naked capitalism

Readers may recall that last we ran a post last week on Lehman’s 2Q earnings announcement, its conference call, and some questions raised by an e-mail from a former Lehman managing director.

To recap: the investment bank reported a gross asset reduction of $147 billion, net of $70 billion (we consider the $70 billion the more significant number, since the balance of the assets were probably subject to agreements to repurchase and thus quite liquid). These reductions were consistently characterized as sales in the conference call. But just as some animals are more equal than others, “sale” can cover a multitude of sins.

Indeed, the ex-Lehman staffer indicated that some assets were sold to parties in which Lehman has an economic interest, newly-formed funds headed by ex-Lehman employees called R3 and One William Street. We indicated that since this information came from a single source, it had to be regarded as a rumor, although it was unusually specific as to details.

Later in the week, a Bloomberg story substantially confirmed the information the source provided about a fund started by the former head of Lehman’s principal strategies group, R3 Capital Partners, particularly that the Lehman did have a stake in the firm and would benefit if the $5 billion of assets sold to the fund appreciated. While this Bloomberg story reported that a very small amount of Lehman assets was sold to One William Street, it, did not indicate whether the investment bank had an ownership stake in that fund. However, an earlier Bloomberg story said that Lehman was expected to back the fund (and perhaps more telling, “One William Street” is the address of the former Lehman headquarters building and has been used in the past as a name for in-house funds. One would imagine that the firm would not permit it to be used casually).

The ex-insider has written to us again to clarify what he initially heard about R3 (we’ll return to that later; he said that several Lehman employees have since told him that a memo circulated internally said, just as the Bloomberg article did, that $5 billion was sold to R3 Capital Partners).

He also passed along another bit of information which he maintains came from “several people who are in a position to know.” Again, however, since I cannot verify this independently, this too must be regarded as a rumor:

LEH is allowing employees of both of these new funds to continue vesting any LEH restricted stock they held when they left. This is extraordinarily generous, and unusual to say the least. Lehman, like most of the brokers, provides 1/3-1/2 of employees’ bonuses in restricted stock that vests after either 3, 4 or 5 years (depending on seniority, size of grant, etc). Ordinarily, when you leave the firm, all unvested stock is forfeited. The expense, to Lehman, of doing this is surely well into nine figures…perhaps quite well into nine figures. Seems an expensive & unnecessary gesture, given that all the employees are now happily ensconced in major hedge funds, and one that LEH can, at the moment, ill-afford.

There aren’t many reasons to provide deferred comp to people who are not longer your employees. In fact, the only good reason to do so is to ensure the former employees’ continued good will, and to have something to hold over them in case they engage in an activity you consider detrimental. It’s hard to think of any detrimental activity that these fine folks might engage in (from LEH’s perspective)…at least, no activity that’d be worth hundreds of millions of dollars of scarce shareholder capital…other than making public details of the arrangements between LEH & the start-ups which LEH might find quite inconvenient to have generally known.

Let me stress that the second paragraph, after the first sentence, is speculation by this former employee. However, I struggle to find a good business rationale for a move like this, assuming the details are accurate, particularly given the magnitude of the costs alleged.

We has noted in our earlier post that pay stood out as an anomaly. Employee compensation rose $418 million in the second quarter, with only $130 million of that severance, when headcount fell by 1,900. To put that in context: the non-severance component of the pay increase equalled 10% of Lehman’s quarterly loss. That begs for explanation.

The severance figure also looks high. I may be woefully out of date, and those more current can correct me, but my understanding is severance is not a common perk on Wall Street. I believe in some countries with tough labor laws, there are minimums based on years of service. Another group that would be eligible were those who had employment contracts; the cost of early termination would be classified as severance. Having said said that, each firm is free to have its own policies, and Lehman’s may simply have been unusually generous. However, $130 million divided by the 1,900 employees dismissed is $68,000 per employee. Any reader input here would be appreciated.

Now some readers may think we are going too harshly after Lehman. Let us stress that in the year and a half this blog has been up, this is only the second source whose information we thought was worth advancing even though we could not independently corroborate it.

In this case, the reason we chose to do so is that while Lehman was extremely forthcoming in what assets it had sold, it was virtually silent as to how it sold them. And “how” is a legitimate question in these difficult credit markets.

In turbulent markets, when firms need to trim their exposures, they try to sell what they can without showing a loss (or much of a loss). Even with positions marked down, a price that is a fair mark for a transaction of $10 million is not what a dealer would realize if trying to unload $500 million in a troubled market with spotty demand. Proof of that conundrum already exists. Investment banks have financed sales of leveraged loan and Alt-A portfolios to avoid realizing further losses. Similarly, a bank took a 25% loss in May on the sale of a $3 billion Alt-A portfolio that it had maintained in an early March regulator filing was impaired by at most 10% using a stress test scenario. Note it had hired an investment bank to market the position, so it clearly could have been sold in smaller pieces had than been more attractive.

Moreover, some of Lehman’s valuations had already come under question (see here, for instance, which includes a discussion of commercial mortgage marks). Having a mark that was, shall we say, generous, would be an impediment to selling an asset, since a disposition would produce a loss.

That’s a long-winded way of indicating that financial firms under stress typically wind up selling their most saleable assets, which makes them more liquid, but often winds up lowering the quality of their remaining holdings.

Yet Lehman seems to have achieved the miraculous, unloading $70 billion of assets, and the only mention of an assist in the conference call was:

The approximately 8 billion of commercial mortgage and real estate held-for-sale assets sold this quarter were across the capital structure to over 170 different client accounts and approximately 80% were outright sales without seller financing.

Note that $8 billion was the only place where I saw the number and nature of end buyers mentioned.

Aside from the mention that “other nonmortgage asset backed exposures… were flat at 6.5 billion this quarter,” the sales were broadly spread across asset types, and as Lehman stressed, “We sold a variety of assets and not just the most liquid….the sales were across sort of, all asset classes, across all the types….They were spread over the whole quarter.” The last is a particularly remarkable achievement, given how disrupted the markets were in March, the month of the Bear meltdown.

So if, as with R3, the firm has an economic interest in the assets sold and an ownership stake in the buyer, that begs the question of whether there is residual risk. As we saw with SIVs, assets were placed in entities that had been thought could be allowed to sink or swim. Yet in most cases, when things went pear-shaped, the sponsoring bank stepped forward to assume control, deeming the potential damage to its reputation and relationships to be too great. Thus while Lehman no doubt structured this relationship so that the asset transfer can be called a sale from an accounting standpoint, the use of that term seems questionable from a common-sense perspective.

So how did the source get one bit of his original information so wrong, namely, that the amount Lehman sold to R3 Capital Partners was $5 billion, when he said the amount R3 acquired was $55 billion? His comment:

The $55 billion figure came from a source directly involved with R3. To be completely fair, the figure related to R3’s entire asset base at its launch (as in, “R3 is spinning out of Lehman, and is launching with $55 billion in assets.”). I don’t recall being told specifically that 100% of the assets came from Lehman…that part was my inference, given that R3 was just starting life as an independent entity.

This illustrates why one has to treat information that cannot readily be verified with caution. But having said that, $55 billion is a very large number for a fund to have at its birth. The number is consistent with the acquisition of a portfolio or several portfolios.

Then again, the person who spoke to the former employee could simply have been wrong….