Bubble Contagion: Mexico City to Tokyo to Bangkok to New York, London, and Reykjavik
Four waves of credit bubbles in 30 years plus a bubble in US stocks in the late 1990s are unique in financial history. Either this succession of waves in a relatively short period was a coincidence, or there was a systematic relationship among several of them. Obviously the first wave – which involved the rapid growth in bank loans to governments and government-owned firms in Mexico and ten other developing countries – was sui generis. Was there a connection between the implosion of this wave and the credit bubble that was centered on Japan in the last half of the 1980s? Similarly was there a connection between the implosion of the asset prices in Tokyo at the beginning of the 1990s and the third wave of bubbles, which involved Thailand and its neighbors in Southeast Asia in the mid-1990s as well as Mexico, Russia, Brazil, and Argentina? Was there a link between the Asian Financial Crisis that began in mid-1997 and the bubble in US stocks in the late 1990s? Finally, was the wave of bubbles in real estate in the United States, Britain, Ireland, Spain, and Iceland between 2002 and 2007 – and in the debt of the governments of Greece and Portugal and Spain in 2008 and 2009 – related to these earlier events?
Asset price bubbles in major industrial countries are infrequent; the previous US stock price bubble had been in the late 1920s. Japan had never previously had an asset price bubble and neither had the other Asian countries. A bubble in three, four or more countries at the same time is extraordinary and suggests a common cause – a Minsky-type displacement or shock, and probably an external one because so many countries were similarly impacted at the same time. Four waves of bubbles in 30 years suggest that the reversal in the direction of cross-border money flows that follows the implosion of one bubble may contribute to the next wave.
Most credit bubbles led to asset price bubbles; every real estate bubble has resulted from the rapid increase in the supply of credit. The credit bubble of the 1970s financed the deficits of the governments and of government-owned firms in Mexico and other developing countries when their GDP growth rates accelerated because of the sharp increase in commodity prices. The bubble in stocks and property in Japan in the second half of the 1980s resulted from the rapid increase in the supply of credit for real estate purchases. Similarly the bubbles in real estate and stock prices in Thailand, Malaysia, and the other Asian countries as well as in Mexico and other Latin American countries in the first half of the 1990s were responses to money inflows and domestic credit expansion. The increases in the prices of residential real estate in the United States, Britain, and other countries after 2002 resulted from a massive increase in the supply of credit that partly followed from the securitization of mortgage loans. The bubbles in the bonds of the governments of Greece and some of its neighbors in the south of Europe followed from a significant increase in money flows to these countries.
One of the three essential components for the development of a credit bubble is a large pool of money that can be accessed to provide loans. A second component is a shock that leads to sharp increases in the anticipated rates of return – or to a significant reduction in risk – on loans to a particular group of borrowers; the shock may involve a reduction in regulation that makes it easier for borrowers to access credit in foreign markets. A third component is a group of investors who are willing to extend more credit to this group of borrowers.
The 1970s surge in bank loans to Mexico et al.
The exceptional feature of the 1970s was the surge in the US and world inflation rates during peacetime; the two previous episodes of sharply rising prices in the twentieth century were during and after the First and Second World Wars. Payments imbalances surged in the late 1960s and early 1970s, which led to a massive increase in money supplies. The 1970s inflation followed from the breakdown in the Bretton Woods arrangement of parities for national currencies, which eliminated the national ‘anchors’ for monetary policies.
One dramatic impact of the increase in the US inflation rate that began in the late 1960s was that interest rates on both short-term and long-term US dollar securities increased. The interest rates that banks in the United States could pay on their deposits bumped into the ceilings set by the Federal Reserve on US dollar deposits. The result was a large shift of US dollar-denominated deposits to offshore banks (including the branches of US banks) in London and other financial centers that were not subject to these ceilings.
The surge in demand associated with rapid growth in money supplies led to sharp increases in the prices of primary products. There were were two massive oil-price shocks and the payments surpluses of some of the oil-exporting countries surged, and their demand for US dollar securities increased rapidly. The increases in the GDP growth rates in the commodity-producing countries meant that their governments were more attractive borrowers.
British, Canadian, Japanese, and other foreign banks sourced the money for dollar-denominated loans to Mexico and other developing countries from the banks in these offshore centers. Previously these borrowers had relied on US banks and on institutions like the World Bank for most of their external financing; the number of lenders willing and able to increase their loans to these borrowers increased dramatically. These non-US banks were able to increase their assets much more rapidly than they could in their domestic economies. The US banks responded to the challenge of foreign banks entering ‘their turf’ by cutting price – reducing interest rates – to minimize the decline in their market share.
The borrowers in these developing countries were eager for the money, and the lenders believed that the credit risk was modest; a wise and respected New York banker said ‘countries don’t go bankrupt’. As long as the money available to the borrowers on new loans was several times larger than the interest payments on the old loans, these borrowers encountered no burden in servicing their external indebtedness. The chatter was that the major international banks were involved in ‘petro-dollar recycling’, which would have enabled the oil-importing countries to finance much larger trade deficits. The surpluses of the oil-exporting countries were recycled, but the increase in the indebtedness of oil-importing countries was much larger than the increase in the holdings of international reserve assets of the oil exporters. The increase in the flow of money to these developing countries enabled them to finance larger trade and current account deficits, which meant that the industrial countries had significantly larger current account surpluses.
The external indebtedness of Mexico and other developing countries increased by 20 percent a year for a decade, while the interest rate on these loans averaged 8 percent a year, although these rates trended upward in the 1970s as the world inflation rate increased. The cash flows from new loans were significantly greater than the interest payments on the outstanding loans, which meant that the currencies of these countries appreciated in real terms, and their trade deficits increased. As long as debt was increasing much more rapidly than income, there was a bubble in the pattern of cash flows, since it was inevitable that at some stage, the lenders would become more cautious, indebtedness then would increase less rapidly, and the currencies of these countries would depreciate.
