Credit Lyonnais Securities Asia MYTH OF ASIAN GROWTH II · 2018-02-21 · Credit Lyonnais Securities Asia Asian Economic Research ASIA UNCOVERED /May 1997 1 ... But while the economist

THE

REA

L M

YTH

OF

ASI

AN

GR

OW

TH I

IMay 1997

Credit Lyonnais Securities Asia

Asian Economic Research

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.

ASIA UNCOVERED /May 1997 1

ASIA UNCOVEREDSECTION I: The emperor has no clothes 2

SECTION II: The informal economy: drag or driver? 4

SECTION III: Macroeconomic measurement: facts and fantasy 8

SECTION IV: Asian data dangers: inconsistencies and anomalies 10

SECTION V: Minda-Wow! The new front door 12

SECTION VI: Interpreting Chinese data: economist or detective? 15

SECTION VII: Optimistic economics revisited 16

SECTION VIII: Asia uncovered: the fuzzy lesson 18

SECTION IX: The message reinteratred 20

Jim Walker, 12 May 1997

CONTENTS

“Into every tidy scheme for arranging the pattern of hu-man life, it is necessary to inject a certain dose of anar-chism.” - Bertrand Russell, Sceptical Essays .

“Patches and shades of gray are the two key ideas offuzzy logic.” - Bart Kosko, Fuzzy Thinking ,

Flamingo 1994, (p.161).

GENERAL SANTOS: The taxi says it all about the Philippines’ so-called “backwater”

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.

ASIA UNCOVERED /May 19972

FIGURE 1: THE FORMAL AND INFORMAL ECONOMIES IN KOREA AND TAIWAN, 1996

18000

16000

14000

12000

10000

8000

6000

4000

2000

0

GNP

per c

apita

(US$

)

KOREA TAIWAN

Upper estimate

Lower estimate

Formal Underground

Sources: Official country data and various underground economy studies

pend disbelief at almost every turn, possi-bly because no alternatives are on offer.The result, rightly, is a rumbling discon-tent with the economics profession for itsfailure to deliver what it so foolishly prom-ises.

In this follow-up to last year’s specialreport on The Real Myth of Asian Growth(The Shapes of Things to Come: Geom-etry of growth, May 1996) we wish toexpose the weaknesses of economic dataand economic analysis in Asia. And weespecially wish to do so at this point ofrelative pessimism among many invest-ment management houses about Asia’s fu-ture. Why now, when the sustainability ofAsia’s growth ‘miracle’ is being ques-tioned? It is precisely because the data arenot clear, on many levels, about economicconditions in the region. For one thing, wecan state quite bluntly that the economicdata available to us understate Asia’s trueincome and wealth as it stands today. In sodoing, the data give rise to doubts aboutthe region’s ability to promote and sustaingrowth internally. Those doubts are mis-placed.

It is our view - one that we outlined inthe Geometry of growth and have sum-marised as Section Seven in this report --that the Asian region is merely at the dawnof its growth upswing, not, as the bearswould have it, approaching dusk. But ourvision of Asia’s future is of a relatively richregion reaching into itsown hinterland in orderto extend and exploit theAsian marketplace.

The internal growthengine, given appropri-ate liberalisation and de-regulation policies, willreplace the export-ledgrowth policies of thelast few decades. Thehuge challenge for Asiangovernments is in rec-ognising the need for apolicy re-orientationaround this goal (thiswill be the subject of thelast part of our trilogyon The Real Myth of

TAIWAN PRODUCES national incomegrowth forecasts to two decimalplaces. The Directorate General ofBudget, Accounting and Statistics cur-rently expects GDP growth of 6.20%in 1997 (the Chung-Hwa Institute, onthe other hand, is looking for 6.26%).Newspaper articles are written andanalysts comments are made on revi-sions of a few one-hundredths of a per-centage point to these forecasts. Theexercise would be laughable were itnot so sad.

Taiwan also has one of the most activeinformal credit markets in Asia - aca-demic articles have been written on its sizeand importance to the economy (all themore significant in view of the underde-veloped banking system). Excellence, amonthly magazine, in an article in 1993estimated that the underground economyaccounted for 40% of overall economicactivities. According to official statistics,the underground economy formed 30% ofGNP. The following quotation is takenfrom an article in Business Taiwan on 20January 1992:

“In 1982, a delegate at a provincialmeeting said that Taiwan has every un-derground activity except an under-ground railway.”

Yet still the authorities insist on pin-point growth forecasts - miraculously,they are rarely out by much. There again,economic accounts, like company ac-counts, can be massaged in a variety ofways (do not forget, national income ac-counts are unaudited). Spurious accuracyis one of the most pervasive and mislead-ing aspects of economic analysis, espe-cially in developing countries. In that re-spect, the economics profession does theinvestment community a huge disserviceday in and day out, not just in Asia but inthe developed markets too. Regardless,minute dissection of even the least robustdata forms the basis of asset allocation de-cisions the world over.

But while the economist as emperor isa trifle exposed, his and her audience arewell and truly party to the deception. Therecipients of economists’ forecasts andanalysis - it would appear - willingly sus-

Asian Growth). The lesser challenge is forthem to recognise the fact that Asia is al-ready rich and diverse enough to supportsuch a strategy.

Asia uncoveredOUR CONTENTION throughout this re-port is that our knowledge of Asian econo-mies is characterised by big patches, inother words, uncertainty prevails.

This is true of income and wealth -which are understated systematically - aswell as growth which, in some instances,is manufactured by economic managerswhile in others, it is inadvertently over-or understated.

What is for sure is that the two-deci-mal place accuracy of Taiwan’s plannersis a bad joke and should be treated as such.Reality, no matter how patchy, is more ap-propriate always than spurious accuracy.Growth ranges and growth trends - thepattern of growth in other words - isabout the best we can hope to forecast withour limited knowledge of Asian economicreality.

To graphically demonstrate the extentof our uncertainty, Figures 1 and 2 show,for the various countries where reportedstudies have been undertaken, the rangesassociated with the underground or infor-mal economy. Invariably this unreportedaspect of reality adds to the measured-GDP numbers.

This is not a trivial point. In all econo-mies - emerged and emerging - there ex-ists an underground economy. Most esti-mates, even in the most developed coun-tries, put it at around 15% of recordedGDP. Often, estimates have greatly ex-ceeded that number.

The estimates of Asian underground ac-tivity range from 8.9% of GDP in Korea(even superficial precision sometimes

“What really counts with fuzzy systems, the real value added, is the tie betweenwords and sets and between knowledge and patches. Certain knowledge, small patch.Uncertain knowledge, big patch. No math model. That’s the advance.”

- Bart Kosko, Fuzzy Thinking , Flamingo 1994, (p.177).

THE EMPEROR HASNO CLOTHES

SECTION I

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


FIGURE 2: THE FORMAL AND INFORMALECONOMIES IN DEVELOPING ASIA, 1996

GNP

per c

apita

(US$

)

INDIA

5000

4000

3000

2000

1000

500

0PAKISTAN PHIIPPINES THAILAND

Formal Underground

Upper estimate

Lower estimate

Sources: Official country data and various underground economy studies

creeps into estimates ofthe informal sector) toas high as 70% in Thai-land. Such restatementof GNP, on conven-tional GNP per capitaterms, would shift Ko-rea into the high-incomecountry bracket andThailand into the upper-middle income bracketas defined by the WorldBank. But there is amore simple message.Throughout the regioneconomies are richerthan they appear at firstsight. This comes as nosurprise to those of uswho have travelled be-yond Asia’s capital cit-ies and tourist resorts(see Section V, a casestudy of such an area).

The message from these charts and thestudies underlying them is that officialGDP is a flawed measure of the Asianeconomies. This is true for three specificreasons which we will examine in moredetail in Sections II and III but, to summa-rise, are: first, the informal economy is,by and large, more extensive in develop-ing countries than in the mature economies.As a starting point we can add on average30-40% to GDP estimates at the outset.

Secondly, the illegal economy - drugs,gambling, prostitution - also figures largein some Asian countries (although it isdwarfed in most instances by the remain-der of the informal economy). Neverthe-less, illegal activities can add up to 15-20%to GNP (as a Chulalongkorn study in Thai-land recently postulated for that economy).

Thirdly, there is incomplete statisticalcapture. The problems of data deficiencyand patchy statistical recording are ad-dressed more fully in Section III. Sufficeto say here that this further blurs the truepicture of growth and income in develop-ing countries.

What are the implications of our analy-sis of the underground and non-capturedeconomies in Asia? First, the stark and easyconclusion is that wealth and income areunderstated in the region.

From the sketchy national studies thatwe have, coupled with an analysis of thebody of literature on informal sector ac-tivities, we would argue that this under-statement is greatest in China, Indonesia,the Philippines and Thailand.

A reasonable guesstimate would add atleast 40% onto existing GNP aggregates

for these countries, possibly considerablymore. We also believe that India and Paki-stan should be added to this list althoughmore formal estimates - one an IMF study- conclude that the informal economy isonly in the region of 20% of GNP in bothcountries. In Section II we will present evi-dence that suggests to us that 20% is fartoo low.

In the more mature Asian economiesthe informal sector is likely to be smaller,although studies in Taiwan suggest that thisis only marginally the case. Official Ko-rea Tax Institute estimates put the under-ground economy at less than 10% in thatcountry although other agencies arrive atmuch higher figures.

There are no studies on Hong Kong andSingapore but the preponderance of inter-national trade activity in both entrepotssuggest that domestic informal activitiesare likely to run at significantly lower lev-els, as a percentage of GNP, than elsewherein Asia.

Likewise, we have found no publishedestimates of the informal economy in Ma-laysia but it probably lies somewhere be-tween the mature and more developingAsian countries.

Secondly, our analysis calls into ques-tion the reported growth rates in Asia. Insome instances, the Philippines for exam-ple, they are clearly too low. However, forother countries, economic growth overthe last decade may well have been inflated.This is most especially true in countrieswhere significant changes to the tax re-gime on domestic sales and foreign tradehave taken place and where more sophis-

ticated data collection techniques havebeen introduced.

A process can be envisaged where hith-erto informal sector activity is steadilycaptured in the formal economy. This in-creases the rate of reported GDP growthin any one year but may be at the expenseof the size of the informal economy. Theresult is fast growth in GDP but slower ex-pansion of the total national income - for-mal plus informal - of the economy. This,in turn, can have implications for corpo-rate earnings potential and macroeconomicpolicy.

Thirdly, and following directly from thelast point, policies based on flawedmacroeconomic growth estimates arelikely to be ineffective or positively harm-ful. The clear lesson here is that domesticpolicy frameworks must be designed withtwo things in mind: flexibility and simplenominal targets. Asian governments haveto take the fuzzy thinking message to heart- macroeconomic data are characterisedby big patches i.e., there is a great deal ofuncertainty surrounding them. In order tobest handle this uncertainty a rule-basedpolicy framework which applies flexiblyand which operates on a variety of signalsis necessary. Fixed price policies - suchas the quasi-fixed exchange rate system inThailand - inevitably lead to fatal rigidities,no matter how impressive the early suc-cesses.

Finally, we believe that our analysis sup-ports the conclusion made in the Geom-etry of growth last year, that Asian growthcan be sustained in the future by the adop-tion of appropriate domestically-orientedpolicies.

These include the liberalisation and de-regulation of internal markets and thesteady absorption of the informal sectorinto the reported economy (which followsdirectly from deregulation). In this waymore appropriate policy signals will be de-termined and Asia’s reliance on OECD ex-port markets will be reduced.

That we know less about Asia’s wealthlevels and its achieved growth rates thanhitherto has been claimed is not an admis-sion of weakness - it is a source of inspi-ration. So much remains to be discovered- uncovered - in Asia’s economic emer-gence.

So much of the region’s potential re-mains to be unleashed. But, at the sametime, we must admit reality into our think-ing; that means less precision, less empha-sis on macroeconomic aggregates, moreunderstanding of data deficiencies andmore patience in realising potential re-turns.

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


sorb the great mass of the urban poor. InPeru poverty existed in the informal sec-tor but so too did riches.

The informal economy studies con-tained in The Silent Revolution concen-trate on Asia and the Near East. A similarpattern was found as in the de Soto book.Avoidance of taxation and, perhaps evenmore important, the avoidance of red tapeand interminably long delays in setting upformal businesses, were the key reasonsfor the burgeoning informal sector.

As we shall see below, the informal sec-tor is an important part of the Asian eco-nomic scene. How important can only beguessed at.

It is likely, however, that growth in thissegment of the economy is faster than inthe domestic formal sector which, in mosteconomies, is still characterised by gov-ernment protection, over-regulation andvested interest oligopolies and monopo-lies.

