/company/bank-of-england

www.facebook.com/bankofengland.co.uk

www.flickr.com/photos/bankofengland/albums

www.youtube.com/user/bankofenglanduk

www.twitter.com/bankofengland

Topical articles

Quarterly Bulletin Topical articles 3

4 Quarterly Bulletin 2016 Q1

• This article assesses how a shock to Chinese growth could affect the UK economy using anempirical model of the world economy that exploits the historical comovement betweeninternational business cycles.

• We find that a 1% slowing in China is likely to reduce UK GDP by around 0.1%. This impact arisesmainly from the increasingly important role of China in the global economy — that is, via theUnited Kingdom’s indirect links with China through its main trading partners.

How could a shock to growth in Chinaaffect growth in the United Kingdom?By Ambrogio Cesa-Bianchi and Kate Stratford of the Bank’s Global Spillovers and Interconnections Division.(1)

(1) The authors would like to thank James Benford, David England, Dan Nixon andJumana Saleheen for their help in producing this article.

Overview

China’s importance to the global economy has grown inrecent years. In 2015, China accounted for around 17% of thelevel of, and a third of growth in, global economic activity.

GDP growth in China has slowed in recent years and somefurther slowing is expected as the Chinese authoritiesrebalance the economy towards consumption and away frominvestment. A key question is: how might that slowingaffect UK growth? And how would the shock be transmitted(through direct bilateral linkages or through other moreindirect channels)?

We find that, today, a 1% negative shock to Chinese GDP islikely to lower UK GDP by 0.1% (summary chart, red line).This estimate captures a combination of trade, financial andconfidence channels as well as an expected offset from loweroil prices. This impact is much larger than direct tradelinkages alone would imply.

By way of comparison, the impact of China slowing on theUnited Kingdom is one third of the size of an equivalentslowing in the euro area, where our trade links are ten timesbigger. Our estimates indicate that the impact is nowfour times larger than it would be if global trade linkageswere at 1990 levels (summary chart, purple line).

We show that this increase in the impact of a Chinaslowdown is due to stronger ties between China and theUnited Kingdom’s traditional trading partners such as theUnited States and the euro area (‘indirect’ effect), rather thanstronger direct linkages between the United Kingdom andChina (‘direct’ effect).

When decomposing this increase into ‘direct’ and ‘indirect’effects, we find that the latter accounts for almost all the fallin UK GDP (summary chart, orange line).

Our best guess is that these estimates are likely tounderstate the overall impact, particularly in the event of asharp slowdown in China, where the spillovers to othercountries and through financial links would probably belarger than our linear model would suggest.

Click here for a short video that discusses some of the keytopics from this article.

–

0.20

0.15

0.10

0.05

0.00

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

2012 (total)

2012 (indirect)

1990

Per cent

Increased impact since 1990 due to greater indirect links

Increased impact since 1990 due to greater direct links

Quarters

Summary chart The changing impact of a negative1% shock to Chinese GDP on UK GDP — direct versusindirect channels

Sources: IMF International Financial Statistics, OECD, Thomson Reuters Datastream andBank calculations.

https://youtu.be/JPmwRlD1rnE


Topical articles Could a shock to growth in China affect growth in the UK? 5

Introduction

GDP growth in China has slowed significantly in recent years.Some further slowdown is expected (Chart 1) as the Chineseauthorities attempt to rebalance the economy towardsconsumption and away from investment. There remainsconsiderable uncertainty around how smooth that rebalancingwill be, though. And there is a risk that growth could slowmore sharply than expected.(1)

Developments in China are likely to have importantimplications for the world economy and, in turn, for theUK economic outlook. Moreover, given the increasinglyimportant role of China in the global economy, the impact ofdevelopments in Chinese economic activity may have changedover time. Indeed, while in 1990 China accounted for around4% of world GDP, China is now the largest economy in theworld, accounting for more than 17% of world GDP.(2) AndChinese imports account for around 10% of global trade.

This article assesses how a shock to Chinese growth couldaffect the UK economy. Note that ‘China’ refers to mainlandChina throughout (ie it does not include Hong Kong). The firstsection sets out the various channels through whichdevelopments in China could spill over to the United Kingdom.The second section describes an empirical model that can beused to estimate the international transmission of shocksoriginating in different countries on the United Kingdom. Andthe final section runs through the results, focusing on how andwhy the impact of China on the United Kingdom has changedover time. A short video discusses some of the key topicsfrom this article.(3)

Through which channels could a shock toChina spill over to the United Kingdom?

A shock to Chinese growth is likely to affect theUnited Kingdom through a number of different channels.In order to think about the overall impact of China on the

United Kingdom, it is helpful to think about how large thesedifferent channels could be. So this section provides anoverview of the different channels that could be important.(4)

Trade channelsA key transmission mechanism through which a shock toChinese GDP might spill over to the United Kingdom isinternational trade. A slowing in Chinese GDP would reducethe demand for goods and services produced by UK firms. Andas the United Kingdom is an open economy, with exportsaccounting for around 30% of GDP, the trade channel couldbe an important part of the transmission of world shocks.

While the trade channel may be theoretically important, theshare of UK exports that go directly to China is smaller thanother trading partners — less than 4% in 2014. So, at firstsight, it may seem that Chinese growth should only have alimited impact on the demand for UK exports and, therefore,GDP growth.

But China is now the world’s largest goods exporter and theindirect trade links (links between China and theUnited Kingdom through other countries) are potentiallysizable. While China only accounts for a small share ofUK exports, it is a far more important source of exportdemand for some of our key trading partners, such as theeuro area and United States: exports to China make up justunder 10% of total goods exports for both the United Statesand euro area. Moreover, roughly 10% of the euro area’sexports go to other Asian economies, who are in turn heavilyreliant on Chinese demand.(5)

The importance of the trade channel is likely to have risensignificantly over the past two decades. In the late 1990s,China only accounted for around 2% of world imports ofgoods and services, and 3% of world exports. But by 2014,those shares had risen to 10% and 11%, respectively, makingChina the world’s second largest importer and the largestexporter. At the same time, UK exports to China have risensignificantly, from around 0.5% through the 1990s to 4%in 2014.

Financial channels and confidence effectsIn addition to trade, financial channels can also play animportant role in the transmission of international shocks.A recent Quarterly Bulletin article found that financial linkages

0

2

4

6

8

10

12

14

16

2000 03 06 09 12 15 18

Percentage change on a year earlier

IMF forecasts

Sources: IMF World Economic Outlook (WEO) October 2015 and National Bureau of Statistics of China.

Chart 1 Chinese GDP growth

(1) See, for example, the November 2015 Inflation Report;www.bankofengland.co.uk/publications/Documents/inflationreport/2015/nov.pdf.

(2) These figures are computed by converting GDP to international dollars usingpurchasing power parity rates.

(3) https://youtu.be/JPmwRlD1rnE.(4) For a discussion of the different channels through which external shocks can affect

the United Kingdom, see Chowla, Quaglietti and Rachel (2014).(5) Additional details on China’s international trade linkages can be found in a box

‘How would a slowdown in China affect the UK economy?’ on page 2 of theNovember 2015 Inflation Report;www.bankofengland.co.uk/publications/Documents/inflationreport/2015/nov.pdf.



are likely to account for the majority of the impact of worldshocks on the United Kingdom since 2007.(1) Given China’ssize in the global economy, we might expect such spillovers tobe significant. In what follows we review some of the mostrelevant channels.

Financial spillovers from China could operate directly throughthe UK banking sector. For example, UK-owned banks havedirect exposures to Chinese financial institutions, corporatesand households through their foreign lending activities. As aresult, UK banks could suffer losses as a result of a weakeningin Chinese demand conditions if it were associated with a risein non-performing loans. Losses abroad could, in turn, beardown on banks’ willingness and ability to lend to the UK realeconomy. They could also lead to an increase in banks’funding costs, particularly if combined with a deterioration infinancial market sentiment. Increased funding costs wouldthen feed through into a higher cost of borrowing forUK businesses and households. In addition, losses suffered byChinese financial institutions could weigh on lending toUK affiliates, or their supply of wholesale funds to UK banks.

This ‘direct’ exposure, however, is relatively small, since only1.6% of UK-owned banks’ foreign claims is directed to Chinesebanks. But, similarly to trade linkages, indirect linkages mayplay an important role. For example, the total exposure whenincluding Hong Kong — which, in turn, has important linkageswith China — amounts to just over US$530 billion, or around16% of UK-owned banks’ foreign claims. For this reason, theBank’s 2015 stress test included a severe scenario for theseeconomies, with sharp reductions in GDP growth, andproperty and equity prices, to ensure that the UK bankingsystem was sufficiently capitalised to withstand the risks onthese foreign exposures.

Financial spillovers to the United Kingdom could also operatethrough non-banking channels. For example, slowing in Chinacould be reflected in falls in asset prices both in China andelsewhere. Such falls would in turn push up UK companies’cost of capital and reduce households’ wealth. And indeed,despite the small share of China in global equity marketcapitalisation (at slightly less than 10% in 2014), financialmarket developments through the summer of 2015highlighted that global asset prices can be very sensitive todevelopments in China.(2)

Financial linkages between China and the rest of the worldhave also become increasingly important. For example, sincethe early 1990s China’s total external liabilities have increasedfrom less than 0.5% of world GDP to around 6%. AndUK-owned banks’ exposures to China have risen roughlytenfold over the past decade, to US$169 billion in 2015 Q3.

A slowing in China could also raise uncertainty about theoutlook for growth there and elsewhere; this could affect

UK households’ and companies’ confidence and so weigh ontheir spending decisions. China is now the largest economy inthe world (on a purchasing power parity basis) and hascontributed more to global growth than all advancedeconomies over the past eight years (Chart 2). So any shockto the Chinese growth outlook might be expected to havesignificant effects on the United Kingdom and other countriesthrough global sentiment and risk aversion.

Commodity pricesIt is likely that slower Chinese growth would weigh heavily oncommodity prices. China has been, by far, the largestindividual driver of demand growth for oil and industrialmetals such as copper and aluminium. For example, China hasaccounted for around one third of the total increase in globaloil demand since 1990. It now accounts for over 10% ofworld oil demand and roughly 50% of copper demand. Duringthe 2009–12 period, China contributed on average 40% toglobal aluminium consumption. As a result, commodity pricestend to be very sensitive to Chinese growth developments.

Unlike the previously discussed channels, a decline in oil andother commodity prices should boost UK GDP growth, giventhat the United Kingdom is a net importer of those goods.Household real incomes should also be boosted by the declinein prices, leading to increased consumption.(3) And UK firmswill benefit from lower commodity prices overall, due to lowercosts of energy and other inputs. In the other direction,though, investment spending of the United Kingdom’sextraction sector is likely to fall. Demand elsewhere in theworld is also likely to benefit, supporting UK exports, although

(1) See Chowla, Quaglietti and Rachel (2014).(2) See, for example, the November 2015 Inflation Report, pages 1–9;

www.bankofengland.co.uk/publications/Documents/inflationreport/2015/nov.pdf.(3) Note here that the United Kingdom is a net importer of oil since the mid-2000s. For

a full discussion of the impact of lower oil prices on the UK economy, see the box onpages 32–33 of the February 2015 Inflation Report;www.bankofengland.co.uk/publications/Documents/inflationreport/2015/feb.pdf.

+

–

4

2

0

2

4

6

8

2000 03 06 09 12 15

Advanced economies

Other emerging market economies

China

World

Percentage changes on a year earlier

Sources: IMF WEO October 2015, OECD, Thomson Reuters Datastream and Bank calculations.

Chart 2 Contributions to world GDP growth


commodity producers would be likely to cut back theirspending in response. Overall, the decline in commodityprices should provide some offset to the drag from trade,financial channels and confidence.

Modelling spillovers from China: a globalvector autoregressive model

To quantitatively assess how a shock to China might affect theUnited Kingdom, it is important to estimate the impact of allchannels of transmission. As explained in the previous section,a shock to Chinese growth is likely to affect theUnited Kingdom through a number of channels. And there arealso complex interlinkages between the United Kingdom andChina via third countries (ie ‘indirect’ effects). All of theselinks need to be captured when attempting to assess how aslowing in China might affect the United Kingdom.

One model that has the potential to capture all of thesevarious channels is the global vector autoregressive (GVAR)model.(1) Unlike other empirical models, the GVAR allowsshocks to propagate across the world economy through manychannels of interdependence, including trade, financial,confidence and third-country effects. And the model alsocaptures the effect of commodity prices.

An overview of the modelThe GVAR approach consists of two steps.(2) First, individualmodels of the economy are built for a large number ofcountries. In these models, a number of key domesticvariables — real GDP, inflation, equity prices, interest rates andthe exchange rate — are all affected by each other, as well asby international variables. The international variables includethe oil price and global measures of each of the domesticvariables — for example, world GDP and world inflation.

Second, the individual country models are linked through theinternational variables to construct a global model of theworld economy. In the GVAR, shocks in one country can spillover to other countries. The intuition is as follows: whengrowth in one country falls, that will lower measures ofworld GDP. As a result, GDP in each of the countries includedin the GVAR model will be affected, as will all the otherdomestic variables. The impact on growth in each country willbe determined by how domestic GDP has tended to respondto movements in international variables over the past.

The GVAR is flexible enough to also take into account thatone country may be important for some countries (such as itsmain trading partners) but not for others. The importance ofone country for another in the GVAR is determined bycountry-specific weights. These weights are used to constructthe international variables, so as to reflect the relativeimportance of each of the foreign economies for eachdomestic economy. In the version of the model used in this

article, export weights are used to aggregate the foreignvariables. So, for the United Kingdom, a large weight is placedon the euro area when constructing the international variables— as a large share of UK exports go to the euro area. But onlya small weight is placed on China, as the direct export links arelimited. These weights also vary over time, to take intoaccount changes in the relative importance of differentcountries over time.

Advantages of the modelThe GVAR model has several advantages relative to someother modelling approaches:

• It is a truly global model: the GVAR models the economiesof more than 30 countries, accounting for more than 90%of world GDP.

• All channels of transmission are captured: the countrymodels estimate the average response over the past ofdomestic variables to international ones. For example, themodel estimates how domestic GDP or equity prices tend torespond to movements in world GDP. This means that themodel will implicitly capture the total impact of a change ininternational variables on domestic ones, rather than justestimating the impact of one channel.

• Shocks can spill over through third countries (ie ‘indirect’effects): the use of country-specific weights means thatspillovers of shocks via third countries are captured in thismodel. For example, if there were a shock to US GDP thiswould initially have a direct impact on the United Kingdomthrough its effect on UK-weighted measures of world GDP,world inflation etc. But the change in US GDP would alsoaffect the economies of all other countries in the model,which would in turn feed back to the United Kingdom.Given this, the GVAR model is able to capture theamplification of shocks through third countries.

• The model captures historical comovement acrosscountries: the GVAR is an empirical model so its results aredriven by the correlations seen in the data. Therefore, themodel captures the scale of spillovers of shocks that havetypically been seen in the past.

Possible weaknesses of the modelWhile the GVAR is a very useful tool for measuring spilloversacross countries, there are at least two possible drawbacks tothis approach. First, the GVAR is linear. This means that, inthe model, the impact on the United Kingdom of a 5% shockto US GDP is five times larger than a 1% shock to US GDP.But it is possible that in periods of crisis, when the shocks arelarger, there could be greater spillovers than a linear model

(1) The model was originally developed by Pesaran, Schuermann and Weiner (2004) andDees et al (2007).

(2) Additional details on the model are provided in the annex.


suggests. And second, the GVAR (as with many otherempirical models) can only distinguish between differentsources of a shock by making a number of assumptions. Inwhat follows the focus of our analysis is on the study of thetransmission of GDP shocks across countries. No attempt ismade to identify the source of the shocks, whether they aredue to demand, supply, productivity or monetary policy whichcould, in practice, affect the impact on the United Kingdom.Uncertainties around the GVAR estimates are discussed at theend of the next section.

The changing impact of China on theUnited Kingdom

Results of the global VARUsing the GVAR, we can estimate what the impact of aslowdown in Chinese GDP growth on the United Kingdommight be. The blue line in Chart 3 shows the impact on thelevel of UK GDP of a permanent negative 1% shock to thelevel of GDP in China. On impact, the model suggests theshock would reduce UK GDP by slightly less than 0.1%. Thateffect would then build over time to a peak of 0.15%, beforefalling back to around 0.1% in the long run. The estimatedimpacts shown here use 2012 trade weights whenconstructing the foreign variables. This means that theyassume that the importance of each country to each other isas it was in 2012.

The GVAR estimates suggest that the spillovers to theUnited Kingdom from China could be much larger than directtrade linkages alone would imply. Simple ready reckoners(ie ‘mechanical’ estimates based on the share of UK exportsgoing to China and the share of UK exports in UK GDP)suggest that the negative shock to Chinese GDP would onlyreduce UK output by around 0.03% through direct trade links.The fact that the GVAR estimates are almost four times larger

than the direct trade channel suggests that additionalchannels and indirect linkages play an important role inamplifying spillovers to the United Kingdom. It is worthnoting, however, that not all these channels work in thesame direction. For example, in the model, the negativeshock leads to a fall in oil prices of about 3%, which supportsUK growth.

To put these results into context, it is helpful to compare theimpact of an equivalent shock to the United States and theeuro area (magenta and orange lines in Chart 3, respectively).A 1% GDP shock in either the United States or euro areawould be expected to have a larger impact on theUnited Kingdom than an equivalent shock in China; thelong-run impact of shocks to these two regions is aroundthree times that of the China shock. Considering just thedirect trade linkages would have suggested a larger differencein the results: the United Kingdom’s trade links with China arearound ten times smaller than with the euro area, andfive times smaller than with the United States. Again, thishighlights the importance of channels other than direct tradelinks.

It is also interesting to note that the long-run UK impact of ashock to the United States and euro area is very similar,despite the much closer trade links with the euro area. Thismay reflect the fact that financial channels are likely to be amore important part of the transmission of shocks from theUnited States than from the euro area. In addition, theUnited States plays a more important role in global demandthan the euro area, amplifying the third-country effects.

An illustrative scenarioHere we consider a simple scenario to illustrate the results ofthe GVAR. As shown in Chart 4, the International MonetaryFund (IMF) is forecasting growth in China to slow graduallyover the next three years to around 6% in 2018. Given thatthere is a great degree of uncertainty around the outlook forthe Chinese economy, we show what might happen toUK GDP if Chinese growth were to fall more sharply in thenear term. We consider an alternative scenario in whichgrowth is assumed to be 1.5 percentage points weaker thanthe IMF’s forecast in 2016 and 2017, before returning to 6% in2018 (Chart 4). This means that, overall, the level ofChinese GDP is around 3% lower by the end of 2018.

Chart 5 shows the impact we would expect such a scenario tohave on UK GDP. If Chinese growth were 1.5 percentagepoints weaker than expected by the IMF in 2016 and 2017, theGVAR suggests this could reduce growth in theUnited Kingdom by around 0.14 percentage points in bothyears. Growth would then pick up as GDP growth in Chinareturns to the rates forecast by the IMF from 2019. Thenegative impact on the level of UK GDP would peak at 0.3%,falling back to around 0.2% by 2019.

–

0.6

0.4

0.3

0.5

0.2

0.1

0.0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

United States

Euro area

China

Per cent

Quarters


(a) The impact is computed using 2012 trade weights.

Chart 3 The impact of a negative 1% shock to GDP inthe United States, euro area and China on UK GDP(a)


How has the impact of China changed over time?It is likely that the impact of China on the United Kingdom andother countries has changed significantly over the past coupleof decades. Indeed, as discussed above, China’s role in theglobal economy has increased markedly since the early 1990s.Given this, a shock to Chinese growth should now have amuch larger impact on the global economy, and in turn theUnited Kingdom, than would have been the case a couple ofdecades ago.

With the GVAR we can answer the question, how havespillovers from China to the United Kingdom changed overtime? Specifically, this section looks at how the impact of aChinese GDP shock today compares to the spillovers we wouldhave expected in 1990. To see how the UK impact varies overtime, we exploit the time-varying nature of the weights usedto construct the international variables in the country models.

Specifically, we simulate the GVAR using the weights from1990 and 2012 to estimate the impact on the United Kingdomof a shock to China at those specific points in time.(1)

The impact on the United Kingdom of a negative 1% shock toChinese GDP using both 1990 and 2012 weights is shown inChart 6 (the 2012 estimate is equal to that shown in Chart 3).The impact of shocks from China has increased significantlyover time. The long-run multiplier is now around four timeslarger than it was in the 1990s. And the short-run impacts aregreater still. That is in contrast to the United States andeuro area, where the impact of shocks has diminished a littleover time.(2)

Why has the impact of China changed over time?It is interesting to ask why the impact of China on theUnited Kingdom has increased so much. In particular, is thisincrease due to stronger bilateral linkages with China (‘direct’effect)? Or is it due to the stronger impact of China on theUnited Kingdom’s largest trading partners (‘indirect’ effect)?

In order to answer this question, we construct a new set ofweights. These are identical to the 2012 country weights, withone exception: the United Kingdom’s export share to China isset back to its 1990 level. This implies that the bilaterallinkages used to simulate the model will be almost identical tothe ones in our baseline, with the exception of the direct linksbetween China and the United Kingdom, which will besignificantly smaller.(3)

0

2

4

6

8

10

12

14

16

2000 03 06 09 12 15 18


IMF forecasts

Scenario

Sources: IMF WEO October 2015 and National Bureau of Statistics of China.

(a) The chart shows the historical path of Chinese GDP and its forecast over the 2016–18 period.The orange line displays a scenario in which growth is assumed to be 1.5 percentage pointsweaker than the IMF’s forecast in 2016 and 2017.

Chart 4 Chinese growth: IMF forecast and illustrativescenario(a)

+

–

+

–

0.3

0.2

0.1

0.0

0.1

0.3

0.2

0.1

0.0

0.1

2016 17 18 19

Per centPercentage points

Impact on level of GDP (right-hand scale)Impact on GDP growth (left-hand scale)


(a) The chart shows the impact on UK GDP of the scenario described in the main text anddepicted in Chart 4 (orange line).

Chart 5 Impact of illustrative scenario on UK GDP(a)

(1) For a more detailed explanation of this procedure, see Cesa-Bianchi et al (2012).(2) A 1% negative shock to US GDP decreases UK GDP by about 0.3% using both 1990

and 2012 weights, while a 1% negative shock to euro-area GDP decreases UK GDP by0.4% using 1990 weights and 0.3% using 2012 weights. These estimates are notreported in the main text but are available from the authors upon request.

(3) The difference between China’s 1990 and 2012 export shares for the United Kingdomwas redistributed proportionally to the remaining countries excluding theUnited States and the euro area (which were left unchanged at their 2012 levels toavoid overestimating the ‘indirect’ effects).

–

0.20

0.15

0.10

0.05

0.00

0 1 2 3 4 5 6 Quarters

7 8 9 10 11 12 13 14 15 16 17 18 19

2012

1990

Per cent

Chart 6 The changing impact of a negative 1% shock toChinese GDP on UK GDP



The purple line and the red line in Chart 7 are the UK impactsobtained under 1990 and 2012 weights (as shown in Chart 6).Therefore, the gap between the two lines represents the totalincrease of the impact of Chinese shocks on UK GDP from1990 to 2012. The orange line, which is obtained under thecounterfactual trade shares described above, shows that muchof the increased impact is due to indirect linkages. In otherwords, if today China still accounted for less than 1% of theUnited Kingdom’s exports (as it did in 1990), the impact of aChinese shock would nonetheless have increased from thepurple line to the orange line given China’s increased links withother countries.

This is clear evidence that the changed linkages between Chinaand the rest of the world are affecting the United Kingdom notonly via stronger bilateral linkages, but more importantlythrough stronger ties between China and theUnited Kingdom’s major trading partners.

Uncertainty around the estimatesWhile the GVAR is a useful model for capturing the spilloversfrom China to the United Kingdom, the results are subject tosome uncertainty. In particular:

• As in any empirical model, the estimated coefficients of theGVAR may change over time. Using the time-varyingweights (export shares) captures the change in thedistribution of exports. But it is possible that the

United Kingdom could have become more sensitive toforeign shocks after the financial crisis. Only a model withtime-varying parameters could properly account for thisissue.