The bubble in the credit flows to the developing countries was punctured by the sharply contractive monetary policy that the Federal Reserve adopted in the autumn of 1979. The ability of these countries to finance trade and current account deficits declined sharply – which meant that the trade surpluses of the industrial countries declined. Moreover there was a sharp flow of money to the United States from other industrial countries, which meant that their currencies depreciated and their trade surpluses surged as the US trade deficit increased.
The European currencies and the Japanese yen depreciated until 1985 and the US trade deficit increased until 1987. The mirror of the increase in the US trade deficit until 1987 was that the trade surpluses of many countries including Japan increased.
The 1980s asset price bubble in Japan
The rapid growth in the supplies of money and credit in Japan in the second half of the 1980s resulted from the reluctance of the Japanese authorities to have the yen appreciate because the country’s trade surplus would decline, with negative impacts on profits and employment in the export industries. The Bank of Japan bought US dollars to resist the upward pressure on the yen, which led to a rapid increase in the reserves of the Japanese banks, which then were able to increase their loans at a rapid rate. Regulations on bank loans for real investment purchases and construction were relaxed in the belief that increased spending on real estate would contribute to economic growth and lead to increases in imports and dampen the tendency toward a stronger yen. Regulations that limited the foreign investments by Japanese firms and banks were relaxed in the belief that the larger money outflow also would dampen upward pressure on the currency.
The 1980s real estate bubble in Japan was so massive that by the end of the decade the chatter in Tokyo was that the market value of the land under the Imperial Palace was greater than the market value of all of the real estate in California. The land area in California is several billion times larger than the grounds of the Imperial Palace, which meant that the price per hectare was a million times higher than the price in California. Not that there had been an auction or even a pseudo-auction for the Imperial Palace grounds. The analyst who made this comparison estimated the value of the palace grounds by multiplying the hundred or so hectares by the recent price per hectare paid for a small plot of land in the nearby Ginza entertainment neighborhood, which was much the most valuable land in Tokyo. The value of Californian real estate was obtained from Federal Reserve data on US household wealth.
All of the financial values in Tokyo were sky-high at the end of the 1980s. The market value of Japanese stocks was twice the market value of US stocks, even though Japanese GDP was less than half of US GDP. The comparison between Japanese and US firms in terms of the ratios of the market value of stocks to profitability was even more skewed. The market value of Japanese real estate was twice the market value of US real estate, even though the land area in Japan is 5 percent that in the United States and 80 percent of Japan is mountainous. The market value of land per capita in Japan was more than four times that in the United States while per capita income in Japan was 60 or 70 percent that of the US.
The Japanese banks were at the top of the hit parades of the world’s banks as measured by assets and by deposits (but not by profits); usually seven of the ten largest banks on this list were Japanese. Similarly the capital of Nomura, Japan’s largest investment bank, was larger than the capital of the five largest US investment banks.
As we mentioned in the first chapter, the Mitsui Real Estate Company paid $625 million for the Exxon building on Sixth Avenue in New York City even though the asking price was $510 million because the company wanted to get into the Guinness Book of World Records. Other Japanese firms were also acquiring trophy properties and buildings in the United States. Mitsubishi Real Estate and a group related to Sumitomo Bank bought the Pebble Beach Golf Course in Northern California. Sony bought Columbia Records and then Columbia Pictures, and Matsushita, its dominant rival in the electronics industry, acquired MGM Universal.
By the 1980s Japan was the second leading industrial power, more economically powerful than Germany. Toyota, Nissan, and Honda were leaders in the global automobile industry. Sony, Matsushita, and Sharp and a seemingly endless list of firms dominated the global electronics industry. Nikon and Canon ‘owned’ the world’s photo-optics industry.
The mandarins in the Ministry of Finance maintained low interest-rate ceilings on both bank deposits and bank lending rates; the interest rates on deposits were below the inflation rate so households had to save a high proportion of their incomes or else their wealth would have declined. The demand for loans from business firms at these low interest rates was much greater than the supply; government officials provided ‘window guidance’ to the banks identifying the firms that were to be given preference. The real rates of return on both real estate and stocks were positive and high.
Japanese banks owned large amounts of real estate and stocks; as the prices of these assets increased, their capital increased. They were then able to increase their loans, and some of these loans went to borrowers who used the money to buy real estate. In the early 1980s the banks established a new set of intermediaries, ‘jusen’, to make housing loans, which traditionally Japanese banks had been reluctant to make. The jusen would get the funds for housing loans by borrowing from the banks – in effect seven of the large banks established these specialized lenders. About the same time the Ministry of Finance established public-sector lending institutions that would also make housing loans. The banks then decided to make housing loans on their own.
In the first half of the 1980s Japan began liberalize its financial regulations, in part because of pressure from the Americans to ‘open up’ the markets in Tokyo so US firms would have access to clients and customers and trading opportunities in Japan comparable to those that Japanese firms had in New York. Interest rate ceilings on deposits and on loans were raised. Window guidance became much less extensive. Moreover the restrictions on the foreign investments of Japanese firms were relaxed and Japanese banks were permitted to increase their foreign branches and subsidiaries.
Financial liberalization meant that the banks could increase their loans to borrowers who wanted to buy real estate and to build new office buildings and apartment buildings and shopping centers. Real estate prices increased. Firms involved in the real estate business accounted for a significant proportion of the market value of all of the firms listed on the Tokyo Stock Exchange. These real estate holding companies were somewhat like mutual funds; when the prices of the properties they owned increased, investors rushed to buy more of their shares, and so the share prices of the real estate companies increased. Increases in real estate prices led to a construction boom as new skyscrapers were constructed. The Japanese banks owned large amounts of real estate and of stocks, and the contribution of the increase in the value of both real estate and stocks to the capital of the banks was much greater than their operating profits. As their capital increased, the banks were able to increase their loans.
