7Finance & Development / March 1996

Stock Markets: A Spur to Economic GrowthR O S S L E V I N E

World stock markets are boom-ing and stock markets in devel-oping countries account for adisproportionately large shareof this boom. Investors are ven-turing into the world’s newestmarkets and some are seeinghandsome returns. But aredeveloping countries themselvesreaping any benefits from theirstock markets? The evidenceindicates that they are.

VER THE past 10 years, the totalvalue of stocks listed in all of theworld’s stock markets rose from $4.7trillion to $15.2 trillion, while the

share of total world capitalization representedby the emerging markets jumped from lessthan 4 percent to almost 13 percent. Tradingin the emerging markets also surged: thevalue of shares traded climbed from less than 3 percent of the world total in 1985 to 17 percent in 1995.

The emerging markets have attracted theinterest of international investors while rais-ing a number of critical questions for policy-makers in developing countries: Do stockmarkets affect overall economic developmentand, if so, how? What is the relationship

between stock markets and banks in fosteringeconomic growth? And, how can developingcountries benefit from stock market growth?

Impact on developmentDo stock markets affect overall economic

development? Although some analysts viewstock markets in developing countries as“casinos” that have little positive impact oneconomic growth, recent evidence suggeststhat stock markets can give a big boost to eco-nomic development.

Stock markets may affect economic activitythrough the creation of liquidity. Many prof-itable investments require a long-term com-mitment of capital, but investors are oftenreluctant to relinquish control of their savingsfor long periods. Liquid equity markets makeinvestment less risky—and more attrac-tive—because they allow savers to acquire anasset—equity—and to sell it quickly andcheaply if they need access to their savings orwant to alter their portfolios. At the sametime, companies enjoy permanent access tocapital raised through equity issues. By facili-tating longer-term, more profitable invest-ments, liquid markets improve the allocationof capital and enhance prospects for long-termeconomic growth. Further, by making invest-ment less risky and more profitable, stockmarket liquidity can also lead to more invest-ment. Put succinctly, investors will come ifthey can leave.

There are alternative views about the effectof liquidity on long-term economic growth,however. Some analysts argue that very liquidmarkets encourage investor myopia. Becausethey make it easy for dissatisfied investors to

sell quickly, liquid markets may weakeninvestors’ commitment and reduce investors’incentives to exert corporate control by over-seeing managers and monitoring firm perfor-mance and potential. According to this view,enhanced stock market liquidity may actuallyhurt economic growth.

The empirical evidence, however, stronglysupports the belief that greater stock marketliquidity boosts—or at least precedes—eco-nomic growth. To see how, consider threemeasures of market liquidity—three indica-tors of how easy it is to buy and sell equities.

One commonly used measure is the totalvalue of shares traded on a country’s stockexchanges as a share of GDP. This ratio doesnot directly measure the costs of buying andselling securities at posted prices. Yet, aver-aged over a long time, the value of equitytransactions as a share of national output islikely to vary with the ease of trading. In otherwords, if it is very costly or risky to trade,there will not be much trading. This ratio isused to rank 38 countries by the liquidity oftheir stock markets in four different groups.The nine countries with the most illiquid mar-kets are in the first group; the nine countrieswith the most liquid markets—that is, withthe largest value-traded-to-GDP ratios—are inthe fourth group; the second and third groups,each of which contains 10 countries, fallbetween the two extremes of liquidity. AsChart 1 shows, countries that had relativelyliquid stock markets in 1976 tended to growmuch faster over the next 18 years than coun-tries with illiquid markets.

The second measure of liquidity is thevalue of traded shares as a percentage of total

O

Ross Levine,a US national, is a Senior Economist in the Finance and Private Sector Development Division of the World Bank’s Policy Research Department.

market capitalization (the value of stockslisted on the exchange). This turnover ratiomeasures trading relative to the size of thestock market. Chart 2 indicates that greaterturnover predicted faster growth. The moreliquid their markets in 1976, the faster coun-tries grew between 1976 and 1993.

The third measure is the value-traded-ratiodivided by stock price volatility. Markets thatare liquid should be able to handle heavy trad-ing without large price swings. As Chart 3shows, countries whose stock markets weremore liquid in 1976—countries with highertrading-to-volatility ratios—grew faster overthe next 18 years than countries with less liq-uid markets. As demonstrated in the series ofpapers on which this article is based (see back-ground note), the strong link between stockmarket liquidity and economic growth contin-ues to hold when controlling for other eco-

nomic, social, political, and policy factors thatmay affect economic growth, and when usinginstrumental variable estimation procedures,various periods, and different country sam-ples. The basic conclusion that emerges fromthis statistical work is that stock market devel-opment explains future economic growth.

