How Goldman Sachs defies gravity

NEW YORK (Fortune) -- It is one of the most stunning bets Wall Street has seen in decades.

As the credit markets fell apart over the summer, causing the prices of hundreds of billions of dollars of mortgage-backed bonds to plunge, Goldman Sachs (Charts, Fortune 500) had already positioned itself so that it would profit massively from a decline in those securities. Thursday, Goldman reported earnings for its fiscal third quarter that were far above expectations.

Goldman made a bold bet as others faced mortgage-related losses.

More from Fortune

Will Mmmhops be a hit? NBA confirms L.A. Clippers sale to ex-Microsoft CEO Steve Ballmer FBI and SEC probe into Carl Icahn and golfer Phil Mickelson

FORTUNE 500 Current Issue Subscribe to Fortune

While several businesses were surprisingly strong in a difficult period, the chief contributor to the earnings blowout were trades that made money from price drops in mortgage-backed securities. Goldman indicated this in its press release when it said that "significant losses" on certain bonds were "more than offset by gains on short mortgage products." (In Wall Street parlance, being "short" a stock or bond means that you will make money if it goes down in price.) "Goldman Sachs showed an ability to not only protect itself from the problems in the market but also to capitalize on them," says Mike Mayo, banks analyst at Deutsche Bank (Mayo rates Goldman a buy.)

When asked on a public conference call Thursday, Goldman's chief financial officer David Viniar declined to give a number for the amount of money Goldman made on its mortgage short in the third quarter.

Goldman doesn't provide enough numbers in its public financials to come up with an informed guess, but the firm's statement that the short trades "more than offset" bond losses that were "significant" is a clear sign that it took time to deliberately set itself up for an expected crash in the market for mortgage-backed bonds. Indeed, Merrill Lynch analyst Guy Moszkowski said Goldman's trading results were $1.7 billion above his forecast in the third quarter, according to a research note released Thursday. (He rates Goldman a buy.)

Amassing a large bearish position in mortgages would have required planning and direction from a senior level. On the conference call, Viniar said the bet was executed across the whole mortgage business, implying that it wasn't the work of one swashbuckling trader or trading desk. Of course, the prescience of the short sale would seem to confirm the view that Goldman is the nimblest, and perhaps smartest, brokerage on Wall Street. Morgan Stanley (Charts, Fortune 500), Goldman's biggest rival, wasn't as well hedged to bond losses, while Bear Stearns' (Charts, Fortune 500) mortgage business suffered considerably in the quarter.

True, from third quarter numbers, it appears that Lehman Brothers (Charts, Fortune 500) also benefited from a short position in mortgages, but its bet wasn't big enough to allow the bank to report earnings that grew from either the previous or year-ago quarters. Goldman's net third quarter profits of $2.8 billion were substantially higher than in both those two prior periods, a notable achievement during a very testing period.

While the short sale allowed Goldman to show outstandingly strong earnings in the quarter, it may actually cause problems for the bank in other ways. First, investors in Goldman's two large poorly performing hedge funds will want to know why the savvy deployed in trading for Goldman's own account was not deployed in their funds. If a strategic decision to be short certain bonds was made high up, why didn't this end up helping the Global Alpha hedge fund and the Global Equity Opportunities fund, which were down 30% and 20%, respectively, in the third quarter alone?

Goldman spokesman Lucas van Praag responds: "We're always disappointed when we don't meet our clients' expectations. We're working hard to adjust our strategies to reflect the lessons we learned in August." In addition, the large gains from the mortgage trade will also deepen investors' desires to get a better handle on how brokerages like Goldman make money.

One of the figures that didn't seem to make sense in Goldman's earnings was a number that estimates the market risk on a broker's balance sheet. This indicator, called Value at Risk, or VaR, moved up only 5% in the third quarter from the second. If Goldman was placing big bets in volatile markets - like the short trade in mortgages - VaR might be expected to move up by more.

In other words, Goldman seems implausibly immune from the general rule in investing that higher returns almost always carry higher levels of risk. Van Praag responds that VaR didn't go up by much because Goldman reduced positions as volatility in the markets went up.

Goldman does seem to have institutionalized a higher level of trading savvy - the third quarter numbers bear that out. The market recognizes that in awarding the broker a valuation that is higher than that of its peers. Quarter in, quarter out, Goldman posts a return on equity in excess of 30%, even though it's highly leveraged, like all brokers. The high leverage should translate into at least some rough quarters. That was the case for Goldman's ailing hedge funds in the third quarter. Why is it never the case for Goldman itself?