Japan had developed the financial equivalent of a ‘perpetual motion machine’. The increases in real estate prices led to increases in stock prices; the increases in both real estate prices and stock prices led to increases in bank capital. As bank capital increased, banks were in a position to increase their loans, and because of financial liberalization, they were much better positioned to increase their loans to groups that had been restricted in their ability to borrow in earlier periods. Because most bank loans were collateralized by real estate, bank loan losses were trivially small as long as real estate prices increased. The profits of the firms that invested in real estate were increasing, and so many of these firms borrowed more in their search for larger profits.
The real estate loans were collateralized by property; traditionally banks would lend up to 70 percent of the appraised value of these properties. Some of the loans made by the jusen went to ‘yakuza’, organized criminal groups, who used their connections to secure exceptionally high property appraisals. Since property prices were increasing by 30 percent a year, a modest ‘error’ made by an appraiser in putting too high a value on the property would soon be corrected.
Real estate prices increased much more rapidly than rents, with the consequence that the rental rate of return declined significantly below the interest rate on the borrowed funds. Investors who had bought properties in the last several years of the 1980s had a negative cash flow – the rental income on their properties after the payment of the operating costs was below the interest payments due to the lenders – but because property prices were increasing so fast, they could raise cash to make the interest payments by increasing the amounts borrowed against properties that had been purchased in earlier years.
The rapid increases in real estate prices and stock prices led to a surge in household wealth – Japanese households owned currency and bank deposits, real estate, and stocks. While many of the stocks in Japanese firms were owned by other firms in the same ‘Keiretsu’ (successors to the Zaibatsu), one-third of the stocks were owned by individuals. In effect each firm was a combination of an operating company and a mutual fund.
As the market value of Japanese stocks increased, investors resident in the United States and Western Europe bought more Japanese stocks. Global stock index funds wanted to own more Japanese shares. The rates of return to non-Japanese investors on their purchases of shares in Japanese firms were high since these investors benefited from the combination of the increase in the price of the stocks and the increase in the foreign exchange value of the Japanese yen.
The bubble in Japan reached its crescendo at the end of 1989. Real estate prices seemed so high that the quip by the much quoted baseball star Yogi Berra that ‘It’s so expensive that no one can afford to live there’ seemed applicable. Banks developed one-hundred-year, three-generation mortgages. The incoming governor of the Bank of Japan was concerned that such high prices for homes would erode social harmony. A new central bank regulation instructed Japanese banks to limit the rate of growth of their real estate loans so that it would be no greater than the rate of growth of their total loans.
Once the rate of growth of bank loans slowed, some recent buyers of real estate developed a cash bind; their rental income was smaller than the interest payments on their mortgages, but they could no longer obtain the cash needed to pay the interest on their outstanding loans from new bank loans. Some of these investors then became distress sellers of their properties. The combination of the sharp reduction in the rate of growth of credit for real estate and these distress sales caused real estate prices to decline; the cliché that the price of land always rises was belied.
Stock prices began to decline at the beginning of 1990; stock prices declined by 30 percent in 1990 and 30 percent again in 1991. The stock price trend in Japan was downward although there were four significant rallies. At the beginning of 2010, stock prices in Japan were in the same ballpark that they were thirty years earlier.
Now the perpetual motion machine began to work in reverse. Property sales led to declines in property prices. The decline in real estate prices and stock prices meant that bank capital was declining; banks were now much more constrained in making loans. Because the value of Japanese stocks was declining while that of US stocks was increasing, the global stock equity funds sold Japanese stocks and bought US stocks.
One of the stylized facts in monetary economics is that the implosion of an asset price bubble is deflationary, the flip-side of the economic boom that occurred during the expansion phase of the cycle. Investment spending in Japan declined in part because the cost of capital had surged and the anticipated growth of profits had been revised downward, and in part because the splurge in investment spending during the expansion phase had resulted in significant excess capacity. Household spending increased much more slowly as millions of families increased their saving from their earned incomes to compensate for the decline in their wealth that followed from the fall in stock prices and real estate prices.
The recession in Japan in 1991 meant that import growth slowed markedly while exports surged; some Japanese firms greatly increased their efforts to sell abroad because the domestic market for their products was growing slowly relative to the growth in their supply capabilities. The result of the slowdown in the growth of imports and the surge in the growth of exports was that Japan’s trade surplus increased. This increase was larger than the increase in the capital flow from Japan and led to the appreciation of the yen, which became a handicap to export-oriented Japanese firms. Many increased their investments in China, Malaysia, and Thailand to take advantage of lower labor costs. The increase in investment spending for firms in the export industries stimulated income growth.
The East Asian economic miracle and the bubble in emerging-market equities
Four innovations contributed to the surge in property prices and real estate prices in Mexico, Thailand, and other emerging-market countries in the early 1990s. The funding of the overhang of bank loans into Brady bonds, which effectively removed these countries from their bankruptcy – and they were re-christened as emerging markets. The discovery of ‘emerging market equities as a new asset class’, index-based mutual funds and pension funds began to acquire these stocks. The expectation was that these countries would industrialize at a rapid rate, and corporate profits would surge. Privatization of government-owned firms in resource extraction, communications, manufacturing, and other industries led to a surge in money inflows. The appreciation of the Japanese yen following the implosion of its bubble led to the ‘hollowing out’ of the Japanese economy and led firms to seek lower-cost production sites in Thailand, Malaysia, Indonesia, and their neighbors.