The scope and consequences ofthe underground economy“After being ignored for many years, theunderground economy has finally workedits way to the center stage of Americanpublic and official attention. The discov-ery of the underground (or subterranean)economy in the United States is attestedby recent editorials and articles in lead-ing newspapers and magazines, hearingsby four congressional committees, re-ports by official agencies (Internal Rev-enue Service and General AccountingOffice), coverage in popular televisionprograms, and professional articles. Inthis discovery the United States has fol-lowed other major countries in which theunderground economy has been the sub-ject of official and unofficial attentionand studies for some time.”

- Vito Tanzi, The Underground Economyin the United States and Abroad , LexingtonBooks, 1982, (p.69).

THE UNDERGROUND economy. Thename conjures up visions of illicit bar-gains and shady deals in the backstreetsand cafes of most major cities. Nowthink of the jeepney driver in Manila,the motorbike taxi in Bangkok, themyriad dealers in the unorganisedmoney markets in Taiwan, the streetvendor in Jakarta, the Mom and Popstores in every city, town and villagein the Philippines, the odd-job man inSingapore, the bicycle repair workshopin Kuala Lumpur - retailers, moneylenders, farmers, entrepreneurs, taxidrivers, couriers and a whole host offaces and activities that we are every-day familiar with. That is the bulk ofAsia’s underground economy.

But there is more - small and mediumsized enterprises, co-operatives, offshootsof state-owned enterprises in China, creditunions. They too form part of the infor-mal economy. Then there is the sleazierside of life - the drug traffickers, prosti-tutes and gambling racketeers. They toooperate underground - by necessity ratherthan choice.

In 1987 Hernando de Soto published astudy on the informal sector in Peru calledThe Other Path. This study turned on itshead the prevailing wisdom - as espousedby the International Labour Organisation- that informal markets were a dysfunc-tional part of the regular economy. Instead,he found that Peru’s underground economywas characterised by strong growth andenterprise.

Often the markets in specific productsor services were highly organised in them-selves although still operating outside thepurview of the authorities (after paying theappropriate bribes, of course).

Certainly pay is low and conditions arepoor - working hours tend to be long -but this part of the real economy existslargely because of the failure of the for-mal sector and government policies to ab-

IT WOULD seem that just as quickly asthe underground economy emerged as atopic in the early 1980s, it submergedagain. Can we take this to mean that the‘problem’ has been solved? The answer isa straightforward no. The fact is that theunderground economy still exists in thedeveloped countries as well as in develop-ing nations. In 1976 three different ap-proaches to measuring the subterraneansector (see the box in this section for ataxonomy of terms) in the US found thatbetween 4-22% of official GNP was un-derground. In studies in Europe the infor-mal sector was found to represent up to30% of GNP.

However, the equally plain fact is thatrecognition of the underground economy’sexistence complicates matters. It compli-cates GDP measurements and what thesesignal to the investment and policy man-agement communities. Moreover, it com-plicates the policy response if the infor-mal sector happens to be growing at a fasterpace relative to the formal economy. In-deed, some researchers have even gone sofar as to postulate that the existence of theunderground economy is the cause of anumber of theoretical and empirical mis-matches i.e., that correct theories havebeen refuted as a result of data deficien-cies (and vice versa).

One of the main reasons for the under-ground economy disappearing from viewagain - except in academic circles - isthat no-one really wants to know. In the de-veloped countries, while this is a pity, it isprobably not too important for investmentdecisions. Indeed, as limitations on gov-ernment’s role in the economy have in-creased over the last two decades - partlyas a result of the Thatcher and Reagan agen-das - the tax burden has grown much moreslowly and, in some cases, even dimin-ished. This ongoing process reduces theneed to operate informally.

The result, in all probability, is that theunderground economy has shrunk in theOECD generally in the 1990s (althoughthe European Union’s Social Chapter mayreverse that trend in western Europe).

Measuring theunderground economyBUT HOW can researchers begin to meas-ure something which, by definition, is notrecorded? A number of methods are avail-able for estimating the informal economy.Direct survey-based techniques have beenused in the US by various statistical bu-reaux, notably the Internal Revenue Serv-ice. These rely on alternative (control)bodies of data e.g., tax returns, being avail-

SECTION II

“Recent studies have shown that microenterprises and the informal sector can be amongthe most dynamic and healthy sectors in developing economies. As such, they fill an essen-tial role in providing jobs for large numbers of urban workers and adapting both to gen-eral patterns of economic growth and to negative shocks. Not only do the informal sector’sentrepreneurs survive in the face of government subsidized large-scale industrialization;their flexibility and adaptability have allowed them to provide essential goods and serv-ices on which large-scale, modern manufacturing enterprises depend, such as clothing,weaving, woodwork, repair services and production of small machines.”- A L Chickering and M Salahdine (eds.), The Silent Revolution , ICS Press, 1991, (p.4).

THE INFORMAL ECONOMY:DRAG OR DRIVER?

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


able to compare with the survey results.Audit sampling is also used to capture taxevaders. These are largely inappropriatetechniques in the developing countries.

A second method is census based. It isassumed that small scale enterprises arethe most likely candidates to go under-ground. Census of Employment data whichdetail firm size by employment level areused to define the reaches of the under-ground economy. This is an arbitrary ap-proach which is dependent on the question-naire design (i.e., what level of employ-ment disaggregation it goes down to) andthe sweeping assumption that companieswith less than, for example, ten employ-ees and self-employed businesses are in-variably informal. Moreover, while thismight be a proxy for the size of the infor-mal labour market, it is less useful as aguide to the magnitude of the informaleconomy in income terms.

Developed country researchers havepaid little attention to inadequate statisti-cal collection techniques in their explana-tion of the size of the informal economy.This is unlike the developing countrieswhere poor statistical capture undoubtedlycompounds the holes created by the exist-ence of informal activities. By definition,there is no way of estimating the short-falls of statistical capture - the data aren’tcaptured. (As we shall show in SectionsIV and VI, it is relatively easy to identifywhere there are anomalies, however.)

In short, the underground economy isassumed to exist because of taxation andeconomic actors’ determination to avoidthe tax liabilities which recognition in theformal sector would bring. (The informalsector that most studies have sought tocapture includes illegal activities within

the definition of informal. Neither the il-legal nor the informal sectors pay formaltaxes on their output but the latter encom-passes activities - such as domestic helpand street vendors - which are far from il-legal per se).

Indeed, the main raison d’être of theunderground economy (as stated above) istwofold: a) the avoidance of taxation, andb) the avoidance of the administrative costsof setting up small businesses or employ-ing small numbers of people. It is the im-portance of the second of these reasonsthat makes the most popular method ofunderground economy estimation morequestionable in developing countries. Weshall detail the reason for this below.

For sure, informal sector activities aregenerally small scale. In developing coun-tries that definition can encompass as muchas 70% of the urban workforce. The othermain characteristic of the informaleconomy - an assumption as much as any-thing - is that it is primarily cash-based (itbeing easier to avoid taxation in this way).

While this is a straightforward notionin the US, in less developed countries itbecomes more complex. There is no rea-son to expect that informal activities willtake place in local currency, especially ifone of the motivating forces of under-ground participation is a distrust of thegovernment.

But it is the cash-based assumption thatprovides researchers with the dependentvariable in most empirical analyses of theunderground economy: the ratio of cur-rency-in-circulation (cash) to M2 (the cur-rency ratio). Variations in this ratio overtime are attributed to the development ofthe economy, the attractiveness of alter-native liquid assets, financial deepening

and, most importantly, to the variation intax rates over time. A multiple regressionequation is constructed to isolate these ef-fects. This is the approach developed byVito Tanzi.

In effect, like all other monetary ap-proaches, the Tanzi method is one of ‘re-sidual’ measurement. After identifying anappropriate multiple regression equationthat models the currency ratio, the tax vari-able is set to zero (or held constant overtime) in order to estimate a currency ratiofor the formal economy only. Once theformal economy currency ratio is identi-fied it is an easy step to work out theamount of cash associated with the under-ground economy: it is the residual gener-ated by taking the difference between theforecasts of the currency ratio in the un-restricted and restricted regression equa-tions. This residual is then multiplied bythe velocity of money in the formaleconomy (the somewhat questionable as-sumption of equivalent monetary veloci-ties in the formal and informal economiesis made) to yield an estimate of under-ground GDP.

This is the method used by the KoreaTax Institute (a unit of the Ministry of Fi-nance and Economics) to estimate theKorean underground economy. It is alsothe basis of the approach used by the IMFin its 1995 study of Pakistan. For Korea,the latest Tanzi estimates suggest an infor-mal economy of just 8.9% in 1995. TheIMF study suggests an average undergroundeconomy level of 22.6% over the period1974-91. We are unconvinced by the ap-plication of the Tanzi method in theseeconomies.

Our attempts to model the Philippinesusing this approach were unsuccessful with

A VARIETY of terms are used to de-scribe the non-official segment of theeconomy. Some are general and othersare more specific. The following list isby no means comprehensive:❚❚❚❚❚ Underground - Despite connota-tions with crime, this is one of the mostgeneral terms encompassing all legaland illegal unrecorded economic ac-tivities. It is equivalent to the little used‘subterranean’ term.❚❚❚❚❚ Informal - Is the most neutral de-scription of the unrecorded economyalthough it gives an impression of casu-alness that is not justified. Informalmarkets are often highly organised and

highly efficient in their own right. Lessemotive than other terms, it is the onemost often adopted in academic re-search.❚❚❚❚❚ Parallel - More specific than eitherof the foregoing as it applies only to le-gal activities which are often unavail-able in formal markets for one reasonor another. A good example are the un-official money or credit markets thatare tapped in developing countries forthe financing of agricultural invest-ments or as capital sources for smalland medium sized enterprises.❚❚❚❚❚ Illegal - Usually reserved for activi-ties that are explicitly prohibited by law

- gambling in some countries, drug traf-ficking, the sex trade. Much of the in-formal sector is involved in anothertype of criminal activity i.e., tax eva-sion, but activities which result in this‘crime’ are not in themselves illegal.❚❚❚❚❚ Black - The associations of this termare with wartime, petty crime and un-scrupulous traders. In reality it is syn-onymous with underground.❚❚❚❚❚ Fragmented - Dysfunctional mar-kets that exploit information inadequa-cies. Not really a subset of our study.❚❚❚❚❚ Hidden, unrecorded, unreported andunofficial are all terms used at timesfor the informal sector.

Underground, informal, parallel, black or what?

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


Secondly, official data only measure of-ficial taxes. We have commented in ourInfofaxes in recent weeks on the growingburden of unofficial taxes in the Chineseeconomy. These are forcing some compa-nies out of business and no doubt, manyothers to flee into the ‘hidden’ economy.

Likewise, bribes and corruption are analternative form of taxation. In a study byRuperto Alonzo (included as a chapter inThe Silent Revolution), he found that oneof the “agonizing” procedures official busi-nesses in Metro Manila had to go throughwas water connection. Contractors nor-mally required a ‘facilitation’ fee to pro-vide the service or else a delay of someweeks would be experienced. There are nomeasures of this alternative ‘tax’ which canbe included in the regression model.

Lastly, the Tanzi method requires a

none of the explanatory variables provingsignificant over the period 1971-94.

Why are we sceptical? For three rea-sons: first, and primarily, because of theimportance of administrative costs as a rea-son for staying underground in developingcountries. Changes in tax rates and the ra-tio of tax to GDP are relatively easy tomodel. However, even if tax rates do notrise the informal economy can still be ex-pected to grow in societies where the bu-reaucracy insists on complicated and slowmoving registration processes. Despite taxincentives for registered companies, tor-tuous procedures and bureaucratic corrup-tion keep most small-scale businesses outof the formal sector in developing coun-tries. Under these circumstances the taxvariable in the Tanzi Method is likely to beflawed.

benchmark period where the undergroundeconomy is thought not to exist or to be ata minimum. In developed economies thebase period is usually chosen to predatethe rise of Keynesianism when big gov-ernment and big taxes became the norm.Data for such periods in the developingcountries do not exist.

Asian evidenceGIVEN THE available data in the region,this leaves us with much less in the way offormal modelling procedures for estimat-ing the underground economies in Asia.

As mentioned in Section I, there havebeen a number of specific country studieswhich have attempted to identify and placea value on the underground economy. Mostof these we have been able to access onlythrough media accounts.

The values placed on the various infor-mal sectors range widely - from 10-70%of formal GNP. Figure 3 contains a sum-mary of these estimates for six countriesin the region.