• As mentioned earlier, the GVAR is a linear model, so itcannot account for the non-linearities that typically arise incrisis periods or in a particularly severe scenario. This meansthat if there were a very sharp slowing in China, oftenreferred to as a ‘hard landing’, then the GVAR model couldunderestimate the spillovers to the United Kingdom.

Given the above, the multipliers implied by Chart 3 are likelyto underestimate the scale of the impacts.

Conclusion

This article asks three questions: first, how large an effectcould a shock to Chinese GDP have on UK output? Second, towhat extent has the impact of shocks emanating from Chinaincreased over time? And third, through which channels —direct bilateral linkages or indirect channels ofinterdependence — has that change come through?

To answer these questions we use a global VAR that includescountry-specific macroeconomic and financial variables, globalvariables and the price of oil. This model implicitly accountsfor many different channels of transmission including trade,financial and confidence channels, and third-country effects.

As expected, we find that shocks emanating from China have abigger impact on the UK business cycle today than in the past.Specifically, if China were to slow down from 6% to 5% overthe next few quarters UK growth would be expected to slowby around 0.1%. By way of comparison, the impact of a 1%shock to China’s GDP is one third of the size of an equivalent1% shock to euro-area GDP.

That said, we view the multipliers implied by Chart 3 as likelyto underestimate the true impact, particularly in the event of asharp slowdown in China, where the spillovers to othercountries are likely to be larger given the associatednon-linearities that our model does not explicitly account for.

Finally, we have also shown that the increased influence ofChina on the United Kingdom is mostly through stronger tieswith the United Kingdom’s traditional trading partners (theUnited States and the euro area), rather than direct linkages.

–

0.20

0.15

0.10

0.05

0.00

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

2012 (total)

2012 (indirect)

1990

Per cent

Increased impact since 1990 due to greater indirect links

Increased impact since 1990 due to greater direct links

Quarters

Chart 7 The changing impact of a negative 1% shockto Chinese GDP on UK GDP — direct versus indirectchannels

Sources: IMF World Economic Outlook (WEO) October 2015, OECD, Thomson ReutersDatastream and Bank calculations.


AnnexA quick overview of the GVAR model

This section presents a simple overview of the GVARmethodology and discusses some of its underlyingassumptions.

The GVAR modelling strategy consists of two main steps.First, each country is modelled individually as a small openeconomy by estimating a country-specific VAR model in whichdomestic variables are related to both country-specific foreignvariables and global variables that are common across allcountries (such as the price of oil). Second, a global model isconstructed combining all the estimated country-specificmodels and linking them with a matrix of predetermined (thatis, not estimated) cross-country linkages.

Consider N + 1 countries in the global economy, indexed byi = 0,1,2,… N. In the first step, with the exception of country 0(which in our application is the United States), all other Ncountries are modelled as small open economies in which a setof domestic variables (xit, to be specified below) is related to aset of country-specific foreign variables, x*it, using anaugmented VAR model (VARX*) specification. Specifically, foreach country i, we set up a VARX*(pi, qi) model in which theki × 1 vector, xit, is related to the vector of country-specificforeign variables, x*it (abstracting from constant, trend, higherlag orders and global variables for simplicity):

where the foreign variables x*it are constructed as:

and where Wij, τ (t) is a matrix that contains the weights of j incountry i at time t, for a given τ(t). Here τ(t) is a generic rulethat indexes the time-varying weights at each time period t.For instance, in our empirical application, for each quarter t,τ(t) refers to a three-year average of weights for the currentyear, t, and the previous two years, t-1 and t-2.

The foreign variables provide the link between the evolution ofthe domestic economy and the rest of the world and, in thecountry-specific model estimations, are taken as (weakly)exogenous — an assumption that is tested and holds in thedata.

In the second step, the GVAR model is set up by stacking theestimated individual country-specific models and linking themwith a matrix of predetermined cross-country linkages.Having estimated the country-specific parameters using thetime-varying weights, the estimated country-specific modelscan now be combined and solved for any given trade weightsbased either on a particular year or on an average of weightsfrom different time periods.

x x x x uit i it i it i it it* *

1 0 1 1φ= +∧ +∧ +− −

x W W xit i t ij t jtj

N*

, ( ) , ( )0

∑( )=τ τ=

References

Cesa-Bianchi, A, Pesaran, M H, Rebucci, A and Xu, T (2012), ‘China’s emergence in the world economy and business cycles in Latin America’,Economia, Vol. 12, No. 2, pages 1–75.

Chowla, S, Quaglietti, L and Rachel, Ł (2014), ‘How have world shocks affected the UK economy?’, Bank of England Quarterly Bulletin, Vol. 54,No. 2, pages 167–79, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q206.pdf.

Dees, S, di Mauro, F, Pesaran, M H and Smith, L V (2007), ‘Exploring the international linkages of the euro area: a global VAR analysis’,Journal of Applied Econometrics, Vol. 22, No. 1, pages 1–38.

Pesaran, M H, Schuermann, T and Weiner, S M (2004), ‘Modeling regional interdependencies using a global error-correctingmacroeconometric model’, Journal of Business and Economic Statistics, Vol. 22, pages 129–62.

Overview

Wage growth is a key indicator of inflationary pressure inthe economy, and is important for the Monetary PolicyCommittee.

This article discusses how the changing characteristics ofthose in employment can affect aggregate measures ofwages. If a disproportionate amount of particularly low orhigh earners enter or leave employment in a given timeperiod, these ‘compositional effects’ can alter the averagelevel of wages and hence the measurement of wage growth.Such effects have been particularly large during the financialcrisis and the ensuing recovery.

Since the mid-1990s the increasing quality of the labourforce has on average added close to half a percentage pointto wage growth per year (summary chart). This growth hasbeen driven primarily by an increase in the average educationlevel of the workforce and a shift towards high-skilledoccupations. These shifts mean workers are moreproductive, which in turn feeds through into higher wages.

At the start of the recession, compositional effects pushedup even further on wage growth relative to normal, aslower-paid employees were laid off and struggled to findnew jobs.

But in 2014 and 2015 these effects went into reverse,slowing wage growth relative to normal. At its peak the dragfrom compositional effects was estimated to be close to1 percentage point. This was caused by both lower-skilledemployees returning to the workforce, and also byhigher-skilled employees exiting the workforce.

For most of 2015 this drag was holding down wage growthby around ¾ of a percentage point relative to normal, but inthe most recent data this has started to dissipate. The dragis likely to continue to dissipate as the labour marketnormalises, leading to an increase in both wage andproductivity growth.

Over the past 20 years, there appear to have beensubstantial changes to how some characteristics affectwages. For example, the estimated returns from having adegree have fallen, as have the returns from job tenure. Thedecline in relative wages for those over 50 compared withyounger age groups, which has been typically observed in theUK labour market, has disappeared since the crisis. Theestimates also suggest that regional and gender pay gapshave narrowed over recent decades.

• Over the past 30 years the composition of UK employment has changed substantially — thesechanges have important implications for wage and productivity growth.

• These ‘compositional effects’ can be more prominent during times of increased labour marketchange and may have dragged down on wages over the past two years.

• The drag from compositional effects is likely to fade as the labour market normalises, pushing upon both productivity and wage growth.

Wages, productivity and the changingcomposition of the UK workforceBy Will Abel, Rebecca Burnham and Matthew Corder of the Bank’s Structural Economic Analysis Division.(1)

Compositional effect

Other

Wage growth(a)

0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

1995–2015 1995–2007 2008–13 2014–15

Percentage points

+

–

Summary chart The contribution of compositionaleffects to average annual wage growth by time period

Sources: Labour Force Survey, ONS and Bank calculations.

(a) Whole-economy total average weekly earnings.

(1) The authors would like to thank Rosetta Dollman for her help in producing this article.


Topical articles Wages, productivity and composition of the UK workforce 13

Wage growth has been weak during the UK economy’srecovery from the financial crisis. An understanding of thereasons for this is important when assessing how muchinflationary pressure there is in the economy, and thus theappropriate stance for monetary policy. One potentialexplanation for weak wage growth is that aggregate wagemeasures have been affected by the changing composition ofthe UK workforce during the recession and its aftermath.

This article explores the extent to which compositional effectscan help to explain weak wage growth in recent years. It setsout how the structure of the UK workforce has changed overthe past 30 years. It then explains how, against this backdrop,the cyclical effects of the changing composition of the labourforce on wage growth can be estimated and applied to therecent recession. The final section looks at how the effect ofsome key characteristics on wage levels has changed over thepast 20 years.(1)

Setting the scene: the persistent weakness ofwage growth

Wage growth is typically thought to be driven by three mainfactors: productivity, inflation expectations and labour marketslack (that is, the level of hours worked relative to potentiallabour supply). In the long run, an individual’s real wageshould track their productivity — the value of what theyproduce over a given period measured in real terms. Nominalwages also need to take account of inflation to preservespending power. As wages are typically set in advance, thismeans factoring in inflation expectations when setting wages.Finally, in the short run at least, if unemployment is high thenworkers are more likely to accept lower wages because theiralternative options, such as searching for a job elsewhere, areworse.

From 2001–07, annual wage growth averaged around 4¼%(Chart 1). This can largely be accounted for by a combinationof steady productivity growth of around 2% and stablehousehold inflation expectations. The unemployment ratewas close to its estimated equilibrium rate over this period andis likely to have exerted little pressure on wage growth.

By contrast, from 2008–12, wage growth averaged lessthan 2%. Some of this fall can be explained by the sharp fall inproductivity growth. But even accounting for this, wagegrowth was weak during the crisis and the immediate recovery.This is likely due to the high level of labour market slack,reflected in the rise in unemployment, reducing the pressureon employers to raise wages to retain and attract staff.

Wage growth from 2013–14 remained weak at around 1¼%.A rise in wage growth might have been expected as workersregained bargaining power due to declining slack in the labour

market, but pay growth remained weak even thoughunemployment began to fall rapidly in 2013.

Wage growth remained below its pre-crisis average in 2015despite unemployment falling back towards its pre-crisis rateof 5%. Although wage growth picked up somewhat during2015, it eased back to around 2% at the end of 2015.

The failure of wage growth to pick up as much as might havebeen expected following the fall in unemployment in recentyears means that other factors could be affecting wagegrowth.

The potential role of changes in thecomposition of the workforce

One explanation is that ‘compositional effects’ have draggeddown on measures of productivity and wage growth over thistime.

Compositional effects in this context mean the change inaggregate measures of wage and productivity growth causedby changes in the mix of characteristics of those inemployment. The timeliest measure of wages in theUnited Kingdom — average weekly earnings (AWE) — iscalculated by dividing the total amount paid as wages in theeconomy by the total number of employee jobs. Similarlyproductivity per worker is simply the total value of all outputproduced in the United Kingdom divided by the number ofworkers. If some types of worker are more productive thanothers and therefore earn more than others, then as peopleenter and exit the workforce, switch industries, undergotraining and gain experience, the resulting changes incomposition will have an effect on the average level of bothwages and productivity.

(1) These calculations only cover 20 years as opposed to the 30 years initially examinedbecause of data limitations.

0

1

2

3

4

5

6

7

8

9

4

2

0

2

4

6

8

2001 02 03 04 05 06 07 08 09 10 11 12 13 14 15

Total wage growth (right-hand scale)

Average wage growth 2001–07 (right-hand scale)

Average wage growth 2008–15 (right-hand scale)

Unemployment rate (left-hand scale)

Percentage change on a year earlierPer cent

+

–

Chart 1 Unemployment and total wage growth 2001–15(a)

(a) Wage growth is measured as average weekly earnings.

To illustrate this, in 2000 about a sixth of workers in theUnited Kingdom had a degree; by 2015 the proportion ofworkers with a degree had risen to about a third. Workerswith a degree tend to earn more than others. Consider asimple example where workers with a degree earn £10 an hourand all other workers earn £7 an hour. In these conditions, adoubling in the proportion of workers with a degree would, allelse equal, raise the average level of wages from £7.50 to £8(Figure 1). This change in the average level of wages is solelythe result of compositional effects; the pay of individualworkers is unchanged.

It might be expected that gradual changes in the compositionof the workforce would not affect cyclical wage dynamics.Even during the rapid recovery in employment seen in 2013and 2014, annual employment growth peaked at only 2.3% —a share of the workforce too small to drag up or down on theaverage wage growth of the entire employed populationsubstantially.

But small changes in employment mask substantial churn inthe labour market. The changes in aggregate employment arethe product of much larger flows into and out of employmentas well as the movement of people between different jobs. Onaverage, just under 3½% of the workforce leave employmentevery quarter, with slightly more flowing in from eitherinactivity or unemployment (this slight gap reflects theincreasing level of employment over time). A further 2%–3%of employed individuals move between jobs each quarter(Chart 2). The cumulative effect of these flows means that,on average, in each year between 1994 and 2007, around onein ten workers had changed employer compared to a yearbefore and one in ten had entered the workforce from eitherinactivity or unemployment: this leaves about eight out of tenwho were at the same employer as they were the previousyear. Even without incorporating the changes caused bypeople being trained and moving roles within the samecompany, this level of churn can generate substantial changesin workforce composition.

While such flows and compositional changes are continuouslyoccurring, the cyclical fluctuations in labour market flows thathave been seen throughout the recession and its aftermath arelikely to make these effects particularly pronounced. Forexample if, at the beginning of a recession, those who losetheir jobs are disproportionately low-skilled workers, averagewage measures would be higher than they would beotherwise, boosting measured wage growth. Thecompositional effect on wages here would be positive.

Understanding the size of compositional effects is importantfor monetary policy decision-making. To the extent that suchcyclical shifts are only likely to affect wage growth temporarily(until the shifts in the mix of employment are complete) theymay mask underlying shifts in pay pressures. Furthermore,wage increases are not always inflationary. If wages andproductivity increase equally, then the wage cost per unit ofoutput, known as the unit wage cost (UWC), will remainunchanged. It is changes in UWCs — rather than nominalwages — which should in theory be most important forcompanies’ pricing decisions and hence inflationary pressure inthe economy.(1) As compositional effects would be expectedto affect both productivity and wage growth in a similar waythey should not have much impact on UWCs or inflationarypressure. But an understanding of compositional effects helpsthe MPC interpret movements in pay and productivityalongside other important factors such as slack in the labourmarket.

Structural changes to the composition of theUK workforce

In order to analyse the compositional changes caused bycyclical behaviour during the financial crisis and subsequent


(1) Technically the primary concern is with unit labour costs, which include non-wagecosts (such as pensions), rather than unit wage costs; for the purposes of this articlethis distinction can be ignored.

£10£7£7£7£7£7

£10£7£7£7£7 £10

2000

2015

Average = £8.00

Average = £7.50

Figure 1 Stylised example of effect of changes in thecomposition of the workforce

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

1993 96 99 2002 05 08 11 14

Flows into employment Flows out of

employment

Job-to-job flows

Per cent of those employed in the previous quarter(four-quarter moving average)

Chart 2 Flows into, out of and within employment everyquarter

Source: Labour Force Survey.

recovery, it is necessary to understand the substantiallonger-term structural trends that have affected thecomposition of the UK labour force.

An individual’s pay and productivity are affected by factorssuch as their level of education, work experience, the industrythey are employed in and their specific job role. Across theeconomy as a whole, there have been large shifts in thedistribution of most of these measures over recent decades.

Probably the most substantial change to the UK labour force,in terms of earnings potential, has been the rapid increase inthe educational attainment of workers. In 1985, less than atenth of workers had a degree: 30 years later, a third have adegree (Chart 3). As the workforce becomes more skilled overtime, this should raise the average level of pay andproductivity.

As the supply of skills has increased in the economy so too hasthe share of high-skilled jobs. Determining the skill level ofany specific job can be difficult, but using a commonlyaccepted approach where managers and those in professionalor technical roles are considered highly skilled, the share ofUK employment in such occupations has increased from 34%in 1992 to 44% in 2015.(1) Roles such as these are likely to bemore productive and so be more highly paid.

The industries in which people work have also changedsubstantially — most notably due to the decline ofmanufacturing employment since the late 1970s andcorresponding increase in the share of service industries inemployment. Manufacturing accounted for a quarter ofemployment in 1978 compared with 8% today, with servicesnow accounting for 83%. The effect of this shift uponearnings is, on the surface, ambiguous. Some of the increasein services employment has been in professional and technicalindustries, which are likely to be more highly paid. However

the largest increases, in terms of employment, have come inthe education and healthcare sectors, neither of which areparticularly highly paid relative to other industries. Thefinancial and information technology industries have notexpanded substantially since the late 1970s, going froma combined share of employment of 6% to 7%. Thiscontrasts with the fact that by other measures theseindustries have grown substantially (for example, stockmarket capitalisation).

Another important shift has been the ageing of theUK population (Chart 4). The ageing of ‘baby boomers’ andthe decline in the UK birth rate has resulted in an oldersociety, with 43% of the population aged 45 and over in 2014compared with 37% in 1981. This shift has fed through intothe workforce: over the same period, the share of those 45and over in employment has risen from 36% to 43%. Olderworkers are likely to have greater levels of work experienceand thus productivity than younger workers, althoughthese effects might not be uniform — for example,productivity may decline when individuals near the end oftheir working lives.

Female participation in the workforce has also increased overtime. The share of women in employment has gone up from41% in the mid-1980s to 47% in 2015. Numerous economicstudies show that there is a gender pay gap faced by women(2)

relative to men which is unexplained by socioeconomic factorslike education and the industries in which they work. To thisextent, women’s increasing share of the workforce might havepulled down on overall average pay. While this effect onindividual pay is statistically significant, the effect is likely tobe substantially smaller than factors such as an individual’seducation, occupation and age.

(1) Measures of occupation are only taken back to 1992 due to changes in the occupationclassifications in the Labour Force Survey.

(2) See, for example, Leaker (2008).

Age

Males

2014

1981

Females

Population (thousands)100 200 300 400 500100200300400500 0 0

2014

1981

05

101520253035404550556065707580

Chart 4 The ageing of the UK population

Degree

Higher education

A-level

GCSE

No qualification

1984 88 92 96 2000 04 08 12

Per cent, as a share of total employment

0

5

10

15

20

25

30

35

Chart 3 Highest education level of those employed(a)

Sources: Labour Force Survey and Bank calculations.

(a) There are slight kinks in the time series due to changes in the survey methodology; wherehighest level of qualification has been recorded as ‘don’t know’ these results are not shown.


The share of people working part-time has also risen, from afifth in the mid-1980s to 27% now. The rise partly reflects therise in female participation mentioned above: while womenaccounted for 47% of all employment in 2015, they onlyaccounted for 37% of full-time employment. Part-timeemployees are more likely to experience lower wages per hourcompared with their full-time counterparts.

The importance of these compositional shifts for aggregatewages will depend on the difference in pay for different typesof workers. Chart 5 shows the average wage in 2015 forgroups of individuals based on different factors, either the skilllevel of their occupation, their highest level of education ortheir gender.(1) The difference in wages between these groups,based on any one criterion alone, is substantial. This explainswhy changes in the share of employment in these differentgroups over time give rise to compositional effects on pay.However, these simple averages can capture a variety ofinter-related effects. For example, those with higher levels ofeducational attainment are more likely to work in high-skilledoccupations. To be able to understand fully the role ofcompositional effects, the marginal effect of differentcharacteristics needs to be identified. This is discussed in thenext section.

Calculating compositional effects

To calculate the size of compositional effects on wage growth,estimates are made of how much different characteristicsaffect wage levels, which are then applied to the changes inthe mix of employment over time.

In order to do this, quarterly anonymised individual-level datafrom the Labour Force Survey (LFS) are used, which includes adetailed set of individual and job-related characteristics,including age, gender, industry, occupation, job tenure andeducation as well as the self-reported level of pay. These dataare available on a quarterly basis from 1994 onwards.(2)

By regressing an individual’s observed hourly wage against arange of these individual and job-related characteristics, it ispossible to calculate for any given quarter how these factorsaffect their wage, on average.(3) These wage equations allowfor a decomposition of observed changes in wage growth intoa combination of ‘explained’ and ‘unexplained’ variation.Here, the ‘explained’ variation is the measurement ofcompositional effects, where changes in pay can be explainedby shifts in the characteristics of employees. All othermovements in wage growth, for example due to changes inproductivity or inflation expectations, are considered‘unexplained’.

The effect of characteristics that are highly correlated can becontrolled for using such regression techniques, which identifythe marginal contribution of a specific factor. For example, inidentifying the effect of industry on wages all other factors willbe held constant, so managers with degrees and the samelevel of tenure are compared with similar individuals acrossindustries, isolating the effect of working in differentindustries. This allows more precise estimates of the effect ofchanges in the mix of employment on pay.

How big are compositional effects typically and whichfactors have been driving them? Changes in the composition of employment have typicallyboosted pay over time. On average these compositionaleffects added just under a ½ percentage point to annual wagegrowth between 1995 and 2015 (summary chart), whichaveraged around 3¼% over this period. This positivecompositional effect is consistent with both the long-termtrends in the workforce described previously, and otherstudies.(4)

The five factors which were found to make the largestcontribution to these compositional effects were education,job tenure, age, occupation (the type of job the employee isdoing) and industry. Other factors which were included in theanalysis, but had much smaller aggregate effects, were gender,region, whether an individual worked full or part-time,whether they were employed in the public sector and whetherthey had a temporary employment contract.

The largest driver of this positive compositional effect overthe past 20 years is the increasing education level of theUK workforce, which explains over 65% of the positive impactof compositional effects on wage growth. The other keycontributor was a shift to more highly skilled occupations,which explains close to 30% of the positive effect. The

(1) These groups overlap so one individual may be counted in several averages shown.(2) Quarterly LFS data is available from 1992, but these data sets do not have all the

variables required for the analysis carried out here.(3) Details of this process are provided in the annex.(4) See, for example, Bell, Burriel-Llombart and Jones (2005).

0 2 4 6 8 10 12 14 16 18

Mean pay

High-skilloccupation

Medium-skilloccupation

Low-skilloccupation

Degree

A-level

GCSE

Men

Women

£ per hour

Chart 5 Average earnings in 2015 for different groups

Source: Labour Force Survey.


changing industrial composition of the UK workforce isestimated to have acted as a net drag on wage growth.(1)

Compositional effects since the financial crisisTo understand the role of cyclical fluctuations incompositional effects, the estimates in a given period shouldbe considered relative to the typical boost they provide.Chart 6 shows the main results of the analysis, measuringcompositional effects relative to their 1995–2010 average.(2)

Bars above the zero line show a positive effect on wage growthfor that factor relative to trend, with bars below the lineindicating a negative effect. The ‘total compositional effect’shows the overall effect of the different factors. By measuringcompositional effects in this way, it is possible to determinewhether they were having an abnormal effect on wage growththroughout the crisis and its aftermath — in particular in thepast two years. At the start of the crisis, compositional effectspushed up strongly on wage growth by around 1 percentagepoint at their peak. In 2014 this effect went into reverse andcompositional effects dragged down on wage growth by againroughly 1 percentage point. This may go some way toexplaining the weak wage growth since that time shown inChart 1.

The positive compositional effect at the beginning of the crisiswas caused by a sharp decline in flows into employment fromunemployment and a decrease in people retiring from theworkforce. Those in unemployment typically have lesseducation and are more likely to have worked primarily inlower-skilled industries. As such, the decline in the flow of theunemployed into employment meant that the average tenure,level of education and share of skilled occupations allincreased substantially. In addition, some individuals are likelyto have delayed retirement during the recession. As thoseclose to retirement are likely to be more experienced andtherefore be receiving higher pay, this too likely pushed up onthe average wage level. These positive effects outweighed adrag caused by a decline in the number of those employed in

relatively highly paid financial and insurance service industriesjobs — shown by the purple bars in 2009–10.

As flows into employment recovered, many of these effectsunwound, causing the compositional effect on wages to turnnegative. For example, the average age of those who exitedthe workforce increased in 2014. It is possible that some ofthose who had previously delayed retirement due to theeffects of the crisis were now exiting the labour force.

