Indicators of financial integration in the euro area, September 2005 · 2006. 9. 15. · Price-based indicators measure discrepancies in asset prices based on their geographic origin

IND IC ATORS OF F INANC IAL INTEGRAT ION IN THE EURO AREA

SEPTEMBER 2005

In 2005 all ECB publications will feature

a motif taken from the

€50 banknote.

I ND I C ATORS OF F INANC I A LINTEGRAT ION IN THE EURO AREA

SEPTEMBER 2005

© European Central Bank, 2005

AddressKaiserstrasse 2960311 Frankfurt am MainGermany

Postal addressPostfach 16 03 1960066 Frankfurt am MainGermany

Telephone+49 69 1344 0

Websitehttp://www.ecb.int

Fax+49 69 1344 6000

Telex411 144 ecb d

All rights reserved. Reproduction foreducational and non-commercial purposesis permitted provided that the source isacknowledged.

ISBN 92-9181-853-4 (print)ISBN 92-9181-854-2 (online)

3ECB c

Indicators of financial integration in the euro areaSeptember 2005

EXECUTIVE SUMMARY

The integration of Europe’s financial marketsis an issue of high importance for theEurosystem. This report provides anassessment of the degree of integration in themain financial segments of the euro area,namely the money, bond, equity and bankingmarkets. It is based on a set of financialintegration indicators that will be publishedsemi-annually on the ECB’s web site. Thereport itself will be published annually, withthe aim of documenting and monitoring theprogress of financial integration in the euroarea. While this first release of these indicatorsalready covers many important dimensions offinancial systems, it is the intention of the ECBto further develop the analysis in the report andthe statistics in the future.

The available evidence suggests that the degreeof integration varies greatly depending on themarket segment. The unsecured money markethas been fully integrated since shortly after theintroduction of the euro. The repo market ishighly integrated albeit to a lower extent.Government bond markets were significantlyintegrated even before the start of EMU,although some yield differentials remain. Theindicators for the corporate bond market,which has grown considerably since the adventof the single currency, point to a high degree ofintegration. Progress has also been made in theintegration of euroarea equity markets, whereequity returns are increasingly determinedby common factors. Banking markets aregenerally much less integrated.

1 INTRODUCTION

A well integrated financial system increasesthe efficiency of the euro area economy byreducing the cost of capital and improving theallocation of financial resources. It alsocontributes to a smooth and effectiveimplementation of monetary policy throughoutthe euro area. Moreover, given the strongconnections between financial integration and

financial stability, a deeper financialintegration may have a potential impact on thestability of the whole financial system, an areaof great interest to central banks. All in all,there is substantial evidence that financialintegration may ultimately support higher andmore sustainable economic growth through avariety of channels. Against this background,the ECB has taken a proactive stance towardsfostering European financial integration withinthe limits of its capabilities and competence.

This report is a new initiative by the ECB. Itprovides an overall assessment of the degree offinancial integration in the different financialmarket segments of the euro area, ranging aswidely as from retail lending to wholesaleequity trading. It is based on a series ofindicators that will be regularly updated andpublished on the ECB’s website.

The report will be produced on an annual basiswith the aim of monitoring the progress offinancial integration in the euro area. The rangeof indicators used may be extended over time,in line with improvements in data availabilityand advances in research. While this firstreport already focuses on a number of crucialdimensions of financial systems, spanningfinancial and credit markets, it is the intentionof the ECB to further develop the dataset in thefuture. In particular, it is envisaged to addindicators on some important underlyingfactors of integration related to financialinstitutions and market infrastructures.Moreover, the ECB will strive to addintegration indicators for markets that couldnot be covered in this edition due to theirrelatively recent emergence.1

1 For example, three different phenomena can be identif ied. Oneis the degree to which the different national markets whichexisted before the launch of the euro have integrated over time.Another is the process by which new market segmentsoriginating in a particular country or f inancial centre havespread and integrated across Europe after the arrival of theeuro. And lastly, there are some important markets that cameinto existence after the launch of the euro, but were integratedfrom the outset. Two signif icant examples of the latterphenomenon are the overnight index swap market and themarket for synthetic collateralised debt obligation (CDOs).The initial set of indicators presented here does not yet coverthese two relatively young markets.

4ECB cIndicators of financial integration in the euro areaSeptember 2005

The market for a given financial instrument isconsidered fully integrated if all economicagents with the same relevant characteristicsacting in that market face a single set of rules,have equal access and are treated equally. Thisdefinition underpins the measures proposed inthe report.2 Two broad categories of indicatorsare considered: price-based indicatorsand quantity-based indicators. Price-basedindicators measure discrepancies in asset pricesbased on their geographic origin. In a perfectlyintegrated market, prices of assets with similarcharacteristics should mostly be influenced bycommon euro area factors. Quantity-basedindicators are used to investigate the extent towhich investors have internationalised theirportfolios. They usefully complement price-based indicators, as in financially integratedmarkets investors will increase their holdings ofnon-domestic assets to fully reap the benefits ofinternational diversification. The indicators areeither “computed” or “model-based”. Computedindicators such as standard deviations, ratiosetc., are summary measures of the underlyingdata and are not model-specific, whereas theother indicators are derived from econometricmodels. Such model-based indicators have beenused when computed indicators do not allow adisentangling of country effects. As theseestimates depend on a specific set of modellingassumptions, they need to be interpreted withcaution. In the report, the focus is on euro-denominated instruments, with particularattention being paid to price differences acrosseuro area countries.

While analysing the developments in theindicators over time, it is worth keeping inmind that in some markets the progress infinancial integration per se is not easilydisentangled from the effects of the eliminationof exchange rate risk and the convergence ofinflation expectations across euro areacountries.

The report is structured as follows. Thefollowing sections analyse money, bond,equity and banking markets. Each sectionpresents an overall assessment of the current

degree of financial integration and discussesthe evolution of the indicators in more detail.There aree two annexes to this report. Annex 1contains all of the indicators while thetechnical documentation for each indicator canbe found in Annex 2.

2 MONEY MARKETS

The available indicators for the euro areamoney market point to an overall high degreeof integration. The unsecured money marketreached a stage of “near-perfect” integrationalmost immediately after the introduction ofthe euro. In addition, the repo market is highlyintegrated albeit to a lower extent.

The money market, broadly defined as themarket for interbank short-term debt ordeposit, consists of various segments. Thefinancial integration of the money market canbe analysed on the basis of the dispersion oflending rates offered by different banks in eachmarket segment. This section discussesindicators for the unsecured interbank marketand the repo market. The data refer to actualtransactions or offered rates by banks includedin the panels for the EONIA (euro overnightindex average), the EURIBOR (euro interbankoffered rate) and the EUREPO (benchmark ratefor secured money market transactions in theeuro area).