The currency ratio itself is worth ex-amining. The trend of the ratio of cash tobroad money should yield some insightsabout the direction of change of the un-derground economy. Moreover, the size ofthe ratio might also yield some relativeinsights about the position of the variouscountries in the region.

In Figures 4-6 (next page) we havegrouped countries into NIEs (Korea, Sin-gapore and Taiwan), developing-ASEAN(Indonesia, Malaysia, the Philippines andThailand) and the sub-continent (India andPakistan). To act as a benchmark in eachchart we have included the currency ratioin Japan (where we would expect cash re-quirements to be relatively small).

Interestingly, the Japanese currency ra-tio has been on a gradually rising trend inthe last few years. Given the difficultiesin that economy with the constraints on thefinancial sector in particular, it is not sur-prising to see cash activities rising. Infor-mal sector activities are also likely to havebecome more important in the Japaneseeconomy in this decade.

The results for the NIEs are somewhatsurprising. Both Taiwan and Korea regis-ter currency ratios lower than Japan (notso surprising is the downtrend exhibitedin all three countries in this sample). Yetboth have underdeveloped banking systems(see our 1st Quarter 1997 Eye on AsianEconomies [Banks in North Asia: Un-derdevelopment exposed]) and both arenoted for active curb markets in unoffi-cial finance. Perhaps the explanation - forTaiwan at any rate - lies in the fact that

Range(%) Comments

Korea 9-42

India 20

Pakistan 22.6

Philippines 25-50

Taiwan 14.4-40

Thailand 25-70

A number of official studies have been published with atten-tion now focussing on the Korea Tax Institute (which usesthe Tanzi method). The same Institute which last publisheda 9% figure for the underground economy (for 1995) wasinvolved in another 1995 study which estimated informal ac-tivities at between 37-42%. 9% would appear to be at thelow end of the scale.

A commonly quoted figure although no formal studies ap-pear available. Probably a huge underestimate.

IMF working paper from 1995 uses a variant of the Tanzimethod to estimate the underground economy. Reliance ontax rates alone, as we have suggested above, is likely tobias estimates of the underground economy downwards indeveloping countries. Too low.

The National Statistical Office admits to a quarter of theeconomy as being informal. Most other estimates rangebetween 40-50%.

Official estimates range between 30-40% of GNP in the in-formal sector. Unlike many other countries the concentra-tion of activities is in the money markets - forex dealing,stock trading and curb market lending.

The latter figure (from a Thai Farmers Research report) ap-pears far too large but a recent Chulalongkorn study put thevalue of illegal operations - drugs, prostitution, gambling,arms trading, oil smuggling and people trafficking - at 17%of GNP. More normal types of informal activities are boundto boost this figure towards the 40-50% mark.

FIGURE 3: ESTIMATE RANGES, THE INFORMAL SECTOR BY COUNTRY

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


cheque discountingforms the main transac-tions payment method inthe unorganised moneymarket.

In Singapore the cur-rency ratio has been de-clining over time as wewould expect with a so-ciety where incomes arerising and financialdeepening is proceed-ing apace. That said,Singaporeans’ prefer-ence for cash is of thesame order as Malaysiaand Indonesia (Figure5), something of a sur-prise.

The developing-ASEAN trend generallyalso suggests that thecash economy is be-coming less importantand, therefore, the un-derground economy isdiminishing in size. Thismay well be true - as-suming, of course, thatthe velocity of circula-tion of cash in the infor-mal sector is the sameas within the measuredeconomy and that USdollars or gold are notplaying an increasinglyimportant role.

Cash activities in thePhilippines remain dis-tinctly higher than in theother developing-ASEAN countries - al-though well off thepeaks of the troubledMarcos period and theearly years of theAquino government.Thailand has experi-enced an upturn in itscurrency ratio in thelast year - a trend thatmay well continue asthe economy hits evenharder times.

Figure 6 showsclearly the importanceof cash in India and Pa-kistan. Moreover, theratios show no system-atic tendency to fall.Currency ratios of40%+ make it difficult to believe that theunderground economy is as low as 20%

is understated more than anywhere else inAsia.

In and of themselves, these charts tellus little about the informal economies inthe region. Two broad conclusions can bedrawn, however. For most countries thecash-economy is diminishing in impor-tance - this is probably partly associatedwith a move of the informal sector intothe measured economy. As we shall showin Section IV, this has implications for thereported growth rates in Asia in recentyears.

Secondly, there is a broad hierarchy ofcurrency ratio countries. The more devel-oped tend to use less cash (although, aswith Taiwan, it is possible that the under-ground economy is operating on slightlymore sophisticated methods of payment)while the least developed - India and Paki-stan - still view cash as king (in all prob-ability actors in these countries have littlealternative).

Unfortunately, the Tanzi method andmeasures of the currency ratio can do lit-tle to help us capture parallel currencymarkets in developing countries. EdgarFeige has pointed out that currency-basedestimates of the size of the undergroundeconomy in the US may be biased upwardsby the fact that a significant proportion ofissued US dollar bills are held as stores ofvalue and transactions media outside theUS.

The corollary of this is that currency-based estimates of the informal sector indeveloping countries (where economic ac-tors may have less faith in government’sand the authorities’ determination to pro-tect the value of the currency) may be bi-ased downwards.

Moreover, US dollars are not the onlyalternative transactions medium. In Indiaand China gold is thought to play an im-portant role in the informal economy.

Where does that leave us? In all hon-esty, still groping in the dark. We cansafely assert that the undergroundeconomy exists. In India, Pakistan, thePhilippines and Taiwan it is large. We sus-pect that this is also the case in China, In-donesia and Thailand. But is it a drag ordriver of growth? Again, the evidence ispatchy but recent studies have found thatthe informal sector is characterised bymore vibrancy than is normally associatedwith the formal economy.

In China, where the state-owned com-panies can best be described as dinosaurs,the small and medium-sized operationswhich exist nowhere in the official data areundoubtedly the dynamic segment of theeconomy.

of GNP in either country. If anything, thewealth potential in both of these countries

FIGURE 4: CURRENCY RATIOS IN THE ASIAN NIEs, 1980-96

0.35

0.30

0.25

0.20

0.15

0.10

0.05

0

%

80

Taiwan

Sources: China Economic Information Centre (CEIC)

81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96

Japan Korea Singapore

FIGURE 5: CURRENCY RATIOS IN DEVELOPING-ASEAN, 1984-96

0.35

0.30

0.25

0.20

0.15

0.10

0.05

0

%

Philippines

84 85 86 87 88 89 90 91 92 93 94 95 96

Japan IndonesiaThailand Malaysia

Sources: CEIC

FIGURE 6: CURRENCY RATIOS ON THE SUB-CONTINENT, 1988-96

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

%

88

Japan

89 90 91 92 93 94 95 96

India Pakistan

Sources: CEIC

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


FIGURE 7: UNCERTAINTY BAND OF NATIONAL INCOME STATISTICS, 1946-1961

450

400

350

300

250

200

150

Log

scal

e do

llars

, bill

ion

Officially reported quarterly national income, seasonally adjusted at annual rate.

Alternative possible path of national income within assumed error limits.

Business cycle turning points as determined by National Bureau of Economic Research

Business cycle turning points according to path ......*

1947 ’53’49 ’51 ’55 ’57 ’59 ’61

**

**

*

*

**

**

*

Source: Morgenstern, op. cit., p.31

ning, it was just the Austrians. Complexphenomena are not easily measured, theyare not predictable and they are certainlynot linear. The Austrians attacked the powerof the positivists on the theoretical level -although the arguments are still valid andcorrect, the assault failed.

Oskar Morgenstern, an Austrian econo-mist but primarily a statistician and bestknown as the co-developer of GameTheory with John von Neumann, under-mined the whole notion ofmacroeconomic analysis and forecastingin his book, On the Accuracy of EconomicObservations (1950, 2nd revised edition1963). His approach was to point out theshortcomings of national income data -largely resulting from sampling errors,infrequent survey updates and poor statis-tical capture. He went on to criticisemacroeconomic analysis on the basis thatit was using fundamentally flawed data.

Morgenstern’s analysis was confined tothe developed countries. Figure 7 is a per-fect example of the problems he identi-fied. This chart shows the path of officiallyreported national income in the US be-tween 1946-61. The unshaded area repre-sents the officially admitted error bandswhich can be calculated around the data.The dotted line is just an assumed alterna-

tive path that can bedrawn within the errorlimits. The worrying as-pect is that business cy-cle turning points caneasily be interpreted asa) either not a turningpoint at all, or b) a turn-ing point in the oppositedirection from that of-ficially sanctioned.

In short, even in thebest equipped country inthe world for data col-lection, no clear inter-pretations can be placedon even historic series.

Observers may wellargue that statisticalcollection techniqueshave improved over thelast forty years. In a waythis is true but cutbacksin government spendinghave ensured that statis-tical collection agen-cies are understaffedand under-resourced.The decline in thenumber of tax and ac-

counting audits carried out in recent yearsby revenue agencies (an incentive to move

and the neo-classical mainstream are nowunder attack from all angles. In the begin-

WE ALL know the jokes about econo-mists: all the economists in the worldplaced end-to-end would never reach aconclusion; what do you get when youput two economists in the same room?Three conclusions, and so on. But thereis a serious side to the inability ofeconomists to agree on theory, on theinterpretation of each new set of dataand on the health of one particulareconomy or another. The simple fact isthat economists, by the very nature ofthe subject matter, deal with shades ofgrey, not black and white.

The positivist approach towards eco-nomic science - what can be termed theneo-classical mainstream - has been in theascendancy for over 100 years. Its plungeinto the quagmire of macroeconomics wasaccelerated by the mathematician, JohnMaynard Keynes.

It should have come as no surprise thatit was a mathematician by training thatwould attempt to portray the world of com-plex phenomena as a series of linear rela-tionships and macroeconomic aggregates.It is just a surprise that so many studentsof the subject have been fooled into be-lieving that this approach is anything otherthan a dead end.

The mathematical economist approach

“The positivist’s claim that it is always the statistics that test the theory, rather thanthe other way round, suggests that the positivist always has more confidence in thedata than he has in his own theory. There is the temptation to agree that his lack ofconfidence in positivist theory is justified. But it could be argued instead that posi-tivist theory tested with statistical aggregates is a case of the blind leading theblind.”

- Roger Garrison, Introduction to Oskar Morgenstern, National Income Statistics ,Cato Press, 1979

“Empirical macroeconomics is now a well-established profession. Respected aca-demic economists are consulted by government and business for their views onmacroeconomic policy, views which usually depend on analysis of official data. Jour-nals are filled, and careers are made, by sophisticated manipulation of these samedata. Theories of increasing complexity are tested using the data, and paradoxes ofthe data are explained by more refined theory. To suggest that the data are badlydeteriorating does not provide comfort to academic macroeconomists.” (emphasisadded)

- Peter Reuter, The Irregular Economy and the Quality of Macroeconomic Statistics,- Ch 8 in Tanzi, op cit., (p125).

“The fuzzy view was just common sense. It was daring and novel at the time becauseyou first had to get your university degrees in the old black-and-white school andthen doubt that school and rediscover what any layman could have told you aboutcommon sense - it’s vague and fuzzy and hard to pin down in words or numbers.”

- Bart Kosko, op cit. (p 161).

SECTION III

MACROECONOMICMEASUREMENT: FACTS

AND FANTASY

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


FIGURE 8: UNCERTAINTY PATCHES IN A NON-LINEAR WORLD

Y

XSource: Kosko, op. cit., p.167

underground?) is a sim-ple testament to thisfact.

Indeed, the radicalrevisions to US eco-nomic data that continueto be made on amonthly, quarterly andannual basis clearlyshow that Morgen-stern’s criticisms of thestatistics are as valid to-day as they were whenhe first wrote aboutthem in 1950. (The cur-rent debate surroundingthe validity of the USCPI measure is a perfect example.) In fact,all we have to show for progress over thelast few decades is more computing powerto process inherently flawed data.

Morgenstern’s critique is given all themore power by Reuter ’s observations.Writing from the perspective of a re-searcher into the informal economy, hesuggested that data quality in the USthrough the 1970s had been deteriorating,not improving. Later studies support thesame conclusion. Pozo’s collection of es-says (S. Pozo (ed),. Exploring the Under-ground Economy, Upjohn Institute, 1996)show that the debate is alive and well andjust as concerning today.

In Asia, we can question the data qual-ity much more comprehensively (as shownin the contradictions and anomalies we citein the next section) and safely postulatethat the underground economy plays amuch bigger role in everyday life. The con-clusion? Asian national income statis-tics and the growth forecasts producedfrom them are nothing short of guess-timates with relatively large errorbands.