While these results are not conclusive, they are consistentwith findings from other data. The Office for NationalStatistics (ONS) has found strong negative effects on meanwages in 2014 due to the difference between those leavingand entering employment.(3) Similarly, in other analysis(4) theONS found a substantial increase in the ‘quality’ of the labourforce at the beginning of the crisis and a sharp decline in 2014,where ‘quality’ seeks to capture factors such as qualificationand age — in line with the results presented here.

In the latest data the impact of compositional effects onwage growth relative to average has fallen to around -¼% in2015 Q4. Compositional effects on wage growth are likely toreturn to their long-run average as the labour marketnormalises.

Summary of the resultsTable A summarises these results, showing the absoluteeffects different compositional factors have had on wagegrowth since the crisis. This shows that changes in

(1) It is important to note that the methodology used to calculate these compositionaleffects has limitations — most importantly it is limited by the available data and thuspotentially exposed to bias by omitting other important variables. For example, thenegative effect of industrial composition on wage growth could be due to certainindustries being correlated with other factors such as union membership, which arenot controlled for in the regression due to lack of quarterly data. If unions are able toincrease worker pay, then a move to less unionised industries would show up in theresult as industrial composition having a negative effect.

(2) These dates are chosen to try to include a full business cycle, while giving a stablebenchmark against which to measure the compositional effects.

(3) See ONS (2015). (4) See Connors and Franklin (2015).

1.5

1.0

0.5

0.0

0.5

1.0

1.5

2008 09 10 11 12 13 14 15

Percentage points

Total compositional effect

+

–

Qualification

Industry Age

TenureOccupation

Other(a)

Chart 6 Compositional effects on annual hourly wage growth relative to 1995–2010 average


(a) ‘Other’ includes gender, region, sector, working pattern and contract type.


qualifications and tenure boosted wage growth at the start ofthe financial crisis (second column) relative to the average(first column). The table also shows the drag compositionaleffects have had in 2014–15, driven by below-averagecontributions from qualifications, job tenure and occupation.

Compositional effects appear to explain a degree of the weakwage growth seen in 2014 and 2015; however they are not acomplete explanation. Compositional effects explain little ofthe weakness of wage growth seen at the end of 2015 becausethe drag from compositional effects started to wane.

Given that compositional effects are also likely to affectproductivity growth, some researchers have looked atcompositional effects to see if they can help to explain thepersistent weakness of productivity growth in theUnited Kingdom since the financial crisis, sometimes called the‘productivity puzzle’.(1) A box on page 19 looks at thisrelationship. It finds that compositional effects make theweakness in productivity early in the crisis more puzzling, butmay help to explain some of the weakness since 2014.

Changing pay premium for different individualor job characteristics

A useful by-product of measuring compositional effects usingthe approach described above is that variations over time inthe effect of different factors on wages can be estimated.

Of the five factors found to have the greatest effect on wagesin the analysis — education, tenure, age, industry and type ofoccupation — the impact of three of these — education,tenure and age — have changed substantially over the 20-yearsample period used for estimating compositional effects.Chart 7 shows the effect of an individual’s highest level ofeducation on pay relative to having no qualifications at all,holding other factors constant. Since 1995, the effect ofhaving a degree on pay has fallen substantially. In 1995, adegree would on average increase wages by 45% relative tohaving no qualifications at all; by 2015 this premium hadfallen to 34%. Over the same period, the wage premium forA-levels and GCSEs also fell, but by far less.

There are a number of possible explanations which would beconsistent with this finding. For example, if demand for highlyskilled workers has not kept pace with an increase in supply, anincreasing number of graduates would also lead to a decreasein the wage premium for those with degrees. Alternatively it ispossible that the large increase in individuals studying for adegree in the United Kingdom has led to a fall in its signallingvalue (the ability of degrees to correctly identify moretalented individuals) and thus the amount of pay which thosewith degrees can command.

The impact of job tenure on pay has also been declining overthe past 20 years (Chart 8). While staying with an employerfor a prolonged period of time does still have substantialbenefits to an individual’s earning potential, the premium forincreased job tenure has declined. The premium for thosewith over 20 years’ tenure has declined by around a third since1995. This decline has happened without any significant shiftin the tenure structure of the economy.

A third finding is that the relative earnings of those over 55have increased since the financial crisis. Traditionalage-earnings profiles have an ‘inverted u’ shape, earningsincrease steeply with age up until the age of about 40, theyare then stable until declining near the end of one’s career.This is the profile found pre-crisis in the United Kingdom(as shown in the magenta bars in Chart 9). However, from2008 onwards, this decline in the earning power of theUnited Kingdom’s oldest workers was rapidly eroded and theresults for 2015 show that there was no significant decline forany age group’s earnings after age 35 (the orange bars inChart 9). These statistical estimates cannot isolate the causeof this sudden change, and at this stage it is unclear howmuch, if any, of the change will prove to persist.

(1) See, for example, Blundell, Crawford and Jin (2013).

0

5

10

15

20

25

30

35

40

45

50

Degree

A-level

GCSE

Per cent pay increase relative to no education(four-quarter moving average)

1995 97 99 2001 03 05 07 09 11 13 15

Chart 7 The declining average effect of having a degreeon wages


Table A Absolute contributions to wage growth fromcompositional changes

Per cent

Average

1995–2010 2008–10 2011–13 2014–15

Total compositional effect 0.5 0.8 0.7 -0.1

Qualification 0.4 0.5 0.4 0.2

Tenure 0.0 0.2 0.1 -0.1

Age 0.0 0.1 0.0 -0.1

Industry -0.1 -0.2 0.0 0.0

Occupation 0.2 0.2 0.2 0.0

Other 0.0 0.0 0.0 0.0


The productivity puzzle and compositionaleffects

This box provides a brief background on the post-crisisslowdown in productivity growth that has been seen in theUnited Kingdom, commonly known as the productivity puzzle.It also discusses to what extent compositional effects mayhelp to explain this puzzle.

Since the financial crisis, labour productivity in theUnited Kingdom has been exceptionally weak. From 1997 to2007 labour productivity (output per hour worked) grew at anaverage of 2.2% per year, whereas from 2008 to 2015 it grewat just 0.2% on average. This can be seen in Chart A.

Measures of productivity can be used to inform estimates ofan economy’s ability to grow without generating excessiveinflationary pressure. If the causes of this slowdown inproductivity were primarily short term and cyclical, then theUnited Kingdom would likely be able to return to trend growthquickly as the economy recovered without it resulting inexcessive inflation. On the other hand, a longer-termslowdown in labour productivity would point to a slowerrecovery, where wage growth, which is directly related toproductivity growth, would be lower.

Most assessments of the productivity puzzle appeal to anumber of factors to explain the apparent weakness.(1) Theseinclude some cyclical factors weakening productivity growthat the start of the crisis, for example as firms held on toworkers at the cost of productivity.

But longer-term factors are also likely to have been important— especially those surrounding the impact of the financialcrisis. Lower investment in capital, barriers to the efficientallocation of capital and the survival of lower-productivity

firms throughout the crisis are all likely to have played somerole in reducing productivity growth for longer than cyclicalvariation in productivity alone would suggest.

Given the likely direct impact of compositional effects onproductivity growth it is worth examining whether they canplay a role in explaining this puzzle. They do not provide ananswer to the question of why productivity growth wassubstantially lower in the aftermath of the crisis. In factcompositional effects were likely to be pushing up onproductivity growth between 2008 and 2013, making theweakness in productivity growth all the more puzzling.

More recently, however, compositional effects do appear toexplain part of the below-trend productivity growth. Chart Bshows productivity growth adjusted for compositional effects.With these adjustments recent productivity growth is closer toits pre-crisis trend. As the drag from compositional effectssubside that could boost productivity growth.

A reduced drag from compositional effects onproductivity growth is one of the reasons the February 2016Inflation Report projects increasing productivity growth in themedium term.

Hourly productivity growth

Compositionally adjusted hourly productivity growth

+

–

6

4

2

0

2

4

6

1997 99 2001 03 05 07 09 11 13 15

Percentage changes on a year earlier

Chart B Adjusting labour productivity for compositionaleffects

Sources: ONS and Bank calculations.

+

–

6

4

2

0

2

4

6

1997 99 2001 03 05 07 09 11 13 15

Hourly productivity growth

1997–2007 average

2008–15 average


Chart A Labour productivity growth declined after thecrisis

Sources: ONS and Bank calculations.

(1) See, for example, Barnett et al (2014).


While the effects of education, tenure and age have changedsubstantially, the role of occupation and industry have beenmuch more constant over time. There is a large and stablepremium in the earnings of high-skilled jobs relative to allothers of between about 35%–50%. This contrasts with amuch smaller gap between medium and low-skilledoccupations of around 6% of earnings.

The pay gaps between industries are similar in size to thosebetween occupations. Relative to a benchmark ofmanufacturing employees, sectors paying a premium includefinance (19%), information, communications and technology(9%) and those in professional services (3%). In contrast,those in sectors such as health (-13%), education (-17%) oraccommodation and food services (-21%) on average earnsubstantially less. In each case this is comparing workers withsimilar qualifications and roles across different industries.These gaps have been fairly stable over time.

While other factors did not play a significant role in thecompositional effects on wages in the past 20 years, they stillprovide useful information. The pay gap between women andmen, for example, has narrowed significantly over timeaccording to these estimates. In 1995 women were paid 16%less than men for a given set of characteristics; by 2015 thisfigure had fallen to 9%. This narrowing of the gap likelyexplains why the increasing participation of women in theworkforce has not been strongly pulling down on aggregatewage growth.

Wage differentials between different regions also appear, to adegree, to have converged. Relative to the South East,employees in London were paid on average 9% more in 1995;this has since declined to 7%. Meanwhile parts of theUnited Kingdom that experienced lower pay comparedwith the South East, such as the North of England, theWest Midlands, Scotland and Wales, have all seen this paygap narrow slightly.

Finally, these estimates suggest that the pay premium forthose working in the public sector fell substantially in the late1990s from a peak of 10% relative to those in the privatesector in 1996, to around 4% in 2002. Since then it hasremained relatively unchanged.(1)

Conclusion

Changes in characteristics such as education, occupation andage can have a significant effect on wage growth. Over thepast 20 years, as the United Kingdom’s workforce has becomemore educated, moved to higher-skilled roles and aged, theseeffects are estimated to have pushed up on annual wagegrowth by an average of roughly ½ a percentage point.

Following the financial crisis, due to large changes in the flowsinto and out of employment, these compositional effectsbecame particularly pronounced. Initially, they pushed up onwage growth by up to a percentage point relative to averageas high earners stayed in employment longer than expectedand lower earners became unemployed. At the end of 2013these shifts went into reverse and compositional effectspushed down wage growth. This helps to explain some of theweakness in wage growth that was seen in 2014 and 2015.

More recently, data for 2015 Q4 show that compositionaleffects have started to subside. The drag on wage growth islikely to dissipate as the labour market normalises and theeffect on wages of changes in the composition of theworkforce returns to normal.

(1) The ONS notes that public sector workers typically work for larger employers whichprovide a pay premium relative to smaller employers. Once this is taken intoaccount, the ONS finds that public sector workers earn less than their private sectorcounterparts. This highlights the sensitivity of some of our results to the controlsthat are used. See ONS (2014).

5

0

5

10

15

20

25

30

35

1995 97 99 2001 03 05 07 09 11 13 15

Per cent pay increase relative to those with 0–3 months’ tenure(four-quarter moving average)

3–6 months

6–12 months

1–2 years

2–5 years

5–10 years

10–20 years

20+ years

+

–

Chart 8 The declining impact of tenure on wages


45

40

35

30

25

20

15

10

5

0

5

10

16 to19

20 to24

25 to29

30 to34

40 to44

45 to49

50 to54

55 to59

60 to64

65 to69

Average percentage change in payrelative to 35–39 age group

1995

2007

2015

+

–

Chart 9 The age-earnings profile relative to thoseaged 35–39



AnnexCalculating compositional effects

In order to calculate the compositional effect on wagegrowth between two given periods a Oaxaca decomposition isused, following the approach used by Blundell, Crawford andJin (2013).

This decomposition splits the wage change between two timeperiods, between observed components, such as workexperience, tenure and industry and a residual that cannot beexplained by these observable characteristics. In this analysis,this unobserved component is a proxy for factors such astechnological change, which are likely to push up onproductivity and pay of the entire workforce over time.Changes in wages between two time periods, which can beexplained by changes in observed characteristics, are recordedas ‘compositional effects’.

Formally, the estimation is based on a linear model wherewages at time t are determined by a vector of characteristicsXt plus an error term:

Yt = Xt’Bt + et

The change in wages between time t and t+n can be written as:

E(∆Yt+n) = Compositional effect+ Coefficient effect+ Interaction term

Where these specific effects can be written as:

Compositional effect = (E(Xt+n) – E(Xt))'Bt

Coefficient effect = E(Xt+n)'(Bt+n – Bt)

Interaction term = (E(Xt+n) – E(Xt))'(Bt+n – Bt)

The coefficient effect and interaction term are combined andtreated as an unexplained component of wage growth in theanalysis presented.

The specific linear model which is estimated is the following:

In (hourly_wages)i= genderi + highest_qualificationi+ industryi + age_groupi + tenurei+ regioni + occupationi+ public_sectori + full_timei+ temporary_contracti

All of the variables in the regression, with the exception ofwages, are discrete variables which are programmed as a seriesof dummy variables described in Table A1.

Hourly wages are used to abstract from the issue of varyingaverage hours worked in the economy.

The model is estimated on LFS quarterly data from1994 Q1–2015 Q4 where individual observations are droppedif any individual characteristics used in the regression aremissing.

Table A1 Characteristics included in compositional effectsregression

Variable Categories

Qualification Degree; Higher education; A-level or equivalent; GCSE or equivalent; Other qualification; No qualification; Don’t know.

Industry SIC2007 Industry section codes.

Age groups 16–19; 20–24; 25–29; 30–34; 35–39; 40–44; 45–49; 50–54;55–59; 60–64; 65–69; 70+.

Tenure < 3 months; 3–6 months; 6–12 months; 1–2 years; 2–5 years;5–10 years; 10–20 years; 20+ years.

Region North; Yorkshire & Humber; East Midlands; East Anglia; London;South East; South West; West Midlands; North West; Wales;Scotland; Northern Ireland.

Occupation Managers, directors and senior officials; Professional occupations;Associate professional and technical occupations; Administrativeand secretarial; Skilled trades; Personal services; Sales andcustomer services; Process, plant and machine operatives;Elementary occupations.

Gender Male; Female.

Working pattern Full-time; Part-time.

Sector Private sector; Public sector.

Contract type Permanent; Temporary.


References

Barnett, A, Batten, S, Chiu, A, Franklin, J and Sebastiá-Barriel, M (2014), ‘The UK productivity puzzle’, Bank of England Quarterly Bulletin,Vol. 54, No. 2, pages 114–28, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q201.pdf.

Bell, V, Burriel-Llombart, P and Jones, J (2005), ‘A quality-adjusted labour input series for the United Kingdom (1975–2002)’, Bank of EnglandWorking Paper No. 280, available at www.bankofengland.co.uk/research/Documents/workingpapers/2005/wp280.pdf.

Blundell, R, Crawford, C and Jin, W (2013), ‘What can wages and employment tell us about the UK’s productivity puzzle?’, IFS Working Paper W13/11, available at www.ifs.org.uk/wps/wp201311.pdf.

Connors, E and Franklin, M (2015), ‘Quality adjusted labour input, estimates to 2014’, available atwww.ons.gov.uk/ons/dcp171766_402986.pdf.

Leaker, D (2008), ‘The gender pay gap in the UK’, ONS Economic and Labour Market Review, Vol. 2, No. 4, pages 19–24.

Office for National Statistics (2014), ‘Public and private sector earnings — November 2014’, available atwww.ons.gov.uk/ons/dcp171776_383355.pdf.

Office for National Statistics (2015), ‘Understanding average earnings for the continuously employed — using the annual survey of hours andearnings 2014’, available at www.ons.gov.uk/ons/dcp171766_404125.pdf.


www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q201.pdf

Topical articles Bank of England notes: the switch to polymer 23

Bank of England notes: the switch topolymerBy Ronan McClintock and Roy Whymark of the Bank’s Notes Directorate.(1)

• The Bank of England is responsible for maintaining confidence in banknotes. • In meeting its banknote issuance responsibilities, the Bank aims to fully exploit innovation. The next £5, £10 and £20 banknotes will be printed on a polymer material.

• The switch to polymer will deliver banknotes that are more secure and better quality, and willsupport confidence in banknotes in the years ahead.

Overview

The issuance of banknotes is probably the Bank of England’smost recognisable function. Having first issued banknotesshortly after it was founded in 1694, the Bank is one of thelongest-standing issuers of physical money in the world.

The Bank of England’s note issuance objectives are to: (i) meet demand for banknotes in the quantities anddenominations required by the public; and (ii) maintainconfidence in banknotes. The key to maintaining confidenceis the distribution of banknotes that are difficult tocounterfeit and easy to authenticate.

During the next few years, the Bank will introduce banknotesprinted on a polymer material and incorporating newsecurity features for the £5, £10 and £20 denominations.These banknotes will also be smaller than their currentcotton-paper equivalents.

The decision to move away from using cotton paper wastaken after a three-year research programme and followedconsultation with a wide range of stakeholders within thecash industry, as well as the public.

The move is not without precedent. In 1988, Australia wasthe first country to issue a banknote using polymer. To date,45 billion banknotes have been produced on polymer, andissued in over 30 countries including Australia, Canada,Gambia, Fiji and New Zealand.

The new banknotes will deliver three benefits. First, thepolymer banknotes, coupled with leading-edge securityfeatures, will deliver a step change in counterfeit resilience.Second, polymer banknotes will be stronger and more

durable, meaning they will last at least two and a half timeslonger than cotton-paper banknotes. And third, the newbanknotes will be cleaner, and the public will enjoy thebenefit of better-quality banknotes in their pockets.

The new polymer £5 banknote, featuring Sir Winston Churchill, will be unveiled on 2 June, and willenter circulation in September 2016. Around a year later, the Bank will launch a new £10 banknote featuring Jane Austen. A new £20 banknote, featuring a character from the visual arts, will be launched by 2020.

This programme of work represents the biggest innovation inBank of England notes for many years. The results will bebeneficial to all users of cash and will maintain publicconfidence in banknotes in the years ahead.

Summary figure Concept designs for the new £5 and £10 banknotes

(1) The authors would like to thank Tom Fish for his help in producing this article.


Introduction

In the next few years, the Bank of England will be launchingnew £5, £10 and £20 banknotes on a polymer material,moving away from the cotton paper that has been used forhundreds of years. The design of the current cotton-paper£50 banknote is comparatively new and no decision has yetbeen taken on whether to replace it with a new polymerbanknote.

This article discusses the programme of work that willculminate in the launch of the new banknotes, considers therationale behind the change, and explains the steps taken indelivering it.

The first section reviews the Bank’s responsibilities withrespect to banknotes, and recent patterns of demand. Thesecond section explores the counterfeit challenge and how theBank responds to the threats to our physical currency. Thethird section outlines the research which identified polymerfor the next generation of banknotes, before the fourth sectionexplains the benefits of polymer banknotes. The fifth sectiondiscusses the cash industry and public consultation processwhich fed into the decision. Finally, the article discusses theprogramme of work put in place to launch the new banknotes.

Bank of England objectives and demand for cash

Banknotes and the Bank of England’s objectivesAs discussed in a previous Bulletin article,(1) there are twotypes of money held by firms and households: cash anddeposits with financial institutions.(2) The vast majority of thismoney — some 97% — is held electronically as deposits, withthe remainder held in physical form as cash. Bank of Englandnotes make up most of the physical cash that circulates in the United Kingdom but, as explained in the box on page 25,cash also comprises coin issued by the Royal Mint, andbanknotes issued by seven commercial banks in Scotland and Northern Ireland.

The Bank’s mission is to promote the good of the people ofthe United Kingdom by maintaining monetary and financialstability. As part of delivering monetary stability, the Bankseeks to maintain the public’s confidence in banknotes.

Banknotes should be of good quality, easy to authenticate,and resilient against the threat of counterfeiting. They shouldalso be readily available, and in the denominations that thepublic needs.

Banknote demandIn the modern economy, consumers are increasingly using awide range of payment methods. However, as shown in

Chart 1, the value of Bank of England notes in circulationcontinues to grow, and has increased by around three quartersover the past decade.

At the end of December 2015, there were around three and ahalf billion Bank of England notes in circulation, across theBank’s four denominations, totalling around £65 billion (Table A).(3)

Cash possesses a unique set of attributes that sets it apartfrom other payment methods — it is tangible, fast andconvenient, almost universally accepted, anonymous, andhelpful for budget management.

As explained in a previous Bulletin article,(4) the Bank expectsdemand for cash to remain strong over the coming years andis investing in the banknotes that it issues.

Addressing the threats to our physicalcurrency

The counterfeit challengeThe counterfeiting of banknotes, and knowingly passing themon, are serious crimes. Since counterfeits are worthless,

(1) See McLeay, Radia and Thomas (2014) for a detailed introduction to money.(2) A third type of money is central bank reserves — deposits placed by commercial

banks with the Bank of England. (3) Note, the value of notes in circulation is always much higher in December than during

the rest of the year. This is due to increased demand for banknotes over theChristmas period.

(4) See Fish and Whymark (2015) for an analysis of how cash usage has evolved and whatmay drive demand in the future.

£50

£20

£10

£5

£1

1975 80 85 90 95 2000 05 10 15

Value of Bank of England NIC, £ billions

0

10

20

30

40

50

60

70

Chart 1 Value of Bank of England notes in circulation(NIC): 1975 to present

Table A Bank of England notes by denomination at end-December 2015

£5 £10 £20 £50 Total

£1.7 billion £8.3 billion £42.1 billion £13.0 billion £65.1 billion

343 million 834 million 2,105 million 260 million 3,542 million notes notes notes notes notes


anyone who accepts one runs the risk of being out of pocket. It is therefore important that the Bank tackles the challengefrom counterfeiting so that people and businesses are able toquickly and easily check that the banknotes they receive aregenuine.

The counterfeit challenge is not new, and is not unique to the United Kingdom. Since the introduction of Bank ofEngland notes in the late 17th century, criminals haveattempted to forge them. However, in recent years, the rapid advancement of digital printing technology hasmade counterfeiting more affordable and accessible. The challenge for the Bank, in common with other centralbanks, is therefore to stay one step ahead of the counterfeiting threat.

Addressing the counterfeit challengeThe Bank’s anti-counterfeiting strategy is based on five pillars.

First, the Bank is committed to developing state-of-the-artcounterfeit resilient banknotes. It is currently developing newpolymer £5, £10 and £20 banknotes which will incorporateleading-edge security features.

Second, as discussed in the box on page 29, the Bank worksclosely with the cash industry so that only high-quality,

authentic banknotes are issued and recirculated. This isbecause manual authentication, and the processing ofbanknotes through machines (such as ATMs), are moredifficult when banknotes are of poor quality.

Third, the Bank has an active education programme that works with businesses and the public to help peopleunderstand how to identify genuine banknotes. The Bank’sBanknote Education Team issues a suite of (free-of-charge)materials including leaflets, posters and a smartphone app.(1) Although some are easier to spot than others, all counterfeits can be identified using guidance supplied by the Bank of England.

Fourth, the Bank — through the Code of Conduct for the Authentication of Machine-Dispensed Banknotes — encourages effective authentication of banknotes that have been locally recycled. Local recycling occurs when a bank or retailer accepts a banknote from a customer and then dispenses it to another customer via a machine (such as an ATM or a self-service checkout). The Code requires that all banknotes dispensed in this way are first authenticated using a device listed in the Bank’s

UK coin and Scottish and Northern Irelandbanknotes

Although cash in circulation in the United Kingdompredominantly comprises Bank of England notes, there are two other components (Chart A).

The United Kingdom’s coin policy is managed by HM Treasury,with coins produced by the Royal Mint.(1) At the end of December 2015, there was £4.4 billion of coin in circulation. In 2017, the Royal Mint will launch a new £1 coin. It will be constructed from two different coloured metals and contain a new security feature, designed to make the coin more counterfeit resilient.