Figures 1, 2 and 3 show the cross-countrystandard deviation for the EONIA, EURIBORand EUREPO, respectively. According toFigures 1 and 2, the cross-sectional standarddeviation of the EONIA and EURIBOR lendingrates across euro area countries plummeted toclose to zero following the introduction of theeuro and remained stable thereafter. All theseindicators paint a consistent picture of a highlyintegrated unsecured money market.

2 For the economic and technical background to measuresproposed in this report, see L. Baele, A. Ferrando, P. Hördahl,E. Krylova and C. Monnet (2004), “Measuring financialintegration in the euro area”, ECB Occasional Paper No. 14.

5ECB c


Figure 3 shows the same type of indicatoras plotted in Figures 1 and 2 for the1-month and 12-month EUREPO rate. Thebehaviour of the 12-month EUREPO rateindicator in particular is characterised by largemovements, in part reflecting the lowerliquidity of this market. The indicators suggestthat there is a relatively high degree ofintegration also in the euro area repo market.

Figure 1 Cross-country standard deviat ionof the average overnight lending ratesamong euro area countries(60-day moving average; basis points)

Source: European Banking Federation, ECB calculations.

0

50

100

150

200

250

300

0

50

100

150

200

250

300

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

1999 2000 2001 2002 2003 2004 2005


0

50

100

150

200

250

300

0

50

100

150

200

250

300

1-month maturity12-month maturity

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

1999 2000 2001 2002 2003 2004 20050

0.2

0.4

0.6

0.8

1

1.2

1.4

0

0.2

0.4

0.6

0.8

1

1.2

1.4

Figure 2 Cross-country standard deviat ionof unsecured lending rates among euro areacountr ies(60-day moving average; basis points)

Figure 3 Cross-country standard deviat ionof repo rates among euro area countries

(60-day moving average; basis points)


1-month maturity12-month maturity

05

101520253035404550

05101520253035404550

2002 2003 2004 2005

2002 2003 2004 2005

2.5

2.0

1.5

1.0

0.5

0.0

2.5

2.0

1.5

1.0

0.5

0.0


3 BOND MARKETS

The available indicators show that with theintroduction of the euro and the removal ofexchange rate risk, government bond yieldshave converged in all euro area countries andincreasingly tend to be driven by commonnews. However, the importance of local factorscontinue to have some influence. This maypartly be explained by differences in liquidityand the availability of developed derivativesmarkets tied to the various individual bondmarkets. Additionally, bond yields in differentcountries also reflect differences in perceivedcredit risks. The impact of these factors,however, should not be seen as an indication ofa lack of integration. The euro area corporatebond market has grown considerably in recentyears. The available evidence suggests thatthis market is already reasonably wellintegrated.

3.1 GOVERNMENT BONDSMeasures of integration in government bondmarkets are based on yield differentials withrespect to a benchmark bond (e.g., the Germanbond for 10-year bonds and the French bond for2- and 5-year bonds). Given comparablematurities and other relevant characteristics,yield spreads for government bonds canprovide a direct measure of the degree ofintegration. A simple measure is to look atdevelopments in the standard deviation of thesespreads over time: the higher the degree ofintegration, the lower the dispersion. Inassessing this indicator, it should be borne inmind that movements related to changes in thecredit risk perceptions by the market do notsignal a variation in the degree of integration.In order to address this problem, additionalindicators have been developed. Anothercommonly used measure, referred to as “beta-convergence”, is based on the economicintuition that the more integrated the market is,the more bond yields should react to commonfactors instead of local factors.3

Figure 4 shows the evolution over time of thestandard deviations of government yieldspreads over German bonds (for 10-year bonds)and French bonds (for 2- and 5-year bonds).The figure shows a significant drop in theseindicators in the run-up to the EMU, which thenremain close to zero from 2001 onwards (theyear in which Greece joined EMU). The sharpdecline of these indicators signals that the euroarea government bond market has reached avery high level of integration.

Figure 4 Standard deviation of governmentbond yield spreads for 2- , 5- and 10-yearmaturit ies(yield dispersion in %)

Source: ECB.Note: As a benchmark, the German government bondyield is taken for the 10-year maturity and the Frenchgovernment bond yields are taken for 2- and 5-yearmaturities. Greece enters the standard deviationcalculations for all maturities at the date of its entry toEMU.

2 year5 year10 year

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

These conclusions are confirmed by theindicators shown in Figure 5, concerning theevolution of the beta-convergence. The plotted“betas” represent the estimated correlation ofchanges in the 10-year government bond yieldof a given country with changes in the German10-year government bond yield. The betasvaried substantially up to 1998 and convergedafterwards towards 1, the perfect integrationlevel. Greek bond yields only converged after2001.

3 An alternative widely used measure is the “sigma-convergence”. This takes into consideration the evolution ofdispersions over time in a regression analysis.

7ECB c


Figure 5 Evolution of beta coef f ic ients forsome selected countries

(10-year government bond yields; 18-month rollingregression)

Source: ECB.Note: Based on a model regressing national 10-yeargovernment bond yields on benchmark German 10-yeargovernment bond yields.

0.85

0.90

0.95

1.00

1.05

1.10

1.15

0.85

0.90

0.95

1.00

1.05

1.10

1.15

2002 2003 2004 2005

-3.0-2.5-2.0-1.5-1.0-0.50.00.51.01.52.02.5

-3.0-2.5-2.0-1.5-1.0-0.50.00.51.01.52.02.5

FranceGreeceItalythe NetherlandsSpain

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

3.2 CORPORATE BONDSThe yield on a corporate bond typicallydepends on a number of factors, such as thecredit rating, time-to-maturity, liquidity andcash-flow structure. With full integration, theimpact of these specific factors should beidentical across all countries. Using the sameset of specific factors, it is possible to obtainmeasures of corporate bond market integrationby investigating whether or not risk-adjustedyield spreads have a systematic countrycomponent. In an integrated market, theproportion of the total yield spread variancethat is explained by country effects should beclose to zero.

Following this approach the indicator showsthat the euro area corporate bond market isquite well integrated. Country effects are seento explain only a very small proportion of thecross-sectional variance of corporate bondyield spreads (see Figure 6).4

4 It should be noted that the total explanatory power of theunderlying regression is relatively low. This is in line withprevious studies on the determinants of corporate bond riskpremia.

Figure 6 Proportion of cross-sect ionalvariance explained by various factors

(percentages)

Source: Bloomberg and ECB calculations.

explained by regressionexplained by common, maturity, coupon, liquidity and sector effectsexplained by rating effectexplained by country effect

-505

10152025303540

-50510152025303540

1998 1999 2000 2001 2002 2003 2004

4 EQUITY MARKETS

The measures of euro area equity marketintegration indicate a rising degree ofintegration. First, since the end of 2000, theadvantages of a sector diversification seem tohave – for the first time in the last 30 years –surpassed those of a geographicaldiversification. Recent data show that thisreversal was short-lived but suggest increasedcorrelations of both country and sectorreturns. Second, equity returns in the euro areacountries are increasingly determined bycommon factors. However, there appears tostill be scope for further integration.