Finally, the attack on the application ofnatural science techniques to the socialscience area is coming from within. Thedevelopment of Chaos Theory and, now,Fuzzy Logic are attacking the heart of thebivalent, linear world which modern phi-losophy and mathematics lives in. If Koskois correct, that not just common sense butthe world, including mechanics, physics,maths, is “vague and fuzzy”, then new mod-els of economic relationships and newframeworks for economic policymakingare urgently required.

Figure 8 is taken from Kosko’s bookand bears a striking resemblance to theMorgenstern chart published 30 years ear-lier. Certain knowledge, small patch. Un-certain knowledge, big patch. Neverstraight lines and never black and white.

The two economists in the same room thatreach three conclusions? The real joke isthat three is all they could come up with.

The fuzziness of Asian growthAsia has been the fastest growing regionin the world economy for the last ten years- fact or fantasy? Well, the macroeconomicstatistics support that contention 100%.We also happen to believe it is true. Butwe are also convinced that the picture is alot less clear than the data suggest.

Let us show by example how develop-ments in the underground economy mighthave distorted the data. Let us assume thatan economy exists with an informal sec-tor that is 50% of reported GNP. Figure 9shows how the growth rate in the formaleconomy compares with that in the total(formal plus underground) over a periodof five years with the following assump-tions made about the behaviour of the un-derground economy.

Scenario 1: Both the underground andformal economies grow at a steady 8%over years 1-5. 50% of the growth in theunderground economy is transferred to theformal economy in Year 1. In each subse-quent year the amount of growth trans-ferred rises by five percentage points. Thisscenario can be likenedto a phase where statis-tical collection tech-niques and official en-forcement are improv-ing slowly.

Scenario 2: Botheconomies grow at 8%but the undergroundeconomy is graduallyshrinking as a result ofderegulation and in-creased confidence inthe government. Startingat 100% of the growthin Year 1, the percent-

age of growth in the informal economytransferred each year rises by five percent-age points (at the end of the process theunderground economy forms only 22% ofGNP).

All of a sudden, with no organic accel-eration in the formal economy at all, meas-ured or recorded economic growth ratesare seen to have jumped sharply. From sim-ply being one of the fast growing econo-mies in the world our hypothetical statecould become the fastest growing countryon earth - in macroeconomic fantasyterms.

Numerous examples can be imaginedwhich simulate transfers from, say, a moreexuberant informal economy to the for-mal sector or one-off deregulation events(such as the introduction of VAT to replacebusiness taxes) which induce a significantshift of activity in one year from the un-derground to formal sectors.

Probably just as well. Fast growth inAsia appears to be an end in itself for gov-ernments. By concentrating on growthrates, especially in times of changing taxstructures and economic deregulation, gov-ernments are in danger of missing an un-derlying malaise in the economy. More at-tention to the microvariables involved inderegulation and liberalisation in Asianeconomies is urgently required.

All this said then, at the level of the to-tal economy conditions can be deteriorat-ing, staying stable or improving - no-onecan really be sure. However, among otherthings, and assuming that our thesis hasplayed out to some extent in the region,this ‘imaginary’ growth in the recordedeconomy can help explain just why GDPgrowth in Asia has, from time to time, solittle correspondence with corporate earn-ings growth. In short, an acceleration inthe official GDP statistics does not nec-essarily mean that the whole economyis growing faster than in the previousyear.

FIGURE 9: FORMAL AND TOTAL ECONOMYGROWTH RATES UNDER SCENARIOS I AND II

* Where FE=Formal economy and TE=Total economy.

FE* TE* FE TEgrowth growth growth growth

(%) (%) (%) (%)

Year 0 8 8 8 8Year 1 10 8 12 8Year 2 10.1 8 11.8 8Year 3 10.1 8 11.5 8Year 4 10.2 8 11.3 8Year 5 10.2 8 11 8

SCENARIO I SCENARIO II

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


FIGURE 11: GNP PER CAPITA AND PPP PER CAPITA, 1994

25000

20000

15000

10000

5000

0

US$

KOREA SINGAPORE

GNP per capita PPP estimate of GNP per capita

AT THIS STAGE, readers might legiti-mately pose the question: isn’t this allold hat? We all know, deep down, thatthe data are distorted and that Asia, inall likelihood, is richer than the fig-ures suggest. Indeed, no less an author-ity than the World Bank has been pro-ducing ‘purchasingpower parity’ esti-mates of Asian GNPper capita for the lastfew years that showthe region to be sig-nificantly richer thancurrent dollar esti-mates. Or so it mightseem to the layman.

In the 1996 WorldDevelopment Report thePPP estimates of GNPper capita have been el-evated from Table 30 toTable 1 (Basic indica-tors). Surely this is an in-dication of how impor-tant and relevant the PPPexercise is for compari-son purposes? If so, thenthe World Bank is per-petrating a gross decep-tion on its readers.While the PPP exercisedoes indicate that Asianper capita GNPs mightbe an understatement oftrue purchasing power(see Figures 10 and 11),its underlying estimates,especially for develop-ing countries, shouldcarry a health warning inthree-foot high letters.

On average, the de-veloping Asian econo-mies (i.e., excludingKorea and Singapore)were reported as havingGNP purchasing power

almost 200% greater in 1994 than currentdollar GNP data would suggest. China’sPPP-based GNP estimate is 370% higherthan the corresponding per capita dollarfigure. But what do these numbers reallymean?

The idea is to compare a fixed basket

of products across countries and estimatethe domestic cost of these products rela-tive to each countries’ domestic incomelevels. As the benchmark, a fixed basketof goods priced in US dollars in the US istaken. In this way, the purchasing power ofindividual countries can be arrived at with-out reference to an arbitrary internationalbenchmark (the prevailing US dollar ex-change rate). Of course, this exercise tellsus nothing about the international purchas-ing power of individual countries as thePPP-estimate units are notional ‘interna-tional dollars’.

In reality, they tell us lots less than that.As with so many academic exercises thedevil is in the footnotes. The whole con-cept is based on the Bank’s InternationalComparison Programme (ICP) - a studyaimed at generating more appropriatecross-country comparisons of income. Inthe 1996 World Development Report thelatest available data, for the OECD coun-tries, was based on 1993 comparisons. Asthe footnotes to the Basic Indicators tabletell us, the results for India, Korea, Paki-stan, the Philippines and Thailand wereextrapolated from the 1985 ICP estimates.For China, which has never participated inthe programme, the estimate is an internalone from the World Bank. In Indonesia,estimates are extrapolated from 1980while in Malaysia they are from 1975. InSingapore, the PPP-estimates are based ona regression equation.

In other words, the methods of estima-tion and the base data used throughout theregion are far from consistent. The resultsof the exercise might not quite be com-parisons of apples and oranges but they arenot far off. So, if international agenciesare guilty of producing grossly mislead-ing statistics, what hope is there for thenational equivalents?

Let your watchword be: take careTHE FACT is that contradictions, incon-sistencies and anomalies abound in Asianmacroeconomic data. We try to point theseout as and when they arise in our dailyAsianomics Infofax but often catching theanomalies is easier than producing an ex-planation for them. In general, investors’expectations about the accuracy and valid-ity of Asian data are just far too high.

Take Malaysia. Reported economicgrowth has been running above 7.5% since1988. The unemployment rate has fallensteadily from 5% in 1990 to just 2.6% in1996. Officials will tell you that even thatnumber is an overstatement of true unem-ployment because it includes those “notactively seeking work”.

FIGURE 10: GNP PER CAPITA AND PPP PER CAPITA, 1994

US$

CHINA

9000

8000

7000

6000

5000

4000

3000

2000

1000

0

INDIA MALAYSIA PHILIPPINES

GNP per capita PPP estimate of GNP per capita

INDONESIA PAKISTAN THAILAND

Sources: World Development Report 1996

“Economists are unique among social scientists in that they are trained only toanalyze, not to collect, data. While psychologists are taught experimental techniques,sociologists learn of the vagaries of interviewing, and anthropologists devote muchof their training to field work, economists are provided only with the tools for dataanalysis. one consequence is a lack of scepticism about the quality of data.”

- Reuter in Tanzi, op cit., 9p.137)

ASIAN DATA DANGERS:INCONSISTENCIESAND ANOMALIES

SECTION IV

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


FIGURE 12: HK GDP GROWTH (TOTAL AND LESS INVENTORIES), % YOY

12.0

10.0

8.0

6.0

4.0

2.0

0.0

-2.0

-4.0

% Y

oY

1Q

GDP less stocks

2Q93 94 95 96

Headline GDP

3Q 4Q 1Q 2Q 3Q 4Q 1Q 2Q 3Q 4Q 1Q 2Q 3Q

Source: Census and Statistics Department, Hong Kong

In the meantime,wage growth per workerhas been running above8% for the last twoyears and M3 moneysupply growth has notfallen below 20% YoYsince November 1995.Yet, from January 1995to March 1997, con-sumer price inflationhas ranged between3.12% and 3.78% YoY.Believable? Not very.

Surely there are noholes in the economicdata for Asia’s model ofefficiency, Singapore?Oh yes there are. Ac-cording to official sta-tistics, no trade is car-ried on between Singa-pore and its largest neighbour, Indonesia.One explanation often cited is that theanomaly dates back to the political tensionsbetween the two countries in the 1960s.Maybe.

Or maybe the official trade flows wouldlook strangely small for such close neigh-bours. A visit to one of the many construc-tion sites around Singapore would uncovernumerous workers smoking kreteks (clovecigarettes). A quick glance at the packswould find that most of these had beensmuggled.

There again there is the national incomedata. All economic growth forecasts inSingapore are produced by analysts usingthe output-based measures in the nationalaccounts (i.e., by estimating industrygrowth rates). Well, we hope they are. Be-cause the expenditure side of the accounts(consumption, investment etc.) are incom-plete. Singapore does not report constantprice data for exports and imports. In thenational accounts analysts are faced withthe entry “Net Exports”.

All forecasts of private consumption,government consumption and investmentexpenditure in Singapore are guesses, pureand simple. They are guesses because thebiggest elements in the expenditure ac-counts, exports and imports, are unknown.

In Hong Kong the story is the reverse.GDP data are available from the expendi-ture side but are only released in currentprice terms on the output side of the ac-counts. Like most Asian countries, neitherSingapore nor Hong Kong even try to pro-duce national income data based on theincome side i.e., wages, profits, interestand rent.

But let’s stick with Hong Kong for a

moment. The above chart (Figure 12) con-tains one of the strangest data anomaliesin Asia in recent years - the case of thevanishing inventories. Given the severeausterity programme that hit China in1994-96, it might have been expected thatthe Hong Kong economy would slow downsharply. That did not happen, according tothe official statistics.

As the chart shows, headline real GDPgrowth held up remarkably well through-out the period (although it did fall to a lowof 3.3% YoY in 1Q96). The main reasonfor the sustained good performance inHong Kong was the build up in invento-ries. In this, the most expensive storagesite in the world, we are asked to believethat companies, people, government andwhoever built inventories.

Without that stockpiling exercise GDP

growth would have looked much more likethe second line on the chart. But growthrates of sub-4%, and even zero, wouldnever have done in the run-up to thehandover to China this year.

Certainly, the PADS project had someeffect on the inventory levels in Hong Kongand we would also put some of it down toinvoluntary inventory build up due to theslow down in China (not that that slow downhas ever been officially recognised in HongKong or Beijing) but the numbers still lookstrange.

Some numbers look strange elsewherein the region, Indonesia for example. Fig-ure 13 shows two data series for importgrowth in Indonesia. One is based on themonthly customs-cleared data (which mostanalysts and fund managers focus on as anearly indicator of trade patterns) and theother is the quarterly Balance of Paymentsdata. Apart from carriage, insurance andfreight costs - which are a pretty constantproportion from year to year - the two se-ries should behave in exactly the same way.

By and large, for most of the time pe-riod under consideration, they do. Only inthe last year or so have they begun to showa quite worrying divergence. At one pointin 1996 the customs-cleared data weresignalling a contraction in import growth- a worrying sign for overall economicactivity. A reason to sell the market?

Not if you had waited for the Balanceof Payments data. Through most of 1996this series was showing an acceleratingtrend - even against a higher base growthrate in the previous year. Imports signal-ling rising or falling economic activity?Depends on where you look.

The explanation for this anomaly ispartly the result of the introduction of free

trade zones in Indonesia.Customs data are notpicking up the through-put from these. The re-sult is confusion and, inaddition, inconsistentsignals on the all impor-tant direction of thetrade balance (accordingto customs data it im-proved YoY in 1996whereas the Balance ofPayments figures nowindicate a trade surplusdeterioration). Fortu-nately, at the moment atleast, investors are notawaiting anxiously thenext trade release.