In 1992 the World Bank published The East Asian Miracle, an expressively descriptive title for the economic performance of the countries in the arc from Thailand to South Korea; the increases in their GDPs were in some ways comparable to the gains that Japan had made in the 1950s and the 1960s. The Korean peninsula had been fractured in the war in the early 1950s; in the mid-1960s South Korea began a remarkable period of economic growth. Singapore had been a fortified swamp in the 1950s but by the 1990s had achieved a first-world standard of living. The change in political leadership in China from Mao Zedong to Deng Xiaoping in 1978 led to a dramatic change from a self-contained, isolated country to one that was open and eager for international trade and international investment; the annual rate of growth averaged nearly 10 percent for more than twenty years and the increase was even more dramatic in the provinces that bordered the seacoast and in the major cities such as Beijing, Shanghai, and Shenzen. Hong Kong had evolved from an outpost for peeking into China in the 1950s, the 1960s, and the 1970s into an entrepôt center for prepping Chinese goods for world markets.
Thus the expansion of the asset price bubbles in the Asian capitals followed from the implosion of the asset price bubble in Tokyo and the surge in the flow of money from Japan at the beginning of the 1990s. Most of the money was Japanese-owned; some had been owned by foreigners who sold Japanese stocks as their prices declined. The flow of money from Tokyo to Thailand and Indonesia and the other Asian countries led to increases in the foreign exchange value of their currencies if they were floating and to increases in the international reserve assets and the money supplies if these currencies were pegged. The rates of growth of domestic income in these countries increased in response to the increases in investment spending and in consumption spending that followed from the increases in the prices of real estate and stocks. The residents in these countries who sold their securities and assets to the Japanese used most of their receipts to buy other domestic securities and real estate.
Mexico was being prepped to join the North American Free Trade Agreement with the United States and Canada. The Salinas government adopted three sets of measures. Several hundred government-owned firms were privatized. The central bank adopted a severely contractive monetary policy to reduce the inflation rate, which had been above one hundred percent; as a result, the real interest rates on Mexican securities were extraordinarily high. Finally, many restraints on competitive business practices were relaxed or removed. The prospect was that Mexico would become a low-cost source of supply for the US and the Canadian markets. There was a massive surge in money flows to Mexico, partly in response to the high real interest rates and partly in response to the opportunity to buy some of the newly privatized firms.
Many of the once-developing countries were re-labeled emerging-market countries in the early 1990s after the overhang of past-due bank loans were funded into long-term bonds under the Brady plan. Some bright investment banker came up with the very profitable idea that emerging market equities were a new ‘asset class’. Every pension fund and every mutual fund that was following an index fund approach toward developing a global portfolio believed it had to acquire stocks that were available in these countries; the sales spiel was that the rates of return would be higher than the rates of return on equities available in the industrial countries because their rates of economic growth were higher. And moreover the pattern of price changes was not correlated with those of equities of the firms headquartered in industrial countries, so their inclusion would reduce the risk of the global portfolios.
In Thailand, Malaysia, and Indonesia in the first half of the 1990s stock prices increased by between 300 and 500 percent and manufacturing activity surged. Stock prices doubled in most of the East Asian countries in 1993 and continued to increase in 1994. Real estate prices soared. The economies boomed. Trade deficits grew. Since the asset price bubbles across these countries were pervasive despite marked differences in economic structures, per capita incomes, exchange rate arrangements, and whether they were international creditor countries like Singapore, Taiwan, and Hong Kong or international debtors like Thailand and Malaysia, there is a strong presumption that the bubbles had a common origin and that it was external.
China, Thailand, and the other East Asian countries were on the receiving end of outsourcing by American, Japanese, and European firms that wanted cheaper sources of supply for established domestic markets. Rapid economic growth was both the result and the cause of the inflow of foreign money, especially from Japan. Japanese investment initially took the form of the construction of manufacturing plants to take advantage of lower labor costs; high valued-added components would be produced in Japan and shipped to the affiliated plants for assembly. From there a large part of the production would be exported, some to the United States, some to Japan and some to third countries. The direct foreign investment by Japanese firms pulled supplier firms and the banks from Japan. The buzzword was export-led growth, which was almost always based on a low value for the country’s currency. Many of these exports were produced by major firms headquartered in the United States, Japan, and Taiwan. Firms headquartered in South Korea began to invest in China and Indonesia because their wage rates were so much lower.
The surge in money flows to Mexico in the early 1990s led to the real appreciation of the peso, and the Mexican trade deficit increased to 6 percent of the country’s GDP. External debt was increasing much more rapidly than GDP. Several political incidents associated with the presidential election in 1994 led to a sharp decline in money flows to Mexico; the Bank of Mexico supported the peso in the currency market by spending its own reserves, and eventually its ability to continue to support the peso was exhausted, and the peso depreciated sharply.
For a while money flows to other countries in Latin America diminished. Then in the winter of 1996 the consumer finance companies in Thailand – many of which had been established by the large Thai banks to circumvent the regulations that limited their ability to make consumer loans – began to experience large losses on their loans. Foreign lenders to Thailand became increasingly concerned about the value of their loans to Thai borrowers, and the flow of money declined. The Bank of Thailand’s ability to support the baht at its existing value was quickly exhausted, and in early July 1997 the baht depreciated sharply.
The depreciation of the baht triggered the contagion effect and within six months the values of each of the currencies on the Asian arc, with the exception of the Chinese yuan and the Hong Kong dollar, had lost 30 percent or more of their values. Stock prices declined by 30 to 60 percent, partly because foreign investors were seeking to cash out, partly because the domestic firms were no longer profitable. Real estate prices declined sharply. Most banks, with the exception of those in Singapore and Hong Kong, failed. The closing of many banks in Indonesia triggered racial strife, and an immense run on the currency which lost more than 70 percent of its value.
