Dr. Mahathir's rhetoric is his alone, but his views are not. In the wake of the financial crises of the past few years, it has become routine to blame hedge funds, which abruptly move large sums of money in and out of nations, for destabilizing economies and impoverishing the innocent. Yet those who are considering the imposition of capital controls on portfolio investments and the regulation of hedge funds should pause before acting.

Such reforms would not eliminate the volatility of global capital flows, because the idea that the actions of hedge funds and other portfolio investors are its prime cause is false. In the recent crises in Asia and Russia, the "hot money" came chiefly from banks, not portfolio investors. Indeed, throughout the 1990s, lending by foreign banks was often a more unstable form of overseas financing than portfolio investments in equities and bonds, for most overseas bank lending takes the form of short-term interbank loans rather than long-term project finance. Furthermore, hedge funds are considerably smaller than other players in the system. Over the long run, as capital markets continue to grow in importance and ultimately replace traditional bank lending as the main source of external financing, international capital flows will probably become less volatile.

These facts should refocus the global debate on the question of how to promote the shift from banks to adequately regulated capital markets in countries that have only just begun to make the transition.

Banks are the problem

In Thailand, capital inflows grew rapidly throughout the 1990s, reaching a peak of $25.5 billion in 1995. Almost three-quarters of this sum took the form of foreign bank loans; the remainder consisted of equity and bond investments and direct investments by companies. By early 1997, it was evident that many of these investments were unprofitable and that Thailand's domestic banking system was saddled with an increasing number of nonperforming loans. Foreign banks responded quickly, turning a $1.9 billion inflow in the first quarter of 1997 into a $6.2 billion outflow in the second quarter—a plunge of more than $8 billion over three months (Exhibit 1). Since this figure does not include unused lines of credit that might have been cut, it may even understate the amount of credit withdrawn from the country.¹

These developments placed enormous pressure on Thailand's currency, the baht, which was devalued in July 1997. Foreign banks continued to pull money out of Thailand throughout the rest of 1997 and during 1998, triggering a massive liquidity crisis that in turn spawned a severe economic recession. In the aftermath of the crisis, domestic capital took flight to the tune of a further $2.5 billion.

What role did foreign portfolio investors play in all this? Unlike foreign banks, they remained calm throughout the crisis, continuing to pour money into Thailand even after the problems began. Immediately following the baht's devaluation, foreign investors went on a buying spree, so that portfolio inflows increased by more than 70 percent during the second quarter of 1997. Foreign portfolio flows to Thailand remained positive, albeit much smaller, for the rest of that year and the first half of 1998. Not until late 1998, when it became clear that the recession caused by the banking crisis would dampen the country's growth prospects for several years to come, did foreign investors begin to withdraw money from Thailand.

Foreign bank lending proved to be the hot money in the rest of Asia and Russia as well. In 1996, the five Asian countries hit by the crisis (Indonesia, Malaysia, the Philippines, South Korea, and Thailand) received a total of $47.8 billion in loans from foreign banks. After the crisis began in 1997, this capital inflow turned into a $29.9 billion outflow—a turnaround of almost $80 billion. Portfolio inflows, meanwhile, fell by half but remained positive. The same pattern held true for individual states, such as Malaysia and South Korea. There it was money from foreign banks, not from portfolio investors, that dried up and fled. Only in the Philippines, the least affected of the five countries in crisis, did the outflow of portfolio investments exceed the outflow of loans from foreign banks.² A year later, in Russia, foreign bank lending again proved to be the money that quit the fastest.

Not only in crises but in general, lending by foreign banks tends to be a more volatile source of financing than portfolio investments

This pattern isn't limited to crises. In general, lending by foreign banks tends to be a more volatile source of financing than portfolio investments. During the 1990s, quarterly swings in the total amount of foreign bank lending were far larger than the swings in portfolio bond or equity flows. Over this period, the volatility of bank loan flows was 82 percent; for portfolio flows it was 50 percent.³ The same pattern applies to many individual countries. From 1992 to 1997, the average volatility of annual foreign bank lending to any one of them was 239 percent, but it was 176 percent for bond flows and 150 percent for equity flows (Exhibit 2).⁴