India suffers a verysimilar trade problem to

FIGURE 13: INDONESIA CUSTOM-CLEARED IMPORT GROWTH VS

BALANCE OF PAYMENTS IMPORT GROWTH

60

50

40

30

20

10

0

-10

% 3

mm

a

Mar-90

Customs-cleared Imports Balance of payments imports

Mar-91 Mar-92 Mar-93 Mar-94 Mar-95 Mar-96

Source: CEIC

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


FIGURE 14: PHILIPPINES REAL GROSS CAPITAL FORMATION, % YOY 1990-96

50

40

30

20

10

0

-10

-20

-30

% Y

oY

Mar90

Dec90

Sep91

Jun92

Mar93

Dec93

Sep94

Jun95

Mar96

Dec96

Source: CEIC

skaters with an unex-pected data release.

The army ofmacroeconomic ana-lysts dissect the latestretail sales numbers forclues as to the directionof the economy, pros-pects for corporateearnings, the state of theconsumer and so on. Amonth later they do thesame thing again whilethe number they reactedto 30 days previouslyhas been revised out ofall recognition.

Often, in retrospect,markets are found tohave reacted in the mostperverse ways to data

that was subsequently revised to show fallsthat of Indonesia with the Reserve Bankand the Commerce Ministry providing sub-stantially different estimates of the tradedeficit each year. In each of the last fouryears the Reserve Bank’s estimate has beenclose to double that of the government -political considerations playing a role oris it all just due to the exclusion of de-fence imports from the Commerce Min-istry’s figures?

For a final example of data inconsist-ency, we turn to the Philippines. Figure 14shows YoY growth rates in fixed capitalformation in real terms. The third quarter1996 stands out like a sore thumb. Whatcaused the boom in investment in this onequarter? Pent-up government expenditure?Perhaps part of the answer. A YoY base ef-fect? Certainly there were weather prob-lems in 1995 but these were mostly in thefourth quarter. No, the statistic just looksodd - as do the value and volume seriesfor exports in the same year. Asian coun-tries have a long way to go before data con-sistency is a term that can be applied withconfidence.

Patchy data, patchy conclusionsWE COULD FILL a book with more ex-amples of Asian data confusions and con-tradictions. The Philippines external ac-counts, GDP revisions in Thailand, the VATeffects on Thai exports in 1996, the re-markably steady growth of headline GDPin Taiwan, the weaknesses in FDI approv-als data, everything in China (see SectionVI). But what would be the point? Only tolearn what we already know but rarely rec-ognise; we are dealing with developingcountries.

In the US, the largest bond and stockmarkets in the world often spin like ice

instead of rises. So much for the modelsof market efficiency and perfect informa-tion. How can the markets be efficientwhen the data are so wrong and so inaccu-rate?

All the more reason to treat the datacautiously when dealing with developingcountries.

The Bank of Thailand was said to have‘lost credibility’ in 1996 as a result of sub-stantial revisions to export data releases.Tosh. The US Commerce Department pro-duces just as large revisions to many ofits widely watched series every month andanalysts still await every release with batedbreath.

Developing country data - apart fromthe immediately observable series such asexchange rates and interest rates - havelarge error terms. They are filled with un-certainty. The conclusions we infer fromthem should be seen in that light.

Mindanao PhilippinesGRDP* GDP

1992 -0.5 0.31993 3.7 2.11994 3.6 4.41995 4.1 4.81996 5.8** 5.5

Note: * Gross Regional Domestic Product** Estimate Source: GEM and NSCB

‘backwater’ of the Philippines (Figure 15shows the official growth rates forMindanao against the Philippines as awhole) in the first place? Principally as aresult of a chance conversation over din-ner with a Philippines’ corporate execu-tive in January 1996. He was talking abouttheir distribution business and how fast itwas growing. “Manila the star performer?”I asked. “No way”, he said, “Mindanao!”.

We now know why. The photographs inthis section are all taken in and around Gen-eral Santos, the tuna capital of the Philip-pines (see map). It is a city of 400,000people. It is booming. The picture on page1 of this report is of one of the city’s taxicompanies, Boomtown Taxis. With an es-timated 12% real growth rate and 6%population growth - mostly migrants from

FERDINAND MARCOS managed onein 20 years. Cory Aquino made it fourtimes in six years. In the five years ofhis tenure so far, Fidel Ramos has doneit 62 times! Done what? VisitedMindanao. We are some way off Presi-dent Ramos’ record but our two tripsto Mindanao convince us that he is onto something big. There is a new starin the making in the Asian firmament.

Yet the popular perception of Mindanaois of the ‘war capital’ or ‘kidnapping capi-tal’ of the Philippines. The US governmentstill has a travel advisory out against visit-ing the island and, indeed, there are stillareas which are unsafe for foreigners.Muslim groups, such as the Moro IslamicLiberation Front (MILF) and Abu Sayyafcontinue their terrorist activities againstgovernment forces in areas of western andcentral Mindanao. Yet our experience hasbeen of a peaceful thriving centre ofagribusiness and industry.

Sensibly, we have stayed out of the trou-bled city of Cotabato which lies at the cen-tre of the small area of Muslim rebel ac-tivity left on the island. To write Mindanaooff because of the Muslim ‘problem’ is likesaying Britain is finished because of thetroubles in Northern Ireland. The relativesizes of the two ‘insurgencies’ are com-parable.

But why would we visit the supposed

MINDA-WOW! THE NEWFRONT DOOR

SECTION V

FIGURE 15: GROWTH IN MINDANAOAND THE PHILIPPINES, % YOY, 1992-96

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


Mindanao’s trade taxes werelowered in 1995, ZamboangaCity immediately experienced aP10m windfall in tariff revenuesfrom smuggling activity thatfound it less expensive (nobribes) to turn legitimate!

General Santos and Davaoare much closer to most Asianmarkets than is Manila. A threehour or less plane journey takesthe passenger (or cargo) to Bru-nei, Jakarta, Hong Kong, KualaLumpur, Papua New Guinea, Sin-gapore, Taipei and even Darwin(yes, the one in Australia).Mindanao is just a stone’s throwfrom Sabah on Borneo. (In fact,it is now just a plane ride;Mindanao Express began ascheduled service fromZamboanga to Sabah in earlyMay. Scheduled service fromGeneral Santos to Manado on

Sulawesi is also due this month.)The market for Mindanao’s agribusiness

products is huge (Japanese airlines are try-ing to access slots to Davao), all that isrequired are the infrastructure links. Thatis where policy comes in. The deregula-tion of the Philippines’ airways has alreadyhad a major impact on Mindanao’s growthrate. Mindanao Express, a start-up com-pany established with help from the GEM(Growth with Equity in Mindanao, a con-sultancy and investment bureau financedby US AID) project, not only flies passen-ger routes that criss-cross the island andinto EAGA but is daily involved in cargoflights from Mindanao to the Visayas toManila and to Subic Bay (and then back).

Grand Air, PAL, Air Philippines andCebu Pacific all have daily (or more fre-quent) flights from Manila to Davao. Onelocal businessman in Davao told us that heused to know all the faces on the singledaily flight from Manila to Davao. Sinceairline liberalisation, he hardly knows any.Fares have tumbled as well. What is goodfor the airways would also be good for theshipping lanes.

At present this is Mindanao’s biggestdisadvantage. The cabotage policy in thePhilippines is outdated and disadvanta-geous to outlying domestic suppliers.Transporting produce from Mindanao toManila is more expensive than bringing itfrom Bangkok to Manila or sending it fromMindanao to the US. That just doesn’t makesense. It also limits the market forMindanao goods.

Open skies, open seas. New roads, newports. The answers to unlocking the wealth

MINDANAO - THE BREAD BASKET OF THE PHILIPPINES

N

CAGAYANDE ORO CITY

Agusandel NorteCamiguin

MisamisOriental

Lanao del NorteZamboanga

del Norte

MisamisOccidental

Zamboangadel Sur

ZAMBOANGA CITY

Basilan

Tawitawi

Sulu

Maguindanao

COTABATOCITY

SaranganiProvince

SultanKudarat

GENERALSANTOS

CITY

DAVAOCITY

Davaodel Sur

DavaoOriental

Surigaodel Sur

Surigaodel Norte

Lanaodel Sur

NorthCotabato

Agusan del Sur

Davao del Norte

Bukidnon

MINDANAO SEA

SULU SEA

elsewhere in Mindanao, the Visayas andLuzon - it is not surprisingly the fastestgrowing city in the Philippines.

And that growth is set to accelerate. Thenew fishing port is due to be finished inlate 1997 (but more likely, in early 1998).This investment will increase tuna handlingcapacity from around 400 tonnes/day to4000 tonnes/day.

To the north of General Santos liesDavao, the largest city in Mindanao with apopulation of 1.5m. The visitor from Ma-nila feels perfectly at home immediatelyin Davao because the traffic is heavy. DavaoLight, the local electricity company, re-ports consumption growth of 12-14% forthe last three years and is projecting 13%growth in 1997. For Mindanao as a whole,which houses almost 25% of the Philip-pines population and accounts for 40% ofits agricultural production, the Universityof Asia Pacific (formerly the CRC) has es-timated a growth rate of between 7-8%over the period 1995-98. This is faster thanwill appear in the official statistics but isprobably, in itself, an underestimate.

But unlike so much fast growth in Asia,the Mindanao story is about agribusiness,not industry. The development ofagribusiness in the mild climate south (theheavier industrial areas are to be found inthe north of the island centred on Cagayande Oro) is at the centre of the island’s suc-cess. As we have argued in the past (seeChapter 8 in the Geometry of growth re-port), agriculture is the forgotten compo-nent of development in most Asian econo-mies. This is distinctly not the case inMindanao.

Tuna (includingsashimi for the Japa-nese market), pine-apples, asparagus(65% of the Japa-nese fresh asparagusmarket from a stand-ing start three yearsago), corn, beef,hogs, bananas, coco-nut oil and rice areamong the crops andproduce that south-ern Mindanao is richin. Okra growing is inthe developmentalstage. Most hearten-

ing of all for the visitor to the region is tolearn that all the programmes and devel-opment plans are well aware of how im-portant a clean environment is to this eco-system. No heavy industry is wanted insouthern Mindanao. The produce, equita-ble land tenure rights, trading links and thepotential are there - all it will take are theright policies to unlock the new front door.

Unlocking the new front doorWHY ‘FRONT’ rather than ‘back’ door?Simple really. Mindanao is the ancient trad-ing heartland of the Philippines. The Mus-lim population are descended from the trad-ers who swept through Indonesia and Ma-laysia long before the Spanish conquestsof Ferdinand Magellan. Trade ties betweenMindanao, Borneo and Sulawesi have beenthriving for centuries - they are now beingformalised in the shape of the East AseanGrowth Area (EAGA). Interestingly, when

TUNA: Rich haul

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


of Mindanao and putting it on the tables ofManila and the rest of the world are in thehands of Malacanang.

Peace and development inthe southern PhilippinesTHAT IS NOT to say that the MalacanangPalace has not already played a major rolein Mindanao in recent years. In its nego-tiations with the Moro National LiberationFront (MNLF) which resulted in the peaceaccord signed last year with the MNLF’sleader, Nur Misuari, the Ramos Adminis-tration has done more to promote peaceon the island than any other. However, aswe have mentioned above, the Muslim in-surgency is not a thing of the past yet -Abu Sayyaf and MILF groups are still op-erational in the Autonomous Region ofMuslim Mindanao (ARMM). Bombingsand kidnappings do still go on in the fourprovinces (out of Mindanao’s 21) thatmake up the ARMM.

But with better economic performancehas come a greater desire on the part ofthe rebels to integrate into Philippines so-ciety. One of the biggest investmentprojects brokered by GEM is a joint-ven-ture between Sumitomo Corp of Japan andex-MNLF guerrillas on the Tawi Tawi Is-lands. This project will become one of theregion’s largest seaweed producing opera-tions within the next few years.

Stumbling blocks still exist to a perma-nent settlement. The promised SouthernPhilippines Council for Peace and Devel-opment, a region encompassing most ofMindanao and to be headed by Nur Misuari,has not taken off. Opposition to the ap-pointed body in Congress and among lo-cal Christians is ensuring a rough passagefor this non-elected quasi-government.

Yet Mindanao quite clearly goes fromstrength to strength. Law and order inDavao is distinctly more tightly enforcedthan in Manila and terrorist activities areconfined to a tiny part of the island cen-tred around the Muslim majority area(Mindanao itself is around 70% Christian).Londoners should be so lucky.