What is important is that other measures ofstock market development do not tell the samestory. For example, stock market size—asmeasured by dividing market capitalizationby GDP—is not a good predictor of economicgrowth (Chart 4), while greater stock pricevolatility does not necessarily predict pooreconomic performance (Chart 5). Empirically,it is not the size or volatility of the stock mar-ket that matters for growth but the ease withwhich shares can be traded.

Countries may be able to garner big growthdividends by enhancing the liquidity of their

stock markets. For example, regression analy-ses suggest that if Mexico’s value-traded-to-GDP ratio in 1976 had been the same as the average for all 38 countries in our sample(0.06 instead of Mexico’s actual ratio of 0.01),the annual income of the average Mexicanwould be 8 percent higher today. This type offorecast does not explain how to enhance liq-uidity, but it does give an indication of thepotentially large economic costs of policy, reg-ulatory, and legal impediments to stock mar-ket development.

Is there really a link between stock marketliquidity and economic growth, or is stock mar-ket liquidity just highly correlated with somenonfinancial factor that is the true cause of eco-nomic growth? Multiple regression proceduressuggest that stock market liquidity helps forecast economic growth even after account-ing for a variety of nonfinancial factors that

Finance & Development / March 19968

0

Chart 1Initial value-traded ratio (1976) and subsequent growth (1976–93)

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Very illiquid

Illiquid

Liquid

Very liquid

0.5 1.0 1.5 2.0 2.5 3.0 3.5

Very illiquid

Illiquid

Liquid

Very liquid

Annual per capita GDP growth (percent)

Annual per capita GDP growth (percent)

Chart 3Initial trading-to-volatility ratio (1976) and subsequent economic growth (1976–93)

Source: International Finance Corporation. Note: Initial liquidity is measured as the ratio, in 1976, of the value of shares traded to GDP.

Source: International Finance Corporation. Note: Initial turnover is measured as the ratio, in 1976, of the value of shares traded to market capitalization.

Source: International Finance Corporation. Note: Initial trading-to-volatility ratio is the ratio, in 1976, of the value of shares traded to volatility, which is measured as a 12-month rolling standard deviation estimate based on market returns. Data are availablefor only 30 countries in the study.

GDP grows faster in economies with liquid stock markets

0.5 1.0 1.5 2.0 2.5 3.0 3.5

Very illiquid

Illiquid

Liquid

Very liquid

Annual per capita GDP growth (percent)

Chart 2Initial turnover (1976) and subsequent economic growth (1976–93)

0

Austria, Colombia, Denmark, Finland,Indonesia, Nigeria, Norway, Portugal, Venezuela

Argentina, Belgium, Greece, Jordan,Luxembourg, Mexico, Spain, Sweden, Thailand, Zimbabwe

Brazil, Chile, France, Germany, India, Italy, Korea, Malaysia, Netherlands, Philippines

Australia, Canada, Hong Kong, Israel, Japan, Singapore,Taiwan Province of China, United Kingdom, United States

Chile, Denmark, Greece, Jordan, Luxembourg, Nigeria, Norway, Portugal, Venezuela

Austria, Belgium, Colombia, Finland, Indonesia, Malaysia, Mexico, Spain, Sweden, Zimbabwe

Australia, Brazil, Canada, France, Germany, Netherlands, Singapore, Thailand, United Kingdom, United States

Argentina, Hong Kong, India, Israel, Italy, Japan, Korea, Philippines, Taiwan Province of China

Argentina, Colombia, Denmark, Greece, Norway, Spain, Sweden, Venezuela

Austria, Belgium, Brazil, Finland, Italy, Mexico, Sweden

Canada, Chile, France, Germany, India, Jordan, Philippines

Australia, Israel, Japan, Korea, Netherlands, Thailand, United Kingdom, United States

Our study, which is based on a total of 38countries, includes both industrial anddeveloping countries. This is because somedeveloping countries have more liquid stockexchanges than countries with higher percapita GDP.

influence economic growth. After controllingfor inflation, fiscal policy, political stability, education, the efficiency of the legal system,exchange rate policy, and openness to in-ternational trade, stock market liquidity is still a reliable indicator of future long-term growth.

Stock markets versus banksIs there an independent link between stock

market development and growth, or is stockmarket liquidity correlated with banking devel-opment—and is the latter the financial factorthat really spurs economic growth? Althoughcountries with well-developed banks—asmeasured by total bank loans to private enter-prises as a share of GDP—tend to grow fasterthan countries with underdeveloped banks(Chart 6), the effects of banks on growth canbe separated from those of stock markets.

To evaluate the relationship between stockmarkets, banks, and growth, our 38 samplecountries were divided into four groups.Group 1 had greater-than-median stock mar-ket liquidity (as measured by the value-traded-to-GDP ratio) in 1976 and greater-greater-than-median banking development.Group 2 had liquid stock markets in 1976 but less-than-median banking development.Group 3 had less-than-median stock marketliquidity in 1976 but well-developed banks.Group 4 had illiquid stock markets in 1976and less-than-median banking development.