As explained in a previous Bulletin article,(2) the Government also authorises seven commercial banks to issue banknotes in Scotland and Northern Ireland (S&NI banknotes). The aggregate value of S&NI banknotes in circulation totalled £7.3 billion at the end of December 2015. The three Scottish commercial issuers — Bank of Scotland, Clydesdale and the Royal Bank of Scotland — have each announced plans to introduce polymer for their £5 and £10 banknotes.(3)

These banknotes will be issued at a similar time to the new Bank of England notes.

The Bank is working alongside both the Royal Mint and theScottish issuers to support a smooth transition for businessesand the public.

0

10

20

30

40

50

60

70

Bank of England notes S&NI notes UK coin

£ billions

Chart A Value of UK cash by component at end-December 2015

(1) See www.bankofengland.co.uk/banknotes/Pages/default.aspx.

(1) See www.royalmint.com/.(2) See Naqvi and Southgate (2013) for a detailed summary of the legal framework for

the issuance of banknotes by commercial banks in Scotland and Northern Ireland.(3) The Northern Ireland issuers are: AIB Group (UK) plc (trades as First Trust Bank);

Bank of Ireland (UK) plc; Northern Bank Limited (trades as Danske Bank); and Ulster Bank Limited. None of these have announced any plans to move to polymerbanknotes. There are also a number of non-UK sterling banknotes that are issued inthe Isle of Man, Channel Islands and Gibraltar. These are issued under currency boardarrangements, with banknotes backed one-for-one with UK pound sterling. However,unlike for the S&NI banknotes, the Bank of England has no role in thesearrangements.

www.royalmint.com/

www.bankofengland.co.uk/banknotes/Pages/default.aspx


Framework for the Testing of Automatic Banknote HandlingMachines.(1)

Fifth, and importantly, the Bank works closely with lawenforcement agencies across the United Kingdom to disruptcounterfeiting operations and to bring the perpetrators tojustice. Where necessary, the Bank will assist the courts byproviding expert witness support.

Counterfeit levelsThe likelihood of an individual being passed a counterfeit isvery low. In 2015, 243,000 counterfeits — simulating Bank of England notes — were discovered and removed fromcirculation (Chart 2). This is equivalent to only a tinyproportion of 1% of all Bank of England notes in circulation.

Choosing polymer banknotes

Research phase Led by its specialist research team of scientists, engineers andother technical experts, the Bank of England commenced athree-year research programme in 2010 examining options forthe next generation of banknotes.

Given the accelerated pace of innovation in recent years,central banks now have a greater choice of materials thatbanknotes can be printed on. The most suitable materials can be split into three broad categories: (i) paper; (ii)polymer; and (iii) hybrid materials — a combination of thetwo. There are various specifications within these categories

0

100

200

300

400

500

600

700

800

2006 07 08 09 10 11 12 13 14 15

Thousands

Chart 2 Volume of counterfeits removed fromcirculation

The Bank of England’s anti-counterfeiting strategy

Newstate-of-the-art

polymerbanknotes

Ensuringhigh-quality

banknotes areissued and

recirculated

Banknotesecurity

feature educationfor the public and

retailers

The local recyclingcode and the

framework forcash-handling

machines

Partnership with law

enforcement

Addressing the counterfeit challenge

(1) See www.bankofengland.co.uk/banknotes/Pages/retailers/framework.aspx.

www.bankofengland.co.uk/banknotes/Pages/retailers/framework.aspx


although, in general, paper banknotes are most commonlymade from cotton fibres, while polymer banknotes are madefrom polypropylene — a transparent and flexible plastic film.Hybrid banknotes will often use multiple layers of the twomaterials (for example, one layer of cotton paper sandwichedbetween two layers of polymer or vice versa).

Technological advances have also provided central banks with a wide choice of security features for incorporation intobanknotes. These range from holograms and micro-lensfeatures through to colour-changing and metallic inks.Choosing security features is a complex process and dependson factors such as the current counterfeiting rate, counterfeitquality, and authentication practices followed by the public,retailers and financial institutions.

At the heart of the Bank’s research was counterfeit resilience.The Bank undertook extensive testing in an attempt tosimulate the efforts of counterfeiters to forge different typesof banknotes incorporating different materials and securityfeatures. The analysis made use of a wide range of techniquesfrom amateur digital printing through to professionalindustrial-scale manufacturing.

With each trial, the Bank developed an understanding of thenecessary skill, time and cost required to counterfeitbanknotes at different levels of sophistication. This enabledthe Bank to rank the combinations of materials and securityfeatures in terms of counterfeit resilience.

Extensive durability testing was also undertaken. This involvedexposing the materials and security features to a wide range ofeveryday household chemicals and activities to measure thewear resistance to abrasion and folding. One of the tests, forexample, involved putting different sample banknotes throughwashing machine cycles.

The research also drew on the experiences of other central banks and included an independent study of theenvironmental impact of both polymer and cotton paper.

The research concluded that there was a strong case forreplacing existing cotton-paper banknotes with newbanknotes printed on polymer. While hybrid materials were also considered, these were ultimately discounted given a lack of available data from real-world deployment.

The benefits of polymer banknotes

Banknotes that are printed on polymer offer three keybenefits, each discussed below.

SecureThe characteristics of polymer have been carefully crafted forbanknotes so that it is difficult for a counterfeiter to source a

similar material which feels the same and is easy to print on.When combined with leading-edge security features, polymerbanknotes can constitute a step change in counterfeitresilience, leading to safer, and more secure, banknotes.In general, printing on polymer is more difficult than printingon paper and requires specialist printing equipment that is noteasily available.

Transparent windows can also be included in polymerbanknotes. These see-through windows can be combined with sophisticated security features which, together, enablethe manual authentication of a banknote from both sides. In addition, these security features can be designed to look different depending on which side of the banknote isbeing inspected — adding another level of counterfeitresilience.

Evidence from countries that have switched to polymer —including Australia, Canada and New Zealand — supports the conclusion that polymer banknotes are more secure. Allcountries that have introduced polymer banknotes havereported a decline — often a significant one — in counterfeitlevels, with very low residual volumes as a result. Althoughcounterfeiting is still attempted, the majority of counterfeitsproduced are of relatively poor quality (with many still beingprinted on paper).

DurablePolymer banknotes are stronger and more durable than theircotton-paper equivalents. The greater durability is due to thephysical structure of the material which makes it moreresistant to damage, such as tearing and folding. This meansthat polymer banknotes last at least two and a half timeslonger than cotton-paper banknotes.

The initial cost of producing a polymer banknote is higher thanfor a cotton-paper one. However, because polymer banknoteslast longer in circulation, fewer replacement banknotes needto be printed over the long term.

The enhanced durability of polymer banknotes is alsobeneficial from an environmental perspective. During theresearch phase, the Bank commissioned an independent studyto assess the environmental impact of both cotton-paperbanknotes and polymer banknotes.(1) The study examined thelife cycle of a banknote from the production of the rawmaterial, and then through the manufacturing, printing anddistribution phases before being destroyed and the wastedisposed. Polymer banknotes can in fact be recycled, with the material used to make other products (for example plant pots) and the Bank will fully separate paper and polymerbanknote waste to facilitate their recycling.

(1) See ‘LCA of paper and polymer bank notes’;www.bankofengland.co.uk/banknotes/polymer/Documents/lcapaperandpolymerbanknotes.pdf.

http://www.bankofengland.co.uk/banknotes/polymer/Documents/lcapaperandpolymerbanknotes.pdf

http://www.bankofengland.co.uk/banknotes/polymer/Documents/lcapaperandpolymerbanknotes.pdf


The study considered seven environmental indicators(1)

including global warming potential and water and energy use.Polymer banknotes scored better than cotton-paperbanknotes for all of the main phases of a banknote’s life cycle,and for five of the seven environmental indicators.(2)

The study concluded that polymer banknotes have a lowerenvironmental impact than their cotton-paper equivalents.

Greater durability will also benefit the cash industry. Sincetatty or damaged cotton-paper banknotes are often the causeof banknote machine jams (and therefore unavailability),better-quality banknotes should lead to greater efficiency in banknote processing.

CleanThe third key benefit of polymer banknotes is that they are cleaner than cotton-paper equivalents. Polymer isimpermeable and of a non-fibrous nature which means that it repels dirt and moisture. Coupled with the improveddurability, polymer banknotes remain in good condition for longer and become less tatty.

Feedback from the public suggests people put a premium onclean, good-quality currency and consider the banknotes intheir pocket as a source of national pride. Importantly,cleanliness and quality also support counterfeit resilience — since it is harder to check security features on a tattybanknote.

Consulting the cash industry and the public

Preliminary consultationDuring the research phase, the Bank undertook twoconsultations to better understand the impact on the cash-handling industry of launching banknotes using a newmaterial, and in a number of different size configurations. Theexercise found that businesses would be able to adapt andthat, although the cash industry identified that there would betransition costs, there was also recognition of the long-termbenefits.

The Bank also sought early feedback from the public. Itcommissioned a number of independently-run focus groupsthat measured the public’s reaction to different banknotematerials. Feedback from many participants was favourabletowards polymer banknotes. The Bank also sought advicefrom special interest groups, including those representingblind and vision-impaired people.

Further consultation on the new £5 and £10banknotesBased on the findings of the research phase, the Bankannounced in September 2013 that it was considering the case for introducing polymer banknotes for the next

series of £5 and £10 banknotes, and would seek the views of the public.

At the same time, the Bank announced that, if polymer werechosen, the new banknotes would be 15% smaller than thecurrent series equivalents (which are large by internationalstandards). The preference for making the banknotes smallermeant less material would be used in their production,reducing manufacturing costs and delivering furtherenvironmental benefits.

Further public consultationTo gauge wider public opinion, the Bank undertook a publicconsultation between September and November 2013. Thisinvolved a ten-week roadshow, which included 47 events heldacross the United Kingdom (including, for example, inshopping centres). These events gave people the opportunityto talk with Bank staff and to look at, and handle, samplepolymer banknotes. To provide more in-depth analysis, theBank also commissioned further independently-run focusgroups and surveys.

Altogether, the consultation included over 18,000 interactionswith the public, with almost 13,000 people providingfeedback. The feedback was overwhelmingly supportive with87% of respondents in favour of polymer banknotes.(3)

Further industry consultationThe cash industry always needs to make changes when a newbanknote is issued — firms need to upgrade cash machines,train staff and modify handling procedures. To support thisprocess, the Bank began consulting more widely with allsectors of the cash industry from September 2013, building onthe results gathered from the earlier engagement.

The Bank consulted financial institutions, ATM providers,retailers (including representative trade bodies) and cash-in-transit firms. The Bank also engaged with banknoteequipment manufacturers to better understand the impact ofthe change on technology.

The consultation process was very valuable. Overall, it gave the Bank confidence that businesses would be able toadapt and that the long-term benefits of the new notes wouldoutweigh the short-term transition costs for the cash industry.The transition costs fall particularly on those businesses thatuse machines to handle cash — for example banks with largeestates of ATMs and large supermarkets that use a lot of self-service checkout machines.

(1) The seven environmental indicators assessed were: (i) primary energy demand — non-renewables; (ii) water consumption; (iii) acidification potential; (iv) eutrophication potential; (v) global warming potential; (vi) human toxicitypotential; and (vii) photochemical ozone creation.

(2) Two indicators were inconclusive: (i) human toxicity potential; and (ii)photochemical ozone creation.

(3) Sixty-nine per cent of respondents were strongly in favour of polymer banknotes with18% somewhat in favour. Only 6% of the people consulted were opposed topolymer banknotes (with 7% neutral). To view the full results of the consultation, seewww.bankofengland.co.uk/banknotes/polymer/Pages/pcp.aspx.


Retailers Consumers

Banknotes are printed at the Bank’s facility in Debden, Essex by the Bank’s partner, De La Rue plc.(1)

The Bank of England checks new banknotes for quality before supplying them toNotes Circulation Scheme (NCS) members.(2)

NCS members supply banknotes to large retailers and commercial banks (who supply smaller banks and retailers).

Banknotes are transported throughout the system by cash-in-transit vans before reaching bank branches, businesses and ATMs.

Businesses increasingly rely on technology for dispensing, accepting, authenticating and counting cash (for example, there are now around 70,000 ATMs and 35,000 self-service checkouts in the United Kingdom).

Once a banknote is deposited at a commercial bank, it will either authenticate and recirculate the banknote (local recycling) or return the banknote to an NCS member.

NCS members check for authenticity and quality — banknotes that are no longer of good quality are returned to the Bank for destruction and replaced by new banknotes. ‘Fit’ banknotes will re-enter circulation.

In delivering confidence in banknotes and meeting demand, the Bank relies on a large and complex banknote supply and distribution chain.

The banknote cycle

(1) The polymer for the new £5 and £10 banknotes will be supplied by Innovia Security. The Bank of England plans to tender for the polymer for the next £20 banknote in Spring 2016.As well as the material, the ink used on the banknotes and the various security features that are incorporated all need to be sourced from the commercial sector.

(2) There are currently four NCS members: G4S Cash Solutions (UK) Ltd; Post Office Ltd; Royal Bank of Scotland plc; and Vaultex UK Ltd (a joint venture between Barclays plc and HSBC plc).

As well as confirming the general support for polymer, it helped to identify specific challenges that had to beaddressed. For example, the feedback raised helpful questions on how to manage the issuance of new banknotesthrough the complex distribution chain which is summarisedin the figure above.

Launching the new banknotes

The decisionHaving assessed the results of the public and industryconsultations, the Bank announced in December 2013 itsdecision that the next £5 and £10 banknotes would be printed


Banknote characters

Since 1970, the Bank has included characters of significance onits banknotes, alongside supporting imagery commemoratingtheir lives. This practice provides an opportunity to celebrateindividuals that have shaped British thought, innovation,leadership, values and society.

The first character to appear on a Bank of England note wasWilliam Shakespeare. He featured on the £20 banknote thatwas launched in 1970. In the four and a half decades since,fifteen characters have appeared on fourteen banknotes.(1)

The way the Bank chooses banknote characters has changed.With the selection of the character to appear on the £20 banknote, for the first time the selection drew heavily onthe views of the public and involved a largely independentBanknote Character Advisory Committee.(2)

The next £5 and £10 banknotesThe new £5 banknote will be unveiled on 2 June and will entercirculation in September 2016. In 2013, the Bank announcedthat it would feature Sir Winston Churchill (Figure A). FormerBank Governor, Mervyn King noted ‘Sir Winston Churchill was a truly great British leader, orator and writer. Above that, heremains a hero of the entire free world. His energy, courage,eloquence, wit and public service are an inspiration to us all’.

In July 2013, the Bank announced that the next £10 banknotewould feature the author Jane Austen (Figure B). It will entercirculation in 2017, around a year after the new £5 banknote.In making the announcement, Governor Mark Carney said‘Jane Austen certainly merits a place in the select group ofhistorical figures to appear on our banknotes. Her novels havean enduring and universal appeal and she is recognised as oneof the greatest writers in English literature’.

The next £20 banknoteIn May 2015, the Bank launched a two-month publicnomination process asking for nominations, from the field of visual arts, for which character should appear on the next £20 banknote. The initiative generated nearly 30,000nominations, covering 590 eligible visual artists.

The Banknote Character Advisory Committee then considered all eligible nominations and, over a number ofstages, produced a shortlist of names. At each stage, theCommittee took into account the contributions made by each visual artist, while also considering the equalityimplications of different choices. The final decision — basedon a shortlist of names proposed by the Committee — is made by the Governor. The chosen character will beannounced in Spring 2016, alongside a concept image showing how the character will appear on the new £20 banknote.

Figure A Concept design for the Sir Winston Churchill £5 banknote

Figure B Concept design for the Jane Austen £10 banknote

(1) Entrepreneur Matthew Boulton and engineer James Watt are jointly commemoratedon the current £50 banknote.

(2) The Banknote Character Advisory Committee is chaired by Ben Broadbent, Deputy Governor for Monetary Policy. The full members are three external experts — Sir David Cannadine, Sandy Nairne and Baroness Lola Young of Hornsey, and theBank’s Chief Cashier — Victoria Cleland. John Akomfrah, Andrew Graham-Dixon andAlice Rawsthorn — all experts in the field of visual arts — have joined the Committeespecifically for the £20 banknote character short-listing process.


on polymer, and that the new banknotes would be smaller. As discussed in the box on page 30, the new £5 banknote willfeature Sir Winston Churchill and the new £10 banknotewill feature Jane Austen. A decision to introduce a newpolymer £20 banknote was taken later and was announced in September 2015.

ProductionWith the material and security features agreed, the Bankcommissioned production for the new £5 banknote. Initially,this involved a number of small-scale manufacturing trials toensure that the proposed design could be produced in volumeto a uniform quality. Following this, two full-scalemanufacturing trials were completed, providing opportunity toresolve any issues in advance of the start of mass production.

During the trial process, the Bank supplied test banknotes,under strict security controls, to banknote equipmentmanufacturers (for example, firms that manufacture ATMs) toassist them with upgrading their equipment in order to accept,dispense and authenticate polymer banknotes. This is anessential part of the process of launching a new banknote andhelps to get cash-accepting and dispensing machines readyahead of the launch date.

De La Rue is currently producing the £5 banknote ready for issue in September 2016. Ahead of the launch, over 400 million banknotes will be produced. This will allow forthe replacement of all the cotton-paper £5 banknotescurrently in circulation as well as those held as stocks either by the industry or the Bank itself. Ahead of its launch in 2017,over one billion £10 banknotes will be manufactured.

Supporting cash industry readinessThe Bank is committed to working collaboratively with thecash industry to support a smooth transition to the new notesand design an approach to the transition that minimises thecosts of change. To support this goal, following the decision in2013 to issue new £5 and £10 notes, the Bank put in place acomprehensive industry liaison programme.

The programme is built on regular dialogue and encouragescollaboration and information sharing across the industry toresolve any challenges that arise. As well as running regularindustry-wide forums, the Bank has commissioned fiveindustry-led working groups, representing all parts of the cashdistribution chain.

Over the past two years, the groups have workedcollaboratively to prepare for the launch in the areas of: (i) wholesale banknote processing; (ii) ATMs; (iii) otherbanknote equipment — for example, vending, self-servicecheckouts; (iv) cash-in-transit; and (v) cash handling andtraining — for example, for retail staff. This collaboration hasbeen invaluable in planning the rollout of the new banknotes.

Feedback from industry participants indicates a high level ofconfidence in how the programme is moving forward.

Public education and awarenessA successful launch will also depend on awareness andacceptance of the new banknotes by the public. From 2 June 2016, the Bank will embark on an extensivecommunications campaign. The campaign will include theBank visiting different regions of the United Kingdom andengaging with the media, the public and local businesses, and providing a range of free educational materials. Theintention is to ensure people are ready, and to build awareness for the launch of the new state-of-the-artbanknotes.

Withdrawal of the current £5 and £10 banknotesHaving launched the new polymer £5 banknote in September 2016, the Bank will — after a suitable period oftime — withdraw legal tender status for the current cotton-paper £5 banknote. This will occur before the newpolymer £10 banknote enters circulation. Subsequently, thecurrent cotton-paper £10 banknote will also be withdrawn.

After legal tender status is withdrawn, customers will nolonger be able to use old series banknotes for purchasinggoods and services. However, the Bank will give plenty ofnotice so that consumers have sufficient time to pay currentseries banknotes into bank accounts or to spend them. Bank of England notes contain a promissory clause stating that the Bank will pay the bearer the sum of the banknotes on demand. In practice, this means that the current cotton-paper banknotes will be exchangeable at the Bank of England for new banknotes (or electronic payment) of the same value forever. For example, even after itswithdrawal in April 2014, the £50 banknote featuring Sir John Houblon can still be swapped by the Bank for thenewer £50 banknote, which features Matthew Boulton andJames Watt.

Conclusion

Demand for cash will likely be resilient in the years ahead. As such, the Bank needs to work with the cash industry andcontinue to invest in banknotes.

Following a comprehensive programme of research anddevelopment, the Bank of England will launch the next £5, £10 and £20 banknotes on polymer. Polymer provides astrong foundation for increasingly sophisticated, leading-edge,security features. The new banknotes will be state-of-the-artand will provide a step change in counterfeit resilience. The banknotes will also be cleaner and more durable, and will deliver confidence in Bank of England notes for theyears ahead.


(1) See www.bankofengland.co.uk/publications/Pages/news/2016/040.aspx.

In the months leading up to the launch of the new polymer £5 banknote in September 2016, the Bank will continue towork with the cash industry to deliver readiness. Victoria Cleland, the Bank’s Chief Cashier, recently remarked‘On Thursday 2 June we will unveil the full design of the newfiver — the Bank of England’s first polymer banknote. The newfiver will bring a step change in counterfeit resilience andquality. We have been working extensively with the cashindustry to ensure a smooth transition to polymer. Now is the

time for retailers and businesses to prepare. Alongside thelaunch, we will release new free-of-charge training materialsto help businesses train their staff, and run an extensive publicawareness campaign to enable everyone to prepare for thenew fiver entering circulation in September. This is an excitingtime for banknotes and we are grateful to the cash industry forhelping us introduce polymer banknotes’.(1)

References

Fish, T and Whymark, R (2015), ‘How has cash usage evolved in recent decades? What might drive demand in the future?’, Bank of England Quarterly Bulletin, Vol. 55, No. 3, pages 216–27, available atwww.bankofengland.co.uk/publications/Documents/quarterlybulletin/2015/q301.pdf.

McLeay, M, Radia, A and Thomas, R (2014), ‘Money in the modern economy: an introduction’, Bank of England Quarterly Bulletin,Vol. 54, No. 1, pages 4–13, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q1.pdf.

Naqvi, M and Southgate, J (2013), ‘Banknotes, local currencies and central bank objectives’, Bank of England Quarterly Bulletin,Vol. 53, No. 4, pages 317–25, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2013/qb130403.pdf.

www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2013/qb130403.pdf

www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q1.pdf

www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2015/q301.pdf

http://www.bankofengland.co.uk/publications/Pages/news/2016/040.aspx

Topical articles Capturing the City: Photography at the Bank of England 33

Figure 2 David Pollard ARPS, One New Change — View of St Paul’s Cathedral dome, 2015.

An exhibition at the Bank of England Museum and an Archive cataloguing project shed light ondifferent aspects of the organisation’s history through photographs.

Viewed from outside, the Bank of England can appear as a formalestablishment, a remnant of past times. We have come a long waysince Montagu Norman’s ‘never explain, never excuse’ era ofcentral banking. In recent years there has been a move to establish‘openness and accountability’ as a core pillar of the Bank’s missionto promote the good of the people of the United Kingdom. TheBank of England Museum’s latest exhibition, Capturing the City:Photography at the Bank of England, aims to show that beyond theaustere stone walls of the Bank’s building on Threadneedle Streetlies a hive of activity incorporating a remarkable architecturalhistory representing the dynamism of the institution, a diverserange of occupations and a vibrant social history.

The Bank of England Archive’s photography collection provides afascinating record of the Bank, its buildings and staff since theVictorian era. Capturing the City looks at the Bank’s past, andexplores the history of a medium that has become so populartoday.(1) This article shows a small selection of images, but thereare many more on show in the exhibition. The display will providean opportunity to see images of the Bank not easily accessible tothe public as well as artworks and artefacts not usually on display.

As part of the exhibition, the Bank of England Museum worked incollaboration with the Royal Photographic Society’s Londonchapter to produce an exciting project based on images from theBank Archive. Members of the Society, as well as the generalpublic, were challenged to choose an Archive image and takephotographs of the same location using as much creativity andimagination as they liked. The results show some of the manychanges in the City over the past century (Figures 1–4).

This article includes sections on life in the Bank and the Bankduring World War II. It also includes a box on the Bank of EnglandArchive’s cataloguing project, which has been running inconjunction with the planning and research of Capturing the City.

Capturing the City: Photography at theBank of EnglandBy Bryony Leventhall of the Bank of England Archive and Anna Spender, formerly of the Bank of England Museum.

(1) Visit the Capturing the City webpage for a summary of the exhibition; www.bankofengland.co.uk/education/Pages/museum/whatson/capturing-the-city.aspx.

