naked capitalism

By Frank Partnoy, Professor of Law and Finance University of San Diego School of Law and author of Fiasco, Infectious Greed, and The Match King

The buzz on the Lehman bankruptcy examiner’s report has focused on Repo 105, for good reason. That scheme is one powerful example of how the balance sheets of major Wall Street banks are fiction. It also shows why Congress must include real accounting reform in its financial legislation, or risk another collapse. (If you have 8 minutes to kill, here is my recent talk on the off-balance sheet problem, from the Roosevelt Institute financial conference.)

But an even more troubling section of the Lehman report is not Volume 3 on Repo 105. It is Volume 2, on Valuation. The Valuation section is 500 pages of utterly terrifying reading. It shows that, even eighteen months after Lehman’s collapse, no one – not the bankruptcy examiner, not Lehman’s internal valuation experts, not Ernst and Young, and certainly not the regulators – could figure out what many of Lehman’s assets and liabilities were worth. It shows Lehman was too complex to do anything but fail.

The report cites extensive evidence of valuation problems. Check out page 577, where the report concludes that Lehman’s high credit default swap valuations were reasonable because Citigroup’s marks were ONLY 8% lower than Lehman’s. 8%? And since when are Citigroup’s valuations the objective benchmark?

Or page 547, where the report describes how Lehman’s so-called “Product Control Group” acted like Keystone Kops: the group used third-party prices for only 10% of Lehman’s CDO positions, and deferred to the traders’ models, saying “We’re not quants.” Here are two money quotes:

While the function of the Product Control Group was to serve as a check on the
desk marks set by Lehman’s traders, the CDO product controllers were hampered in
two respects. First, the Product Control Group did not appear to have sufficient
resources to price test Lehman’s CDO positions comprehensively. Second, while the
CDO product controllers were able to effectively verify the prices of many positions
using trade data and third‐party prices, they did not have the same level of quantitative sophistication as many of the desk personnel who developed models to price CDOs. (page 547)

Or this one:

However, approximately a quarter of Lehman’s CDO positions were not affirmatively priced by the Product Control Group, but simply noted as ‘OK’ because the desk had already written down the position significantly. (page 548)

My favorite section describes the valuation of Ceago, Lehman’s largest CDO position. My corporate finance students at the University of San Diego School of Law understand that you should use higher discount rates for riskier projects. But the Valuation section of the report found that with respect to Ceago, Lehman used LOWER discount rates for the riskier tranches than for the safer ones:

The discount rates used by Lehman’s Product Controllers were significantly understated. As stated, swap rates were used for the discount rate on the Ceago subordinate tranches. However, the resulting rates (approximately 3% to 4%) were significantly lower than the approximately 9% discount rate used to value the more senior S tranche. It is inappropriate to use a discount rate on a subordinate tranche that is lower than the rate used on a senior tranche. (page 556)

It’s one thing to have product controllers who aren’t “quants”; it’s quite another to have people in crucial risk management roles who don’t understand present value.

When the examiner compared Lehman’s marks on these lower tranches to more reliable valuation estimates, it found that “the prices estimated for the C and D tranches of Ceago securities are approximately one‐thirtieth of the price reported by Lehman. (pages 560-61) One thirtieth? These valuations weren’t even close.

Ultimately, the examiner concluded that these problems related to only a small portion of Lehman’s overall portfolio. But that conclusion was due in part to the fact that the examiner did not have the time or resources to examine many of Lehman’s positions in detail (Lehman had 900,000 derivative positions in 2008, and the examiner did not even try to value Lehman’s numerous corporate debt and equity holdings).

The bankruptcy examiner didn’t see enough to bring lawsuits. But the valuation section of the report raises some hot-button issues for private parties and prosecutors. As the report put it, there are issues that “may warrant further review by parties in interest.”

For example, parties in interest might want to look at the report’s section on Archstone, a publicly traded REIT Lehman acquired in October 2007. Much ink has been spilled criticizing the valuation of Archstone. Here is the Report’s finding (at page 361):

… there is sufficient evidence to support a finding that Lehman’s valuations for its Archstone equity positions were unreasonable beginning as of the end of the first quarter of 2008, and continuing through the end of the third quarter of 2008.

And Archstone is just one of many examples.

The Repo 105 section of the Lehman report shows that Lehman’s balance sheet was fiction. That was bad. The Valuation section shows that Lehman’s approach to valuing assets and liabilities was seriously flawed. That is worse. For a levered trading firm, to not understand your economic position is to sign your own death warrant.