The volatility of lending by foreign banks reflects the current nature of their activities. Many people assume that bank loans come in the form of long-term, project-based finance that by definition can't be withdrawn suddenly. In fact, international bank lending today often takes the form of short-term interbank loans. At the end of 1997, after the Asian crisis had begun, more than 55 percent of foreign bank loans around the world had maturities of less than one year. Over one-third were interbank loans, which were then used for many purposes, including the financing of long-term domestic lending. These proportions were even higher in many emerging markets. In Thailand, two-thirds of loans had maturities of less than one year, and most foreign bank lending went to banks and finance companies, where it remained.⁵ In better days, foreign banks would roll over most short-term loans as soon as they expired, often indefinitely. This became so routine that many banks and finance companies in emerging markets imprudently began to rely on these loans to fund new loans of different duration. But when trouble appeared on the horizon, foreign banks stopped rolling over the loans and instead demanded repayment, precipitating the crisis and prompting the outflow of capital.

Incentives and competitive pressures

What makes the lending activities of foreign banks so volatile? The explanation isn't that they are irrational or fail to understand the risks they incur; it lies rather in their incentives and the competitive pressures they face. During the 1990s, margins on traditional bank lending shrank. Under pressure from shareholders, many banks sought to bolster their returns by lending to emerging markets. The capital requirements of the Bank for International Settlements (BIS) inadvertently encouraged this trend by making it less expensive for banks to lend money to some emerging-market governments and to other banks than to the largest, most stable corporations.⁶ Once banks invest their money in these markets, however, they have an incentive to pull loans quickly in times of trouble. This combination of competitive pressures and incentives has thus caused banking institutions to drive the hot money in global financial markets. The fact that all banks have the same incentives and tend to act in unison amplifies the effect on the market.

The nature of the loan contracts is also to blame. Bank loans are mostly illiquid, fixed-price assets. Once booked, they are not repriced to reflect new information about borrowers, except in cases of breach of contract or default. Because the price of a loan doesn't automatically adjust itself to changing market conditions, banks adjust the quantity of lending instead. Thanks to the short maturities common today, a bank that is monitoring its loan portfolio can avoid the defaults of borrowers it believes to be in trouble by simply ceasing to lend. Bonds and equities, in contrast, adjust to changing market conditions through price rather than quantity; they reprice continuously to reflect new information as it emerges. Losses are realized immediately, so an investor can't avoid them by selling. Instead, investors have an incentive to hold on to their investments and wait for prices to rise.

Paradoxically, advances in bank risk management techniques exacerbate the volatility of foreign bank lending. With bank loans, it is the banks themselves (and their shareholders) that bear the full risk of a loss. As the debt crisis of the 1980s showed, the loss from foreign lending incurred by a single bank can be large; Citibank, for example, wrote off $3.5 billion in loan losses on Latin American debt.⁷ To manage these risks, banks have developed sophisticated risk management tools, such as the value-at-risk (VAR) model. When the market or credit risk in a country increases more than expected, these models indicate that banks must supply extra capital to cover the risk or reduce their exposure to the market in question, either by selling their securities from that country or by cutting their lending to it. Because securities are repriced immediately, banks are more likely to cut lending, since doing so allows them to avoid losses. Thus, as the credit risk of Asian banks increased over 1997, risk management principles called for foreign banks to reduce their level of lending. The adoption by most banks of similar risk management models has increased the impact of these actions.

The largest portfolio investors, however, usually don't employ this risk management technique, in part because risk and losses are not concentrated in mutual and pension funds but are instead passed on to a large number of individual investors. These funds' investment managers have little reason to pull out of a market abruptly unless withdrawals by individual investors force them to do so. But the strategies of such investors have been remarkably stable, mainly because emerging markets typically account for only a small percentage of their overall portfolios. Indeed, customer withdrawals from emerging-market mutual funds didn't increase notably during the Asian crisis.⁸ Portfolio investment managers thus have more flexibility in reacting to risk and less incentive to withdraw funds when market conditions change.

Despite the banks' skittishness, a drop in the profitability of domestic lending in developed markets over the past decade has made lending to emerging markets attractive. Such loans offer high returns, especially after the lower capital requirements of some of this lending is factored in. Although the Group of Seven (G-7) and other countries are proposing changes, the Basle Accord currently requires a bank to hold less capital against a sovereign loan to Mexico, for example, than against a loan to General Motors.⁹ And banks have carefully structured their lending activities to emerging markets to minimize the risk to themselves. Short maturities give banks the flexibility to pull loans if economic conditions change, and the vast majority of loans to emerging markets are denominated in US dollars, thereby eliminating currency risk. (Three-quarters of all foreign loans to Thailand were made in that currency, though they came largely from Japanese and European banks.) Such precautions, while increasing the risk for borrowers, have paid off for banks. Most of the losses that US ones, at least, sustained in Asia arose from trading or other kinds of business, such as underwriting, not from bad loans.