Figure 1 New Change Accountant’s Department (nowOne New Change shopping centre), view of St Paul’s Cathedraldome from an 8th floor window, 1958 (15A13/2/36).

http://www.bankofengland.co.uk/education/Pages/museum/whatson/capturing-the-city.aspx

Bank life

Early photography Early photographic portraits relied heavily on the same principles as the painted portrait. Victorianphotographers used pillars and swathes of drapery to enhance the backgrounds of theirphotographs, just as portrait painters had. As smaller, less expensive cameras began to enter themarketplace, it generally became easier for photographers to leave the studio to produce informalimages. This resulted in the popular appeal of photography that is familiar today.

One of the Bank’s earliest photographic albums dates to the 1840s and depicts images of Governorsand Directors. Looking through the sepia-toned pictures, one can find the earliest photograph of aGovernor — William Cotton, who was in office 1842–45 (Figure 5). These portraits demonstrateone of the earliest photographic processes — the salt print. Salt prints were developed in 1840, andwere popular until the 1860s when they were replaced by the albumen print. Most salt prints weremade by soaking writing paper (giving them their characteristic matte appearance) in a dilutesodium chloride solution, and then left to dry. The paper was then treated with silver nitrate toform silver chloride which created light-sensitive paper. The paper was put directly beneath anegative and exposed to sunlight for up to two hours to produce the final image.

Figure 3 Princes Street from King William Street, 1890s(15A13/1/3/30).

Figure 4 Valerie McGlinchey, Princes Street fromKing William Street, 2015.


Figure 5 Salt print of William Cotton, Governor 1842–45(15A13/18/2/8).


Figure 9 Consols Office, 1894 (15A13/1/13).

Figure 6 Photograph of William Henry Clegg, c. 1930s. Notethe grid lines to help Clausen plan his painting (Figure 7).

Figure 7 William Henry Clegg, Director of the Bank ofEngland 1932–37, by Sir George Clausen (0605).

Figure 8 Digital photograph of Mark Carney, Governor2013–present (© Bloomberg 2013).

With the invention of the camera, portrait painting aided by photography grew in popularityduring the 19th century.(1) Photographs were often used as a reference tool to reduce the timespent sitting for an artist. A newly discovered photograph in the Bank Archive reveals that theartist Sir George Clausen (1852–1944) used this technique when painting his portrait ofWilliam Henry Clegg, Director of the Bank of England 1932–37 (Figures 6 and 7).(2)

Advances in photographic equipment and techniques over the past century have givenphotographers the ability to capture images with much shorter exposure times and to makeportraits outside the studio (Figure 8). Once a pursuit for a select group of photographers instudios, photography is now a pastime enjoyed by the general public, whether on highspecification SLR cameras, compact digital cameras or their smartphones.

Office technologyTechnology in banking, as well as in photography, has changed significantly over the years and onecan chart its progression through the Bank of England Archive photography collection. This isparticularly evident in the evolution of the Bank’s Accountant’s Department.(3) The originalCharter of 1694 entrusted the Bank of England as Registrar, keeping books to record stocktransfers, paying dividends and other aspects of stock management. In a photograph of theConsols Office, dated 1894, one can see numerous stock ledgers on countertops all of which weremeticulously handwritten (Figure 9). In stark contrast is a photograph recording the use of theProgramme Controlled Computer (PCC) in the Dividend Preparation Office, 1962 (Figure 10).Prior to the 1930s, clerks were expected to manually calculate and handwrite every warrant untilthe punched-card system was introduced in 1934. The PCC was introduced in the late 1950s.While still using the punched-card system to prepare the warrants for stockholders, the PCC’shigh-speed memory and mass storage meant that the process was faster and much more accurate.

The mid-1950s had seen a period of discussion and research into computerisation throughout thewhole of the UK banking industry. Following several proposals the Bank decided to trialPowers-Samas equipment.(4) It was described by the Chief Accountant as ‘a revolutionarydeparture from present methods and habits of thought’. The pilot scheme was the Bank’s firstdirect experience of computers and was installed at its new building on Cheapside, New Change(now the site of One New Change shopping centre).(5)

(1) Writing to the politician Sir Charles Dilke in 1873, the portrait painter George Frederick Watts said that photographs ‘help to make oneacquainted with peculiarities and shorten the sittings necessary’. The letter is now in the National Portrait Gallery’s Archive, ref. NPG 1827 (2a).

(2) Clausen was Professor of Painting at the Royal Academy of Arts and an official war artist during the First World War. During the 1880sClausen developed an interest in cameras, which were still a primitive technology. He would photograph his subjects but also recognisedthat the images were no substitute for observation and combined the medium with sketches. Clausen would have referenced hisphotograph of Clegg in collaboration with a sketch which is now in the Royal Academy of Arts collection, ref. 05/2996.

(3) The Accountant’s Department was renamed the Registrar’s Department in 1980 and was closed in 2004.(4) Powers-Samas was a British company which sold unit record equipment.(5) For further information about technical developments in the Registrar’s Department see Bank of England (1985), ‘The Bank of England as

registrar’, Bank of England Quarterly Bulletin, September, pages 415–21; www.bankofengland.co.uk/archive/Documents/historicpubs/qb/1985/qb85q3415421.pdf.


Figure 10 New Change Dividend Preparation Office,Programme Controlled Computer, 1962 (15A13/2/9).

http://www.bankofengland.co.uk/archive/Documents/historicpubs/qb/1985/qb85q3415421.pdf


Figure 12 Men’s sleeping huts exterior at Hurstbourne Camp(15A13/15/136).

The Bank during World War II

Early in 1937 it was recognised within the Bank that war with Germany was certain. For the nexttwo years the Bank made preparations in what was code-named ‘Zero’. When war was declared in1939 all the activities of the Bank were able to continue with minimum disruption.

Bank staff who did not have to retain constant contact with the City were evacuated to Hampshireand by 1940 1,098 staff were billeted in the neighbouring villages of Hurstbourne Priors andWhitchurch. The first move towards the Bank’s wartime evacuation resulted from a decision, takenin 1938, to transfer the printing of banknotes at St. Luke’s (Old Street) and the DividendPreparation Office from London to the village of Overton, in order that they should be in closecontact with Portals paper mill.(1) The Bank of England has been issuing banknotes for over300 years, and maintaining confidence in the currency is central to the Bank’s mission.(2) Two largebuildings, which became known as the ‘shadow factories’, were erected adjoining Portals atOverton; one to house part of the Printing Works (Figure 11) and the other the DividendPreparation Office.

The Accountant’s Department and Establishment Department(3) were evacuated toHurstbourne Mansion, then in the occupation of Patrick Donner, MP, who had consented to reservethe tenancy for the Bank for a small retaining fee, and only 48 hours’ notice needed to be given inthe event of war. The house was suitably located close to Overton but, though large, was not largeenough to accommodate both staff quarters and offices. A temporary camp was constructed forstaff containing offices, a canteen and living quarters for men (Figure 12). Female members ofstaff stayed at the Mansion (Figure 13). Conditions were cramped: with only seven bathrooms,Hurstbourne Mansion accommodated as many as 309 women at one time.

In the spirit of ‘digging for victory’ some members of Bank staff started a small farm at theiraccommodation at Whitchurch and bought bees, two dozen chickens, two pigs, two goats, and apony and trap. This project was so successful that the Bank took charge of the livestock, andpoultry keeping was extended to ten other hostels. All labour was voluntary and all productionwent into the canteens. During the war staff produced 75,000 eggs, 350 gallons of goats’ milk and300 lbs of honey. The pony and trap were used by the gardeners for transporting vegetables.

Figure 11 Large Note Printing Machine, Overton(15A13/15/121).

(1) Portals Ltd had supplied banknote paper to the Bank of England from 1724 until 2003 when the Bank first outsourced its banknoteprinting to De La Rue, based at the Bank’s facility in Debden, Essex.

(2) Another article in this edition of the Bulletin discusses the plans to switch to polymer banknotes. See McClintock, R and Whymark, R(2016), pages 23–32; www.bankofengland.co.uk/publications/Pages/quarterlybulletin/2016/q1/a3.aspx.

(3) The Establishment Department looked after staff at the Bank. In effect, its purpose was that of a modern Human Resources department.

Work continued at Threadneedle Street and emergency accommodation was created to enablestaff to work below ground. A first aid station and several first aid posts, an operating theatre andgas contamination centre were installed in the vaults with advice from St Bartholomew’s Hospital(Figure 14). As fire raids began in 1940, the Bank was prepared with a team of male and femaleVolunteer Fire Guards (Figure 15). During a raid on the night of 9/10 September 1940 the Bankwas damaged by two bombs, one of which fell in Threadneedle Street and the other on the roofnorth of the Garden Court.(1) In January 1941, 111 people were killed when a bomb fell on Bankunderground station (Figure 16).

Capturing the City: Photography at the Bank of England opened on 18 January 2016 and will rununtil the end of 2016. The Bank of England Museum is open 10:00–17:00 on weekdays (see theMuseum’s website for special opening hours; www.bankofengland.co.uk/education/Pages/museum/whatson. Admission is free of charge.

To find out more about the Archive and to search our catalogue, please visit our websitewww.bankofengland.co.uk/archive. Any enquiries can be sent to [email protected] is also an associated digital slide show www.bankofengland.co.uk/education/Documents/museum/whatson/capturingthecity_qb.pdf and the Capturing the City Flickr pagehttps://www.flickr.com/photos/bankofengland/sets/72157663360744376, both of which includeadditional photographs from the Bank Archive.

Visitors inspired by the exhibition are invited to share their own images of the City with the Bankof England Museum via their Twitter feed @boemuseum — using the hashtag #CitySnapsBoE.Each month the Museum’s Curator will select a favourite image, which will join the exhibitiondisplays for four weeks, instantly becoming part of the Bank’s history.

Figure 13 Women’s sleeping quarters at HurstbourneMansion (15A13/15/49).

Figure 14 Medical Officer, Dr Norris and Sister Neave in theEmergency Operating Theatre in the Sub-Vault atThreadneedle Street (15A13/1/11/5).


Figure 15 Female staff stirrup pump training. St Paul’sCathedral can be seen in the background (15A13/1/11/28).

Figure 16 Bomb damage on Threadneedle Street, 1941. TheBank can be seen on the left, the Royal Exchange on the right(15A13/1/11/40).

(1) For further information about the Bank during the Second World War, the Bank of England Archive has digitised an unpublished historybegun by John Osborne, Adviser to the Governors, in 1943 (ref: M5/533–539); www.bankofengland.co.uk/archive/Pages/digitalcontent/archivedocs/warhistoryww2.aspx.

http://www.bankofengland.co.uk/education/Pages/museum/whatson/default.aspx

www.bankofengland.co.uk/education/Documents/museum/whatson/capturingthecity_qb.pdf

www.bankofengland.co.uk/education/Documents/museum/whatson/capturingthecity_qb.pdf

http://www.bankofengland.co.uk/archive/Pages/digitalcontent/archivedocs/warhistoryww2.aspx


The Archive cataloguing project

The Bank of England Archive’s remit is to collect and provide access to the historically importantrecords of the Bank. The information contained in these records can be accessed by internal andexternal researchers. The Archive currently holds over 80,000 ledgers, files and individual records.In 2014 the Archive’s collection was significantly enhanced by taking over responsibility for theBank’s historic photographic collection. This diverse collection represents a huge range ofactivities and subjects. For example, while exploring the collection in the early stages of theproject, an album of Quarterly Bulletin editors (1960 to 1985) was unearthed. Figure 17 shows theearliest image in the album, Editor John Standish Fforde. From the photographic collection we canglean much about subjects including architecture, technology and fashion, not to mention thehistory of photography over the past two centuries. The collection will, therefore, be of interest toresearchers beyond those focused on economic history. A one-year project is currently under wayto review, catalogue, preserve and provide access to the photographic collection which is made upof an estimated 45,000 items. This section briefly explains the different aspects of the project.

The reviewing process involves identifying which elements of the collection should be kept.‘Original’ photographic items will be kept and any duplicates or modern facsimiles removed.Where the original is a negative, a print is also kept as an ‘access copy’.

Archivists apply two key principles when cataloguing: provenance and original order. Provenancecaptures the ownership history of a collection and original order means keeping records in theorder that they were created or used. Preserving these aspects retains the authenticity andcontext of the records, which helps researchers trust and interpret them. It is particularlyimportant with a collection of photographs because it is in most cases unclear what motivation laybehind producing many of the images or who commissioned them. Often there is littleinformation besides a description of the image subject and a date. These may all be criteria afuture researcher uses when searching the Archive and so wherever possible this detail is capturedin the catalogue.

In addition, because a single image can capture so much, a thesaurus of keywords has been addedto identify themes which may be of interest to future researchers. Searching these keywords willallow users to identify items across the collection which may be of interest. Figure 18 is a goodexample. It is an image of Threadneedle Street during the Second World War. The intendedsubject is the gas detectors at work. Incidentally, the image also captures parts of the roof notfound elsewhere in the collection, particularly the lion-head rain spouts. As a result of thesaurusterms entered, searches for ‘chemical warfare’ and ‘Second World War’ will return this image in theresults, but so too will ‘architecture’. This and other metadata added to the catalogue makes eachrecord easier to locate.

The collection comprises a range of different photographic media produced across two centuriesincluding albumen prints, gelatin prints, cyanotypes, glass-plate negatives and positives (alsoknown as lantern slides) and cellulose acetate negatives. These all have unique preservationrequirements. Figure 19 shows an unusual Victorian leather photograph album. The front covercontains a (sadly empty) compartment with a blue velvet lining. In addition to the individualphotographs in the album, the materials in such covers degrade over time. The Archive’s storagefacilities have a controlled climate which will decelerate the deterioration of these fragilematerials. Any apparent damage is recorded while cataloguing which will expedite identifyingprofessional conservation requirements after the project.

The Archive is already facilitating access to the collection, responding to all photographicenquiries, and has collaborated with the Museum on the current exhibition. An online gallery ofsome further hidden gems has also been created.(1) In time the collection will also be madesearchable via our online catalogue for researchers to browse.

(1) Visit the Capturing the City Flickr page to view an online gallery of additional photographs from the Bank Archive; www.flickr.com/photos/bankofengland/sets/72157663360744376.

https://www.flickr.com/photos/bankofengland/sets/72157663360744376

Figure 18 Gas detectors at work on the roof of ThreadneedleStreet during the Second World War (15A13/1/11/12).

Figure 17 J S Fforde, Quarterly Bulletin Editor (1960), andEditor in chief (1961–63) (15A13/18/2/11).

Figure 19 Victorian leather photograph album, front cover opening(15A13/18/2/12).


Topical articles The small bank failures of the early 1990s 41

• Prior to the recent global financial crisis, the Bank of England last provided emergency liquidityassistance to banks in the early 1990s. This was intended to prevent contagion from a group ofsmall banks to larger, systemically important financial institutions.

• The Bank of England is now in a better position to guard against many of the vulnerabilities thatled to the small banks crisis. History suggests that regulators should continually look for earlywarning signs of heightened risk in the financial system, such as rapid credit growth, a decline inunderwriting standards and large shifts in business models.

The small bank failures of the early1990s: another story of boom and bustBy Kushal Balluck and Artus Galiay of the Bank’s Financial Stability Strategy and Risk Directorate, Gerardo Ferrara ofthe Bank’s Financial Market Infrastructure Directorate and Glenn Hoggarth of the Bank’s International Directorate.(1)

(1) The authors would like to thank Ian Bond, Eric Engstrom and Alex Merriman forcomments and George Barton and Tiago Dos Santos for their help in producing thisarticle.

(2) Governor’s speech to the Chartered Institute of Bankers (1993).

Overview

‘There are several questions banks should ask themselves.Do they really pay attention to the lessons of history —for example the property crisis of the early 1970s? Didthey really monitor the credit criteria which had servedthem well in the past?’ — Robin Leigh-Pemberton,Governor of the Bank of England (1993).(2)

The early 1990s witnessed the failure of a large number ofsmall banks in the United Kingdom. Although these bankswere not by themselves systemically important, the episodewas serious enough for the Bank of England to provideemergency liquidity assistance to a few of them in order toprevent contagion to larger more systemically importantbanks.

The origins of the crisis lay in the banking system’s responseto the deregulation of the UK retail banking system. Acombination of rapid growth in lending to households andlower underwriting standards as banks competedaggressively for business left them exposed to heightenedrisk of loan losses. Losses crystallised after the UK economyexperienced a downturn in the early 1990s, exacerbated bytight monetary policy. Banking system credit losses in theUnited Kingdom in the early 1990s were over three timeshigher than they were in the recent financial crisis.

As well as being vulnerable to loan losses, many small bankswere overly reliant on short-term wholesale funding, in

particular from local authorities. This left them exposed tofunding risk which materialised once market participants lostconfidence in their creditworthiness. Loan losses and fundingoutflows led to the eventual failure of 25 small banksoperating in the United Kingdom at that time.

The Bank of England is now in a better position to guardagainst the vulnerabilities that manifested themselvesalmost a quarter of a century ago. There have beenimprovements in the regulatory framework for banksfollowing the recent financial crisis; for instance, tougherminimum liquidity and funding requirements have beenintroduced to reduce banks’ exposures to an outflow ofwholesale funding. Macroprudential oversight can help spot risks, including rapidly increasing credit provision and funding vulnerabilities and the Financial PolicyCommittee has been given a range of powers to helpmitigate those risks.

Nonetheless, history tells us that regulators should remainvigilant. The events leading up to the small banks crisis in the early 1990s are not dissimilar to those that haveresulted in many crises before and since. Looking for early warning signs of heightened risk in the system, such as fast credit growth, decline in underwriting standards, andrapid changes in business models, will help the Bank ofEngland protect the financial system against such risks.


Between 1990 and 1994, 25 small banks operating in theUnited Kingdom failed in some sense.(1) A concentration ondomestic property-related loans, coupled with reliance onwholesale markets for funding, left the small bank sectorvulnerable to the recession at the time. Although these bankswere very small — even collectively — relative to theaggregate level of UK economic activity — the episode wasserious enough for the Bank of England to provide emergencyliquidity assistance to a number of them in order to avoidcontagion to the wider banking system.

This is one of a series of Quarterly Bulletin articles focusing on historical episodes of financial instability in the United Kingdom.(2) The purpose of examining this and otherhistorical episodes is to build a better understanding of thecauses of financial crises, and to disseminate this knowledge to a wide audience. This article aims to draw out the lessons from the early 1990s episode that may help the Bank of England to achieve its objective of maintainingfinancial stability.

The first section of the article explains how a number offactors — the pre-crisis liberalisation of the financial sector, itsimpact on competition and the overall macrofinancialenvironment — contributed to the crisis in the small bankssector. The second section then describes how the crisisunfolded, explaining the causes of bank failures, and the Bank of England’s response. The last section considers thelessons for avoiding and managing banking crises today.

Origins of the small banks crisis

A raft of measures was implemented in the 1970s–80s aimedat liberalising the UK financial sector. These changes led to anincrease in competition in the provision of financial services,which resulted in banks lowering their lending standards andcontributed to a boom in credit.

Financial liberalisation and credit expansionPrior to 1971, the Bank of England determined the maximumamount of credit that banks could extend. In 1971, theauthorities introduced a ‘new approach to Competition andControl’, primarily to promote competition in the highlycartelised banking sector. This new approach aimed atcontrolling credit supply by setting its price rather than bylimiting its quantity. Early teething problems with supervisinga more liberalised financial system were witnessed with theSecondary Bank Crisis in the early 1970s. A number of smallerand medium-sized ‘fringe’ banks had funded themselves fromrecently deregulated wholesale markets to finance propertylending. But many of them failed once the property priceboom went into reverse.

Even then, some credit constraints remained, but banks werefreed up further to increase credit provision following the

removal of the so-called ‘corset’ in 1980, which allowed banksto increase credit provision. The corset was a mechanism thatpenalised banks for rapidly expanding their deposits, thusconstraining the rate at which banks could increase theirdomestic lending.(3) But it became less effective after theabolition of capital controls in 1979 that meant credit could beprovided directly to UK borrowers from abroad.

Banks responded to the removal of the corset by expandingtheir domestic mortgage lending — a market which had untilthen been dominated by building societies and which was seenas profitable and ‘low risk’. Banks’ share of new mortgagelending consequently increased from around 10% in 1980 to40% in 1988.

A rapid boom in domestic lending over the decade ensued.The supply of credit, no longer limited, responded to highdemand during a housing and commercial real estate (CRE)boom. Mortgage demand was supported by interest rate taxrelief and endowment products that at the time appearedattractive. Between 1985 and 1989 residential property pricesincreased by nearly 80% and CRE prices by almost 90%.(4)

Throughout the 1980s, lending to households was strong andoverall lending accelerated in the late 1980s as CRE lendingincreased materially. In aggregate, non-bank private sectordebt doubled relative to GDP during the 1980s (Chart 1).

(1) Throughout this article, we define failure as entering administration or liquidation,relying on liquidity support from the Bank of England, or closing down after havingauthorisation revoked.

(2) Previous articles looked at the failure of the City of Glasgow Bank in 1878 (Buttonand Knott (2015)) and lessons from Japan’s banking crisis in the 1990s (Nelson andTanaka (2014)).

(3) See McLeay, Radia and Thomas (2014) for a further explanation of how extendingcredit can increase deposits in the banking system.

(4) See Zhu (2002).

0

20

40

60

80

100

120

140

160

180

200

1970 75 80 85 90 95 2000 05 10 15

Households

PNFCs

Per cent of GDP

Sources: Bank of England, Office for National Statistics (ONS) and Bank calculations.

(a) ONS data are not available before 1990. Before then, ONS household and private non-financial corporations (PNFCs) debt series are assumed to grow at the same rate as theBank of England’s household and PNFC lending series.

Chart 1 Non-financial private sector debt(a)


Competition and lending standardsMortgage lending and retail deposit-taking had traditionallybeen the preserve of building societies, so increasedcompetition from banks put pressure on their business models.In response, the government decided to allow buildingsocieties to broaden their activities to include the provision ofnon-mortgage loans. They were also allowed to accesswholesale funding markets, which were at the time cheaperthan retail deposit funds. A number of building societiesconverted into banks. These changes exposed banks to morecompetition from building societies, as they became freer torespond to market conditions.

Banks also faced competition from specialist mortgage lendersthat relied on low wholesale funding costs.(1) This forcedbanks to reduce their margins to remain competitive. Forexample, among other innovations, for the first time in 1985UK banks offered interest on current accounts. The netinterest margins of the four largest clearing banks fell from anaverage of 5.7% in 1985 to 4.7% in 1989.(2)

As lower margins put pressure on banks’ profits, they soughtto increase lending by reducing their lending standards. Thiswas evident on two metrics in mortgage lending — the loan tovalue (LTV) ratio and the loan to income (LTI) ratio, whichmeasure the amount lent as a proportion of the value of theproperty and of the borrower’s income respectively. Both ofthese ratios rose during the 1980s (Chart 2). By the late1980s, almost half of new mortgages advanced had LTV ratiosof over 90%.

Monetary policy in 1990In October 1990, the United Kingdom joined the ExchangeRate Mechanism (ERM) of the European Monetary System.The Deutsche mark (DM) was the de facto anchor currency

against which sterling was pegged. In Germany at the time,inflationary pressures following the reunification resulted inthe Bundesbank tightening monetary policy. Although therewere already signs of a domestic recession, UK interest rates —which had been raised from 1988 in preparation for the ERM— were kept high in order to keep sterling within its exchangerate band against the DM.

This exacerbated the slowdown in the economy. Annual GDPgrowth fell from over 4% in June 1989 to minus 1% two yearslater. As growth slowed there was a precipitous fall inproperty prices — residential and commercial property pricesfell by 14% and 27% from peak to trough respectively.Mortgage borrowers faced the twin impact of rising mortgagecosts from higher interest rates, and a reduction in the valueof their property. The unemployment rate — which even atthe peak of the boom was 7% — rose steadily, to 10.5% by1993.

Against this background, worsening lending standards leftbanks vulnerable to credit losses from property-relatedexposures. As the economy slowed and interest rates rose,borrowers struggled to make repayments on their loans.Consequently, banks’ write-offs on their mortgage lendingincreased significantly.