In an integrated equity market, country-specific factors should be of less importance.The more integrated the market, the greater thebenefits of diversification through sector-based equity investment strategies rather thanthrough country-based ones. By comparing thecross-sector dispersion with the cross-countrydispersion of equity returns, it is thereforepossible to derive indicators of financialintegration.


Figure 7 Fi ltered country and sectordispers ions of euro area equity returns

(percentages)

Source: Thomson Financial Datastream and ECBcalculations.Note: The f igure shows Hodrick-Prescott f iltered cross-sectional dispersions of monthly euro area country andsector equity returns.

country dispersionsector dispersion

1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 20050

1

2

3

4

5

6

7

0

1

2

3

4

5

6

7

Figure 7 plots the country and sectordispersions of monthly equity returns overtime. For nearly the whole sample, the countrydispersion has been higher than the sectordispersion. To the extent that dispersion andcorrelation are inversely related, this findingsuggests that a country diversification strategyhas been superior to a sector diversificationstrategy for most of the time. The differencebetween country and sector dispersionsnarrowed in the late 1990s and the sectordispersion has even slightly exceeded thecountry dispersion from 2001 onwards,suggesting a possible shift in the assetallocation paradigm from country-based tosector-based strategies. However, both countryand sector dispersions have strongly decreasedin the most recent years. Obviously, thiscomplicates an interpretation of the results interms of the advantage of sector diversificationrelative to country diversification.

Alternative indicators of financial integrationin equity markets can be derived from factormodels. Under the assumption that equityreturns in the euro area countries react to a localand a global factor (proxied, respectively byshocks in aggregate euro area and US equity

markets), it is possible to estimate the intensity(beta) with which euro area and global shocksare transmitted to national equity markets. Thepart of local return fluctuations that is notexplained by the common factors can beinterpreted as the reaction to purely local news.Thus, a higher spillover intensity suggests ahigher degree of equity return co-movementsacross countries (i.e. a higher integration).

Figure 8 shows that the average sensitivity tocommon shocks has increased significantlyover the last decades, in particular as regardseuro area-wide return shocks. This can beinterpreted as an indication of increasedintegration of euro area equity markets. It canbe noted, however, that the contemporaneousincrease in US shock spillover intensitiessuggests that the increased integration mayalso be a global phenomenon. In this context,the slight decrease of US shock spilloverintensities and the continued increase ofeuro area spillover intensities since theestablishment of the euro are worth observing.

It is also possible to look at the proportion ofthe total domestic equity volatility that can beexplained by euro area-wide and US shocks

Figure 8 Euro area and US shock spi l loverintens ity

Source: Thomson Financial Datastream and ECBcalculations.Note: For each period, the f irst column shows theunweighted average intensity with which euro area-wideequity market shocks, other than those from the UnitedStates, are transmitted to local euro area equity markets.The second column shows the unweighted averageintensity with which US equity market shocks aretransmitted to local euro area equity market returns.

euro area shock spillover intensitiesUS shock spillover intensities

0.00.10.20.30.40.50.60.70.80.9

0.00.10.20.30.40.50.60.70.80.9

1973-1985 1986-1991 1992-1998 1999-2005

9ECB c


respectively (“variance ratios”). Ceterisparibus, a higher variance ratio that can beassociated with euro area wide changes is anindication of a more integrated euro area equitymarket, signalling that national stock marketreturns are increasingly driven by commonnews. Figure 9 shows that the variance ratios

Figure 9 Proportion of variance in localequity returns explained by euro area andUS shocks(percentages)

Source: Thomson Financial Datastream and ECBcalculations.Note: For each period, the f irst column shows theunweighted average of the relative importance of euroarea-wide factors, other than US equity market shocks,for the variance in individual euro area countries’ equitymarket returns (“variance ratio”). The second columnshows the unweighted average of the relative impor tanceof US equity market shocks for the variance in euro areaequity market returns.

euro area shocksUS shocks

1973-1985 1986-1991 1992-1998 1999-20050

5

10

15

20

25

30

35

40

0

5

10

15

20

25

30

35

40

have increased over the past 30 years withrespect to both euro area-wide and US shocks,although the increase has been the strongest forthe former type of shocks. This suggests thatregional euro area integration has proceededmore quickly than the world wide integration.However, the still relatively low level ofvariance explained by euro area wide factors(about 35%) reveals that local shocks are stillrelatively important, indicating that furtherintegration is possible.

5 BANKING MARKETS

The available evidence suggests that the euroarea banking market remains highlyfragmented, as indicated by a high cross-sectional dispersion of the same types ofinterest rates and low cross-border bankingactivity. However, the euro area interbankmarket has shown signs of increasingintegration.

Integration in the retail bank market can bemeasured directly by looking at the dispersionof interest rates on consumer credit, lending forhouse purchase and deposits with agreedmaturity. Figure 10 suggests that cross-country

Figure 10 Cross-country standard deviat ionof MFI interest rates: loans to householdsand deposits with agreed maturity

Source: ECB.Note: The measure is based on MFI interest rates on newbusiness.

consumer creditlending for house purchasedeposits with agreed maturity

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Jan. Mar. May July Sep. Nov. Jan. Mar. May Jan. Mar. MayJuly Sep. Nov.2003 2004 2005

Figure 11 MFI loans to non-MFIs: outstandingamounts by residency of the counterparty as a share oftotal loans granted by MFIs, excluding the Eurosystem(percentages)

Source: ECB.Note: This indicator displays the geographicalcounterparty diversif ication of loans granted by euro areaMFIs to non-MFI counterpar ties resident in other euroarea countries and non-euro area EU Member States.

other euro area countriesrest of EU

Sep.1997

Mar.2005

Mar. Sep.1998

Mar. Sep.1999

Mar. Sep.2000

Mar. Sep.2001

Mar. Sep.2002

Mar. Sep.2003

Mar. Sep.2004

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0


Figure 12 MFI loans to MFIs: outstanding amounts byresidency of the counterparty as a share of totalloans granted by MFIs, excluding the Eurosystem(percentages)

Source: ECB.Note: This indicator displays the geographicalcounterpar ty diversif ication of loans granted by euro areaMFIs to MFI counterparties resident in other euro areacountries and non-euro area EU Member States.