Ironically, the banks that were in the worst shape had the greatest incentive to lend to emerging markets. Faced with a mountain of bad debt at home and low overall returns, Japanese banks became the largest lenders to Thailand and the rest of Southeast Asia. By June 1997, they had extended $97.2 billion in loans to the region, while US banks had extended only $23.8 billion. Large French and German banks—prompted by stagnant domestic markets and by pressure from nimble, smaller competitors on the margins of more traditional products—also became prominent lenders to Southeast Asia.

For a time, this lending was highly profitable. Japanese banks could raise money domestically at 1 percent and then lend it to Thai or South Korean banks at 4 or 5 percent. But difficulties at home made Japanese institutions highly sensitive to potential losses, prompting the massive withdrawal of credit at the start of the crisis. Of the $17.5 billion decline in lending to Southeast Asia from June to December 1997, $10.5 billion was withdrawn by Japanese banks, which lacked the financial strength to renew credit to borrowers in difficult times.

It should therefore surprise no one that foreign bank lending can be highly volatile. Changes in the types of lending activity, the nature of the loan contract, risk management techniques, and competitive pressures all prompted banks to make high-risk loans to emerging markets in a way that posed minimal risk to themselves.

What about hedge funds?

At first glance, hedge funds would seem to be prime suspects in the search for the causes of volatility in financial markets. These funds are largely unregulated, with few disclosure requirements and a penchant for secrecy. They locate off-shore or in outlying areas instead of clustering in global financial centers with the large banks and brokerage houses. They use highly dynamic trading strategies, darting in and out of markets to arbitrage price differences. They labor under very few investment restrictions. They can build highly concentrated portfolios and invest in the most exotic instruments, taking long, short, or derivative positions. And as the near bankruptcy of Long-Term Capital Management showed, hedge funds occasionally make highly leveraged market bets. Some people fear that this makes it possible for the funds to manipulate markets and to cause financial crashes.

Hedge funds and other participants in the market, such as proprietary trading operations at commercial banks, are not very different

But the impact of hedge funds on international markets is greatly exaggerated. To begin with, hedge funds are not fundamentally different from other market participants, such as proprietary trading operations at commercial and investment banks, which have been using the same types of trading strategies and techniques as hedge funds for years. By now, there is essentially no difference in the activities of these various institutions; in fact, traders from leading banks started many hedge funds. In 1997, it is estimated, the proprietary trading operations of banks as a group had about twice the assets of hedge funds.

Moreover, hedge funds are relatively small. Collectively, pension funds, mutual funds, and insurance companies control assets of almost $25 trillion, but hedge funds command assets of only $800 billion to $1 trillion, making those funds about 4 percent of the size of the institutional asset managers (Exhibit 3). Although Long-Term Capital Management infamously took positions that were 20 or more times the size of its capital base, such risk taking is not typical. Van Hedge Fund Advisors International reports that about one-third of hedge funds use no leverage and that the average leverage ratio for all hedge funds is only between 4:1 and 7:1. Even counting leveraged assets, hedge funds are therefore significantly smaller than institutional investors. And while hedge funds, owing to their greater appetite for risk, are likely to have more of their investments in foreign markets than do institutional asset managers, they are again much smaller contenders there. The size of hedge funds thus suggests that on their own, they are not likely to have been the source of excessive volatility in most markets.

As events in Thailand show, that volatility results from the behavior of many market participants acting in unison. On July 2, 1997, the Thai government gave up its exchange-rate peg and allowed the baht to float, starting a depreciation that was to wipe out 56 percent of the currency's value by January 1998. When the exchange-rate peg was abandoned, the Bank of Thailand estimated that hedge funds held at most a quarter of the $28 billion in forward contract sales of the baht (the main instrument used for short selling). Most of the pressure on the baht came from short sales by other market operators, the massive outflow of foreign bank loans, the liquidation (by domestic companies and banks) of securities positions, and last-minute attempts by corporations and banks to hedge previously unhedged foreign loans. Hedge funds took their positions on the baht only after significant pressure had already built up, and they exited the market early.¹⁰ Many hedge funds started building long positions in the baht and other Asian currencies in August and September—and lost money.