Countries with both liquid stock marketsand well-developed banks grew much fasterthan countries with both illiquid markets andunderdeveloped banks. Furthermore, greaterstock market liquidity is associated withfaster future growth no matter what the levelof banking development. Similarly, greater

banking development implies faster growthno matter what the level of stock market liquidity. Thus, it is not a question of stockmarket development versus banking develop-ment—each, on its own, is a strong predictorof future economic growth.

Why might stock markets and banks both,independently of each other, boost economicgrowth? Although the empirical evidence isconsistent with the view that stock marketsand banks promote economic growth indepen-dently of each other, the reasons are not fullyunderstood. One argument is that stock mar-kets and banks provide different types offinancial services. Stock markets offer oppor-tunities primarily for trading risk and boost-ing liquidity; in contrast, banks focus onestablishing long-term relationships withfirms because they seek to acquire informa-tion about projects and managers and enhance

9Finance & Development / March 1996

Chart 4Initial market size (1976) and

subsequent economic growth (1976–93)

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Very large

Large

Small

Very small

Annual per capita GDP growth (percent)

Source: International Finance Corporation. Note: Initial market size is measured as the ratio, in 1976, of market capitalization to GDP.

0 0.5 1.51.0 2.52.0

Size and volatility of stock markets are less important in forecasting GDP growthChart 5

Initial volatility (1976) and subsequent economic growth (1976–93)

Very volatile

Volatile

Stable

Very stable

Annual per capita GDP growth (percent)

Source: International Finance Corporation. Note: Initial volatility is measured as a 12-month rolling standard deviation estimate based on market returns in 1976. Data are available for only 32 countries in the study.

Chart 6Initial banking development (1976) and subsequent economic growth (1976–93)

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Very developed

Developed

Less developed

Under- developed

Annual per capita GDP growth (percent)

Source: International Monetary Fund. Note: Initial banking development is measured as a percentage of GDP in 1976 represented by bankloans to enterprises. Data are available for 37 countries in the study.

Chart 7Initial liquidity (1976) predicts high rates of subsequent capital accumulation and

productivity growth (1976–90)

0 1 2 3 4Annual growth of capital per person and productivity (percent)

Sources: International Monetary Fund and International Finance Corporation. Note: Initial liquidity is measured as the ratio, in 1976, of value of shares traded to GDP. (See chart 1 for countries.)

5

Very illiquid

Illiquid

Liquid

Very liquid

Productivitygrowth

Capitalgrowth

Banks also spur growth

Canada, Chile, Hong Kong, Japan,Jordan, Luxembourg, Singapore,United Kingdom, United States

Australia, Belgium, Brazil, Greece, Israel, Malaysia, Netherlands, Spain, TaiwanProvince of China, Zimbabwe

Colombia, Denmark, Finland, France, Germany, Korea, Mexico, Norway, Philippines, Venezuela

Argentina, Austria, India, Indonesia, Italy, Nigeria, Portugal, Sweden, Thailand

Argentina, Brazil, Chile, Korea, Mexico, Philippines,Thailand, ZimbabweGreece, Israel, Italy, Jordan,Norway, Spain, UnitedKingdom, Venezuela

Australia, Belgium, Canada, Colombia, Denmark, France, Pakistan, Sweden

Austria, Finland, Germany, India, Japan, Netherlands,New Zealand, United States

Austria, Finland, Germany, HongKong, Japan, Portugal, Singapore, Spain, Taiwan Province of China

Canada, France, Israel, Italy, Jordan, Korea, Netherlands, Norway, Sweden, United States

Australia, Belgium,Denmark, Greece,Malaysia, Philippines, Thailand, United Kingdom, Venezuela

Argentina, Brazil, Chile, Colombia, India, Indonesia, Mexico, Nigeria, Zimbabwe

corporate control. (There is, of course,some overlap. Like stock markets,banks help savers diversify risk andprovide liquid deposits. Like banks,stock markets may stimulate theacquisition of information aboutfirms, because investors want tomake a profit by identifying under-valued stocks to invest in; stock mar-kets may also help improve corporategovernance by simplifying takeovers,providing an incentive to improvemanagerial competency.)

Is greater stock market liquidityassociated with more or better invest-ment? Both. Chart 7 shows that coun-tries that had more liquid stockmarkets in 1976 enjoyed both fasterrates of capital accumulation andgreater productivity gains over thenext 18 years.

However, although liquid equitymarkets imply more investment, newequity sales are not the only source offinance for this increased investment.Most corporate capital creation isfinanced by retained earnings andbank loans. Although this phe-nomenon is not wholly understood, greaterstock market liquidity in developing countriesis linked to a rise in the amount of capitalraised through bonds and bank loans, so that corporate debt-equity ratios rise withmarket liquidity. Stock markets tend to com-plement—not replace—bank lending andbond issues.