Sep.1997

Mar.2005

Mar. Sep.1998

Mar. Sep.1999

Mar. Sep.2000

Mar. Sep.2001

Mar. Sep.2002

Mar. Sep.2003

Mar. Sep.2004

0

5

10

15

20

25

0

5

10

15

20

25

Figure 13 MFI holdings of securities issued by non-MFIs:outstanding amounts by residency of the issuer as ashare of total holdings, excluding the Eurosystem(percentages)

Source: ECB.Note: The indicator shows securities held by euro areaMFIs and issued by non-MFIs resident in other euro areacountries and non-euro area EU Member States.


Sep.1997

Mar.2005

Mar. Sep.1998

Mar. Sep.1999

Mar. Sep.2000

Mar. Sep.2001

Mar. Sep.2002

Mar. Sep.2003

Mar. Sep.2004

0

5

10

15

20

25

30

35

40

0

5

10

15

20

25

30

35

40

dispersions of bank interest rates haveremained relatively high (compared, forexample, with the government bond market)since January 2003.5 However, this may notonly be due to an incomplete integration. TheECB, together with the NCBs is undertaking acomprehensive analysis of cross-countryinterest rate dispersion.

Looking at the share of cross-border activity inthe total lending activity of MFIs, thesefindings are broadly confirmed. Cross-borderretail bank lending activity in the euro arearemains very limited (around 3.5% of the total),suggesting a fragmented retail banking market(see Figure 11).

Although domestic loans still account for morethan 50% of the total, the euro area interbankmarket has shown signs of increasingintegration as cross-border activity among theeuro area countries has increased substantially(see Figure 12). Furthermore, the share ofMFIs’ holdings of non-domestic (euro area inparticular) securities in total holdings hasincreased markedly in recent years, whichpoints to a higher degree of capital marketintegration in the euro area (see Figure 13).

5 Data prior to 2003 could be shown once historical standardisedMFI interest rate data become available. Using unharmoniseddata available up to January 2003, L. Baele, A. Ferrando,P. Hördahl, E. Krylova and C. Monnet (2004) showed thatcross-country dispersions of bank interest rates have declinedsomewhat since the early 1990s (partly due to the strongconvergence of inflation rates). Analogous to developments ingovernment bond yields, the declining trend has come to a haltin recent years, which may point to a slowdown in the processof integration in this market segment.

11ECB c


Money market indicators

N Indicator Description Computed/model- Sourcesbased

Government bond market indicators

1

2

3

4

5

6

7

8

Computed

Computed

Computed

Computed

Computed

Model-based

Model-based

Model-based

EBF/ECB

EBF

EBF

ECB

ECB

ECB

ECB

ECB

Cross-country standarddeviation of the averageovernight lending ratesamong euro areacountries

Cross-country standarddeviation of unsecuredlending rates among euroarea countries

Cross-country standarddeviation of repo ratesamong euro areacountries

Standard deviation ofgovernment bond yieldspreads for 10-yearmaturity

Standard deviation ofgovernment bond yieldspreads for 2-and 5-yearmaturities

Evolution ofbeta coefficients

Average distance ofintercept/beta fromvalues implied bycomplete integration

Variance ratio

The measure is based onaverage overnight ratesfor each of the euro areacountries, as reported byEONIA banks

The measure is based onthe unsecured 1- and12-month lendingrates

Based on the quotesreported by EUREPOpanel banks for 1- and12-month repo rates

Based on euro areacountry yields on10-year governmentbonds

Based on euro areacountry yieldson 2- and 5-yeargovernment bonds

Based on 18-monthsrolling regression ofchanges in country yieldsfor 10-year governmentbonds with respect tochanges in yields on thebenchmark (German) 10-year government bond

Based on the same modelas indicator 9


ANNEX 1 I ND I C ATOR S O F F I N ANC I A LI N T E G R AT I O N


Corporate bond market indicators


9

10

11

Model-based

Model-based

Model-based

Bloomberg

Bloomberg

Bloomberg

Proportion of cross-sectional varianceexplained by variousfactors

Estimated coefficients ofcountry dummies

Cross-sectionaldispersion of countryparameters

Based on the MerrillLynch EMU corporatebond index. The measureis derived by running aregression of spreadsrelative to a set ofvariables includingrating



12

13

14

Model-based

Model-based

Model-based

Thomson FinancialDatastream



Filtered country andsector dispersions of euroarea equity returns

Euro area and US shockspillover intensity

Proportion of variancein local equity returnsexplained by euro areaand US shocks

Based on monthlycross-sectional Hodrick-Prescott filtered totalequity returnson country and sectorindexes

The indicator is derivedfrom a model specifyingeuro area-wide and US(global) shocks toestimate the averagesensitivities of countryreturns to commonfactors

The indicator is derivedfrom a model specifyingeuro area-wide and US(global) shocks toestimate the proportion oftotal domestic equityvolatility explained bycommon factors

Equity market indicators

13ECB c



Banking market indicators

15

16

17

18

19

20

Computed

Computed

Computed

Computed

Computed

Computed

ECB

ECB

ECB

ECB

ECB

ECB

Cross-country standarddeviation of MFI interestrates on loans to non-financial corporations

Cross-country standarddeviation of MFI interestrates on loans tohouseholds and depositswith agreed maturity

MFI loans to non-MFIs:outstanding amounts byresidency of thecounterparty as a share oftotal loans granted byMFIs, excluding theEurosystem

MFI loans to MFIs:outstanding amounts byresidency of thecounterparty as a share oftotal loans granted byMFIs, excluding theEurosystem

MFI holdings ofsecurities issued bynon-MFIs: outstandingamounts by residency ofthe issuer as a share oftotal holdings, excludingthe Eurosystem

MFI holdings ofsecurities issued byMFIs: outstandingamounts by residency ofthe issuer as a share oftotal holdings, excludingthe Eurosystem

Based on MFI (monetaryfinancial institutions)interest rate statistics

Based on MFI interestrate statistics

Based on BSI (balancesheet items) statistics

Based on BSI statistics




MONEY MARKE TChart 1 Cross-country standard deviat ion ofthe average overnight lending rates amongeuro area countries(60-day moving average; basis points)


0

50

100

150

200

250

300

0

50

100

150

200

250

300

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

1999 2000 2001 2002 2003 2004 2005

Chart 2 Cross-country standard deviat ion of unsecured lending rates among euro areacountr ies(60-day moving average; basis points)


0

50

100

150

200

250

300

0

50

100

150

200

250

300

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

1-month maturity

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1999 2000 2001 2002 2003 2004 2005

0

50

100

150

200

250

300

0

50

100

150

200

250

300

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

12-month maturity

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1999 2000 2001 2002 2003 2004 2005

15ECB c


Chart 3 Cross-country standard deviat ion of repo rates among euro area countries

(60-day moving average; basis points)


05

101520253035404550

05101520253035404550

2002 2003 2004 2005

1-month maturity

0.0

0.5

1.0

1.5

2.0

2.5

0.0

0.5

1.0

1.5

2.0

2.5

2002 2003 2004 2005

05

101520253035404550

05101520253035404550

2002 2003 2004 2005

12-month maturity

0.0

0.5

1.0

1.5

2.0

2.5

0.0

0.5

1.0

1.5

2.0

2.5

2002 2003 2004 2005


GOVERNMENT BOND MARKE TChart 4 Standard deviat ion of governmentbond yield spreads for 10-year maturity

(yield dispersion in %)

Source: ECB.Note: As a benchmark, the German government bondyield is taken. Greece enters the standard deviationcalculations for all maturities at the date of its entry toEMU.