When the crises occurred, the play script was a reprise of similar events in Japan in the previous decade. The chatter about the East Asian miracle disappeared and new buzzwords arose – crony capitalism, spontaneous privatization, and destabilizing speculation.
The sharp depreciation of the currencies led to significant losses for those firms that had borrowed US dollars or Japanese yen or some other foreign currency. The banks that had lent to these firms also incurred losses; the banks in most of the Asian countries toppled into bankruptcy because of their own revaluation losses and the losses of the firms to which they had made loans.
The depreciation also resulted in a very quick reversal of the Asian countries’ trade balances, from large deficits to large surpluses. The counterpart of the large swings in the trade and current account balances of the Asian countries was the mirror-image increase in the US trade deficit.
Irrational exuberance and US stock prices in the dot.com boom
Between 1982 and 1999 US stock prices increased by a factor of thirteen, the most remarkable run of annual increases in stock prices in the two hundred years of the American republic. In the very long run, US stock prices have declined every third year; in the last two decades of the last century, stock prices fell in only one year, and then only by 5 percent. The market value of US stocks increased from 60 percent of US GDP in 1982 – an exceptionally low ratio – to 300 percent of GDP in 1999, an extraordinarily high ratio.
The increase in US real estate prices during this period was modest for the country as a whole, although there were significant regional increases in areas that were experiencing large increases in per capita income or in the number of employed individuals, including Silicon Valley and the broader San Francisco Bay area, Washington, DC, and Boston and New York.
The US economy boomed in the 1990s. The inflation rate declined from above 6 percent at the beginning of the 1990s to less than 2 percent at the end of the 1990s, the unemployment rate declined from 8 percent to less than 4 percent, the rate of economic growth increased from about 2.5 percent to 3.5 percent, and there was a remarkable increase in US productivity. The US Treasury’s annual fiscal balance changed by more than 5 percent – from a deficit of nearly $300 billion at the beginning of the 1990s to a surplus of nearly $200 billion at the end of the decade.
One of the ‘negatives’ in terms of US economic performance was that the annual US trade deficit surged to $500 billion. Another was that the household saving rate declined to a new low.
The remarkable aspect of the boom was the focus on the ‘new economy’ and especially the role of information technology, the computer, dot.coms, and the firms that provided both the hardware and the software or exploited these developments to serve traditional needs. These technological developments led to sharp declines in the cost of sending and storing information. Ebay provided a nationwide auction market for tens of thousands of different products. Amazon developed the technology for the sale of books and electronic products. Peapod allowed individuals to shop for most of their groceries at home. Millions of accounts were established at the discount broker Charles Schwab and at its competitors. Firms were established that enabled investors to trade stocks using the computer at extremely low transaction costs. ‘Day traders’ emerged: individuals who quit their regular jobs to trade stocks either from their computers at home or from desks in specially designed shops. Priceline enabled airlines and hotels to sell seats and rooms at sharply discounted prices.
Entrepreneurs were able to get the cash to develop these ideas from venture capitalists (VCs) who provided seed money. The VCs developed a portfolio of investments in different firms in the hope that within three to five years they would be able to sell these shares – and make their profit – when the firms made their first public offerings of stock. The VCs’ rates of return would depend on whether the firms in which they had invested were successful in their technological challenges and on the selling prices of their shares and the lengths of the holding periods.
As stock prices increased, the high rates of return earned by the VCs attracted lots of money, and the capital available to the VCs as a group surged by a factor of five. No investors wanted to be left behind. The money was there, so a large number of ideas were funded; money was chasing ideas and concepts. Three or four years on, the new firms would have an Initial Public Offering (IPO). The IPO would follow the traditional road-show, the occasion on which the investment banks would parade the company to asset managers around the country seeking to induce them to buy its shares.
At the end of the road-show, the investment bankers would estimate the amount of the shares that they might sell at the IPO, and set both the price and the quantity. In 99.46 percent of the cases, the share price at the end of the first day of trading was significantly higher than the IPO price, and those fortunate enough to be able to buy at the initial offering price would make a significant capital gain.
One impact of this price pop was that more and more investors clamored to buy at the IPO price. The second impact was that the demonstration of the strength of the demand meant that more and more people wanted to get a piece of the action – the entrepreneurs were attracted to the immense wealth they might earn with a successful innovation, the VCs were attracted to the large profits they could gain by identifying the entrepreneurs that were likely to be successful, and the investment bankers wanted the fees from bringing a large number of firms to the public. The investors wanted the large capital gains associated with the pop between the IPO price and the price of the same shares at the end of the first day of trading, the first week of trading, and the first month of trading.
The size of the price pop was like dynamite or nitroglycerine or maybe even a nuclear explosion. The investment bankers appeared to set the price for the IPOs so as to maximize the price pop on the first day of trading – and not maximize the cash received by the shareholders who were selling. From this point of view, a lower price for the IPO might be preferable to a higher price, for the demand for the stock would increase – at least for a while – as the pop increased. The entrepreneurs sold only a small part of their total shares at the IPO; they calculated that the larger the pop, the greater their wealth. They were more interested in the apparent value of the shares they owned at the end of the first day’s trading than they were with the amount of cash they might obtain from the IPO.
On some IPO days the number of shares traded was three or four times the number of shares that had been sold at the IPO. Since many of the buyers at the IPO had been told to hold their shares, the float was much smaller than the number of shares sold, and so these shares in the float might have been traded five or six times in the course of the day.
The United States seemed to have its own perpetual motion machine, one designed to enrich the fortunes of hundreds of thousands of families. The larger the price pop on the first day of trading, the greater the number of investors that were attracted to IPOs. The stronger the demand for IPOs, the larger the number of venture capitalists that were willing to back the entrepreneurs. The more capital that the entrepreneurs were willing to put in play, the larger the number of entrepreneurs that would seek their fortunes by breaking away from established firms.
