Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Strategies & Market Trends : Hedge Funds -- Ignore unavailable to you. Want to Upgrade?

To: Marty Rubin who wrote (4)	9/10/1998 9:52:00 PM
From: Marty Rubin	Read Replies (1) \| Respond to of 120

COMMENTARY: WHEN COMPUTER MODELS SLIP ON THE RUNWAY The wreckage is everywhere as hedge funds, banks, and securities firms rack up big losses. This time, the losses are not sustained by free-wheeling high-rollers betting with their gut, but by nerdy computer mavens betting with their supercomputers and software--and losing. Despite state-of-the-art arbitrage trading strategies, virtual reality has failed to come to their rescue. Market-neutral strategies, in large measure, have failed. Their strategies may have been perfect on paper. But the market refused to cooperate. The only surprising thing about the recent spate of high-tech trading failures is that they have come as a surprise to so many--even to some of the super-traders running these firms. Today, it is John W. Meriwether's Long-Term Capital Management. But in 1997, it was UBS and Natwest Capital Markets. And a few years before that, it was Askin Capital Management. The list goes on and on--a largely forgotten litany of disappointments (table). BIG FLOPS. Trading models work well most of the time. But we should never forget that they are prone to calamitous failure. There's even a term of art for this phenomenon: model risk. Indeed, perhaps the most important lesson to be drawn from the recent blunders of some of the smartest people on Wall Street is that, sooner or later, some of their most promising models are bound to fail--and fail big. And there's not much that can be done about it. Model failure can be a simple case of bad pricing data--''garbage in, garbage out''--such as the mispricing that touched off the collapse of David Askin's $600 million hedge-fund family in 1994. Or it could be the apparent cause of Long-Term Capital's sickening 44% decline in August, as its computer models became overwhelmed by an unexpected combination of market forces. Whatever the reason, the latest travails give an illustration of what some observers have been saying for a long time. Quite simply, these models can be too complex for their own good. They are akin to an aircraft engine breaking down because it inhaled a seagull. ''Every model is based on all sorts of assumptions, and one of hundreds of assumptions in sophisticated models can go wrong. It gives people a false sense of security,'' notes Joe Kolman, editor of Derivatives Strategy, a New York-based newsletter. Derivatives experts such as Kolman have been offering warnings about the dangers of model risk for years. Long before the latest crisis, Derivatives Strategy ran as its June, 1997, cover story, an article entitled ''Controlling Model Risk.'' It contained a jarringly prescient exposition on the problems lurking in multifaceted computer models. ''Riskier derivatives often include very high degrees of leverage, exotic cash-flow patterns or are based on obscure markets,'' derivatives consultant Tanya Styblo Beder said in the October, 1994, issue of Corporate Cashflow Magazine. Valuing such derivatives, she noted, often depends ''on sophisticated mathematical models, creating the risk that the model's value may be different than that ultimately obtained in the market.'' HARSH TEST. Even a master of the model universe such as John Meriwether seems to have forgotten that his financial strategies are far from unsinkable. In a Sept. 2 letter to investors, Meriwether pointed to the fund's record of stellar performance and noted that while he anticipated a setback, ''I did not anticipate, however, how severe the test would be.'' After explaining the fund's dreadful recent performance, Meriwether noted that the fund was giving its investors the opportunity to put in more money. A sucker bet? Not necessarily. The opportunities for hefty profits, which sustained Long-Term Capital and other rocket scientists in recent years, have certainly not gone away. But Meriwether, and the rest of us, have been taught yet another lesson in the drawbacks of these delicately choreographed high-tech trading strategies. They're an alluring way of making money, no question about it. And they will continue to be--until the next disaster strikes. By Gary Weiss Copyright 1998, by The McGraw-Hill Companies Inc. All rights reserved. (Got to #reply-5719343 for "A Rogue's Gallery of Failures.")