Lehman sits on the hedge

NEW YORK (Fortune) -- It's disconcerting when a public company won't satisfactorily explain how a large loss came about. And it was doubly disturbing Tuesday to hear Lehman Brothers' finance chief dance around questions aimed at finding out how the Wall Street firm came up with a net $700 million hit to bond and lending revenues in its third quarter, ended Aug. 31.

During a conference call Tuesday morning about Lehman's earnings, Chief Financial Officer Christopher O'Meara was clear and emphatic that Lehman had weathered the credit crunch well, stating that, "the worst of the credit correction is behind us." But when discussing the $700 million reduction in fixed-income business revenue, O'Meara was vague, offering up only partial explanations of how Lehman had come up with a decline of that size.

As a result, expect the already intense speculation about the health and accuracy of balance sheet numbers at Lehman and other brokers to increase, not wane. Also reporting third quarter results this week are Morgan Stanley (Charts, Fortune 500), Bear Stearns (Charts, Fortune 500) and Goldman Sachs (Charts, Fortune 500).

Lehman made $1.54 per share in the quarter, sailing past the $1.47 per share expected by the main analysts who follow the company. A big reason for beating analysts' estimates was an unexpected decline in Lehman's tax rate, but stronger-than-expected investment banking and equity revenue also contributed. Lehman's stock jumped $2.37, or 4%, Tuesday afternoon, to $60.99, helped in part by the Fed's decision to cut the federal funds rate by half a percentage point. Lehman declined to comment.

The recent sharp pull back in corporate lending (also known as "leveraged loans") and massive dislocation in the market for bonds backed with mortgages were expected to hurt Lehman, which has high exposure to both businesses. No surprise, then, that the brokerage felt it necessary Tuesday to report reductions in the value of leveraged loans and bonds during the third quarter, when things turned ugly. Indeed, the net reduction of $700 million seemed sizable enough to satisfy some analysts at first.

But as the conference call progressed, it became clear that Lehman was not overly keen on disclosing how it arrived at that number. The bank did allow that mark downs on leveraged lending far exceeded $1 billion. But that was about it. One analyst asked O'Meara to further break out the contributions to the $700 million, but the CFO responded, "It's just hard."

Hard, in part because of the way the company's hedging strategy works. Broadly speaking, hedging is buying or selling a financial instrument as insurance against an adverse move in the price of another financial instrument. In Lehman's press release, it said that losses in the fixed income business were "partially offset by large valuation gains on economic hedges and other liabilities." In other words, Lehman had some sizable increases elsewhere on its balance sheet.

But where? In theory, Lehman could have been supremely hedged before the third quarter for a big decline in the fixed income business. It might be hard to get good hedges for leveraged loans, so the hedges presumably were in the area of mortgage-backed securities. But could Lehman have been sufficiently hedged to make gains that would offset losses in its mortgages book? That would seem implausible, given that Lehman had nearly $80 billion of mortgage-backed securities on its books at the end of the second quarter. (Even if Lehman did profit from a huge bet that mortgages would decline, how repeatable is such a bet? )

So, if we assume that very large hedging gains related to mortgages were unlikely, the losses in fixed income seem to have been offset chiefly by the other source Lehman flagged -- reductions in the value of Lehman's liabilities (in accounting, a decline in a certain type of liability translates into a gain for revenue).

Why would it be an issue if Lehman benefited from these sorts of liability gains? Because these sorts of liability gains are arguably a very low-quality contributor to earnings. Following O'Meara's remarks on the call, the accounting appears to work like this: When Lehman's perceived creditworthiness declines in the market and its costs of borrowing go up, certain liabilities appear to get marked down, producing a gain for the fixed income business.

But a gain from a drop in perceived creditworthiness is not an earnings source most companies would hanker after. Moreover, some of the liabilities that produced the putative gain may not trade in liquid markets, which means Lehman itself would have to provide best estimates on their worth.

Indeed, they might be called "level three" liabilities, the classification given to assets and liabilities that barely trade and whose values are largely the result of informed guesswork by the bank's own employees. As noted in a recent Fortune column, there has been much speculation about the contribution of level three balance sheet items to earnings at the brokerages. And one intriguing factoid O'Meara did offer up was that level three assets increased to 10%-11% of financial assets at Lehman in the third quarter, from 8% at the end of the second quarter.

This is definitely a negative development, as more of Lehman's assets depend on internal guesswork for valuation, rather than external market prices. Why? Because it implies that markets for certain assets have dried up and deprived Lehman of the price inputs it needs, pushing level two assets, which have some market reference points, down into level three.

In other words, Lehman's black box just got bigger. Add to that the CFO's reluctance to expand on its losses, and there is little reason to buy into Lehman's boast that it has left the credit crunch behind.