10-year maturity

0.0

0.5

1.0

1.5

2.0

2.5

0.0

0.5

1.0

1.5

2.0

2.5

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Chart 5 Standard deviat ion of governmentbond yield spreads for 2- and 5-yearmaturit ies(yield dispersion in %)

Source: ECB.Note: As a benchmark, the yield on French governmentbonds is taken. Greece enters the standard deviationcalculations for all maturities at the date of its entry toEMU.

2-year maturity5-year maturity

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Chart 6 Evolution of beta coef f ic ients


Source: ECB.Note: Based on a model regressing national 10-yeargovernment bond yields on benchmark German 10-yeargovernment bond yields.

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

AustriaBelgiumFinlandFranceGreece

IrelandItalythe NetherlandsPortugalSpain

0.85

0.90

0.95

1.00

1.05

1.10

1.15

0.85

0.90

0.95

1.00

1.05

1.10

1.15

2002 2003 2004 2005

Chart 7 Average distance of intercept/betafrom values impl ied by completeintegrat ion(10-year government bond yields; 18-month rollingregression)

Source: ECB.Note: See Chart 6.

0.000.020.040.060.080.100.120.140.160.180.20

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 20050.0

0.2

0.4

0.6

0.8

1.0

1.2

dispersion in intercepts (left-hand scale)dispersion in betas (right-hand scale)

0.002

0.004

0.006

0.008

0.010

0.012

0.014

2002 2003 2004 20050.040.050.050.060.060.070.070.080.08

17ECB c


Chart 8 Variance rat io


Source: ECB.Note: See Char t 6. Proportion of the variance of national10-year government bond yields that is explained by thevariance in the benchmark German 10-year governmentbond yield.

0102030405060708090

100

0102030405060708090100

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

90

92

94

96

98

100

90

92

94

96

98

100

2002 2003 2004 2005

AustriaBelgiumFinlandFranceGreece

IrelandItalythe NetherlandsPortugalSpain


CORPORAT E BOND MARKE TChart 9 Proportion of cross-sect ionalvariance explained by various factors

(6-month averages; percentages)

explained by regressionexplained by common, maturity, coupon, liquidity and sector effectsexplained by rating effectexplained by country effect

-505

10152025303540

-50510152025303540

1998 1999 2000 2001 2002 2003 2004

Chart 10 Est imated coeff ic ients of countrydummies

(6-month averages)

Sources: Bloomberg and ECB calculations.

AustriaGermanySpainFranceIrelandthe NetherlandsItaly

-20

-15

-10

-5

0

5

10

15

20

-20

-15

-10

-5

0

5

10

15

20

1998 1999 2000 2001 2002 2003 2004

Chart 11 Cross-sect ional dispers ion ofcountry parameters

(basis points)


0

2

4

6

8

10

12

14

16 16

0

2

4

6

8

10

12

14

1998 1999 2000 2001 2002 2003 2004 2005


19ECB c


EQU I T Y MARK E TChart 12 Fi ltered country and sectordispers ions of euro area equity returns

(percentages)

Source: Thomson Financial Datastream and ECBcalculations.Note: The char t shows Hodrick-Prescott f iltered cross-sectional dispersions of monthly euro area country andsector equity returns.

country dispersionsector dispersion

1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 20050

1

2

3

4

5

6

7

0

1

2

3

4

5

6

7

Chart 13 Euro area and US shock spi l loverintens ity

Source: Thomson Financial Datastream and ECBcalculations.Note: For each period, the f irst column shows theunweighted average intensity with which euro area-wideequity market shocks, other than those from the UnitedStates, are transmitted to local euro area equity markets.The second column shows the unweighted averageintensity with which US equity market shocks aretransmitted to local euro area equity market returns.

euro area shock spillover intensitiesUS shock spillover intensities

0.00.10.20.30.40.50.60.70.80.9

0.00.10.20.30.40.50.60.70.80.9

1973-1985 1986-1991 1992-1998 1999-2005

Chart 14 Proportion of variance in localequity returns explained by euro area andUS shocks(percentages)

Source: Thomson Financial Datastream and ECBcalculations.Note: For each period, the f irst column shows theunweighted average of the relative importance of euroarea-wide factors, other than US equity market shocks,for the variance in individual euro area countries’ equitymarket returns (“variance ratio”). The second columnshows the unweighted average of the relative impor tanceof US equity market shocks for the variance in euro areaequity market returns.

euro area shocksUS shocks

1973-1985 1986-1991 1992-1998 1999-20050

5

10

15

20

25

30

35

40

0

5

10

15

20

25

30

35

40


BANK I NG MARKE TChart 15 Cross-country standard deviat ionof MFI interest rates on loans to non-f inancial corporations

Source: ECB.Note: The measure is based on MFI interest rates on newbusiness.

floating rate and up to 1 year initial rate fixationover 1 year initial rate fixation

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2


Chart 16 Cross-country standard deviat ionof MFI interest rates on loans to householdsand deposits with agreed maturity

Source: ECB.Note: The measure is based on MFI interest rates on newbusiness. Deposits with agreed maturity include depositsfrom non-f inancial corporations and households.

consumer creditlending for house purchasedeposits with agreed maturity

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2


Chart 17 MFI loans to non-MFIs: outstanding amountsby residency of the counterparty as a share of totalloans granted by MFIs, excluding the Eurosystem(percentages)

Source: ECB.Note: This indicator displays the geographicalcounterpar ty diversif ication of loans granted by euro areaMFIs to non-MFI counterparties in other euro areacountries and non-euro area EU Member States.


Sep.1997

Mar.2005

Mar. Sep.1998

Mar. Sep.1999

Mar. Sep.2000

Mar. Sep.2001

Mar. Sep.2002

Mar. Sep.2003

Mar. Sep.2004

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Chart 18 MFI loans to MFIs: outstanding amounts byresidency of the counterparty as a share of totalloans granted by MFIs, excluding the Eurosystem(percentages)

Source: ECB.Note: This indicator displays the geographicalcounterpar ty diversif ication of loans granted by euro areaMFIs to MFI counterpar ts resident in other euro areacountries and non-euro area EU Member States.