The crisis unfolds

Much of the banking sector’s difficulties in the early 1990sreflected the recession at the time, which was shared by anumber of advanced economies, following the late 1980sboom. In this fragile environment of high loan losses, thereliance of a number of small banks on short-term wholesalefunding left them vulnerable to a liquidity shock (ie banksneed to quickly refinance maturing debt at a time whencreditors are less willing to lend). When these funds dried upin 1991, the Bank of England intervened to prevent the failureof a number of small banks and potential contagion to the rest of the banking sector. This section sets out theseevents, starting with a description of the structure of the UK banking system in the early 1990s, and following on withlosses on banks’ loan portfolios, funding problems and finallythe authorities’ reaction.

Structure of the banking system in the early 1990sThe UK banking system at that time can be divided into a fewbroad categories: (i) the large retail and investment banks, (ii) branches of foreign-owned banks, and (iii) small andmedium-sized banks. Table A shows the number of banks in

0

10

20

30

40

50

60

1979 84 89 94 99 2004 09 14

LTV ≥ 90

Per cent (a)

LTI ≥ 3

Sources: Council of Mortgage Lenders (CML), Financial Conduct Authority (FCA) Product SalesData and Bank calculations.

(a) Data from the FCA’s Product Sales Database (PSD) are only available since 2005 Q2. Datafrom 1979 to 2005 Q1 are from the discontinued Survey of Mortgage Lenders (SML), whichwas operated by the CML. The two data sources are not directly comparable and shares areillustrative prior to 2005 Q2.

Chart 2 New mortgages advanced for house purchaseby LTV and LTI

(1) In general, specialist mortgage lenders had access to more funding sources (eg oftenin the form of residential mortgage-backed securities) than small banks. Thespecialist mortgage lenders have traditionally focused their attention on one or morespecialised markets, such as buy-to-let or self-certified mortgages, or lending tocustomers with adverse credit histories.

(2) Callen and Lomax (1990).


each category from February 1990 up to February 1994. Ofthe 540 banks operating in the United Kingdom in 1990, justless than a quarter were small or medium-sized banks.(1) Themajority of the lending to UK residents was done by thelargest UK retail banks and building societies (Figure 1).Building societies had a much larger market share of mortgagelending than they do currently, accounting for around 60% ofthe stock of mortgage lending. Lending to companies,however, was largely done by banks. The small banks whosesolvency was threatened in the early 1990s played a small partin the aggregate provision of credit, accounting for less than1% of the stock of UK lending to the non-financial privatesector.

As well as doing a considerable amount of lending to the UK private sector, foreign and British investment banks wereamong the most active in UK money markets. Nearly 60% ofinvestment banks’ sterling-denominated loans were made inwholesale markets (Chart 3). Japanese banks also lent asubstantial amount of money in these markets. At the end of1989, they had £15 billion of loans outstanding in moneymarkets — marginally less than investment banks, andequivalent to half their lending in the United Kingdom. Theweakness of Japanese banks as a result of a domesticdownturn in the early 1990s, and their eventual withdrawalfrom UK markets likely exacerbated the tightening in sterlingmoney market conditions.

Credit losses and problems at small banksUK banks’ aggregate losses on domestic household lending inthe early 1990s were high. Cumulative write-offs onresidential mortgages between 1991 and 1995 were similar tothose in Sweden, which suffered an outright systemic bankingcrisis with the failure of its largest banks. They amount to overthree times the size of credit losses in the recent financial crisis(Chart 4).

The larger UK banks had enough capital to absorb these creditlosses without threatening their solvency. But problems aroseat a number of small banks. The degree of their exposure to

Major British retail banks £207 billion

Building societies £156 billion

Otheroverseas

banks£47 billion

Medium-sizedand small

British banks£22 billion

Japanesebanks

£14 billion

Britishmerchant

banks£12 billion

US banks£10 billion

Vulnerablesmall banks£4 billion(b)

(a) Stock of sterling-denominated loans to UK residents, excluding financial institutions.(b) Estimate based on data available for the banks on the Bank of England’s vulnerable bank

watchlist.

Figure 1 UK lending to households and companies atend-1989(a)

(1) They accounted for less than 5% of UK lenders’ total assets.

0

10

20

30

40

50

60

70

Britishmerchant

banks

Japanesebanks

Americanbanks

Otheroverseas

banks

OtherBritishbanks

MajorBritishretailbanks

Per cent of assets

(a) Chart shows major sterling-denominated lenders grouped by type of bank.

Chart 3 Share of sterling-denominated loans in UK wholesale markets in December 1989(a)

0

2

4

6

8

UnitedKingdom2008–12

Sweden1991–93

UnitedKingdom1991–95

Ireland2011–13

United States2008–12

Per cent of total loans

Sources: Bank of England, Central Bank of Ireland, Council of Mortgage Lenders, FDIC, Glas Securities, MIAC-Acadametrics, Nordiska Ministerrådet, Prudential Regulation Authority,Statistics Sweden, published accounts and Bank calculations.

(a) For the United Kingdom, 1991–95: banks and building societies and 2008–12: for a sampleof the largest mortgage lenders as of end-2007. Swedish losses based on housing creditinstitutions. Irish losses are from 2011 to mid-2013 and refer to Allied Irish Bank, Bank ofIreland and Permanent TSB. United States losses refer to all FDIC-insured institutions. Alllosses are scaled to the share of loans outstanding at the beginning of the period.

Chart 4 Banks’ cumulative write-offs(a)

Table A Number of authorised banks in the United Kingdom

1990 1991 1992 1993 1994

Commercial and investment banks 75 70 72 73 71

Branches of foreign banks 340 336 328 332 360

Small and medium-sized banks 125 116 111 96 80

Total 540 522 511 501 511


property-related loans was uncertain but thought to besubstantial, which meant credit losses could potentiallythreaten their solvency. In the autumn of 1990, Authority Bank, a small lender, went into administration as a result of a large exposure to a group of property and leasing companies. In February 1991, Chancery Bank, whichalso had large exposures to property-related lending, alsofailed.

The bad debt problems experienced by small banks in late1990 and early 1991 were well publicised and contributed togrowing nervousness in money markets. The failure ofAuthority Bank closely followed the failure of British andCommonwealth Merchant Bank (BCMB), which went intoadministration earlier that year due to difficulties faced by itsparent company. BCMB was unable to replace funding,including from short-term money markets, as counterpartieswithdrew due to uncertainty following the failure of theparent company and fears about losses from property-relatedloans.

Funding problems and contagion BCMB’s reliance on wholesale funding markets was notunique. Perhaps as a result of generalised low interest rates inwholesale funding markets in the late 1980s, many smallbanks were heavily dependent on these markets, and inparticular on larger banks and local authorities who providednearly half of their funding. Local authorities were attractedby the higher deposit rates that small banks often paid relativeto their larger peers. In June 1990, a quarter of local authorityfunds placed with banks were held at 40 small banks.

Such dependence on funding from local authorities meantthat small banks were vulnerable to deposit flights in theevent of a confidence shock. A handful of small banks reliedon local authorities for more than one fifth of their funding.The failures of three banks — BCMB, Chancery Bank andEdington Bank (a small investment bank that failed inNovember 1990) — had already served as a warning to localauthorities, whose funds at these failed banks had been lockedin administration.

The closure of Bank of Credit and Commerce International(BCCI) in July 1991 — due to an instance of fraud —represented a turning point for local authorities. A number ofthem lost access to funds placed with BCCI. At this point,local authorities began to withdraw virtually all their short-term deposits from some small banks. There was a‘flight to quality’, with funds redeposited with larger clearingbanks. A timeline representing the crisis is shown in Chart 5.The share of small banks’ deposits from local authorities fellfrom 21% in December 1990 to only 6% nine months later inSeptember 1991 (Chart 6).

Larger banks, including foreign banks, also began to reappraisetheir exposure to the small banks’ sector during this period.Foreign banks, in particular, retrenched from interbankmarkets. As a result, UK banks’ total wholesale borrowing fellby nearly 15% during 1991 (Chart 7). Both US and Japanesebanks were experiencing problems in their home markets,which may have exacerbated their reaction to the UK downturn. Within four years, 25 small banks had in somesense, failed.

2

1

0

1

2

3

4

5

6

7

8

1987 88 89 90 91 92 93 94

Annual growth in GDP

Per cent

Sterling officially joins ERM

UK interest rate increases

BCMB entersadministration

Authority Bankenters administration Chancery

Bank fails BCCI fails. Local authorities withdraw deposits

Bank of England provides liquidityassistance to three small banks

19 small banks fail

+

–

Chart 5 Timeline of the crisis


The succession of small bank failures was not in itself ofsystemic importance — the very small combined size of thebanks that failed meant that the loss of financial servicesprovision in the United Kingdom was negligible. The assets ofthe small bank sector as a whole were equivalent to only 2.3%of GDP in 1991, about 35 times less than those held by themajor British banks. The small banks that failed had totalcombined assets of only 0.2% of GDP — equivalent to thoseof Dunfermline Building Society, which failed in 2009. Thisstands in sharp contrast to big bank failures in 2008 (Figure 2). But the actual and prospective failures had led toan environment of uncertainty and fear in wholesale markets,and there was a possibility of contagion to larger, moresystemically important banks.

There are both direct and indirect channels of contagion frombank failure. In terms of ‘direct’ contagion, larger banks’exposures to small banks were generally small in thisparticular case, and so this channel did not pose a material riskto large banks in 1991.(1)

However, there was greater evidence of (and potential for)contagion through ‘indirect’ channels. First, banks can beaffected through confidence effects that cause investors to reassess risks from banks, and reduce or cut the funding theyprovide them with. The environment of uncertainty created ininterbank and wholesale funding markets after the closure ofBCCI in the summer of 1991 made contagion through thischannel a potentially important risk.

Another indirect channel of contagion is that the failure of oneor more financial institutions — even if small — can causeliquidity providers to reconsider their lending to biggerfinancial institutions which they think have a similar businessmodel and/or vulnerability to the same adverse shock (so-called ‘information based’ contagion).(2) This occursbecause collecting information on the creditworthiness ofindividual banks is costly. The failure of one bank cantherefore be sufficient for market participants to draw stronginferences about the viability of other banks — even if thoseinferences turn out to not be warranted by the underlyingfinancial positions of those other banks. Such contagion wasalso possible in 1991, as medium-sized banks also relied onlocal authority and bank funding, although not to the sameextent as small banks. There was also some concern thatcontagion might even spread to one or more of the biggest UK banks.

The UK authorities’ reactionIn the weeks following the closure of BCCI, the Bank ofEngland considered that the risk of contagion through theseindirect channels had become large enough to have widerimplications for financial stability. In particular, the Bankworried that risk aversion in funding markets could spread tolarger banks, and may even consequently cause problems atlarge systemically important banks, if problems in wholesalefunding markets were to worsen. As an internal July 1991 Bankreport put it: ‘the systemic danger of National Home Loans

0

5

10

15

20

25

0

200

400

600

800

1,000

1,200

Dec. June Dec. June

£ millions Per cent

Deposits outstanding from local authorities (left-hand scale)

Share of deposits from local authorities (right-hand scale)

1989 90 91Mar. Mar.Sep. Sep.

Chart 6 Small banks’ deposit funding from localauthorities

(1) See Ferrara et al (2016).(2) See Santor (2003).

20

40

60

80

100

120

140

160

180

1988 89 90 91 92 93 94

£ billions

Loans to UK banks

Total wholesale borrowing

Failure of BCCI

0

Chart 7 UK banks’ sterling-denominated wholesaleunsecured lending and borrowing

Royal Bankof Scotland

(2008)158% of GDP

Lloyds Banking Group(2008)

75% of GDP

Northern Rock

(2007)8% of GDP

Barings(1995)

1% of GDP

DunfermlineBuilding Society(2009)

0.2% of GDP

25 smallbanks

(1991–94)0.2% of GDP

Sources: Published accounts and Bank calculations.

(a) Bubbles show total assets as a percentage of annual GDP in the year of failure.

Figure 2 Size of bank failures in the United Kingdomsince 1990 (per cent of GDP)(a)


[a vulnerable small bank] is that it has a substantial £600 million of commercial paper in the market and it has£90 million of funding from quality corporate names’. Inother words, the Bank feared that the small banks’ problemswould spread to a different sector of the funding market, and asector on which some of the larger banks were heavilyreliant.(1)

At this point, the Bank of England was already playing acentral role in managing the small banks crisis by acting as amediator between banks, encouraging the larger clearingbanks to lend to the small banks’ sector. The Bank did this byarranging meetings between the key stakeholders concerned.This approach had been successful during the 1970s SecondaryBanking Crisis when the Bank had set up a lifeboat operation— whereby larger clearing banks were asked to continue toprovide funding to smaller ‘secondary’ banks with a liquidityproblem to avoid a systemic crisis. Perhaps because of the lesssystemic nature of the small banks in the early 1990s and themore competitive environment in the banking sector, effortsto facilitate the provision of liquidity from the clearing banksproved unsuccessful.

The Bank decided instead to arrange official liquidity supportwhen three more small banks — National Home Loans(NHL),(2) City Merchants Bank,(3) and East Trust — ran intosevere funding problems. The Bank did this either by providingfunding directly to the small banks in trouble, or byguaranteeing credit lines from larger clearing banks. The Bankonly disclosed the provision of official support over a yearlater.

In addition, the Bank kept a further 40 small banks under closesupervisory review.(4) These banks were generally dependenton wholesale markets for more than one fifth of their funding.The banks were required to provide regular additionalinformation, especially relating to their liquidity and cash flow.The Bank continued to do this even after it had providedliquidity support and conditions in money markets eased. Thisis in part because small banks continued to close down until1994. In some cases, the Bank helped institutions to winddown their business in an orderly manner.

The economic cost of the failure of these institutions wassmall. Together, the small banks that failed had assets ofapproximately £1 billion — less than 0.2% of GDP in 1991. Incontrast to many banking crises around the world, there wasno government injection of capital into the banking system,although the Bank of England took a small loss on the credit that it extended (Chart 8). In 1993, the Bank disclosedthat it had made provisions of £115 million against possiblelosses.

Lessons for today

There are a number of lessons that can be drawn from theearly 1990s small banks crisis. The Bank of England is nowbetter equipped in a number of ways to protect the financialsystem against the series of events which occurred then —through an improvement in micro and macroprudentialregulation as well as a regime to resolve failing banks, amongother things. If institutional memory is maintained,regulators’ experiences of banking crises should continuallyimprove their understanding of how to reduce their likelihoodand impact. Some of the lessons from the early 1990s episodeare discussed below.

Rapid change brings challenges for bank managers andtheir supervisorsAs outlined in the first section of this article, the financialliberalisation of the 1970s and 1980s resulted in substantialchanges to the structure of the UK banking system. Thisopened up the banking sector to more competition andinnovation, as well as rapid credit growth. It is clear ex postfrom the credit losses during the recession that banks took ontoo much risk during the boom period.

(1) Bank of England Archives (Reference 9A257/1).(2) NHL relied heavily on local authority funding, losing much of this in March/April 1991.

By mid-May, only one third of the next two months’ maturities were expected to rollover. Following BCCI’s collapse most of this was lost. The clearers agreed to providea £200 million facility, but only with an authority’s guarantee. By February 1992, theBank of England was committed to meeting all NHL’s liabilities and bought it outrightin 1994 for a nominal sum in order to facilitate control over asset realisation.

(3) City Merchants’ liquidity difficulties were less severe but client money funding was atrisk. The Bank again guaranteed a facility, allowing City Merchants time to run downthe business. After discussion with the Bank, NatWest and Lloyds agreed to provide a£30 million facility for two months (with possible extension up to twelve months)provided the Bank of England would cover the risk.

(4) See Bank of England (1993).

0.10

0.16

10

0

10

20

30

40

50

60

70

80

Uni

ted

King

dom

(199

1)

Nor

way

(199

1)

Swed

en (1

991)

Uni

ted

Stat

es (1

988)

Uni

ted

Stat

es (2

007)

Spai

n (2

008)

Finl

and

(199

1)

Japa

n (1

997)

Uni

ted

King

dom

(200

7)

Sout

h Ko

rea

(199

7)

Turk

ey (2

000)

Thai

land

(199

7)

Icel

and

(200

8)

Indo

nesi

a (1

997)

Gre

ece

(200

8)

Irela

nd (2

008)

+

–

Per cent of annual GDP

Sources: Hoggarth and Soussa (2001), International Monetary Fund (2015) and Laeven andValencia (2012).

(a) United Kingdom (1991): peak of Bank of England provisions.(b) Other crises: cumulative gross financial outlays of central government on financial sector

restructuring. Magenta bars show, where available, costs after recovery.

Chart 8 Fiscal costs of banking crises(a)(b)


Rapid change and innovation may often create challenges forbankers and regulators alike. Bank managers may feeltempted, under competitive pressures, to take on more risksto maintain or increase their returns. Banks can do this bychanging their business models, for example to lend inlucrative markets that they are less familiar with. These riskscan sometimes be misunderstood or underestimated. In thelate 1980s, this risk manifested itself in a rapid expansion oflending and a lowering in lending standards. Ahead of themost recent crisis, rapid innovation came in the form of sometypes of securitisation that were systematically mispriced.

For supervisors, spotting these risks when they are building upcan be difficult, but rapid and large-scale responses to shockson the part of banks may act as a warning sign that they maybe taking on excessive risk. In the United Kingdom, moredetailed regulatory data are now collected on a regular basisfrom banks than in the early 1990s, and the PrudentialRegulation Authority (PRA) — the Bank of England’ssupervisory arm — is better equipped to identify and tacklethe build-up of risks in banks’ balance sheets.

Rapid growth in credit may be a sign of excessive risk-takingAs the 1990s small banks’ experience highlighted, rapidlending growth can be an indicator of a decline in lendingstandards. The pattern of a generalised rapid expansion ofcredit provision fuelling a property boom, followed by aneventual crash and large credit losses is one that has oftenrepeated itself in banking crises in the past and around theworld.(1) It was also a feature of the recent global financialcrisis in many countries, including the United Kingdom.

Macroprudential regulation can guard against this risk by developing and monitoring measures of excess aggregatecredit growth. In the United Kingdom currently, the Financial Policy Committee (FPC) looks at a number of coreindicators of potential excess credit to do this, including thecredit to GDP gap.(2) This indicator measures the deviation ofaggregate credit to households and non-financial companiesrelative to GDP from its long-run trend. Interestingly, if theUK authorities had been using the credit to GDP gap indicatorin the 1980s, it would have likely been flashing a warning signof a pending downturn. The level of credit rose to 20% abovetrend ahead of the recession and associated small banks crisisin the early 1990s compared to 10% in the run-up to theglobal financial crisis (Chart 9). That said, at the currentconjuncture, the FPC has indicated that there are a number ofdrawbacks with the credit gap measure and that there is nosimple mechanistic link between this indicator and the FPC’spolicy.(3)

Even at an individual bank level, rapid credit growth can be apowerful predictor of failure. This was the case in the early1990s bank failures. On average across the banks that went

on to fail, credit growth was 30% at its peak in 1989, almosttwice the rate of the banks that survived (Chart 10). Of thesmall banks in the fastest-growing quartile, 40% went on tofail. Unsurprisingly, the banks that failed reduced their lendingmaterially in 1990 as losses started to build up, perhaps in abid to improve their liquidity positions and their lendingportfolios.

(1) See, for example, Laeven and Valencia (2012).(2) These core indicators are published in the Bank of England’s semi-annual

Financial Stability Report.(3) See Bank of England (2015a), page 36.

Percentage points

1970 75 80 85 90 95 2000 05 10 1530

20

10

0

10

20

30

+

–

UK recessions(b) Actual (right-hand scale)

Trend (right-hand scale) Gap (left-hand scale)

Percentage points

0

20

40

60

80

100

120

140

160

180

200

Sources: British Bankers’ Association, ONS and Bank calculations.

(a) The credit to GDP gap is calculated as the percentage point difference between the credit toGDP ratio and its long-term trend.

(b) Recessions are defined as two consecutive quarters of negative quarterly real GDP growth.

Chart 9 Domestic credit relative to GDP(a)

15

10

5

0

5

10

15

20

25

30

35

–

+

Small banks that survived the mid-1991–94 (‘crisis’) period

Small banks that failed during mid-1991–94 (‘crisis’) period

All banks and building societiesPer cent

1988 89 90 91

Source: Logan (2000).

Chart 10 Annual growth in bank lending to UK non-bankprivate sector


Undiversified lending and funding can leave banksvulnerable to shocksA clear lesson from the small banks crisis was that overrelianceon a single source of funding can leave banks vulnerable to awithdrawal of funds. When the local authorities withdrewtheir funds, small banks struggled to find short-term financeto pay out these deposits, and eventually failed or relied onthe Bank of England for liquidity support. This risk crystallisedin a similar way in the recent crisis: Northern Rock eventuallyfailed in 2007 when it could not repay lenders following itsoverreliance on funding from short-term wholesale markets.

Two regulatory requirements are being introduced, followingthe financial crisis of 2008, to guard against vulnerabilities toliquidity runs. First, the Liquidity Coverage Ratio (LCR)requires banks to hold enough liquidity to cover a short-termoutflow of funds. Second, the Net Stable Funding Ratio(NSFR) will require them to fund long-term illiquid assets withlong-term stable liabilities, and should ensure that banks areless vulnerable to funding shocks in the first place.

Lack of diversification was also a problem in the small banks’lending decisions. Logan (2000) shows that the banks thatfailed during the 1990s had less diversified lending portfoliosthan those that survived — the median bank (in terms of size)that failed had nearly half of its assets concentrated amongten counterparties. The banks that ended up failing also hadmuch lower levels of capital, leaving them more vulnerable toa few of their creditors defaulting. The regulatory capitalframework now guards against such risks. When settingcapital requirements for banks, the PRA assesses risks from theconcentration of exposures.(1)

The need to deal with ‘too big to fail’ issues While the early 1990s crisis was limited to small UK banks, theBank of England was concerned at the time that stress inwholesale markets may also affect bigger banks’ ability tofund themselves, thereby potentially destabilising them.Indeed, larger UK banks were also hurt by the recession, andthree of the major UK banks (Barclays, NatWest and MidlandBank) lost their triple-A credit ratings by end-1991.

In the event, limited contagion in funding markets meant thatthe UK authorities did not at the time have to grapple with thereality of a large bank failure. The hypothetical question ofhow to deal with a large bank failure without using taxpayerfunds was left unanswered in the aftermath of the crisis. Inthe recent global financial crisis, some of the largest UK bankswere bailed out at enormous public cost.

The Bank is now better placed to resolve bank failures througha framework called the Special Resolution Regime.(2) In thecase of a small bank whose failure would have no systemicconsequences, the Bank of England would put it into a bankinsolvency procedure — ensuring that insured deposits are

promptly paid out while winding down the bank. If this courseof action could give rise to systemic consequences, other toolsare available to the Bank to stabilise the firm. For larger banks,the Bank of England would use its powers to carry out a ‘bail-in’ to absorb losses and restore solvency using the bank’sown resources — shareholders and unsecured creditors arewritten down and/or converted to equity to restore the bank’scapital position.(3) Regulatory standards are being introducedinternationally to ensure that bail-in resolution strategies canbe carried out in the event of a bank failure.(4)

Prompt and covert official liquidity provision can stopcontagion through financial marketsIt is difficult to measure the success of the Bank of England’sintervention in financial markets in 1991. Conditions in thewholesale funding markets did not continue to worsen — butit is difficult to attribute this solely to the Bank’s provision ofofficial support. It is likely that the covert nature of theliquidity support helped to calm financial markets withoutalarming participants about specific firms. However, thefailure of the Bank of England’s original plan to set up a‘lifeboat’ operation slowed down the official response to thecrisis — at the cost of worsening conditions in fundingmarkets.

The Bank of England now has a clear framework for providingliquidity insurance facilities. The Bank’s Discount WindowFacility (DWF) and Contingent Term Repo Facility (CTRF) areboth available to banks and building societies that experiencea liquidity shock. Firms facing an idiosyncratic or market-wideshock would borrow from the Bank of England via the DWF,while the CTRF would be activated by the Bank in response toactual or prospective market-wide stress of an exceptionalnature. In addition, the Indexed Long-Term Repo operation isavailable to banks to meet predictable liquidity needs. Banksand building societies are able to position collateral with theBank of England in advance of a shock to facilitate drawingfrom these facilities in a quick and efficient manner.