More importantly, with economic suc-cess comes a greater willingness for Mus-lims and Christians to co-exist - bothgroups benefit. With land costs in Davao(the most expensive city in Mindanao) atless than half the level of Metro Manilaand electricity cheaper by 35%, the con-ditions are in place for the island to attractinvestment. As Asia’s next growth hot spot- but different in that it will be a more en-vironmentally-friendly one - that invest-ment is set to pour in through Mindanao’snewly-opened front door.FISHING BOATS: Labour-intensive, low-tech and highly profitable

MAKAR WHARF: 8,500 dwt is peanuts for the new Gen San port

Okra planting, the next cash crop?

NEW TUNA PORT: Tuna landing capacity would rise from 380 to 4,000 tonnes/day

AGRI-EXPORTS: Asparagus for Japan

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


Sources: Census and Statistics Dept., Hong Kong and China’s Latest Economic Statistics

FIGURE 17: HK RE-EXPORTS VS CHINA EXPORTS

70

60

50

40

30

20

10

0

-10

% Y

oY, 3

mm

a

Mar

Chinese export growth

Jul92

HK re-export growth

92Nov92

Mar93

Jul93

Nov93

Mar94

Jul94

Nov94

Mar95

Jul95

Nov95

Mar96

growth. Consider this: in 1995 China ex-perienced its highest ever level of realisedforeign direct investment, the economygrew at the fastest recorded rate in Asia(10.2%), export growth touched 23% YoY,food imports were at an all time peak andthe renminbi appreciated slightly in nomi-nal terms and strongly in real terms for thesecond year in a row.

All five of these conditions supportrapid import growth. The result? Importswere up 14.2% YoY, the second slowestincrease in the 1990s (behind 1994) andthe lowest increase in Asia. But is the fig-ure so unbelievable? Yes.

The import content of Chinese exportsis conservatively estimated at 40%. Theimport content of FDI flows was calcu-lated by the World Bank at 68% in 1994.Assuming that both of these proportionsheld in 1995 we would estimate the in-crease in imports from export and FDIsources alone to be US$14.5bn.

In the event, imports increased by onlyUS$16.4bn for the full year. This wouldleave very little scope for import growthgiven real economic expansion of 10%+and increased food imports. The figuresjust do not add up unless it is realised thatChina’s borders are extremely porous andthat official data collection is not that re-liable.

Again, this time on the export side, oneof the contributory factors to China’s sup-posed 10.2% GDP growth rate in 1995was a 23% surge in nominal export growth(even as the real exchange rate was appre-ciating rapidly). Figure 17 shows China’sexport growth since 1992 plotted againstHong Kong’s re-export growth.

While Hong Kong’s dominance as the

IN OUR 1st Quarter 1996 Eye on AsianEconomies we dropped all point fore-casts for China’s GDP growth rates. Wereplaced the numbers with three let-ters - U (up), D (down), F (flat). In ourview, the quality of the data just couldnot support spreadsheet analysis. Some-times it is a mistake to theorise evenwhen one has data.

We expected a few complaints fromclients (to expect a flood would have beenaltogether too vain given what we knowclients do with most brokers’ research).We have received none. On marketingtrips, our explanations for dropping thenumbers have been accepted with someamusement and much resignation.

But China should not be judged tooharshly in this respect. While we feel thatBeijing willingly accepts output data thatare quite clearly inflated by local cadresor factory managers, we equally believethat it is misled just as often as the rest ofus.

Remember, China is new to the nationalincome accounting game. Prices are stillnot entirely free and measurement tech-niques and statistical collection proce-dures are rudimentary at best. The authori-ties there have a wall of obfuscation toclimb in attempting to untangle the truepicture of state-owned enterprises.

That said, Beijing does itself few fa-vours. Releasing full year GDP growthestimates just before the end of the sameyear stretches credibility to breaking point.Not that this is always clear in analysts’reports. China’s official GDP numbersappear to be swallowed hook, line andsinker by most economists in the regionor to be accompanied with a warning aboutdata quality. Of course, once the warningis given the analyst seems free to use thedata as if nothing were wrong any more.Fund managers need to be more critical ofthis approach.

In our 3rd Quarter 1996 Eye on AsianEconomies (After the hard landing:OUCH! Where next?) we examined themain Chinese data series and outlined theanomalies in them. As examples of howpoor the data can be in Asia, there are none

“It is a capital mistake to theorize before one has data.”- Sherlock Holmes in Sir Arthur Conan Doyle’s, Scandal in Bohemia

INTERPRETING CHINESEDATA: ECONOMIST

OR DETECTIVE?

SECTION VI

better. The following is a reproduction ofpart of the Viewpoint section from thatquarterly.

Flying blindTHERE ARE 88 industrial production cat-egories reported in the official publicationChina’s Latest Economic Statistics eachmonth. In 1995 industrial output grew onaverage by 16% YoY. Figure 16 shows abroad classification of these categories bygrowth rate.

While it is possiblethat the fast growing cat-egories (which includeradio sets, motor cyclesand micro-computers)could make up for con-tractions in areas suchas energy production,the overall balance ofgrowth rates in indi-vidual categories doesnot feel right when com-pared with average out-put growth.

Another series ofanomalies which do notsupport the fast growthcontention are con-tained in the externaltrade data, especiallywith respect to import

FIGURE 16: CHINA’S INDUSTRIALPRODUCTION BY

GROWTH BANDS, 1995

Average gross industrial outputgrowth - 16%. Of which:

❚ 18 categories registeredcontractions

❚ 51 categories grew between0-16%

❚ 14 categories grew between17-32%

❚ 5 categories grewby more than 32%

Source: China’s Latest EconomicStatistics, January 1996

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


trans-shipment centre for Chinese exportsis gradually being eroded (as activity onthe mainland becomes more robust aroundShanghai in particular) it is still in that po-sition at present.

How then do we explain the pattern ofexport and re-export growth depicted inthe chart? Three main points are worth stat-ing:❚ Hong Kong data are of better qualitythan Chinese data,❚ Changes to the VAT rebate system forexports in China resulted in front-end load-ing of reported exports in 1995. Paper-work was completed for exports beforephysical movement was effected in orderthat companies would receive 17% rebatesrather than 13%.❚ The incentive to under-record exportsin China has diminished. This last point isthe most important. During 1993 in par-ticular there was every incentive for Chi-nese exporters to under-invoice exportsin China and realise earnings externally,particularly in Hong Kong. In turn, thisearnings flow could either be re-routedback into China (in the form of FDI withall the special privileges that FDI attractsrelative to domestic investment) or justparked in US dollar-based assets.

So, while Hong Kong re-export growthin that year was around 20% China’s ex-port growth meandered along at a sup-posed sub-9% rate.

In 1995 the reverse story was the case.Chinese companies had an incentive tomove money back into China - and theonly official way of doing so is in the formof export receipts - either to finance in-ternal operations because of the scarcityof bank loans or to take advantage of highinterest rates in the unofficial money mar-kets (where returns in excess of 25% oncapital could be achieved).

Hong Kong’s 17% re-export growthsits uneasily with the reported 23% risein Chinese exports. Moreover the patternof export growth bears little resemblanceto the Hong Kong data.

China’s fast export growth in 1995 wasjust as fictitious as its slow export growthin 1993. However, at least this makes theslow growth in imports somewhat morebelievable since ‘real’ i.e., actual, exportswere undoubtedly significantly less thanthe figure reported for the year. In turn,this allowed for greater import growth tosatisfy real demand and food requirements.

So, where does that leave us on theeconomy? Over the last two years we be-lieve that the Chinese economy has beenthrough a much rougher patch than the of-ficial figures would suggest. Individual

data series paint a very flawed picture ofthe ups and downs of the economy. How-ever, taking a broader view of the data webelieve that economic growth was some-where in the range of 5-8% in 1995. Fivefactors lead us to this conclusion:❚ The collapse in corporate earnings inthe listed A, B and H shares suggested ahigh cost economy with falling volumegrowth which induced a major marginsqueeze on the business environment. Fur-thermore, the collapse in recorded infla-tion last year is not indicative of 10%+growth but rather of sharply falling demandand overcapacity in the economy.❚ Hong Kong’s economic growth fell tosub-5% YoY. This is a direct result ofslower growth and poorer profits in Chinafor the HK-owned companies based there(much of whose output is actually destinedfor the domestic Chinese market ratherthan export markets).❚ The 14% import growth does not sit

easily with a developing economy appar-ently growing in double-digits and experi-encing record levels of foreign direct in-vestment.❚ The data and commentaries from sen-ior politicians point to industrial outputgrowth being overstated.❚ The institutional framework, which al-lows for the positive accounting of worth-less output in national income statistics,builds in an upward bias to official GDPgrowth. In other words, stockbuilding(which accounts for over 20% of GDP) isin part made up of goods which, if pro-duced in a capitalist economy, would bevalued at zero. In the Chinese ‘socialistmarket’ economy shadow pricing tech-niques are used to apportion values togoods which will never be used. GDPgrowth is biased upwards as a result.

Endnote: And all that was before westarted researching the undergroundeconomy!

LAST YEAR, in the first part of The RealMyth of Asian Growth, we developed theidea of endogenous growth - wheregrowth causes high savings and in it-self is the driving force for moregrowth. It is not a simple concept andis widely rejected by the neo-classicalmainstream stuck in their linearworld.

Fuzzy thinking is all about non-linearities, complexities and surprisingdiscoveries. Last year we used de Bono’s

notion of lateral thinking to ask clients tounlock the ‘concept prisons’ that conven-tional macroeconomic approaches haveplaced them in. The rest of this section re-visits the Young thesis of growth beget-ting itself and underlines the Kosko con-clusion that system (macroeconomic)complexity exceeds subsystem(microeconomic) complexity.

Optimistic economics restatedIN THE TWELVE months since our first

“System complexity exceeds subsystem complexity. That still baffles atomic physi-cists who use big chunks of matter to study and puzzle over the finest divisions ofmatter. It baffles medical students who cut up cadavers and wonder where the mo-tion and emotion went. It baffles econometricians who try to predict employmentand interest rates with simple linear models of an economy. As Bertrand Russellsaid, every man has a mother but mankind does not have a mother.”

- Kosko, op cit., p.108

“For the whole economy increasing returns are the rule. To think otherwise is tocommit the fallacy of composition in applying to growth theory a concept appropri-ate only to microeconomics.”

- Lauchlin Currie, Implications of an Endogenous Theory of Growth Based on AllynYoung’s Macroeconomic Concept of Increasing Returns, History of Political Economy ,

(forthcoming - Fall 1997)

“The dynamic element in growth is not saving but the increase in real demand whichin the longer term is dependent on improvements that reduce real cost or improvequality. The latter increases the supply and offer of goods which in turn becomesthe demand for other goods, which in turn generates the profits to finance the rela-tively small amount of capital necessary to maintain the current rate of growth.”

- Currie, ibid.

OPTIMISTIC ECONOMICSREVISITED

SECTION VII

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


look at Allyn Young’s optimistic growthmodel some things have changed. (AllynYoung died in 1928. He was an economistin the Classical tradition.)

The most reported of these is a seriesof studies, one from the IMF, which sug-gest that the original analysis of total fac-tor productivity (TFP) growth in Asia waswrong.

These new studies suggest that TFP, theresidual increase in output not explainedby increases in the inputs, has been muchhigher in Asia than previously thought.

We do not intend to enter that debate.After our comments about the data qualityin Asia in Sections II-IV it is our strongbelief that the estimates emanating fromthe growth accounting technique (which,like all neo-classical models, are ex-tremely data hungry in the testing phase)could be massaged to tell us anything. Inour view it is better to stick with theoryand decide whether or not themacroeconomy is a macrocosm of themicroeconomy and that the same rulesapply.

We think the fuzzy logic people andChaos theorists would put more credenceon Young’s thesis than on the growth ac-counting analysis. So do we.

What is the endogenous growth theorythen? Unlike conditions prevailing in themicroeconomic world of the firm (wherediminishing marginal returns exist), inYoung’s macroeconomic world increas-ing returns are the norm. Here, naturalgrowth in output leads to greater demand.Why? Because the act of producing morereduces the cost of any good in terms ofother goods.

This last point is key. Young’s endog-enous growth world does not rely on a be-nevolent central bank which injects everincreasing amounts of liquidity (the Aus-trians have effectively dynamited thatflawed route to growth in their analysis ofthe business cycle). It is purely the factthat the demand for most goods is elasticand that, as more of any one good is pro-duced, its price will fall relative to the priceof other goods.

In turn, this will increase the effectivedemand for the good and for all others. Inorder to improve profits the producer mustoffer more of the good at a lower price(the technological revolution of recentyears being a prime example of this).