Sep.1997

Mar.2005

Mar. Sep.1998

Mar. Sep.1999

Mar. Sep.2000

Mar. Sep.2001

Mar. Sep.2002

Mar. Sep.2003

Mar. Sep.2004

0

5

10

15

20

25

0

5

10

15

20

25

21ECB c


Source: ECB.Note: The indicator shows securities held by euro areaMFIs and issued by non-MFIs resident in other euro areacountries and non-euro area EU Member States.


Sep.1997

Mar.2005

Mar. Sep.1998

Mar. Sep.1999

Mar. Sep.2000

Mar. Sep.2001

Mar. Sep.2002

Mar. Sep.2003

Mar. Sep.2004

0

5

10

15

20

25

30

35

40

0

5

10

15

20

25

30

35

40

Chart 19 MFI holdings of securities issued by non-MFIs:outstanding amounts by residency of the issuer as ashare of total holdings, excluding the Eurosystem(percentages)

Chart 20 MFI holdings of securities issued by MFIs:outstanding amounts by residency of the issuer as ashare of total holdings, excluding the Eurosystem(percentages)

Source: ECB.Note: The indicator shows securities held by euro areaMFIs and issued by MFIs resident in other euro areacountries and non-euro area EU Member States. .


Sep.1997

Mar.2005

Mar. Sep.1998

Mar. Sep.1999

Mar. Sep.2000

Mar. Sep.2001

Mar. Sep.2002

Mar. Sep.2003

Mar. Sep.2004

0

5

10

15

20

25

30

0

5

10

15

20

25

30


23ECB c


MONEY MARKET INDICATORS (INDICATORS 1-3)

COMPILATIONThe ECB avails itself of business frequency(daily) data at the level of individualinstitutions for both unsecured and securedinterbank short-term debt or deposits. The dataare made available by the European BankingFederation (EBF) and cover the EONIA andEURIBOR (unsecured lending) as well as theEUREPO2 for different maturities.

For each dataset, the indicator is theunweighted standard deviation (D

t) of the

average daily interest rates prevailing in eacheuro area country. The number of euro areacountries (n in the formula below) reflects thenumber of countries having adopted the euro inthe reference period:

∑ −=c

ttct rrn

D 2, )(1

(1)

where rc,t

is the unweighted average of theinterest rate r

c

i,t reported by each of the m

c panel

banks at time t in a given country c:

∑=i

cti

ctc r

mr ,,

1(2)

The euro area average rt is calculated as the

unweighted average of the average interestrates r

c,t.

The data are smoothed by calculating a60-(business) day centred moving average ofthe standard deviation, transformed intomonthly figures taking the end-of-monthobservation of the smoothed series.

ADDITIONAL INFORMATIONThe individual rates of the banks of the EONIApanel are collected by the ECB on behalf of theEuropean Banking Federation. Every day onwhich TARGET (the Trans-EuropeanAutomated Real-Time Gross SettlementExpress Transfer system) is open, each EONIApanel bank reports to the ECB its aggregatevolume of intra-day unsecured lendingtransactions and the weighted average lendingrate for these transactions. All lending

ANNEX 2 ME THODOLOG I C A L NOT E S 1

transactions carried out before the closing ofTARGET at 6.00 p.m. (C.E.T.) have to bereported. The panel of reporting banks includesonly the most active banks located in the euroarea and beyond. Reported bid rates areconsidered to be national rates of country X, ifthe reporting bank has its headquarters incountry X. However, the counterparty of thetransaction is not known and the reportedinterest rate could actually (in part) refer totransactions with a bank outside country X.

EURIBOR and EUREPO rates do not relate totransactions conducted during the day but areindicative rates quoted by banks which aremembers of the respective panels. Every panelbank delivers its quotations for theseinstruments directly to Moneyline Telerate nolater than 10.45 a.m. (C.E.T.) on each day thatTARGET is open. Moneyline Telerate noresponsible for computing the aggregateEURIBOR and EUREPO indices and forproviding the underlying data to the EBF. Datafor EUREPO rates start in March 2002 whenthe EUREPO index was introduced.

1 For additional technical background, see L.Baele, A. Ferrando,P. Hördahl, E. Krylova and C. Monnet (2004), “Measuringfinancial integration in the euro area”, ECB Occasional PaperNo. 14.

2 EONIA stands for euro overnight index average. It is theeffective overnight reference rate for the euro and is computedas a weighted average of all overnight unsecured lendingtransactions undertaken in the interbank market, initiatedwithin the euro area by the contributing banks. The EONIA iscomputed with the help of the European Central Bank. Thebanks contributing to the EONIA are the same as theEURIBOR panel banks (composed of banks resident in theeuro area and in other EU Member States as well as someinternational banks). The EURIBOR (euro interbank offeredrate) is the benchmark rate of the large euro money market thathas emerged since 1999. The EUREPO is the benchmark rate ofthe euro repo market that has emerged subsequent to theintroduction of the euro in 1999. It is the rate at whichone prime bank offers funds in euro to another prime bankwhen the funds are secured by a repo transaction usinggeneral collateral. For further information, see http://www.euribor.org/default.htm and http://www.eurepo.org/.


GOVERNMENT BOND MARKET INDICATORS(INDICATORS 4-8)

Standard deviations of government bond yieldspreads (indicators 4-5)

COMPILATIONThe cross-country standard deviations ofgovernment bond yield spreads for 2-, 5- and10-year maturities are calculated on the basisof daily data for the government bond yieldspreads relative to the government bond yieldin the country selected as a benchmark for thecalculation (Germany for the 10-yearmaturities and France for 2- and 5-yearmaturities). The standard deviation (S

t) takes

the following form:

∑ −=c

tbtct yyn

S 2,, )(1

(3)

where yc,t

denotes the yield on the governmentbond of euro area country c with the relevantmaturity on day t and y

b,t is the yield on the

government bond of the country selected as abenchmark for that maturity (i.e. Germany orFrance).

In the second step, data are smoothed bycalculating a 60-(business) day centred movingaverage of the standard deviation, transformedinto monthly figures taking the end-of-monthobservation of the smoothed series.

The standard deviation of 10-year governmentbond yield spreads is based on bonds fromBelgium, Greece, Spain, France, Ireland, Italy,the Netherlands, Austria, Portugal and Finland.For the 5-year maturity, the government bondsof Belgium, Germany, Greece, Spain, Ireland,Italy, the Netherlands, Austria, Portugal andFinland are used. For the 2-year maturity, themeasure is based on bonds from Belgium,Germany, Greece, Spain, Italy, theNetherlands, Austria, Portugal and Finland.Greece enters the standard deviationcalculations for all maturities at the date of itsentry to EMU. For Luxembourg no benchmarkbond for the residual maturity of close to two,five or ten years exists.