Too little and too much competition can both bedamaging to financial stabilityThe early 1990s small banks crisis shows that althoughcompetition imposes a welcome and healthy pressure on thebanking system which can benefit customers (for example inthe form of easier access to credit), competitive pressures canlead to the build-up of risks in the banking system. The lessonfor regulators is to monitor banks’ lending behaviour to ensurethat they do not respond to these pressures in a way thatendangers financial stability.

(1) See Bank of England (2015b).(2) See Bank of England (2014).(3) Chennells and Wingfield (2015) provide an introduction to the resolution of bank

failure through bail-in. An indication of the type of tools that are likely to be used fordifferent types of bank can be found in Table 1 of Bank of England (2015c).

(4) See Financial Stability Board (2015).


Nowadays, the Bank of England should be better equipped todeal with the market failures that may arise from either ‘toolittle’ or ‘too much’ competition. The PRA was given thesecondary objective of facilitating ‘effective competition’,which Dickinson et al (2015) define as being achieved when‘market or regulatory failures are either not significant or elsehave been addressed’. To deal with risks arising fromheightened competition, such as the weakening of lendingstandards seen in the 1990s small banks crisis, the PRA canrely on its judgement-based approach and supervisory powersto identify excessive risk appetite.

Conclusion

The credit boom in the late 1980s proved costly and resultedin a subsequent deep economic downturn in the early 1990s.This, in turn, contributed to the small banks crisis at the time and the need for the Bank of England to provideemergency liquidity assistance to a number of banks.However, the small banks crisis did not have a scarring effecton the United Kingdom’s economy. And although the crisisthreatened to spread to larger banks, the banks that failedwere too small to have any lasting effect on the structure ofthe banking system. But the lessons from a familiar story ofboom and bust in the property market, and a subsequentbanking crisis, remain relevant today. It would be impossibleand undesirable to make the financial system failure proof, butbearing these lessons in mind — both in spotting risks andvulnerabilities in the banking system, and on dealing withthem — should help the Bank of England better safeguardfinancial stability in the United Kingdom in the future.


References

Bank of England (1991), Banking Act report for 1991/92.

Bank of England (1992), Banking Act report for 1992/93.

Bank of England (1993), Report and accounts.

Bank of England (2014), The Bank of England’s approach to resolution, October, available atwww.bankofengland.co.uk/financialstability/Documents/resolution/apr231014.pdf.

Bank of England (2015a), Financial Stability Report, December, available at www.bankofengland.co.uk/publications/Documents/fsr/2015/dec.pdf.

Bank of England (2015b), ‘The PRA’s methodologies for setting Pillar 2 capital’, Prudential Regulation Authority Statement of Policy, available atwww.bankofengland.co.uk/pra/Documents/publications/sop/2015/p2methodologies.pdf.

Bank of England (2015c), ‘The Bank of England’s approach to setting a minimum requirement for own funds and eligible liabilities (MREL)’,December, available at www.bankofengland.co.uk/financialstability/Documents/resolution/mrelconsultation2015.pdf.

Button, R and Knott, S (2015), ‘Desperate adventurers and men of straw: the failure of City of Glasgow Bank and its enduring impact on the UK banking system’, Bank of England Quarterly Bulletin, Vol. 55, No. 1, pages 23–35, available atwww.bankofengland.co.uk/publications/Documents/quarterlybulletin/2015/q102.pdf.

Callen, T S and Lomax, J W (1990), ‘The development of the building societies sector in the 1980s’, Bank of England Quarterly Bulletin, Vol. 30, No. 4, pages 503–10, available at www.bankofengland.co.uk/archive/Documents/historicpubs/qb/1990/qb90q4503510.pdf.

Chennells, L and Wingfield, V (2015), ‘Bank failure and bail-in: an introduction’, Bank of England Quarterly Bulletin, Vol. 55, No. 3, pages 228–41, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2015/q302.pdf.

Dickinson, S, Humphry, D, Siciliani, P and Straughan, M (2015), ‘The PRA’s secondary competition objective’, Bank of England Quarterly Bulletin,Vol. 55, No. 4, pages 334–43, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2015/q402.pdf.

Ferrara, G, Langfield, S, Liu, Z and Ota, T (2016), ‘Systemic liquidity risk in the interbank network’, Bank of England Working Paper Series,forthcoming.

Financial Stability Board (2015), ‘Principles of loss-absorbing and recapitalisation capacity of G-SIBs in resolution’, available atwww.financialstabilityboard.org/wp-content/uploads/TLAC-Principles-and-Term-Sheet-for-publication-final.pdf.

Hoggarth, G and Soussa, F (2001), ‘Lender of last resort and the changing nature of the banking industry’, Financial Stability and Central Banks —a global perspective, Central Bank Governors Symposium Series, Routledge.

International Monetary Fund (2015), Fiscal Monitor, October.

Laeven, L and Valencia, F (2012), ‘Systemic banking crises database: an update’, IMF Working Paper No. 12/163.

Leigh-Pemberton, R (1993), ‘Speech to the Chartered Institute of Bankers in Scotland’, Bank of England Quarterly Bulletin, Vol. 33, No. 1, pages 103–05, available at www.bankofengland.co.uk/archive/Documents/historicpubs/qb/1993/qb93q1103105.pdf.

Logan, A (2000), ‘The early 1990s small banks’ crisis: leading indicators’, Bank of England Financial Stability Review, Issue 9, December, pages 130–45, available at www.bankofengland.co.uk/archive/Documents/historicpubs/fsr/2000/fsrfull0012.pdf.

McLeay, M, Radia, A and Thomas, R (2014), ‘Money creation in the modern economy’, Bank of England Quarterly Bulletin, Vol. 54, No. 1, pages 14–27, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q1prereleasemoneycreation.pdf.

Nelson, B and Tanaka, M (2014), ‘Dealing with a banking crisis: what lessons can be learned from Japan’s experience?’, Bank of England QuarterlyBulletin, Vol. 54, No. 1, pages 36–48, available at www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q104.pdf.

Santor, E (2003), ‘Banking crises and contagion: empirical evidence’, Bank of Canada Working Paper 2003–1.

Zhu, H (2002), ‘The case of the missing commercial real estate cycle’, Bank for International Settlements Quarterly Review, September, Vol. 21,pages 56–66.

Recent economic andfinancial developments

Quarterly Bulletin Recent economic and financial developments 53


• The first increase in the US federal funds rate since the start of the financial crisis proceededsmoothly, with the move having been well communicated in advance by the Federal Reserve andwidely anticipated by investors.

• There was a material deterioration in sentiment across asset markets at the start of the year,prompted by a combination of factors, including concerns about global growth.

• Developed-country interest rates across short and long tenors fell as a result, and there was asharp decline in the prices of a range of risky assets. Much of the decline in interest ratespersisted, despite a stabilisation in sentiment.

• The sterling ERI fell by 7.5% over the review period. In part the depreciation reflected changes inrelative interest rates. Contacts also attributed a portion of it to uncertainty ahead of thereferendum on UK membership of the European Union.

Overview

Around the start of the review period, on 16 December, theFederal Open Market Committee tightened US monetarypolicy by raising the target range for the federal funds rate by25 basis points (to a 0.25%–0.50% range). This had beenwell communicated in advance and was widely anticipatedby the market. There was a fairly limited reaction across abroad range of asset prices. Meanwhile, the EuropeanCentral Bank (ECB) loosened policy in December, loweringthe interest rate on its deposit facility and extending thehorizon for its asset purchase scheme. Shortly after the datacut-off, the ECB announced further measures to loosenpolicy.

Following the turn of the year, Chinese equities fell sharply,and there was a depreciation of the renminbi. There was alsoa further fall in the price of oil, and various other commodityprices. This confluence of factors led to a broad-baseddeterioration in confidence about the outlook for the globaleconomy and a generalised increase in risk aversion. Againstthat backdrop, the Bank of Japan unexpectedly announcedthe introduction of negative interest rates on some reservesbalances. And there was a material loosening in theexpected path of central bank policy in the United Kingdomand United States, with declines in interest rates at bothshort and long tenors.

The prices of many risky assets declined markedly during theperiod of heightened volatility from the start of the year. In

part, this was due to increased risk aversion, and concernsabout global growth. There were also some specific issuesassociated with the banking sector, including worries aboutearnings growth and the impact of exposures to the energysector, with the equity and debt of financial institutions hitparticularly hard.

Towards the end of the review period, however, there was astabilisation of financial markets, helped in part by a modestrise in the price of oil. This fostered a recovery in the pricesof many risky assets, although expectations for internationalmonetary policy remained considerably looser than at thebeginning of the review period.

The sterling exchange rate index declined by 7.5% over thecourse of the review period. Much of the depreciationreflected changes in relative interest rates against tradingpartners of the United Kingdom. And the generalised rise inrisk aversion might have contributed as well. Contacts alsopointed to a role for heightened uncertainty associated withthe coming vote on UK membership of the European Union.Contacts reported that uncertainty was evident in foreignexchange options, with option prices indicating that marketparticipants anticipated a rise in volatility around the time ofthe referendum, as well as some downside risk to the valueof sterling.

Markets and operations

Recent economic and financial developments Markets and operations 55

In discharging its responsibilities to ensure monetary andfinancial stability, the Bank gathers market intelligence fromcontacts across a range of financial markets. Regular dialoguewith market contacts provides valuable insights into howmarkets function, and provides context for the formulation ofpolicy, including the design and evaluation of the Bank’s ownmarket operations. The first section of this article reviewsdevelopments in financial markets between the 2015 Q4Quarterly Bulletin and 3 March 2016. The second section goeson to describe the Bank’s own operations within the SterlingMonetary Framework.

Monetary policy and interest ratesAt the start of the review period on 3 December, theEuropean Central Bank (ECB) cut the deposit rate by0.1 percentage points to -0.3% and extended the time horizonof asset purchases under its Public Sector PurchaseProgramme by six months. Although significant action wasexpected by contacts, there had been no clear consensusregarding specific measures. In the event, both short andlonger-term European interest rates rose significantly after thedecision, implying markets had been expecting moresignificant easing.

In contrast, the Federal Open Market Committee (FOMC)tightened US monetary policy by raising the target range forthe federal funds rate by 25 basis points (to a 0.25%–0.50%range) on 16 December. The move was clearly communicatedby the Federal Reserve in advance, and as a result was widelyanticipated by the market. As might be expected, theimmediate financial market reaction — across a broad range ofasset prices — was fairly muted. Contacts noted that theaccompanying communication regarding the likely future pathof rates was also in line with expectations.

While the first increase in the federal funds rate since the startof the financial crisis progressed smoothly, much of 2016 hasbeen dominated by sporadic bouts of risk aversion, reflecting acombination of growing worries about the prospects for globalgrowth and inflation, along with a generalised rise inuncertainty, fuelled in part by specific concerns related to theenergy sector and financial institutions. The touchpaper forthe deterioration in sentiment appeared to be a combinationof renewed declines in the price of oil — which dropped tothirteen-year lows in January — along with a depreciation ofthe Chinese renminbi, which renewed concerns about both apotential slowing in growth and the possibility of a change inexchange rate policy there.

Against this backdrop of heightened risk aversion and concernsaround the prospects for global growth, the Bank of Japan(BoJ) introduced a negative interest rate of -0.1% on somereserves balances. Contacts reported that the adoption ofnegative interest rate policy came as a surprise to manyinvestors. There were large declines in developed-country

interest rates in the days following the announcement, whichsome contacts suggested reflected a change in perceptionsabout how likely it was that such policies would be adoptedmore widely, as well as a shift in views about the effectivelower bound in central bank policy rates.

One-year, one-year forward overnight index swap (OIS) ratesin the United Kingdom fell by 42 basis points between27 November and 3 March (Chart 1), in line with moves in theUnited States. And the market-implied expectation for thetiming of the first increase in Bank Rate shifted from 2017 Q1to 2020 Q1. There was also a substantial slowing in themarket-implied pace of tightening, with a marked flattening inthe expected path of developed-economy interest rates(Chart 2). Shortly after the data cut-off, the ECB announcedthat the rate on its deposit facility would be lowered furtherinto negative territory, along with several other measures toloosen the stance of monetary policy.

1.0

0.5

0.0

0.5

1.0

1.5

2.0

2.5

2015 16 17 18 19 20

Sterling

US dollar

Euro

Per cent

Dashed lines: 27 November 2015

Solid lines: 3 March 2016

+

–

Sources: Bloomberg and Bank calculations.

(a) Instantaneous forward rates derived from the Bank's OIS curves.

Chart 2 Instantaneous forward interest rates derivedfrom OIS contracts(a)

0.75

0.50

0.25

0.00

0.25

0.50

27 Nov. 11 Dec. 25 Dec. 8 Jan. 22 Jan. 5 Feb. 19 Feb.

Per cent

United Kingdom

United States

Euro area

ECB announcement

BoJ announcement FOMC meeting

2015 16

+

–


(a) Instantaneous forward rates derived from the Bank’s OIS curves.

Chart 1 Cumulative change in one-year OIS rates,one year forward since 27 November 2015(a)

Despite a modest improvement in sentiment toward the endof the review period, expectations for policy remainedconsiderably looser than prior to the period of mostpronounced volatility. Broadly speaking, contacts havestruggled to reconcile the extent of the downward moves ininterest rates with the actual news flow over the period, withfew material downside surprises in the economic data to pointto that might have shifted expectations (Chart 3).

Further along the yield curve, UK, US and euro-area ten-yeargovernment bond yields fell by around 30–40 basis pointsover the review period (Chart 4). The decline can beattributed to falls in both expected real rates and inflationcompensation. Contacts suggested that this was broadlyconsistent with heightened concerns about both globalgrowth and, perhaps, about potential limits to the ability ofmonetary policy makers to counter global disinflationarypressures.

Five-year, five-year forward measures of inflationcompensation have declined further since the start of 2016 inthe United Kingdom, United States and euro area (Chart 5). Infact, the level of five-year, five-year forward inflationexpectations fell to all-time lows in both the United Statesand euro area. Towards the end of the review period,market-implied inflation expectations stabilised somewhat,and recovered part of the earlier falls.

Contacts attribute much of the decline in US market-impliedmeasures of inflation expectations to lower compensation forinflation risk. But contacts thought that expected inflationmight also have decreased. UK inflation compensation fell byless than in the United States, with contacts continuing toattribute the relative resilience of UK long-term inflationexpectations to the hedging activities of liability-driveninvestors such as insurers and pension funds.

Foreign exchangeThere were large movements in a number of major foreignexchange rate pairs during the review period, including insterling. Between 27 November and 3 March the sterlingexchange rate index fell by 7.5% (Chart 6). That was drivenlargely by depreciation against the euro and dollar, given thelarge trade weights of those currencies. In part, the decline inthe ERI was a reflection of changes in relative interest ratedifferentials, with UK interest rates falling by relatively morethan those in the euro area in particular. But the move wasbroad-based, with sterling falling against a number ofcurrencies.

Some of the recent weakness of sterling was perhaps relatedto a generalised deterioration in sentiment during the period.In the past, the sterling ERI has tended to fall during periods ofextreme volatility in financial markets (although, the particularnature of the risk is important, with sterling havingappreciated during the euro-area sovereign crisis, for


150

100

50

0

50

100

United Kingdom

United States

Euro area

China

Jan. Mar. May July Sep. Nov. Jan. Mar.2015 16

+

–

Indices

Source: Citigroup.

Chart 3 Citi Economic Surprise indices for theUnited Kingdom, United States, euro area and China

0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Jan. May Sep. Jan. May Sep. Jan.

United Kingdom

United States

Germany

Previous Bulletin

Per cent

+

–

2014 15 16


(a) Yields to maturity derived from the Bank’s government liability curves.

Chart 4 Selected ten-year government bond yields(a)

1.0

1.5

2.0

2.5

3.0

3.5

4.0


United Kingdom (RPI)

United States (CPI)

Previous Bulletin

Per cent

0.0

2014 15 16

Euro area (HICP)


(a) Swap rates derived from the Bank’s inflation swap curves.

Chart 5 Selected five-year inflation swap rates, fiveyears forward(a)


example). The often-observed tendency of sterling todepreciate during periods of risk aversion may be becausesterling assets have sometimes been a destination for carrytrades, which tend to be unwound when volatility rises. So,while sterling-denominated assets themselves may beconsidered ‘safe’, during periods of extreme volatility thecurrency may behave less like a ‘safe haven’ than do someothers, such as the yen, which strengthened during the recentturbulence (Chart 6).

In addition, contacts attributed at least a portion of thedepreciation to a rise in the risk premium in sterling exchangerate pairs, due to uncertainty associated with the forthcomingreferendum on UK membership of the European Union (EU). Itwas suggested that the most straightforward way ofexpressing views on the risks surrounding the vote was via theexchange rate.

Contacts also reported that uncertainty was evident in foreignexchange option prices, with an increase in the impliedvolatility of some sterling pairs at maturities spanning the dateof the referendum (Chart 7). And there had been an increasein the relative cost of hedging against a depreciation ofsterling (given by the implied volatility from a put option)compared with the cost of hedging against an appreciation(given by the implied volatility from a call option) (Chart 8).

Towards the end of the review period, there was a modestappreciation of sterling, and a fall in both implied volatilitiesand the relative cost of protection against a depreciationversus an appreciation. Contacts reported that someshort-term investors had reduced the size of theirshort sterling positions. And UK corporates had also beenobserved entering the market, with a view to hedging future

foreign-currency receipts at relatively favourable price levels.These factors had helped to provide some support to thecurrency.

Elsewhere, both the onshore and offshore renminbidepreciated against the US dollar. Commentators noted thatChinese foreign exchange reserves had fallen to the lowestlevel since 2012, perhaps indicating ongoing support for therenminbi by the Chinese authorities. People’s Bank of ChinaGovernor Zhou dismissed concerns about capital outflows anddeclining foreign exchange reserves, stating that there was nobasis for further deprecation of the renminbi.

Corporate capital marketsChanges in equity prices and corporate bond spreads wereinfluenced heavily by wider shifts in risk sentiment over theperiod. There were material falls in developed-economyequity indices, with particularly pronounced declines in theTopix and Euro Stoxx, which dropped by 14.1% and 12.1%,

85

90

95

100

105

110

115 Indices: 27 November 2015 = 100

Sterling

US dollar

Euro

Yen

July Oct. Jan.2015 16

8

7

6

5

4

3

2

1

0

1Per cent

+

–

Previous Bulletin

Other

Emerging markets(a)

Japanese yen

Euro US dollar

Sterling ERI

Sources: Bloomberg, ECB, Thomson Reuters Datastream and Bank calculations.

(a) The emerging market currencies in the narrow sterling ERI are: Chinese renminbi,Czech koruna, Indian rupee, Polish zloty, Russian rouble, South African rand and Turkish lira.

Chart 6 Selected exchange rate indices (ERIs) andcontributions to change in the sterling ERI since the startof the review period

One

mon

th

Option maturity

Two

mon

ths

Thre

em

onth

s

Four

mon

ths

Six

mon

ths

Nin

em

onth

sO

neye

ar

Two

year

s

Thre

eye

ars

Four

sye

ars

Five

year

s

11

13

15 Per cent

Dollar-yen

Euro-dollar

Sterling-dollar

Euro-sterling

9 0

Chart 7 Term structure of option-implied volatility ofselected exchange rates as of 3 March 2016

2.0

1.5

1.0

0.5

0.0

0.5

1.0

1.5

2.0

2.5 4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0

0.5

1.0


–

+

2014 15 16

Euro-sterling (left-hand scale)

Sterling-dollar (right-hand scale)

+

–

Per cent (inverted scale) Per cent

(a) Inverted scale for euro-sterling as sterling is the price currency.

Chart 8 Implied volatility of a four-month call optionminus the implied volatility of an equivalent put option(a)


respectively (Chart 9). The Shanghai Composite index alsoended the period 16.8% lower. Sentiment improved duringthe latter part of February, helped by a stabilisation in theprice of oil, resulting in a modest boost to equity prices. Butindices remained somewhat lower than at the start of thereview period.

Bank equities in developed markets declined by even morethan headline indices over the review period (Chart 10). Inaggregate, the banking industry group of the S&P 500 andFTSE All-Share declined by around 14.3% and 14.6%(compared with falls of 4.6% and 3.6% in the respectiveheadline indices). Even larger moves were seen in Europeanand Japanese banking stocks. In part, the decrease in the priceof banking stocks was simply a function of the fact thatfinancials tend to move by more than the market as a whole.But there were also specific concerns related to the exposuresof particular institutions to the energy sector, the impact ofnegative interest rate policy on earnings and idiosyncraticinstitutional factors at some euro-area lenders.

Concerns about slowing global growth and challenges facingthe financial sector weighed on expectations for bankearnings, and there was a rise in senior unsecured bank spreadsover the review period (Chart 11). In addition, the additionalTier 1 (AT1) debt market came under particular scrutiny, withsome investors becoming concerned that coupons on theseinstruments could be cancelled and that banks might choosenot to exercise options to redeem callable AT1 debt early,instead deferring repayment of the principal.

Spreads on private non-financial corporate bonds alsoincreased for both high-yield and investment-grade debt(Chart 12). Particularly sharp increases in US high-yield bond

spreads were partly driven by falls in the price of oil aroundthe start of the year, along with the broader shift in sentiment.But high-yield spreads subsequently fell back somewhat, asboth the price of oil and financial market confidence stabilised.

Amid fragile risk sentiment, there was rather less primarymarket high-yield corporate bond issuance than usual for thetime of year. Contacts attributed the fall in issuance toheightened market volatility and impaired liquidity, and theresulting increase in the premium demanded by investors tobuy new bonds rather than invest in comparable instrumentsalready available in the secondary market. Nevertheless,US dollar and euro-denominated investment-grade corporatebond issuance remained robust.

60

70

80

90

100

110

120

130

140

Jan. Mar. May July Sep. Nov. Jan. Mar. May July Sep. Nov. Jan. Mar.

DJ Euro Stoxx

FTSE All-Share S&P 500 Topix MSCI Emerging Markets index

2014

Previous Bulletin

Indices: 6 January 2014 = 100

15 16


(a) Indices are quoted in domestic currency terms, except for the MSCI Emerging Markets index,which is quoted in US dollar terms.

(b) The MSCI Emerging Markets index is a free-float weighted index that monitors theperformance of stocks in global emerging markets.

Chart 9 International equity indices(a)(b)

50

60

70

80

90

100

110

120

July Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar.

S&P 500

DJ Euro Stoxx

FTSE 350

Topix

Indices: 27 November 2015 = 100

2015 16

Previous Bulletin

Source: Bloomberg.

Chart 10 Selected bank equity sub-indices

0

20

40

60

80

100

120

Jan. Apr. July Oct. Jan. Apr. July Oct. Jan.

United Kingdom(d)

United States(c)

Euro area(b)

Basis points above mid-swaps

Previous Bulletin

2014 15 16

Sources: Bloomberg, Markit Group Limited and Bank calculations.

(a) Constant-maturity unweighted average of secondary market spreads to mid-swaps of banks’five-year senior unsecured bonds, where available. Where a five-year bond is unavailable, aproxy has been constructed.

(b) Average of Banco Santander, BBVA, BNP Paribas, Crédit Agricole, Credit Suisse,Deutsche Bank, ING, Intesa, Société Générale, UBS and UniCredit.

(c) Average of Bank of America, Citi, Goldman Sachs, JPMorgan Chase & Co., Morgan Stanleyand Wells Fargo.

(d) Average of Barclays, HSBC, Lloyds Banking Group, Nationwide, Royal Bank of Scotland andSantander UK.

Chart 11 Indicative senior unsecured bank bond spreads(a)


Operations

Operations within the Sterling Monetary Frameworkand other market operationsThis section provides an update of the Bank’s operationswithin the Sterling Monetary Framework (SMF) over thereview period, as well as its other market operations.Collectively, these operations help implement the Bank’smonetary policy stance and provide liquidity insurance toinstitutions when deemed necessary.

The aggregate level of central bank reserves is closelymonitored by the Bank. The level of central bank reserves isaffected by (i) the stock of assets purchased via the AssetPurchase Facility (APF); (ii) the level of reserves supplied byoperations under the SMF; and (iii) the net impact of othersterling flows across the Bank’s balance sheet. Over thereview period, aggregate reserves remained around£315 billion, but had fluctuated due to the redemption andsubsequent reinvestment of gilts held in the APF (discussedbelow).