Certainly, a few large hedge funds profited handsomely from the initial devaluation of the baht. Nonetheless, it is clear that they did not play a leading role in the Thai crisis.¹¹ On the contrary, it was the overwhelming pressure exerted by the collective actions of all market participants that precipitated the problems. As one trader in Thailand noted, "Who wasn't betting against the baht?"

Indeed, hedge funds can play a critical role in the market by providing liquidity. Investment strategies abound in the hedge fund universe. Relative-value funds arbitrage price differences between securities; macro funds take bets on the direction of the market; global funds assess the prospects for individual companies. Often, different hedge funds end up taking opposing positions in a market.¹² Hedge funds are also likely to be contrarian investors; that, too, helps ensure that there are enough buyers and sellers. Although empirical evidence on this point is not available, several academic studies tend to confirm it.¹³ The idea makes sense given the incentives of hedge fund managers, who are rewarded not for matching an industry benchmark but for their total returns—an approach that encourages them to find overlooked market opportunities rather than follow the herd. It is the enormous flexibility of the managers of hedge funds in choosing their investment strategies, and the restrictions the funds place on customer deposits and withdrawals, that give them the freedom to be contrarian.¹⁴

In addition, hedge funds can play an important role in promoting price efficiency, as we learned from interviews with a variety of capital market participants. Relative-value hedge funds, which earn profits by correcting anomalies in the relative prices of assets, help ensure that risk is priced more similarly and accurately throughout the world. Macro hedge funds promote price efficiency by correcting the price level within a market when it is out of equilibrium. In developed capital markets, the activities of hedge funds (and the nearly identical activities of proprietary traders at banks) have greatly increased the efficiency of pricing, as evidenced by the decline in arbitrage opportunities. Thus, hedge funds and proprietary traders generally play a positive role in soundly functioning capital markets by ensuring that prices accurately reflect market fundamentals and that risks are priced uniformly.

It is too little, rather than too much, reliance on capital markets that has landed countries in trouble. Instead of focusing on hedge funds' activities, policy makers should turn their attention to improving disclosure, transparency, and, possibly, investor protection.

Around the world, capital markets are replacing banks as the main capital intermediaries. Today, the countries with the deepest capital markets, such as Switzerland, the United Kingdom, and the United States, are furthest along in this transition. Other countries lag behind because of a lack of economic development (in the case of emerging markets) or a failure to adapt their infrastructure, legislation, and regulatory regimes (as in Japan and much of Europe, for example). Nonetheless, most countries are embarking upon the transition, however slowly. As it progresses, bank intermediation will become less important.

This shift to capital markets has important benefits for individual countries and for the world. It will increase the stability of financial markets as more financial instruments begin to make adjustments through price rather than quantity. Risks will be disaggregated and spread across a larger investor base, cutting the cost of capital for issuers. And competition among intermediaries will continue to intensify, improving the efficiency of intermediation.

In the Halifax initiatives following the 1994-95 Mexican crisis and in the 1999 finance ministers' report "Strengthening the International Financial Architecture," the G-7 countries have promoted more open and transparent markets and stronger national financial systems. But some governments and policy makers now speak of slowing down, rather than speeding up, this transition. Talk of imposing capital controls and strictly regulating hedge funds abounds. Countries hit by the crisis are busy bailing out (and in some cases restructuring) weak domestic banking institutions. Policy makers and international institutions have focused on improving domestic banking systems, but though these actions are needed and appropriate, they are not a comprehensive solution.

Two questions must be added to the agenda: how to promote the development of domestic capital markets and how to encourage the move from banks to capital markets. While many countries opened fledgling markets to foreign investment in the 1990s, subsequent crises showed that their underlying infrastructure was woefully underdeveloped. Private firms and government institutions in these countries should focus on strengthening the "intangible'' structure of modern markets by adopting emerging global standards for accounting and reporting, which in the case of government bodies, such as the treasury, would include reporting on the real level of usable reserves. The adoption of such standards would also entail more rigorous market supervision and regulation, the protection of majority and minority shareholders, and the protection of creditors, even in the event of a borrower's insolvency. Elsewhere, measures of this sort have proved efficient and effective.

A role remains for foreign banks, which can introduce more rigorous credit assessment and loan-monitoring practices to emerging markets, where poor skills and low standards are the norm. More sustained foreign direct investment in companies would also improve the supervision of management. Both reforms would ultimately diminish the level of volatility.