ADDITIONAL INFORMATIONNot all government debt in the euro area is fullysubstitutable as regards, e.g. perceived creditrisk or liquidity of the relevant bonds. Thismight affect the yields on the selected bondsand thus the computed indicator.

Evolution of beta coefficients (indicator 6)

COMPILATIONIf bond markets are fully integrated andcountry-specific changes in perceived creditrisks do not occur, bond yields should onlyreact to news common to all markets. That is,changes in the bond yields of individualcountries should react exclusively to commonnews, which is reflected in a change of thebenchmark government bond yield. To separatecommon from local influences, the followingregression is run:

tctgertctctc RR ,,,,, εβα +∆+=∆ (4)

where a denotes a country- and time-varyingintercept; b is a country- and time-dependentbeta with respect to the benchmark (German)bond yield; DR is the change in the bond yieldand e is a country-specific shock.

The conditional betas are derived by estimatingthe above regression using the first 18 monthsof monthly averages. Subsequently, the datawindow is moved one month ahead and theequation is re-estimated, until the lastobservation is reached. A time series for b

c,t is

thus obtained.

ADDITIONAL INFORMATIONThe outcome of the econometric specificationdepends on the selection of the mostappropriate benchmark bond, in this case the10-year German government bond. In addition,one should not expect that common factorsfully explain changes in local bond yields as“local news” concerning credit and liquidityrisks will continue to have an impact on localyields.

25ECB c


Average distance of intercept/beta from valuesimplied by complete integration (indicator 7)

COMPILATIONThis indicator is derived using regression (4),as for the previous indicator. From theindividual country regressions, the unweightedaverage a

c,t and b

c,t values are calculated and

measured in proportion to the values impliedby complete market integration (0 and 1,respectively). The analysis is based on monthlyaverages of government bond yields.

ADDITIONAL INFORMATIONSame as for indicator 9.

Variance ratio (indicator 8)

COMPILATIONThis indicator measures the proportion of thevariance in local (country-specific) yields thatis explained by the variance in the benchmark(German) 10-year government bond yield; i.e.the “variance ratio”. The indicator is derivedfrom the same 18-month rolling regression asfor the previous two indicators (see equation(4) above). The total variance in local yields isgiven by

( ) ( ) ( )tctbtctc VarRVarRVar ,,2,, εβ +∆=∆ (5)

and the variance ratio by:

( )( )tc

tbtctc RVar

RVarVR

,

,2,

, ∆∆

=β

(6)

Hence, a variance ratio close to one is obtainedwhen the beta approaches one and when thevolatilities of the local and the benchmark bondyield changes are of a similar magnitude. Theanalysis is based on monthly averages ofgovernment bond yields.

ADDITIONAL INFORMATIONSame as for indicators 9 and 10.

CORPORATE BOND MARKET (INDICATORS 9-11)

Proportion of cross-sectional variance explained byvarious factors (indicator 9)

COMPILATIONThis indicator is derived by estimating thefollowing equation using an Ordinary LeastSquares (OLS) technique:

( ) rti

K

rtrtt

irc CRztSP ,

1,, ,, ∑

=

++= γατ

ti

N

ctcitc

itt

sti

sts eCzS ,

1,,,,

2

1, ∑∑

==

+++ βϕδ (7)

where ( )ti

rc ztSP ,,, τ is the yield spreadfor corporate bond i at time t issued in country cwith τ years to maturity, with credit rating r andset of instruments z

t. a is an intercept common

to all corporate bonds, CRri,t

is a rating dummywhich takes a value of one when corporate bondi belongs to rating category r at time t and zerootherwise, and Ss

i,t is a sector dummy which

takes a value of one for financial corporationsand zero for non-financial corporations. Theparameter vector u groups the sensitivities ofthe various corporate bonds to the instrumentscontained in zi

t, namely time-to-maturity,

liquidity and coupon of the ith bond. As a proxyof liquidity, we use the ratio of the number ofdays that the bond has been traded to the totalnumber of trading days within every timeinterval. tciC ,, is a country dummy that equalsone when corporate bond i belongs to country cat time t, and zero otherwise.

The sample is composed of 1,990 individualbonds that are used in the Merrill Lynch EMUcorporate bond index, which incorporates euro-denominated investment-grade bonds with aminimum size of issue of €100 million. Bondsrated below investment grade and asset-backedbonds are excluded from the analysis. Inaddition, bonds with less than one year tomaturity and bonds which were traded less thanonce per week in a given four-week time intervalwere excluded. All euro-denominated bonds notissued in a euro area country were eliminated aswell as data for countries that do not have at least


ten corporate bonds in every time interval. Thus,the analysis is based on a sample of bonds issuedin seven countries: Austria, France, Germany,Ireland, Italy, the Netherlands and Spain. Italyhas been included in the regression analysisfrom June 2003.

The indicator represents the proportion ofcross-sectional variance explained by thevarious components (common, rating, sector,maturity, liquidity coupon and country effects)over time.

Estimated coefficients of country dummies overtime (indicator 10)

COMPILATIONAs a test for integration, we test whether thecountry parameters b

c,t in equation (7) are zero,

or at least converge towards zero.

Cross-sectional dispersion of country parameters(indicator 11)

COMPILATIONThis indicator is derived by calculating theaverage size of the estimated country dummiesderived from regression (7). An overalldecrease in the dispersion of the country effectswould be an indication of increasingintegration of the corporate bond market.

ADDITIONAL INFORMATIONBefore June 2003 the indicators werecalculated without Italy.

EQUITY MARKET INDICATORS(INDICATORS 12-14)

Filtered country and sector dispersions of euro areaequity returns (indicator 12)

COMPILATIONThis indicator is derived by calculating thecross-sectional dispersion of both sector andcountry index returns for the euro areacountries.3 Data refer to the EMU global sectorindices provided by Datastream and arecalculated on a weekly basis from January 1973

onwards. They include (reinvested) dividendsand are denominated in euro.

The cross-sectional dispersions are filteredusing the Hodrick-Prescott smoothingtechnique, which provides a smooth estimate ofthe long-term trend component of the series.

ADDITIONAL NOTESThe indicator is useful for uncoveringstructural changes in the aggregate euro areaequity market, but is less informative aboutsuch changes in individual markets.

Euro area and US shock spillover intensity(indicator 13)

COMPILATIONThis measure is equivalent to the bond marketnews-based indicators (e.g. indicator 9).However, empirical evidence suggests thatequity returns are significantly driven byglobal factors. For that reason, both euro area-wide shocks and US shocks (as a proxy forglobal factors) are included in the assessmentof common news.

