Doomed to repeat it? 400 years later, the rules are little changed

Doomed to repeat it?
By Gillian Tett
Copyright The Financial Times Limited 2007
Published: August 27 2007 03:00 | Last updated: August 27 2007 03:00


A few years ago, Robert Bruner and Sean Carr, two American academ- ics at a Virginia business school, decided to write a comprehensive study of a financial market crash that took place on Wall Street back in 1907.

In normal, calmer times, their worthy opus might have attracted only limited attention. After all, the 1907 turmoil has never carried the fame of 1929. But in a happy burst of brilliant timing - and sheer luck - their work is due to be published this autumn, and the two authors are unexpectedly finding their insights in hot demand from the financial world. "We had no expectation that a crisis would sprout this summer," admits Professor Bruner. "[But] our re-search taught us that the drivers of crisis are always present [so] we thought the lessons of 1907 would be immensely relevant to investors, CEOs and regulators."

It is a sentiment Wall Street and the City of London seem to share. For as investors and financiers recoil in shock from this summer's violent market swings, and as a crisis in the subprime mortgage lending sector has triggered gyrations in stocks, many are now reaching for the history books with a newfound enthusiasm - or desperation - to assess how this crisis will play out. "Everyone is muttering about 1987, 1998 or 1929," says one senior hedge fund manager. "I don't know much about 1907, but probably I should."

From some perspectives, this sudden fascination with the past marks something of a U-turn. After all, the financial sector has spent much of this decade operating with a short-term view that was focused on the future, not the past. Indeed, as recently as this spring, it was rare to find any financial trader who spent much time pondering events more than a decade old - or beyond the data points typically found on a trading terminal.

That partly reflected the fact that financial traders are often too young to remember many economic cycles. However, more importantly, many of the instruments that have been in the eye of the recent market storm have only risen to prominence this decade. Thus the "historical" data bankers feed into their computer models to assess market swings, or measure their levels of risk-taking, is often based on just a few years of records. That can potentially distort the way these computer models work, since it means that bankers are effectively presuming that the future will be similar to the past - but based purely on very recent experience. "What is remarkable is that the risk models currently applied [in some markets] do not reflect the experience of the autumn of 1998, only a few years ago," says Harald Malmgrem, a Washington-based economist.

However, the other reason for the recent lack of interest in history is that many bankers have believed - at least until recently - that this decade's burst of market innovation had rewritten the rules of finance. For as financiers have created products that distribute credit risk across the capital markets, this has altered the way the financial system works. That in turn, may have changed the way the credit cycle works - or so some optimists believed until very recently.

However, this summer's market swings are now blowing apart many of these cosy assumptions. As a result, the indifference towards the past is being replaced by a violent thirst for historical knowledge, as financiers reacquaint themselves with the unpalatable truth that almost every bubble is accompanied by a belief that innovation has changed the rules - a belief that typically proves to be false. "This neo-modern credit market is not very dissimilar after all from its classical predecessors," says Jack Malvey, an analyst at Lehman Brothers. "The catalysts differ, but market reactions feel similar [to crises before] . . . in our view long-term economic and capital markets history is the best teacher and best model [to understanding the present]."

T he one big problemthat confronts men such as Mr Malvey, however, is that "long-term history" offers a truly dazzling array of models to peruse. Indeed, Lehman Brothers itself estimates there have been more than 60 market crashes since 1622. As a result, pundits disagree sharply about which particular crash might offer the "best" analogy for today, depending on whether they are optimistic about the macro-economic outlook or not.

One parallel provoking much discussion, for example, is the collapse of the internet bubble at the start of this decade. That seems similar to this year's events because the crash came after a bout of equity market exuberance that anticipated recent conditions in credit markets. Most notably, in the late 1990s, debt levels were rising in the system as equity investors threw caution to the winds amid a widespread belief that innovation had changed the investment rules. (Most notably, in 1999, the rise of the internet was perceived to have altered the business cycle.)

However, when equity investors suffered a sudden loss of confidence in overvalued technology stocks, they scrambled to slash debt, in a bout of deleveraging similar to that of this summer. "Deleveraging an overleveraged system is always dangerous," said Credit Suisse, in a recent note to clients. "The last big time around [this occurred] was the tech bust, when corporate sector leverage was the main culprit, but equity investors were also running far riskier portfolios than they normally do."

Seven years ago, these events sparked concern that a serious recession was looming, which was only partly averted after the Federal Reserve cut interest rates. The deleveraging also eventually contributed to a corporate credit crunch in the commercial paper market - or the sector where entities raise short-term finance. "During 2002, corporate commercial paper markets closed, forcing back-up liquidity providers to fill the breach, leading to large increases in credit risk for the banking system," says Jeffrey Rosenberg, analyst at Bank of America, noting that a similar phenomenon erupted this month in parts of the commercial paper market again.

However, as Mr Rosenbergalso notes, one striking difference between now and 2002 is that it is financial companies - not, say, manufacturers - that face a funding squeeze. And that highlights an even more crucial distinction: whereas in 2000 the list of companies that were overladen with debt included mainstream names, this time it is financial players, such as hedge funds or banks, that are excessively leveraged.

That may help to explain why the current turbulence has largely originated in the credit markets, rather than the equity world (as it did in 2000). But it also implies that today's market turmoil has a less negative impact on the "real economy" than seven years ago, since it is not hitting mainstream companies - yet.

Consequently, some analysts suspect that a much better model to analyse today's events lies further back - namely in those of 1997 and 1998, when a storm in the Asian financial markets triggered an unexpected default on Russian debt, which eventually lead to the near implosion of Long-Term Capital Management hedge fund. One reason this parallel is alluring to some traders is that the market movements this summer have been so violent they suggest some large - unidentified - institutions are in such distress that they are conducting fire sales of their portfolios. If so, that echoes the pattern in August 1998, when LTCM and its counterparties also tried to liquidate portfolios quickly, triggering seemingly bizarre price swings and a collapse in trust between the banks.