With higher growth, supply increasesand prices fall - not the model the US Fed-eral Reserve believes in at present butthere again, its senior members (with theexception of Greenspan) represent the ul-timate in mainstream neo-classical econo-

mists. It is, however, either implicitly orexplicitly the model of the pro-growthlobby in the US.

The crucial concept in the Youngian the-sis is an old one - specialisation throughthe division of labour. It is the process ofspecialisation that raises the prospect offurther gains to existing resources (withspecialisation comes more roundaboutprocesses of production which not only addto the complexity of the economy but alsoto the productivity of existing labour re-sources).

Within this process - and this worldview incorporates time and the notion ofdevelopment as a process, not the staticequilibrium paradigm of the growth ac-countants - increasing returns are the rule.There are three main tenets to the endog-enous growth model:❚ Growth is self-generating in character.Left to the free play of the market it willexpand continuously.❚ Traditional factors of production arenot the source of growth but, rather, in-puts of capital and labour are a conse-quence of growth itself. This has impor-tant, positive connotations for the Philip-pines (see below).❚ Endogenous and exogenous elementsinterplay with each other. In this waygrowth rates can be enhanced by policieswhich facilitate access to new markets e.g.,open trade policies, trade tax liberalisationetc.

What are the implications from thesetenets? First, policies and events which actto extend the market will prolong and pro-mote growth in and of themselves.

As economies open up and infrastruc-ture development takes place the wholeprocess will be enhanced, if not acceler-ated. This takes place because of the in-creased opportunities to specialise anddivide labour.

Secondly, with growth independent offactor inputs, efforts to explain the phe-nomenon in terms of labour and capital(then apportion the residual to what isknown as productivity or technologicaladvance) are wrong-headed.

TFP measures are meaningless (evenwhen they are not calculated with flaweddata). Moreover, investment does not ex-plain growth but rather, is explained by it.That is not to say that Young’s theory de-nies any importance to investment. His ar-gument is simply that investment would beneither profitable nor efficient if the endmarket for the products of such investmentdid not exist in the first place.

In the context of the Philippines, this isimportant. Criticisms abound about the

country’s low savings rate relative to therest of Asia.

In Young’s eyes there is no surprise inthe difference - the Philippines has beengrowing much more slowly than the restof Asia over the last 10-20 years. How-ever, this will change.

The evidence is widespread in the de-velopment economic literature that, aseconomic growth accelerates, the incomesof the high saving segments of societiesrise faster than the average. Savings ratesincrease because the incomes of the sav-ers rise relative to national income. Thespringboard for financing growth will takecare of itself.

Thirdly, international trade and govern-ment expenditure can be viewed asexogenous growth generators i.e., they arenot dependent on growth itself.

Consumption and investment, there-fore, are consequences of growth -whether it be endogenous or exogenous.Public investment programmes, especiallywhere they facilitate future growth by pro-viding infrastructure to extend the market,can be viewed as promoting endogenousgrowth as a second round effect. Liberaltrade policies are exogenous growth pro-moters.

Putting this theoretical construct in theempirical Asian context would lead us toconclude that far from being at the latestages of the growth process, Asia hasmerely scratched the surface so far. Themost important points in this respect arethat growth in the region to date has beenconcentrated in limited geographical cen-tres, among a small percentage of the re-gion’s population and has been largely de-pendent on external sources.

As the internal market extends - withsubstantial help from domestic deregula-tion and liberalisation policies - the pros-pects for further division of labour, andhence specialisation, in Asia are enor-mous.

But isn’t this process limited by thetransfer of existing technology? No. Thetechnological frontier is being pushed outconstantly. In the 15 years to 1982 onestudy identified 2047 equipment innova-tions in that oldest of industries, textilemaking.

The microchip revolution has turned ariver of new processes, products, inven-tions and innovations into a flood.

Specialisation and innovation know nobounds. They are governed not by the dy-namics of mechanics but by the infinitelymore complex phenomenon of human in-ter-relationships. Demand and growth arelimited only by the market.

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


WHAT HAVE WE truly uncoveredabout Asia in the course of this report?First and foremost, that we know a lotless about the region than we pretendto. Secondly, that sizeable undergroundeconomies exist that lead to a system-atic understatement of income andwealth in the area.

Thirdly, we know that themacroeconomic data are not only flawedas a direct consequence of the existenceof the informal sector but are also inad-equate as a result of poor statistical meas-urement, inconsistent accounting classi-fications and a shortage of data collectionresources (both physical and human). Ifcountries in the region cannot collect taxesfrom their richest inhabitants, what chancehave they of collecting the type of infor-mation that make up national income ac-counts. More than that, the more fool usfor believing that they can.

Fourthly, assuming that the undergroundeconomy is gradually being ‘formalised’,economic growth rates - which always re-fer to the measured or reported economy- are being exaggerated as a result. Asia isricher than it appears at first sight but theoverall economy (underground plus for-mal) may have been growing more slowlythan has been suggested for the last fewyears.

Finally, this last point underlines the dan-gers for macroeconomic policy makers.If governments and central banks are run-ning policy on the basis of growth targetsand, at the same time, are using past expe-rience as the benchmark for what is andwhat is not possible, then painful mistakesare possible if not likely. If, for example,the absorption of the undergroundeconomy into the measured sector has ac-counted for 2-3 percentage points of GDPgrowth over the last five years, then as-sumptions made about the growth-inflationrelationship or the growth-external bal-ance relationship over this period are un-likely to hold when the absorption proc-ess stops.

The fuzzy lesson: hierarchiesof uncertaintyAS EINSTEIN said, reality is characterisedby uncertainty. But more than that, reality

is characterised by a hierarchy of uncer-tainties in all areas of study. Take philoso-phy to begin with.

The laws of the ‘hard’ natural sciencessuch as physics, chemistry and maths aremore certain than the laws of astronomy,medicine and biology which, in turn, aremore certain than the laws for complexsocial phenomena such as economics, his-tory and psychology. The laws of thermo-dynamics are characterised by smallpatches whereas those of economics arecharacterised by large patches, sometimesvery large patches.

Our own predisposition to the Austrianschool of thought, where the idea of theimmeasurability of complex social phe-nomena and the folly of positivism is atits strongest in economics, leads us to fa-vour pattern predictions rather than pointforecasts. Convention among fund manag-ers demands ever increasing amounts ofthe latter.

Staying with the economics subjectarea, let us move to the sphere ofmacroeconomic data. We can think interms of a hierarchy of uncertainty of eco-nomic measurement. In theory, identifica-tion of the extent of the formal economyshould be better in the OECD countriesthan in the developing countries. In turn,the more mature developing countries, forexample, the NIEs of Asia, should producemore certain economic estimates than,say, the low income countries such asChina, India and Pakistan. Country size canalso play a role.

With this in mind, we can return to es-timates of the informal economy in Asia.Given the evidence from the literature onthe black economy, indications from thecurrency ratios shown in Section 2, esti-mates from reported studies around theregion and just what we know ourselvesabout the Asian economies, we can arriveat a hierarchy of potential undergroundeconomies in the region. Figure 18 con-tains our subjective ranking of the under-ground economy hierarchy in Asia and theextent to which we believe the informalsector might exist.

Are these estimates too high? We don’tthink so. In an article in The Economist (3May 1997) a recent study of the ‘shadow’

FIGURE 18: HIERARCHY OFPOTENTIAL UNDERGROUND ECONOMYCONTRIBUTIONS IN ASIA (% OF GNP)

❚ China, India, Pakistan50-60%

❚ Indonesia, Philippines, Thailand30-50%

❚ Korea, Malaysia, Taiwan20-30%

❚ Hong Kong, Singapore10-20%

economy in the OECD countries was out-lined (the study used the Tanzi method de-scribed in Section II). On average, the au-thor found that the underground economyaccounted for 15% of GDP in the richcountries - ranging from about 7% inSwitzerland to 26% in Italy. For us to placeSingapore and Hong Kong around the av-erage of the OECD countries seems, ifanything, conservative.

Taking the mid-points of these ‘under-ground ranges’ and grossing up 1996 offi-cial GNP numbers would result in an ag-gregate gross national product for Asia ofUS$3.8 trillion, 40% higher than the re-ported total and just over half the size ofthe (formal) US economy. Japan, the sec-ond largest economy in the world, had anofficial GDP of around US$4.6bn in 1996(which would now be US$4bn at the pre-vailing exchange rate) - only 21% higherthat the Asian ‘total’ economy. In the un-derground economy study for the OECDboth the US and Japan were found to haveinformal sectors amounting to around 10%of GDP.

To say the least, the Asian figure repre-sents a sizeable internal market. It shouldbe noted that we have excluded most ofIndochina (Burma, Cambodia, Laos andVietnam) from the exercise as well asBangladesh and Sri Lanka. The prevailingmindset, despite much of the political bra-vado, is one of dependency rather than eco-nomic independence. Yet the policy em-phasis in Asia remains outward looking.

While the export-orientation is notcounter to Young’s thesis of endogenousgrowth, this type of strategy opens econo-mies to the vagaries of international de-mand swings and diverts attention awayfrom the easiest growth source, the inter-nal market. (Of course, market constrainedeconomies such as Hong Kong, Singaporeand, to a lesser extent, Malaysia are forcedto adopt primarily exogenous growth strat-egies).

There is a worrying twist to this obser-vation. In de Soto’s work on Peru he even-

“As far as the laws of mathematics refer to reality, they are not certain. And so far asthey are certain, they do not refer to reality.”

- Albert Einstein, Geometry and Experience .

ASIA UNCOVERED:THE FUZZY LESSON

SECTION VIII

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


tually concluded that there had been almostno capitalism in Peru. Instead, protectionfor domestic industries - for example, re-tailing, distribution, banking, steel produc-tion, cement production, and so on - hadresulted in a mercantilist, formal privatesector, not a capitalist one. In his view, theonly true capitalists in the country werethe informals and the poor.

It is very easy to think of examples inAsia where liberalisation of trade policieshas outstripped deregulation of the domes-tic economy - Korea, Taiwan, Indonesia,Thailand, Malaysia, China, India, Pakistanand, above all, Japan. Indeed, it is also thecase that, apart from multinational enter-prises, the small and medium sized com-pany sector tends to dominate the exportmarket in most of developing-Asia (this isnot true of Korea).

The acid test for Asian governments iswhether or not they are willing to crackopen the hard nuts - the vested interestgroups, the connected families, the oldmoney and the ‘big hands’ that control andoperate domestic industry and servicesthroughout the region. For its efforts al-ready in the telecommunications, oil andairline sector we have left the Philippinesout of the list in the preceding paragraph.The government there would be the firstto admit that there is a lot further to go.

Policy lessons and investment rulesIF THE REAL WORLD is characterisedby multivalence and shades of grey; byfuzzy facts and hierarchies of uncertainty,then so too is the world of economicpolicymaking, economic analysis and as-set allocation.

The Washington-based Institute of In-ternational Finance stated in a press re-lease on 24 April 1997 that:

“Emerging markets are showing con-tinued improvement in the area of datatransparency, a development that prom-ises to sustain private capital flows tothese regions.”

Later in the same article we find out thatthe report examined the timeliness andcomprehensiveness of data coverage butnot the quality of the data. So there we are:in a data hungry world, moulded by the phi-losophy of positivism and its prevailingwisdom of simplified models and testablehypotheses, all we need are lots of quicklyreleased economic data points. The factthat they might be garbage or subject to10% or 20% or 50% error bands doesn’tmatter. Well, in our book, it does matter.

The first basic step in handling emerg-ing country macroeconomic data is to rec-ognise the extent of our unknowledge.

Daily observations of the exchange rateagainst the US dollar or swap or moneymarket interbank rates are characterised byhigh degrees of certainty. It is, however,all downhill from there. The monetary sta-tistics should be relatively accurate assum-ing that the central bank has properly de-fined and classified all of the entries in theaccounts. That is a big assumption.

Moreover, since monetary data are col-lected by survey we have to assume thatthe survey is well designed and not opento interpretation by completing banks andfinancial institutions. There are standardinternational guidelines on monetary sur-veys but they are not always followed inthe less developed countries. Of course,even perfect monetary data leave us withthe next question - which series are rel-evant and why? Even monetarists can’tagree on that one.

Trade data are not as accurate as observ-ers might expect. Scams to avoid taxes orattempts to claim back more VAT than isdue can result in either under- or overstate-ment of exports.

Likewise with imports. The existenceof free trade zones, as we have shown inSection IV, can distort the monthly cus-toms-cleared data if the authorities’ col-lection procedures are adequate. The Bal-ance of Payments data are fraught with evenmore problems.