In December 1996 Chairman Greenspan of the Federal Reserve Board first used the term ‘irrational exuberance’; the Dow Jones was at 6400 and change and the NASDAQ was at 1300. Greenspan was cautious and careful with data; it seems highly unlikely that he would have commented on stock prices unless he believed they were then over-valued by a minimum of 15 or 20 percent. At the end of December 1999 the Dow Jones was at 11700 and the NASDAQ was at 5400, and the market value of NASDAQ stocks was 80 percent of the market value of stocks traded on the New York Stock Exchange.
In the late 1990s the prices of stocks representing the new economy – the dot.coms, e-commerce, fiber optics, servers, chips, software, IT, telecoms – which were traded on the NASDAQ had increased much more rapidly than the prices of old-economy stocks, those of firms such as GE and GM, AT&T and Time-Life, that were traded on the New York Stock Exchange. But there was more than a spillover effect, for the enthusiasm about the future that was characterizing the new economy stocks was infectious and also led to increases in the prices of the old-economy stocks.
It seemed as though developments in information technology were driving finance. Computers were becoming much more powerful and less expensive. The costs of transmitting and storing information and data were declining rapidly. Moore’s Law came into play, and the cost of a unit of computing power declined by 30 percent a year. The World Wide Web was developing fast, and markets in separate centers were becoming linked. Computers were replacing humans in trading stocks. Individuals could order their airline tickets over the web. Fiber optics were linking the US east and west coasts and the prices of long-distance phone calls were declining to the level of local calls. Servers were big, and so was storage capacity. Thousands – tens of thousands – of new firms had been established to move information or data or to store information or data, and the profits earned by the venture capitalists that had funded them were so high that more money flooded in from pension funds, university and charitable endowments, and wealthy families. The demand for newly issued shares at the IPO prices was much larger than the supply, which led the investment bankers to engage in the process of ‘spinning’ – of allocating a goodly number of shares to the heads of the major firms that would bring them investment banking business. The investment banks grew rich; they had a lot of product to sell and a public convinced that share ownership would bring them large profits.
There is no easy answer to the question of when rational exuberance morphed into irrational exuberance. The idea that there might be an asset price bubble in US stocks occurred to different investors at different dates. The first date for the onset of the bubble in US stocks is the spring of 1995, eighteen or twenty months before Greenspan’s ‘irrational exuberance’ comment. Stock prices had increased at an annual rate of 34 percent in 1995, and at 25 percent a year in the first eleven months of 1996.
The surge in stock prices in 1995 and 1996 can be attributed to two different aspects of the Mexican financial crisis of 1994; one was direct and one was indirect. The direct effect was the sharp depreciation of the Mexican peso which resulted in a sudden shift in the Mexican trade balance from a deficit of $20 billion in 1994 to a surplus of $7 billion in the next year; the mirror image was that the US trade deficit increased by about $25 billion, since the United States was much the largest trading partner for Mexico. The counterpart of this change in the Mexican trade balance was that there was a money flow to the United States (a modest preview of the events that would occur in 1997 following the Asian Financial Crisis and the massive turnaround in the trade balances of the Asian countries). In effect the flow of money from Mexico to the United States led to an increase in the prices of US securities. The second aspect was that the Federal Reserve eased its monetary policy and reversed its tightening policy of 1994.
An alternative starting date for the onset of the bubble is the summer of 1998, following the Asian Financial Crisis, the financial debacle in Moscow and the collapse of Long-Term Capital Management. The sharp depreciation of the Asian currencies led to an increase in the US trade deficit of more than $150 billion. Moreover, the Federal Reserve again eased policy, partly because of concern with the fragility of the monetary arrangements following the crisis in Long-Term Capital Management, until then the most professional and sophisticated of the many US hedge funds.
In the twelve months after the end of June 1998 the market value of the stocks traded on the New York Stock Exchange increased by 40 percent, from $9005 billion to $12,671 billion. The market value of NASDAQ stocks increased by 90 percent.
The surge in the flow of money to the United States had the same impact that the increase in money flows had on other currencies; the US dollar appreciated and the US trade deficit increased. The money from abroad went into the asset markets. Asset prices, especially stock prices, continued to increase, and brought about an increase in domestic investment and a dramatic decline in the domestic saving rate (which is identical with an increase in domestic consumption).
Those who moved funds to the United States then bought US dollar securities, which led to increases in their prices; US financial wealth surged. The Americans who sold some of the securities they owned to foreign investors then had to decide what to do with the money they received from the sale. They used most of the money to buy more securities from other Americans, but they also increased their purchases of US goods as their wealth objectives were achieved. The decline in the US saving rate and the increase in the US trade deficit were inevitable outcomes of the increase in the flow of money to the United States.
The data suggest that between 95 and 97 percent of the increase in household wealth that followed from the flow of savings from other countries to the United States was used to buy other US securities and only 3 to 5 percent was used to buy consumption goods. Yet to the extent each of the sellers of the securities spent some of the money receipts on consumption goods, the domestic saving rate declined.
The spending on consumption goods is like a ‘leakage’. The smaller the amount spent on consumption goods, the larger the amount spent to buy securities and real assets and hence the larger the increase in their prices.
During 1999 the Federal Reserve, the banks and the country at large became obsessed with the Y2K problem – a neurosis that the economy would break down because some computers would not be able to change the date when the year changed to 2000. Precautionary behavior by the Federal Reserve led to an expansion of bank liquidity. Once again banks increased their loans in response to the increase in liquidity.
The increase in stock prices attracted European investors, and the euro depreciated. Because of the decline in US import prices and the surge in the trade deficit, inflationary pressures declined in the United States. Consequently, the Federal Reserve felt no need to adopt more contractive monetary policies.