But another crucial parallel is that the 1998 turbulence was also centred on the financial world, not mainstream companies. As a result, LTCM's problems triggered few immediate adverse effects on the "real" economy.

That parallel prompts some observers to also conclude that this month's market chaos will do little tangible economic damage. The catch, however, is that losses in the subprime world now appear to be hitting a much broader range of investors than in 1998 - partly because they have been so widely scattered around as a result of financial innovation. Meanwhile, the pain is also potentially much larger: whereas the LTCM episode threatened to create a $3bn-odd hit to hedge fund investors and banks, estimates for the size of the losses from subprime mortgages range from from $50bn to $200bn. "The consensus remains sanguine on the outlook for global growth and is convinced we are seeing a repeat of the 1998 LTCM crisis," notes Jan Loeys, of JPMorgan. "But we are aware that history never fully repeats itself."

Nevertheless, that does not stop some analysts groping even further back in history for better lessons. Thus far, few pundits have attempted to suggest that the current events are a replay of the best-known drama of all - 1929. After all, the global economy still appears to be in rude health and the scale of market swings still looks extremely small compared with 1929 or other crashes. (Indeed, were it not for the fact that levels of market volatility have been unusually low this decade, some observers might hesitate to use the word "turmoil" at all.)

However, some analysts see parallels with another well-known crash - the events of 1987, when equity markets tumbled 22 per cent in a day and 60 brokerages went bust. The linkage partly revolves around the use of trading models. In the run-up to the 1987 crash, Wall Street had adopted the use of so-called "portfolio trading" models, which effectively exacerbated the downward move when equities started to fall. Similarly, some observers suspect that the widespread use of a new breed of trading models in 2007, known as quantitative - or "quant" - strategies has also magnified the market swings this summer, and enabled a shock in a narrow segment of the credit markets to infect numerous asset classes.

However, as Mr Bruner and Mr Carr demonstrate in their well-timed book, history shows that contagion has predated computers. The event that triggered the crash of 1907 in New York was the suicide of Charles Barney, the deposed president of the Knickerbocker Trust Company, who had attempted to corner shares in the United Copper company. Though this death seemed an isolated event - or a "contained" surprise, in modern parlance - as the implications spread, it set off a chain reaction throughout the banking world, fuelled by the fact the leverage levels were high and the economy had been weakened by a recent earthquake.

"Crises are like hurricanes," notes Mr Bruner. "Each is unique, yet we know enough about them all to be able to generalise - our big generalisation [from 1907] is that explanations come from a convergence of causes, most of which are always present in the global economy. [But] when these causes click into the right combination, financial crisis follows."

That may not offer definitive help for investors who want clear guidance about whether the current turbulence is just a passing storm - or not. But as economists thumb through the history books, the one thing that is crystal clear is that this summer's turmoil will not be the last.

On the contrary, as Mr Malvey notes, the lesson from the history books is that these episodes occur with striking regularity - typically, at least once a decade - whenever excess leverage, innovation and investor hubris collide. "Markets have always moved between three phases: pessimistic wariness, complacency and exuberance," he notes. "Well in advance, beware of a credit crunch in 2017. [But] sadly by then, the market will once again behave as if the turbulence of 2007 never took place."


400 years later, the rules are little changed
By Gillian Tett
Copyright The Financial Times Limited 2007
Published: August 27 2007 03:00 | Last updated: August 27 2007 03:00


Though financial markets have only existed for a few centuries, they have already delivered a truly dazzling array of storms, writes Gillian Tett. Lehman Brothers, for example, reckons that the first financial crash in the Western world can be dated back to 1622, when the Holy Roman Empire debased its coins, triggering the equivalent of a modern banking panic.

That was followed by the 1637 tulip boom-and-bust in Holland and the 1720 South Sea bubble - an event that spawned the first literary analysis of crashes.

Eight more crashes then occurred in the 18th century, culminating in the Hamburg commodities bubble of 1799. Then in the 18th century, the pace of financial disaster speeded up, with 18 financial storms erupting in Europe, and increasingly in the US too.

These included bank crises that hit the US in 1819, 1837, 1847, 1857, 1873, 1884, 1890 and 1896, several of which were linked to booms and bust in the commodities sphere, or railroad sector.

However, these events contributed to regular bank runs in Europe too, and a financial crisis even struck Australia in 1893.

It was the 20th century, however, that produced the most dazzling array of financial turbulence, with 33 market storms, according to Lehman Brothers. The best known of these are the stock market crashes of 1929 and 1987; however, the Japanese bank turmoil of the 1990s also caused deep economic distress, as did the savings and loan crisis in America in the 1980s - and the bank run of 1907.

Meanwhile, another, lesser-known period that some analysts think carries lessons for banks today was a stock market crash in London in 1974, which was accompanied by a banking crunch.

By the standard of many of these market storms, the recent turmoil still looks pretty modest. Indeed, it is arguable that the only reason the price swings have generated so much shock among investors is that markets have been unusually quiet during most of this decade.

Nevertheless, the one factor that does make this summer's events different from previous cycles is the degree to which the events have played out in the capital markets, rather than the banking world. That is largely because credit risk has been distributed to a much wider pool of investors than before.

Another crucial distinction is the sheer speed at which shocks now tend to move round the world, and between asset classes, as computing power links markets more closely than ever before. Back in the 19th century, for example, it could take days for news about railroad problems in America's Midwest to reach London; but this summer, bad news of subprime losses has often hit trading screens round the world in seconds.