Tips for policymakersGiven the important role well-functioning

stock markets seem to play in economicgrowth, what can countries do to promotethem? Fully answering this question is wellbeyond the scope of any single article. Legal,regulatory, accounting, tax, and supervisorysystems influence stock market liquidity. Theefficiency of trading systems determines theease and confidence with which investors canbuy and sell their shares. And the macroeco-nomic and political environments affect mar-ket liquidity.

Consider the impact of one particular policylever: liberalizing controls on internationalcapital flows. Liberalization may involve eas-ing restrictions on capital inflows or reducingimpediments to repatriating dividends or cap-ital. In either case, reducing barriers to cross-border capital flows can affect the functioningof emerging stock markets: first, by enhancingthe integration of emerging markets intoworld capital markets, thereby bringing theprices of domestic securities into line withthose elsewhere; and, second, by forcing

domestic firms seeking foreign investment toupgrade their information disclosure policiesand accounting systems. Moreover, the entryof more foreign investors into emerging mar-kets may lead to pressure to upgrade tradingsystems and modify legal systems to supportmore trading and the introduction of a greatervariety of financial instruments.

Through all of these channels, the removalof barriers to foreign investment can improvethe operation of domestic capital markets.This is consistent with recent evidence. Asshown in the table, stock market liquidity rosesignificantly in 12 out of 14 countries that lib-eralized controls on international capitalflows. Chile, for example, liberalized restric-tions on the repatriation of dividends by for-eign investors in January 1988. None of the 14 countries experienced a statistically sig-nificant drop in liquidity following liberaliza-tion. In conjunction with the earlier findingsthat market liquidity boosts economic growth,these results suggest that liberalizing inter-national capital flow restrictions can acceler-ate economic growth by enhancing stock mar-ket liquidity.

The table also indicates, however, that stockmarket volatility rose in 7 out of 11 countriesfollowing liberalization. Volatility did not fallsignificantly in any of the 11 countries follow-ing liberalization. Thus, while easing interna-tional capital flow restrictions may increaseliquidity, it may also increase volatility. Thisshould not be of much concern in the long run,

however. As shown in Chart 5, volatil-ity does not have any measurableeffect on long-term growth. Thus, ifpolicymakers have the patience toweather some short-run volatility, lib-eralization offers expanded opportuni-ties for long-run economic growth.

Does every country need an ac-tive stock market of its own?Unfortunately, there is not much evi-dence available to answer this ques-tion. In principle, all countries do notneed domestic stock markets. Theydo, however, need easy access to liq-uid stock markets where residents anddomestic firms can buy, sell, and issuesecurities. It is the ability to trade andissue securities easily that facilitateslong-term growth, not the physicallocation of the market. In other words,there is little reason to believe thatCalifornia would grow faster if theNew York Stock Exchange weremoved to Los Angeles.

When should policymakers reallypush stock market development? Thisis even more uncertain. The availableinformation suggests that policymak-

ers should remove impediments—tax, legal,and regulatory barriers—to stock marketdevelopment. But there is not strong evidenceto support interventionist policies—like taxincentives—that artificially boost stock mar-ket size and activity.

While much work remains to be done, agrowing body of evidence suggests that stockmarkets are not merely casinos where playerscome to place bets. Stock markets provide ser-vices to the nonfinancial economy that are cru-cial for long-term economic development. Theability to trade securities easily may facilitateinvestment, promote the efficient allocation ofcapital, and stimulate long-term economicgrowth. Furthermore, the evidence suggeststhat stock market liquidity encourages—or atleast strongly forecasts—corporate invest-ment, even though much of this investment isfinanced through retained earnings, bankloans, and bonds, rather than equity issues.Policymakers should consider reducingimpediments to stock market development.Easing restrictions on international capitalflows would be a good place to start.

Finance & Development / March 199610

This article is based on 12 papers presented at aWorld Bank conference, “Stock Markets, CorporateFinance, and Economic Growth,” organized by AsliDemirgüç-Kunt and Ross Levine. These papers areavailable from the author on request.

Year ofliberalization Size Liquidity Volatility

Argentina 1989

Brazil 1983

Chile 1988

Colombia 1989–91

India 1990–92

Jordan 1987

Korea 1981–92

Malaysia 1986 n.a.

Pakistan 1990

Philippines 1988

Portugal 1988 n.a.

Thailand 1988

Turkey 1990 n.a.

Venezuela 1988

Source: Ross Levine and Sara Zervos, 1995, “Capital ControlLiberalization and Stock Market Development,” unpublished (WorldBank).

Note: Arrows indicate increases; dashes indicate no significantchange; n.a. indicates data were not available.

Stock markets after liberalization

F&D