With the advent of the new millennium, the Fed began to withdraw liquidity. Stock prices began to decline. The aggregate decline in the stock market between 2000 and 2003 was 40 percent and the decline in the market value of stocks traded on the NASDAQ was 80 percent. The decline in stock prices meant that the costs of capital to the firms increased, and the decline in household wealth led to less rapid increases in household spending. Once again the implosion of a stock price bubble was followed by a recession, which was dated from the beginning of 2002. Because the bubble in asset prices had not been based on increases in credit, the subsequent decline in stock prices had no significant impact on the banks.
The global bubble in real estate
The remarkable feature of the first years of the third millennium is that there significant increases in real estate prices in the United States, Britain, Ireland, Spain, Iceland, South Africa, Australia, and New Zealand, which began in about 2002. The price increases in the United States were in sixteen states that account for about half of the country’s GDP. Most of these states were in the South and along both the Atlantic and Pacific coasts, and were among those that were growing most rapidly – but not all of the high growth states experienced significant increases in real estate prices. Spain and Ireland were members of the European Monetary Union, and hence did not have their own currencies.
Each of these countries had a current account deficit; each experienced an increase in money inflows and in most cases, their currencies appreciated. One exception was that the US dollar depreciated, and hence the increase in money flows to the United States dampened the appreciation of the euro that was underway after the implosion of the stock price bubble of the late 1990s.
The ratio of the increase in household wealth to GDP was much larger in Iceland than in any of the other countries that experienced sharp increases in asset prices. Moreover, the ratio of the increase in the flow of money to GDP was much higher for Iceland than for any other country. The unique feature of the bubble in Iceland was that the increase in stock prices was three to four times larger than the increase in real estate prices. The flow of money to Iceland during several years was more than 20 percent of the country’s GDP. The Icelandic krona appreciated by more than 30 percent. The balance of payments accounting identity is that the changes in the value of a country’s capital account balance and changes in the value of its current account balance were necessarily equal as long as currencies were floating. Thus the autonomous increase in the foreign demand for Icelandic securities led to increases in the value of the Icelandic krona and to increases in asset prices in Iceland.
After the implosion of the real estate bubble in the United States, and prices of homes and commercial real estate began to decline, and banks and other financial institutions began to collapse, the press chatter turned to the cause or causes of the bubble. The list is long and varied and includes excessively high leverage, the repeal of the Glass-Steagall Act, the corruption of the credit rating agencies, compensation based on increases in revenues rather than on increases in profits after adjustment for loan losses, ‘derivatives’, the opacity of the over-the-counter market in derivatives, the inflow of money from China, the proclivity of Chairman Greenspan to keep interest rates too low for too long, and so on.
This list is peculiarly US-centric, and ignores the bubbles in real estate markets in Britain, Ireland, Spain and Iceland, as well as in several other countries; the list also ignores the three previous waves of bubbles.
Sloshing money and asset bubbles
Just as water flows downhill, money flows across national borders toward higher anticipated rates of return. Both increases in stock prices in a country and increases in the value of its currency attract money from abroad. The surge in money flows to the United States following the Asian Financial Crisis illustrates how the change in the pattern of cross-border money flows is the link from one wave of an asset price bubbles to the next.
One factor that explains why there were four waves of bubbles in a thirty-year period is that there had been a surge in the pool of global money as a result of the large payments imbalances that developed in the late 1960s and continued into the 1970s and the next several decades. Moreover the floating currency arrangement meant that asset prices in a country would increase in response to an increase in money inflows. In addition the efforts of banks headquartered in different countries to grow their loans in foreign markets contributed to the increase in the supplies of credit available to borrowers in various countries. Each wave of bubbles begins when the lenders become much more willing to extend credit to a group of borrowers, perhaps because their incomes or anticipated incomes increased, or because the regulatory environment had become less restrictive and the lenders were no longer prohibited from extending credit to these borrowers. Alternatively, the change in the regulations enabled borrowers to reduce their anticipated costs of credit by tapping into foreign markets.
The second feature of the period since the early 1970s is that the differences in interest rates and in anticipated rates of return on similar securities denominated in different currencies have become much larger once countries were no longer committed to parities for their currencies. The divergences in national inflation rates were larger – although the departure from the Bretton Woods system of adjustable parities occurred because of differences between the United States, and Germany and several other European countries on the maximum acceptable inflation rate. Cross-border money flows occur when the differences in the anticipated rates of return on similar securities available in different countries does not correspond with the anticipated rates of change in the currency values. In the long run, these cross-border flows reduce the differences in anticipated returns in various countries. However, because there may be an almost immediate feedback from the movement of money across a border to an increase in the economic performance in the destination country, the movement of money may initiate a bubble – and then the extension of credit can become self-fulfilling, at least in the short run, because higher rates of growth led to higher rates of return.
The onset of a bubble can be illustrated by the case of Iceland in 2002. Rates of return on Icelandic kronor securities were high and attractive to foreign investors at a time when interest rates were relatively low in most industrial countries. The increase in the flow of money to Iceland led to both the appreciation of the kronor and an increase in the prices of kronor securities, the feedback from the money flow to the value of the currency was almost immediate because the market for the kronor was thin. Hence the money flow that led to an appreciation of the currency and an increase in the price of Icelandic securities was small relative to the global pool of money and also small relative to the capital of the investors that acquired Icelandic securities.
The currencies of most of the countries that experienced an asset-price bubble appreciated in response to an increase in money inflows; the major exception was that the euro and the Swiss franc and other currencies appreciated in 2002 and the next several years even as money flows to the United States increased. (The surge in stock prices in Tokyo in the second half of the 1980s attracted foreign money, but the Japanese purchases of foreign securities were larger than the money inflows and the yen appreciated as the Japanese trade surplus increased.) The increase in asset prices in each country as its currency appreciated was an integral part of the adjustment process to the money inflows to ensure that the change in the country’s trade deficit would correspond with the increase in money inflows.
