O Efeito do Quantitativo: quando o problema é o modelo

Reportagem da The Economist (25/10/2007 Heart of darkness) discute um problema do sistema financeiro mundial, baseado, de forma significativa, nos quant, designação dos modelos baseados em métodos quantitativos desenvolvidos por especialistas da área de exatas.

Segundo a revista "eles usam a alavancagem. Isto significa que em lugar de fornecer liquidez na crise, os quants adicionam instabilidade." Abaixo, a reportagem:

Heart of darkness
The peril for markets when computers miscalculate

WHEN markets wobbled in August, almost all the media attention was focused on the credit crunch and the links to American mortgage loans. But at exactly the same time, another crisis was occurring at the core of the stockmarket.

This crisis stemmed from the obscure world of quantitative, or quant-based, finance, which uses computer models to find attractive stocks and to identify overpriced shares. Suddenly, in August, the models went wrong.

The incident revealed a problem at the heart of the financial system. In effect, the quant groups were acting as marketmakers, trading so often (some are aiming for transaction times in terms of milliseconds) that they set prices for everyone else. But unlike traditional marketmakers, quant funds are not obliged to make markets come rain or shine. And unlike marketmakers, they use a lot of leverage. This means that instead of providing liquidity in a crisis, the quants added to instability. There is a lesson there.

In a way, the crisis stemmed from the quants' success. Many firms, such as the American hedge fund Renaissance Technologies, had done fantastically well and had been able to charge hefty fees. But if one firm can hire top mathematicians and use the latest technology, so can others. An arms race developed, with some trading faster and faster—even siting their computers closer to the exchanges in order to cut the time it took orders to travel down the wires.

And as the computers sifted through the data, some strategies became overcrowded. A paper* by Amir Khandani and Andrew Lo of the Massachusetts Institute of Technology back-tested a proxy for a typical strategy, involving buying the previous day's losing stocks and selling the winners. Such a strategy would have delivered a daily return of 1.38% before (substantial) costs in 1995 but the return fell steadily to 0.15% a day last year.

In the face of declining returns, the authors reckon, the natural response of managers would have been to increase leverage. But that, of course, increased their vulnerability when things went wrong.

Both the MIT academics and a paper by Cliff Asness of AQR Capital Management, a leading quant group, agree that August's problems probably began when a diversified, or multi-strategy, hedge fund experienced losses in the credit markets. The fund sought to reduce its exposures but its credit positions were impossible to sell. So it cut its quant positions instead, since that merely involved selling highly liquid stocks.

However, that selling pressure caused other quant funds to lose money as their favoured stocks fell in price. Those that were leveraged were naturally forced to reduce their positions as well. These waves of selling played havoc with the models. Quant investors thought they were aware of the risks of their strategy and had built diversified portfolios to avoid it. But the parts of the portfolio that were previously uncorrelated suddenly fell in tandem.

In theory, quant funds could have been bold and borrowed more; after all, the stocks they thought were cheap had become even cheaper. The traders who took on the positions of Long-Term Capital Management (LTCM), after the hedge fund failed in 1998, ended up making money. But the example of LTCM, which went bust before it could be proved right, argued in favour of a more cautious approach. “We could have rolled the dice but that would have risked the business,” said one quant-fund manager. “I don't know of anyone that did so.”

Avoiding that trap simply led quant investors into another. On August 10th, the stocks that quants had favoured suddenly rebounded. Those who had cut their positions most could not benefit from the rally. That category clearly included Goldman Sachs's Global Alpha hedge fund, which lost a remarkable 23% on the month.

If it were just a few hedge funds, backed by rich people, losing money, it might not matter. But the funds had become too important: rather than adding stability, as marketmakers are supposed to do, they added volatility.

Quants will adjust their models and clients will become more discerning; AQR's Mr Asness says his firm will look harder for “unique” factors, that is, not used by other fund managers. But regulators should also reflect that markets are less stable than they assumed. The presence of leveraged traders such as quants at their heart means conditions can now turn, at the flick of a switch, from stability to panic.