For example, in few countries are com-prehensive data collected on foreign com-pany retained earnings. In principle, theseshould be recorded twice in the externalaccounts, once as an investment incomeoutflow (through the current account) andthen as a foreign direct investment inflow(through the capital account).

For investors stuck in the conventionalmacroeconomic wisdom that the currentaccount balance is an all important stockmarket indicator, this accounting princi-ple (which at no point results in a flow ofmoney to or from the country) should helpthem lose money. Malaysia watchers takenote.

There is so much confusion over theaccounting for overseas contract workerremittances in the Philippines that it cansafely be said that no-one is clear aboutwhat is going on. The gloom-mongers caninterpret the holes negatively while theoptimists (of which we are one) see thediscrepancies in much more benign light.There is no guaranteed benchmark of whois right or wrong; only time will tell.

The GDP accounts are the ultimate inthe large-patch, uncertainty hierarchy ofmacroeconomic measurement in theLDCs. We have shown in Section III how

changes in the underground economy overtime might distort official GDP growthrates. Add to that all the other uncertain-ties mentioned in the statistical measure-ment context and we arrive at a set of datathat should be handled with extreme cau-tion.

For policymakers the rules are clear:keep policy flexible and choose targets thatare relatively easily observable. Inflationrates could qualify as one of the more cer-tain indicators but this is entirely depend-ent on the government’s attitude to con-trolling prices.

Alternatively, there is the exchange rate.A fixed exchange rate certainly anchorspolicy but it also neuters one of the mostimportant price signals in the economy.When this particular price is fixed, the lee-way in domestic monetary and fiscal policyis constrained dramatically.

We have never hidden the fact that webelieve this to be a particularly bad policyoption. More flexible exchange rate tar-geting - particularly when it avoidsmispricing capital for long term invest-ments - is worth considering, however.

What about credit growth or basemoney targets? Certainly these are rela-tively easily monitored but the question ofthe appropriate target range within specificeconomies is more difficult to resolve.Assuming that money supply growth tar-gets are set to support a given level ofnominal GDP growth, governments haveto be clear in what they are dealing with inrespect of the latter? Is it formal economyGDP growth or total (formal plus under-ground) economy GDP growth? The dif-ference might well be enormous and leadto all sorts of unintended distortions if thetargets are set inappropriately.

So where does that leave us as inves-tors? Completely in the dark? No. The con-clusion of this report is a positive one:Asia is richer than we think and more ca-pable of self-sustaining growth than iscommonly believed. To that extent, themonthly, quarterly and annual data don’treally matter. The invariably importantinvestment criterion in developingcountries is the direction of policy.Here are our five golden rules for invest-ment in Asia.

Rule 1: The government is moving for-ward on liberalisation and deregulationpolicies. This is particularly important inthe domestic sphere where the greatestpotential lies. The Youngian thesis ofgrowth begetting growth will take over asthe market is extended. From this, the re-turns to investment will flow. Trade liber-alisation policies are also good but may

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.


send a misleading signal if they are notsupported by domestic deregulation. Acombination of exogenous and endog-enous growth stimulation is required (seeSection VII).

Rule 2: The government is operatingflexible, rule-based policies. Rules can andshould be fuzzy, along with everythingelse. In the light of our (and the authori-ties’) unknowledge, flexibility is para-mount. In a world increasingly character-ised by global capital flows fixed pricepolicies of one sort or another (especiallythe exchange rate), will inevitably lead todistortions and eventual policy breakdown.Moreover, the quality of data do not sup-port rigid policy models.

Rule 3: The role of government in theeconomy is diminishing. This process cantake a number of forms such as the movetowards balanced budgets or surpluses,decreasing levels of public expenditure asa percentage of GDP, privatisation of gov-ernment monopolies and a general trendtowards market signals replacing govern-ment direction of resources. (It is this lat-ter feature that has made us more positiveon China over the last twelve months.)

Rule 4: Microeconomic evidence con-firms macroeconomic data. Do bank loangrowth data, retail sales, electricity sales,industry sales revenues and business sur-vey evidence support official data re-leases? If not, believe the former and dis-card the latter. The uncertainty patches aresmaller the more disaggregated and directthe data.

Rule 5: Think long term. Investing indeveloping economies cannot be aboutquarterly performance. The depth and in-dustrial variety in these markets are justnot there to support such a strategy. Un-derstanding the trend in development andthe emphasis of policy is paramount inemerging market investment. Identifyingpolicy weakness yields good early warn-ing signals but too much faith in flaweddata results merely in violent market fluc-tuations as investors panic about nothing.

Fuzziness is a fact of life. In investmentterms, the degree of fuzziness increaseswith underdevelopment. That requires adifferent investment mentality from thatfollowed in the mature economies. Policydirection is paramount - appropriatepolicy actions will ensure that the flow ofreturns in the economy increases but notnecessarily overnight.

The trick for investors once they havepicked their long term winners is to bepositioned in the right stocks - fortu-nately, this aspect isn’t one of the econo-mist’s direct responsibilities.

IN THE FIRST section of this report,we said that we wanted to raise theissue of the uncertainty surroundingAsian macroeconomic measure-ments particularly at this time -when everything is doom and gloom.From our vantage point we see con-fused investors, besieged fund man-agers and asset allocators sharpen-ing their scythes.

But the message we have is ulti-mately a positive one because there isa way out of the current impasse forAsia - the endogenous growth route ofYoung’s optimistic economics. Wehave to admit, however, that there willbe a few fund manager coronaries alongthe way if Asia decides to take our pre-ferred path.

Endogenous growth is about govern-ments in the region rejecting their im-plicit dependency on the OECD andlooking to the development of their ownmarkets as the key to future growth. Itis about unlocking domestic opportu-nities and tackling vested interests. DeSoto’s study in Peru found the domes-tic formal sector to be mercantilist,hooked on government protection, notcapitalist. His conclusion: the only truecapitalists were the informals.

This segment of society makes up50-70% of the urban labour force indeveloping countries and a much higherproportion in the rural areas. Asia ispopulated by capitalists who do not havea dependency culture because their eco-nomic survival is all about living in thereal world and not about state protec-tion.

Our plea to Asian governments: em-power them, learn from them, make lifeeasier for them. That, in turn, means lib-eralisation.

The escape route from the gloom inAsia is in the hands of policymakers.The strategy has to switch fromexogenous (export-led) to endogenous(domestic market extension and de-regulation-led) growth.

The mainstream economic wisdomis that Asian growth is export depend-ent and that is the way it must stay. Ifexports slow, growth dies and invest-ment funds must leave.

In our view these funds will be leav-

ing at just the point in history that eco-nomic forces are turning Asian coun-tries to their most profitable market -the internal, Asian market.

So, is the Asian growth story over?As we argued in the Geometry ofgrowth last year, the Asian growth storyis far from over, indeed it hasn’t evenreally begun. But to unleash the forcesof strong domestic growth governmentsaround the region have still to realisejust what their economic strategy mustbe. That strategy consists of abandon-ing the export and investment-ledgrowth models of the World Bank andIMF (Malaysia and Singapore are ex-ceptions in this respect) and offocussing on domestic conditions.

This is by no means an advocacy ofprotectionism, quite the reverse. Gov-ernments in Asia have to crack open thevested interest groups. They have to re-ject the mercantilism that permeates theregion.

For sure, this will mean that somefavourite ‘core holdings’ of fund man-agers will have to be ditched. Some-times the economist’s agenda and thatof investment managers does not al-ways coincide. But it is our strong con-tention, and we believe the irrefutableevidence for this is staring us in the face,that production for the domestic mar-ket is inherently more profitable thanproduction for export. That is the his-tory of the developed, albeit Western,countries.

We equate domestic-led growth withprofits. Johnson & Johnson, Ciba-Geigy, Nestle, IBM, Motorola, GeneralMotors and the next 500 largest com-panies in the world obviously see ex-actly the same thing. It would do fundmanagers no harm to follow their strat-egy. Overweight the domestic plays.Seek out and buy the consumptionstocks. That is where the next wave -and it is a tidal one - of growth is com-ing from.

In the meantime, the countrieswhich are leading the pack in policymomentum (whether by luck or design)are China (incorporating Hong Kong),India, Indonesia and the Philippines.These are our macro market picks forthe future.

THE MESSAGE REITERATEDSECTION IX

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.

Hong KongCredit Lyonnais Securities (Asia) Ltd33/F Lipppo Centre,EIE Tower89 QueenswayTel: (852) 2810-9338Fax: (852) 2868-0189Tlx: 81678 CLSA HX

China - ShanghaiCredit Lyonnais Securities (Asia) LtdRoom 403B, Central Place16 Henan Road SouthShanghai 200002Tel: (86) 21 6374-819/92/93Fax: (86) 21 6355-8010

JapanCredit Lyonnais Securities (Japan)Hibiya Kokusai Building 7th Floor2-2-3 Uchisaiwai-choChiyoda-ku, Tokyo 100Tel: (81) 3-5510-8650Fax: (81) 3-5512-5896Tlx: J33110 CLSECTOK

PhilippinesCredit Lyonnais Securities Philippines, Inc17/F, Trafalgar Plaza105 H. V. dela Costa StreetSalcedo VillageMakati City, Metro ManilaTel: (63) 2 848-3772 thru 81Fax: (63) 2 848-3697

USACredit Lyonnais Securities (USA) IncCredit Lyonnais Building1301 Avenue of The AmericasNew YorkNew York 10019Tel: (1) 212 408-5888Fax: (1) 212 261-2502

China - ShenzhenCredit Lyonnais Securities (Asia) Ltd19/F International Finance BuildingTien She RoadShenzhenTel: (86) 755 223-6542Fax: (86) 755 222-7452

KoreaCredit Lyonnais Securities (Korea)7/F Youone Building75-95 Seosomun-DongChung-ku, Seoul 100-110Tel: (82) 2-3708 7300Fax: (82) 2-771-8583

SwitzerlandCredit Lyonnais Securities (SE) Asia Limited2/F, 7 Rue de la Croix D'Or1204 GenevaTel: (41) 22-818-3030Fax: (41) 22-311-0807

United KingdomCredit Lyonnais Securities LtdBroadwalk House5 Appold Street, BroadgateLondon EC2A 2DATel: (44) 171-696-9190Fax: (44) 171-214-5401Tlx: 888397/8

IndiaCredit Lyonnais Securities India Pvt Ltd8/F Dalamai HouseNariman PointBombay 400 021Tel: (91) 22 284-1348Fax: (91) 22 284-0271

MalaysiaCredit Lyonnais Securities (Malaysia) Sdn BhdSuite 15-2 Level 15Menara PanGlobal8 Lorong P RamleeOff Jalan P Ramlee50250 Kuala LumpurTel: (60) 3-232-4288/466-3242Fax: (60) 3-238-4868

TaiwanCredit Lyonnais Securities (Asia) Ltd6/F, No. 117, Sec. 3Min-sheng E. RoadTaipeiTel: (886) 2-717-0737Fax: (886) 2-717-0738

SingaporeCredit Lyonnais Securities (Singapore) Pte Ltd16 Collyer QuayHitachi Tower #32-00Singapore 049318Tel: (65) 534-3268Fax: (65) 533-8922Tlx: 43286 CLSECS

IndonesiaPT Credit Lyonnais Capital IndonesiaWISMA GKBI Suite 1501Jl. Jendral Sudirman No.28Jakarta 10210Tel: (62) 21- 574 - 2626/

574 - 2323Fax: (62) 21-574-6920

PakistanCredit Lyonnais Securities (Asia) LimitedNo. 410 4/F Clifton CentreKhayaban-e-RoomiClifton, KarachiTel: (92) 21587-1074/587-1011Fax: (92) 21 586-2802

ThailandCredit Lyonnais Securities (Asia) Ltd10th Floor, Ploenchit Center BuildingTwo, Sukhumvit Road (Soi 2)Bangkok 10110Tel: (662) 656 8628Fax: (662) 656 8640

This report is produced by Credit Lyonnais Securities Asia for private circulation. All information, views and advice are given in good faith but without legalresponsibility. Credit Lyonnais Securities Asia or companies or individuals connected with it may have used research material before publication and may have

positions in or may be materially interested in any of the stocks in any of the markets mentioned.

MITA (P) No. 115/06/96

!"#$%"#& '(" ) *+ ,-.(/#012##3#"453,%05(.

Documents

Credit Lyonnais Securities Asia MYTH OF ASIAN GROWTH II · 2018-02-21 · Credit Lyonnais Securities Asia Asian Economic Research ASIA UNCOVERED /May 1997 1 ... But while the economist