The increases in the external indebtedness of the countries involved in the first, third, and fourth waves were more rapid than the increases in their GDPs; similarly the rates of increase in money inflows were much higher than the interest rates on the indebtedness. Hence it was inevitable that these currencies would depreciate when the money inflows would slow. Similarly it was inevitable that the growth of credit for the purchase of real estate in Japan would slow, and that some of the recent buyers of real estate would encounter cash flow problems because their interest payments would be larger than their rental incomes. The increase in the domestic indebtedness of those who bought real estate in Japan in the 1980s was much more rapid than the increase in their incomes, and it was inevitable that the lenders would become more cautious in the extension of credit.
The increases in the supplies of credit often have been so rapid that the less creditworthy borrowers become attractive to the lenders. Often new or recent entrants to the credit market may cut prices to increase market share, and then the larger, more established lenders may respond by reducing price – that is, the interest rate markups over the costs of funds.
The first wave of credit bubbles involved sharp increases in the loans from the international banks to governments and government-owned firms in Mexico and other developing countries at a time when the world inflation rate was accelerating. The demand inflation led to rapid increases in commodity prices and in the anticipated growth rates in the primary producing countries. Moreover the pool of global savings surged as a result of large payments imbalances, the rapid growth in offshore dollar deposits, and the large payments surpluses of Saudi Arabia, Kuwait, and several other oil-exporting countries.
The link between the implosion of the first wave of credit bubbles and the bubble in Japan was the extensive intervention by the Bank of Japan to limit the appreciation of the yen in the second half of the 1980s. The Japanese trade surplus had surged in the 1970s and the first half of the 1980s. The appreciation of the yen in the second half of the 1980s would have led to a decline in profit rates and employment in the tradable goods sector; the Bank of Japan bought dollars to dampen the appreciation of the yen. Moreover the restrictions that had limited bank credit for real estate, which had been in place for more than thirty years, were relaxed and the rapid increase in these loans led to a surge in real estate prices, a boom, and euphoria.
The appreciation of the yen that followed from the implosion of the bubble in Japan led to an increase in the anticipated returns on investments in Thailand and other emerging-market countries that would supplement and replace factories in Japan as sources of supply. The yen appreciated because the dramatic slowdown in the Japanese economy led to a decline in imports and a surge in exports as productive capacity was diverted from the domestic market to the foreign markets. Japanese firms began to invest extensively in China, Thailand, and their neighbors in response to the ‘hollowing out’ of the Japanese economy. About the same time, four regulatory changes contributed to the increase in the demand for securities available in these countries. The discovery of ‘emerging market equities as a new asset class’ followed from the remarkable increase in stock prices in Japan in the 1980s. Moreover the Brady bond initiative of 1988 funded the overhang of bank loans to the developing countries into long-term bonds, and these countries became ‘bankable’. The privatization of government-owned firms, both in the industrial countries and in the emerging-market countries, led to a flow of money to Mexico, Brazil, Argentina, and other countries to purchase these firms and shares in these firms. One factor was Mexico-specific and involved the policy initiatives to stabilize the economy to prepare it for entry into the North American Free Trade Agreement; an extremely tight monetary policy adopted to reduce the monthly inflation rate led to high real interest rates on short-term peso securities that attracted lots of money from the Unites States.
Banks headquartered in Thailand, Malaysia, South Korea, and Indonesia increased the amounts borrowed in the offshore money market as their creditworthiness increased; the interest rates in the offshore market were lower than those in their domestic market, although the offshore borrowing meant that the banks were incurring a currency risk, either directly or at one remove if they on-lent to domestic borrowers in the foreign currency.
When the bubbles in Thailand and Malaysia popped, their currencies depreciated sharply in the second half of 1997; there was a surge in the money flows to the United States as they repaid loans. The Federal Reserve reduced interest rates three times within several months to dampen the flow of money from the Asian countries – which contributed to the acceleration of the bubble in US stocks.
The implosion of the bubble in US stocks led to a reduction in the flow of money to the United States and many foreign currencies – the euro, the Swiss franc, the Japanese yen – appreciated, which led to a reduction in the US trade deficit and contributed to an increase in the profit rate in US manufacturing. At about the same time, an autonomous event – an increase in the Chinese trade surplus – led to increases in the US trade deficit and the trade deficits of other industrial countries. The money flows from China went to the countries with the most attractive security markets.
The pattern is that the implosion of a bubble leads to change in the cross-border money flows. The implosion of the credit bubble in Mexico and other developing countries in the early 1980s would have led to an appreciation of the currencies of the industrial countries; the desire of the Japanese to dampen the strengthening of the yen led to an increase in the money supply and the credit available for purchase of real estate. The increase in the flow of money from Japan in the early 1990s contributed to the increase in asset prices in various emerging market countries; similarly the flow of money from these countries contributed to the bubble in the United States.
Whereas the implosion of the bubble in Japan led to an appreciation of the yen and the increase in anticipated profit rates in manufacturing in Thailand and Malaysia, the implosion of these other three waves led to a depreciation of the currencies because money flows to these countries diminished.
One feature of the period is that there is a large – an increasingly large – volume of money that can move from one country to others – more precisely, from the dollar to some other currency – at relatively low cost. Another feature is a succession of shocks, including policy shocks like the shift from a pegged currency arrangement to a floating currency arrangement or the relaxation of regulations that limited bank loans for real estate in Japan or the relaxation on borrowing abroad by banks in Southeast Asia. The implosion of several of the bubbles induced a change in the pattern of cross-border money flows and contributed to the increase in asset prices in the countries that experienced an increase in these inflows.