To calculate the relative importance of euroarea-wide and US stock market fluctuations forlocal stock market returns, the stock marketreturns of individual countries are modelled ashaving an expected component, and anunexpected component, e

c,t.4 The unexpected

component is then decomposed into a purelylocal shock (e

c,t) and a reaction to euro area

news (eeu,t

) as well as world (US) news (eus,t

):

tusus

tcteueutctctc e ,,,,,, εβεβε ++= (8)

where b represents the country-dependentsensitivity to euro area and US market changes

3 This indicator is based on an approach presented byK. Adjaouté, and J.-P Danthine (2003), “European f inancialintegration and equity returns: a theory-based assessment”, inV. Gaspar, P. Hartmann and O. Sleijpen (eds. 2002).“Thetransformation of the European financial system”.

4 The expected return is obtained relating euro area and USreturns to a constant term and to the returns in the previousperiod. The conditional variance of the error terms is governedby a bivariate asymmetric GARCH (1,1) model.

27ECB c


(of the unexpected component of equityreturns), respectively.

In order to investigate the development of thebetas over time, three dummy variables areintroduced for the periods 1986-1992, 1992-1998, and 1998-2005.

For each period, the indicators report theunweighted average intensity with which euroarea-wide equity market shocks, other thanthose from the United States, are transmittedto local euro area equity markets and theunweighted average intensity with which USequity market shocks are transmitted to localeuro area equity markets.

Data refer to the EMU global sector indices andare calculated on a weekly basis from January1973 onwards.

ADDITIONAL INFORMATIONTo be able to distinguish global shocks frompurely euro area shocks, it is assumed that euroarea equity market developments are partlydriven by events in the US market. It isfurthermore assumed that the proportion oflocal returns not explained by common factorsis entirely due to local news.

Proportion of variance in local equity returnsexplained by euro area and US shocks (indicator 14)

COMPILATIONTo compare the relevance of euro area and USshocks across average changes in countryreturns, the indicators report the varianceratios, i.e. the proportion of total domesticequity volatility explained by euro area and USshocks, respectively. The indicator is derivedby assuming that the total variance inindividual country-specific returns is given by

( ) ( ) 2,

22,

2,

2, tus

ustteu

euttctc h σβσβσ ++= (9)

where hc,t

is the variance of the local shockcomponent. The euro area variance ratio is thengiven by

( )2,

2,

2

,tc

teueuteu

tcVRσ

σβ= (10)

and correspondingly for the US. Theconditional variances are obtained from astandard asymmetric GARCH (1,1) model.

For each period, the indicators report theunweighted average of the relative importanceof euro area-wide factors, other than US equitymarket shocks, for the variance in individualeuro-area countries’ equity market returns(“variance ratio”), and the unweighted averageof the relative importance of US equity marketshocks for the variance in euro area equitymarkets returns.

Data refer to the EMU global sector indices andare calculated on a weekly basis from January1973 onwards.

ADDITIONAL INFORMATIONThe variance ratio is derived assuming thatlocal shocks are uncorrelated across countriesand that they are also not correlated with theeuro area and US benchmark indices.

BANKING MARKET INDICATORS(INDICATORS 15-20)

Cross-country standard deviations of MonetaryFinancial Institutions (MFI) interest rates(indicators 15-16)

COMPILATIONThe price measures for credit marketintegration are based on MFI interest rates(MIR) on new business reported to the ECB (inaccordance with the Regulation ECB/2001/18),at a monthly frequency as from January 2003.

In detail, the following instrument categoriesare considered:

– loans to non-financial corporations otherthan bank overdrafts, with a floating rate andup to one year of initial rate fixation;


– medium- and long-term loans to non-financial corporations, i.e. loans with overone year of initial rate fixation;

– consumer loans to households;

– housing loans to households; and

– time deposits, i.e. deposits by householdsand non-financial corporations with anagreed maturity.

For the purpose of measuring financialintegration, it would be preferable to computethe dispersion of rates as measured by thestandard deviation using unweighted interestrates at the level of individual monetaryfinancial institutions. However, these data arenot available at the ECB and, therefore,weighted rates and standard deviations arecalculated instead.

The following general notation is used for eachof the above categories of loans or deposits:

rc,t

= the interest rate prevailing in country c inmonth t,

bc,t

= business volume in country ccorresponding to r

c,t .

t

tctc B

bw ,

, = is the weight of country c in the total

euro area business volume B: ∑c

tcb ,

The euro area MFI interest rate is computed asthe weighted average of country interest ratesr

c,t taking the country weights w

c,t

∑=c

tr tc,tc, rw (11)

The euro area weighted standard deviationtakes the following form:

∑ −=c

tcttc wrrM ,2

, )( (12)

ADDITIONAL INFORMATIONSince the harmonised MFI interest rates haveonly recently been established and a long time-

series is not yet available, the time horizon istoo short to make a meaningful analysis overtime (less than two and a half years of monthlyobservations are available at present). This isalso the reason for not having smoothed thedata. Work is currently underway to estimatehistorical data for MFI interest rates so as to beable to show the evolution of financialintegration on the basis of these indicators overa longer period.

Cross-border loans and securities holdings(indicators 17-20)

COMPILATIONThese indicators display the geographicalcounterparty diversification of loans grantedby euro area MFIs (excluding central banks)to non-MFI counterparties and other MFIs,respectively, that is, within the same country(domestic), other euro area countries, non euroarea EU Member States5 and the rest of theworld. Similar indicators are computed forsecurities held by euro area MFIs and issued bynon-MFIs and MFIs, respectively.

ADDITIONAL INFORMATIONThese four indicators are built on the basis ofthe national aggregated MFI balance sheet datareported to the ECB, at a monthly and quarterlyfrequency (in accordance with RegulationECB/2001/13).6

These balance-sheet items are transmitted on anon-consolidated basis. This means that thetransactions with foreign counterpartiesinclude those with foreign-controlled branchesand subsidiaries.

5 Since May 2004, the group has comprised Denmark, Sweden,the United Kingdom and the new EU Members States. BetweenJanuary 2001 and May 2004, Denmark, Sweden and the UnitedKingdom formed this group while, before January 2001,Greece was also included.

6 These data cover the MFI sector excluding the Eurosystem andalso include data on money market funds (MMFs). Thederivation of indicators strictly referring to banking markets isnot yet possible. Consequently, as MMFs typically invest ininter-MFI deposits and short-term securities, three of the fourindicators considered are somewhat affected by the MMFsbalance sheet items. Only for the indicator showing loans tonon-MFIs, MFI and credit institutions data coincide.

=tB

Documents

Indicators of financial integration in the euro area, September 2005 · 2006. 9. 15. · Price-based indicators measure discrepancies in asset prices based on their geographic origin