Operational Standing FacilitiesSince 5 March 2009, the rate paid on the OperationalStanding Deposit Facility has been zero, while all reservesaccount balances have been remunerated at Bank Rate. As aconsequence, there is little incentive for reserves accountholders to use the deposit facility. Reflecting this, the averageuse of the deposit facility was £0 million in the three monthsto 14 January 2016.(1)

The rate charged on the Operational Standing Lending Facilityremained at 25 basis points above Bank Rate. However, giventhe large aggregate supply of reserves, there was no demand

from market participants to use the lending facility. Theaverage use of the lending facility was also £0 million over thequarter to 14 January 2016.

Indexed Long-Term Repo operationsThe Bank conducts regular Indexed Long-Term Repo (ILTR)operations as part of its provision of liquidity insurance tobanks, building societies and broker-dealers. During thereview period, the Bank offered a minimum of £5 billion viasix-month repos in each of its ILTR operations on 8 December2015, 5 January 2016 and 9 February 2016 (Table A).

Participation in, and usage of, ILTR operations has continuedto remain higher than during the same period last year, withsome participants using the facility as a source of term repoliquidity. Nonetheless, the total amount allocated in eachoperation remained below the minimum £5 billion on offer(Chart 13). Over the review period, a total of £6.1 billion ofILTRs matured and £8.1 billion of new ILTRs were allocated,resulting in a net increase in central bank reserves of around£2 billion.

The Bank announced on 7 March that it will offer threeadditional ILTR operations in the weeks around theEU referendum. These operations are to be held on 14 June,21 June and 28 June 2016. The Bank will continue to monitormarket conditions carefully and stands ready to takeadditional action if necessary.

Contingent Term Repo FacilityThe Contingent Term Repo Facility (CTRF) is a contingentliquidity facility that the Bank can activate in response to

High-yield (US dollar) (right-hand scale) High-yield (euro) (right-hand scale) High-yield (sterling) (right-hand scale) Investment-grade (US dollar) (left-hand scale) Investment-grade (euro) (left-hand scale) Investment-grade (sterling) (left-hand scale)

Previous Bulletin

80

120

160

200

240

280

320

0

100

200

300

400

500

600

700

800

900

Jan. Apr. July Oct. Jan.

Basis points Basis points

2015 16

Source: BofA Merrill Lynch Global Research.

Chart 12 International corporate bond option-adjustedspreads

Table A Indexed Long-Term Repo operations(a)

Total Collateral set summary

Level A Level B Level C

8 December 2015 (six-month maturity)

Minimum on offer (£ millions) 5,000

Total bids received (£ millions) 3,600 3,550 10 40

Amount allocated (£ millions) 3,600 3,550 10 40

Clearing spread (basis points) 0 5 15

5 January 2016 (six-month maturity)




Clearing spread (basis points) 0 n.a. 15

9 February 2016 (six-month maturity)




Clearing spread (basis points) 0 5 15

(a) The minimum amount on offer is the size of the operation that the Bank is willing to allocate, in aggregate,across all collateral sets at the minimum clearing spreads.

(1) Operational Standing Facility usage data are released with a lag.


actual or prospective market-wide stress of an exceptionalnature. The Bank reserves the right to activate the facility as itdeems appropriate. In light of market conditions throughoutthe review period, the Bank judged that CTRF auctions werenot required.

Discount Window FacilityThe Discount Window Facility (DWF) is a bilateral on-demandfacility provided to institutions experiencing a firm-specific ormarket-wide liquidity shock. It allows participants to borrowhighly liquid assets in return for less liquid collateral inpotentially large size and for a variable term. The Bankpublishes quarterly data of DWF usage with a lag. The averagedaily amount outstanding in the DWF in the three months to30 September 2014 was £0 million.

Other operationsFunding for Lending SchemeThe Funding for Lending Scheme (FLS) was launched bythe Bank and HM Treasury on 13 July 2012. The initialdrawdown period for the FLS ran from 1 August 2012 until31 January 2014. The drawdown period for the FLS extensionopened on 3 February 2014 and will run until 31 January 2018,following the extension beyond January 2016 announced on30 November 2015.(1)

The quantity current participants can borrow in the FLS islinked to their lending to the UK real economy from 2013 Q2to 2015 Q4, with the incentives currently skewed towardssupporting lending to small and medium-sized businesses.From 1 August 2016, borrowing allowances will reduce by

25%, and by the same amount every six months thereafter,phasing the scheme out gradually by 31 January 2018.

US dollar repo operationsThe Bank conducts seven-day US dollar liquidity-providingoperations and will continue to do so until further notice. Thenetwork of bilateral central bank liquidity swap arrangementsprovides a framework for the reintroduction of furtherUS liquidity operations if warranted by market conditions.There was no use of the Bank’s US dollar facilities throughoutthe review period.

Bank of England balance sheet: capital portfolioThe Bank holds an investment portfolio that is approximatelythe same size as its capital and reserves (net of equityholdings, for example in the Bank for InternationalSettlements, and the Bank’s physical assets) and aggregatecash ratio deposits. The portfolio consists ofsterling-denominated securities. Securities purchased by theBank for this portfolio are normally held to maturity, thoughsales may be made from time to time, reflecting, for example,risk or liquidity management needs or changes in investmentpolicy. The portfolio currently includes around £5.6 billion ofgilts and £0.1 billion of other debt securities.

Asset purchasesIn the publication of the Inflation Report on 5 November 2015,the Monetary Policy Committee announced that it expects tomaintain the stock of purchased assets at £375 billion,including reinvesting the cash flows associated with allmaturing gilts held in the APF, at least until Bank Rate hasreached a level from which it can be cut materially.

A total of £14.7 billion of cash flow associated with giltredemptions was successfully reinvested in gilts across thecurve during December 2015 and January 2016. £6.3 billion ofwhich was invested in December 2015 and £8.4 billion inJanuary 2016.

The total stock of gilts outstanding in the APF, measured asproceeds paid to sellers, remains at £375 billion. The stock ofgilts comprised of £56.8 billion of purchases in the 3–7 yearsresidual maturity range, £150 billion in the 7–15 years residualmaturity range and £168.1 billion with a residual maturity ofgreater than 15 years (Chart 14).

Gilt lending facilityThe Bank continued to offer to lend gilts held in the APF viathe Debt Management Office (DMO) in return for otherUK government collateral. In the three months to31 December 2015, the daily average value of gilts lent, as part

0

10

20

30

40

50

60

70

0

1,000

2,000

3,000

4,000

5,000

6,000

Jan. July Jan. July Jan. July Jan. July Jan. July Jan.

Amount allocated (£ millions)

Clearing spread (basis points)

2011 16

Three-month Level A allocated (left-hand scale)

Three-month Level B allocated (left-hand scale)

Six-month Level A allocated (left-hand scale)

Six-month Level B allocated (left-hand scale)

Six-month Level C allocated (left-hand scale)

Level A clearing spread (right-hand scale)

Level B clearing spread (right-hand scale)

Level C clearing spread (right-hand scale)

12 13 14 15

(a) Where there has not been any allocation to a collateral set, no clearing spread is marked.

Chart 13 ILTR reserves allocation and clearing spreads(a)

(1) For more details, see www.bankofengland.co.uk/publications/Pages/news/2015/096.aspx.

www.bankofengland.co.uk/publications/Pages/news/2015/096.aspx

www.bankofengland.co.uk/publications/Pages/news/2015/096.aspx


of the gilt lending facility, was £316 million. The average dailylending in the previous quarter was lower at £152 million.

Corporate bondsThere were no purchases of corporate bonds during the reviewperiod. Future purchase or sale operations through thescheme will be dependent on market demand and conditions,which the Bank will keep under review in consultation with itscounterparties. Reflecting the recent lack of activity, thescheme currently holds no bonds.

Secured commercial paper facilityThe Bank continued to offer to purchase secured commercialpaper backed by underlying assets that are short term andprovide credit to companies or consumers that supporteconomic activity in the United Kingdom. No purchases weremade during the review period.

0255075

100125150175200225250275300325350375400

Feb. Feb. Feb. Feb. Feb. Feb. Feb. Feb.

15+ years7–15 years0–7 years

£ billions

2009 10 11 12 13 14 15 16

(a) Proceeds paid to counterparties on a settled basis.(b) Residual maturity as at the date of purchase.

Chart 14 Cumulative gilt purchases by maturity(a)(b)

Quarterly Bulletin Appendices 63

Appendices


The articles that have been published recently in theQuarterly Bulletin are listed below. Articles fromDecember 1960 to Winter 2005 are available on theBank’s website at:

www.bankofengland.co.uk/archive/Pages/digitalcontent/historicpubs/quarterlybulletins.aspx.

Articles from Spring 2006 onwards are available at:

www.bankofengland.co.uk/publications/Pages/quarterlybulletin/default.aspx.

Articles

2012 Q3– RAMSI: a top-down stress-testing model developed at the Bank of England– What accounts for the fall in UK ten-year government bond yields?– Option-implied probability distributions for future inflation– The Bank of England’s Real-Time Gross Settlement infrastructure– The distributional effects of asset purchases– Monetary Policy Roundtable

2012 Q4– The Funding for Lending Scheme– What can the money data tell us about the impact of QE?– Influences on household spending: evidence from the 2012 NMG Consulting survey– The role of designated market makers in the new trading landscape– The Prudential Regulation Authority

2013 Q1– Changes to the Bank of England– The profile of cash transfers between the Asset Purchase Facility and Her Majesty’s Treasury– Private equity and financial stability– Commercial property and financial stability– The Agents’ company visit scores– The Bank of England Bank Liabilities Survey– Monetary Policy Roundtable

2013 Q2– Macroeconomic uncertainty: what is it, how can we measure it and why does it matter?– Do inflation expectations currently pose a risk to the economy?

– Public attitudes to monetary policy– Cross-border bank credit and global financial stability– The Old Lady of Threadneedle Street– Central counterparties: what are they, why do they matter and how does the Bank supervise them?– A review of the work of the London Foreign Exchange Joint Standing Committee in 2012

2013 Q3– Macroprudential policy at the Bank of England– Bank capital and liquidity– The rationale for the prudential regulation and supervision of insurers– Recent developments in the sterling overnight money market– Nowcasting world GDP and trade using global indicators– The Natural Rate Hypothesis: an idea past its sell-by date– Monetary Policy Roundtable

2013 Q4– SME forbearance and its implications for monetary and financial stability– Bringing down the Great Wall? Global implications of capital account liberalisation in China– Banknotes, local currencies and central bank objectives– Banks’ disclosure and financial stability– Understanding the MPC’s forecast performance since mid-2010– The financial position of British households: evidence from the 2013 NMG Consulting survey– What can company data tell us about financing and investment decisions?– Tiering in CHAPS– The foreign exchange and over-the-counter interest rate derivatives market in the United Kingdom– Qualitative easing: a new tool for the stabilisation of financial markets

2014 Q1– Money in the modern economy: an introduction– Money creation in the modern economy– The Court of the Bank of England– Dealing with a banking crisis: what lessons can be learned from Japan’s experience?– The role of business model analysis in the supervision of insurers– Nowcasting UK GDP growth– Curiosities from the vaults: a Bank miscellany– Monetary Policy Roundtable

Contents of recent Quarterly Bulletins

www.bankofengland.co.uk/publications/Pages/quarterlybulletin/default.aspx


www.bankofengland.co.uk/archive/Pages/digitalcontent/historicpubs/quarterlybulletins.aspx

www.bankofengland.co.uk/archive/Pages/digitalcontent/historicpubs/quarterlybulletins.aspx


2014 Q2– The UK productivity puzzle– The Bank of England as a bank– Credit spreads: capturing credit conditions facing households and firms– Assessing the risk to inflation from inflation expectations– Public attitudes to monetary policy– How have world shocks affected the UK economy?– How has the Liquidity Saving Mechanism reduced banks’ intraday liquidity costs in CHAPS?– Risk managing loan collateral at the Bank of England– Sterling Monetary Framework Annual Report 2013–14– A review of the work of the London Foreign Exchange Joint Standing Committee in 2013

2014 Q3– Innovations in payment technologies and the emergence of digital currencies– The economics of digital currencies– How might macroprudential capital policy affect credit conditions?– Household debt and spending– Enhancing the resilience of the Bank of England’s Real-Time Gross Settlement infrastructure– Conference on Monetary and Financial Law– Monetary Policy Roundtable– Changes to the Bank’s weekly reporting regime

2014 Q4– Bank funding costs: what are they, what determines them and why do they matter?– Why is the UK banking system so big and is that a problem?– The interaction of the FPC and the MPC– The Bank of England’s approach to resolving failed institutions– The potential impact of higher interest rates on the household sector: evidence from the 2014 NMG Consulting survey

2015 Q1– Investment banking: linkages to the real economy and the financial system– Desperate adventurers and men of straw: the failure of City of Glasgow Bank and its enduring impact on the UK banking system– Capital in the 21st century– The Agencies and ‘One Bank’– Self-employment: what can we learn from recent developments?– Flora and fauna at the Bank of England– Big data and central banks

2015 Q2– Mapping the UK financial system– Banking sector interconnectedness: what is it, how can we measure it and why does it matter?– The prudential regulation of insurers under Solvency II– A bank within a bank: how a commercial bank’s treasury function affects the interest rates set for loans and deposits– Do inflation expectations currently pose a risk to inflation?– Innovations in the Bank’s provision of liquidity insurance via Indexed Long-Term Repo (ILTR) operations– A review of the work of the London Foreign Exchange Joint Standing Committee in 2014

2015 Q3– How has cash usage evolved in recent decades? What might drive demand in the future?– Bank failure and bail-in: an introduction– Insurance and financial stability– How much do UK market interest rates respond to macroeconomic data news?– Estimating market expectations of changes in Bank Rate– Over-the-counter (OTC) derivatives, central clearing and financial stability– Monetary Policy Roundtable

2015 Q4– Bonus regulation: aligning reward with risk in the banking sector– The Prudential Regulation Authority’s secondary competition objective– Trends in UK labour supply– The potential impact of higher interest rates and further fiscal consolidation on households: evidence from the 2015 NMG Consulting survey– BoE-HKMA-IMF conference on monetary, financial and prudential policy interactions in the post-crisis world

2016 Q1– How could a shock to growth in China affect growth in the United Kingdom?– Wages, productivity and the changing composition of the UK workforce– Bank of England notes: the switch to polymer– Capturing the City: Photography at the Bank of England– The small bank failures of the early 1990s: another story of boom and bust


The Bank of England publishes information on all aspects of its work in many formats. Listed below are some of themain Bank of England publications. For a full list, please referto our website:

www.bankofengland.co.uk/publications/Pages/default.aspx.

Staff working papers

An up-to-date list of staff working papers is maintained on the Bank of England’s website at:

www.bankofengland.co.uk/research/Pages/workingpapers/default.aspx

where abstracts of all papers may be found. Papers publishedsince January 1997 are available in full, in portable documentformat (PDF).

No. 575 Long-run priors for term structure models(December 2015)Andrew Meldrum and Matt Roberts-Sklar

No. 576 A global factor in variance risk premia and local bondpricing (December 2015)Iryna Kaminska and Matt Roberts-Sklar

No. 577 Adaptive models and heavy tails (January 2016)Davide Delle Monache and Ivan Petrella

No. 578 The varying coefficient Bayesian panel VAR model(January 2016)Tomasz Wieladek

No. 579 What can Big Data tell us about the passthrough ofbig exchange rate changes? (January 2016)John Lewis

No. 580 Centralized trading, transparency and interest rateswap market liquidity: evidence from the implementation ofthe Dodd-Frank Act (January 2016)Evangelos Benos, Richard Payne and Michalis Vasios

No. 581 Policy and macro signals as inputs to inflationexpectation formation (January 2016)Paul Hubert and Becky Maule

No. 582 How does labour market structure affect theresponse of economies to shocks? (January 2016)Aurelijus Dabusinskas, Istvan Konya and Stephen Millard

No. 583 A Bayesian VAR benchmark for COMPASS(January 2016)Sílvia Domit, Francesca Monti and Andrej Sokol

No. 584 Macroprudential policy under uncertainty(January 2016)Saleem Bahaj and Angus Foulis

No. 585 Output gaps, inflation and financial cycles in theUnited Kingdom (February 2016)Marko Melolinna and Máté Tóth

External MPC Unit discussion papers

The MPC Unit discussion paper series reports on researchcarried out by, or under supervision of, the external membersof the Monetary Policy Committee. Papers are available fromthe Bank’s website at:

www.bankofengland.co.uk/monetarypolicy/Pages/externalmpcpapers/default.aspx.

The following papers have been published recently:

No. 44 The spillovers, interactions, and (un)intendedconsequences of monetary and regulatory policies(January 2016)Kristin Forbes, Dennis Reinhardt and Tomasz Wieladek

No. 45 Monetary policy and the current account: theory andevidence (March 2016)Ida Hjortsoe, Martin Weale and Tomasz Wieladek

Monetary and Financial Statistics

Monetary and Financial Statistics (Bankstats) contains detailed information on money and lending, monetary andfinancial institutions’ balance sheets, banks’ income andexpenditure, analyses of bank deposits and lending, externalbusiness of banks, public sector debt, money markets, issues of securities, financial derivatives, interest and exchange rates, explanatory notes to tables and occasional relatedarticles.

Bankstats is published on a monthly basis, free of charge, onthe Bank’s website at:

www.bankofengland.co.uk/statistics/Pages/bankstats/default.aspx.

Further details are available from the Statistics and RegulatoryData Division, Bank of England: telephone 020 7601 5432;email [email protected].

Bank of England publications

www.bankofengland.co.uk/statistics/Pages/bankstats/default.aspx

www.bankofengland.co.uk/statistics/Pages/bankstats/default.aspx

www.bankofengland.co.uk/monetarypolicy/Pages/externalmpcpapers/default.aspx

www.bankofengland.co.uk/monetarypolicy/Pages/externalmpcpapers/default.aspx




Articles that have been published in recent issues of Monetary and Financial Statistics can also be found on theBank’s website at:

www.bankofengland.co.uk/statistics/Pages/ms/articles.aspx.

Financial Stability Report

The Financial Stability Report is published twice a year underthe guidance of the Financial Policy Committee (FPC). Itcovers the Committee’s assessment of the outlook for thestability and resilience of the financial sector at the time ofpreparation of the Report, and the policy actions it advises toreduce and mitigate risks to stability. The Bank of Englandintends this publication to be read by those who areresponsible for, or have interest in, maintaining and promotingfinancial stability at a national or international level. It is ofespecial interest to policymakers in the United Kingdom andabroad; international financial institutions; academics;journalists; market infrastructure providers; and financialmarket participants. The Financial Stability Report is availableat:

www.bankofengland.co.uk/publications/Pages/fsr/default.aspx.

Handbooks in central banking

The series of Handbooks in central banking provide concise,balanced and accessible overviews of key central bankingtopics. The Handbooks have been developed from studymaterials, research and training carried out by the Bank’sCentre for Central Banking Studies (CCBS). The Handbooksare therefore targeted primarily at central bankers, but arelikely to be of interest to all those interested in the varioustechnical and analytical aspects of central banking. TheHandbook series also includes ‘Technical Handbooks’ which areaimed more at specialist readers and often contain moremethodological material than the Handbooks, incorporatingthe experiences and expertise of the author(s) on topics thataddress the problems encountered by central bankers in theirday-to-day work. All the Handbooks are available via theBank’s website at:

www.bankofengland.co.uk/education/Pages/ccbs/handbooks/default.aspx.

The Bank of England’s Sterling MonetaryFramework (the ‘Red Book’)

The ‘Red Book’ describes the Bank of England’s framework forits operations in the sterling money markets, which is designedto implement the interest rate decisions of the Monetary

Policy Committee while meeting the liquidity needs, and socontributing to the stability of, the banking system as a whole.It also sets out the Bank’s specific objectives for theframework, and how it delivers those objectives. Theframework was introduced in May 2006. The ‘Red Book’ isavailable at:

www.bankofengland.co.uk/markets/Documents/money/publications/redbook.pdf.

Cost-benefit analysis of monetary andfinancial statistics

The handbook describes a cost-benefit analysis (CBA)framework that has been developed within the Bank to ensurea fair balance between the benefits derived from good-qualitystatistics and the costs that are borne by reporting banks.Although CBA is a well-established approach in othercontexts, it has not often been applied to statistical provision,so techniques have had to be adapted for application to the Bank’s monetary and financial statistics. The handbook alsodiscusses how the application of CBA has enabled cuts in boththe amount and the complexity of information that is requiredfrom reporting banks.

www.bankofengland.co.uk/statistics/Pages/about/cba.aspx.

Credit Conditions Survey

Developments in credit conditions are of key interest to theBank of England in its assessment of economic conditions.This quarterly survey of bank and building society lenders is aninput to this assessment. The survey covers secured andunsecured lending to households and small businesses; andlending to non-financial corporations, and to non-bankfinancial firms. Copies are available on the Bank’s website at:

www.bankofengland.co.uk/publications/Pages/other/monetary/creditconditions.aspx.

Bank Liabilities Survey

Developments in lenders’ balance sheets are of key interest tothe Bank of England in its assessment of economic conditions.The aim of this quarterly survey of banks and building societylenders is to improve understanding of the role of lenders’liabilities and capital in driving credit and monetary conditions.The survey covers developments in the volume and price ofbank funding; developments in the loss-absorbing capacity ofbanks as determined by their capital positions; anddevelopments in the internal price charged to business unitswithin individual banks to fund the flow of new loans,

www.bankofengland.co.uk/publications/Pages/other/monetary/creditconditions.aspx

www.bankofengland.co.uk/publications/Pages/other/monetary/creditconditions.aspx

www.bankofengland.co.uk/markets/Documents/money/publications/redbook.pdf

www.bankofengland.co.uk/markets/Documents/money/publications/redbook.pdf

www.bankofengland.co.uk/education/Pages/ccbs/handbooks/default.aspx

www.bankofengland.co.uk/education/Pages/ccbs/handbooks/default.aspx


sometimes referred to as the ‘transfer price’. Copies areavailable on the Bank’s website at:

www.bankofengland.co.uk/publications/Pages/other/monetary/bls/default.aspx.

Credit Conditions Review

This quarterly publication presents the Bank of England’sassessment of the latest developments in bank fundingconditions and household and corporate credit. It drawsmainly on long-established official data sources, such as theexisting monetary and other financial statistics collected bythe Bank, and other data sources such as surveys of businessesand data from other organisations. The analysis also draws onthe results of the Bank of England’s Bank Liabilities Survey andCredit Conditions Survey. Copies are available on the Bank’swebsite at:

www.bankofengland.co.uk/publications/Pages/creditconditionsreview/default.aspx.

Quarterly Bulletin

The Quarterly Bulletin explores topical issues relating to theBank’s core purposes of monetary and financial stability.Some articles present analysis on current economic andfinancial issues, and policy implications. Other articlesenhance the Bank’s public accountability by explaining theinstitutional structure of the Bank and the various policyinstruments that are used to meet its objectives. TheQuarterly Bulletin is available at:

www.bankofengland.co.uk/publications/Pages/quarterlybulletin/default.aspx.

Inflation Report

The Bank’s quarterly Inflation Report was first published in1993. The Report sets out the detailed economic analysis and inflation projections on which the Bank’s Monetary Policy Committee bases its interest rate decisions, andpresents an assessment of the prospects for UK inflation. TheInflation Report is available at:

www.bankofengland.co.uk/publications/Pages/inflationreport/default.aspx.

www.bankofengland.co.uk/publications/Pages/inflationreport/default.aspx

www.bankofengland.co.uk/publications/Pages/inflationreport/default.aspx



www.bankofengland.co.uk/publications/Pages/creditconditionsreview/default.aspx

www.bankofengland.co.uk/publications/Pages/creditconditionsreview/default.aspx

www.bankofengland.co.uk/publications/Pages/other/monetary/bls/default.aspx

www.bankofengland.co.uk/publications/Pages/other/monetary/bls/default.aspx

© Bank of England 2016ISSN 0005-5166Printed by Park Communications Limited