The lessons of Lehman for other banks
Lest we forget, Lehman Bros was regarded as one of the world's most sophisticated, well-managed investment banks, just a year or so before it went belly-up.
Investors loved the stock, valuing the bank at more than $30bn as late as January 2008. Financial institutions, including the world's most lauded banks and hedge funds, lent it hundreds of billions of dollars. Regulators trusted that it had the appropriate systems to control the risks it was taking.
But it turns out that those at the top of the bank were - to an extent - flying blind about the risks being taken by Lehman. And so too, therefore, were US and British regulators.
That is the inescapable conclusion of the 400-page valuation section of the recent report on the collapse of Lehman by the examiner for the New York bankruptcy court.
That section hasn't as yet received much media attention, because it is much less sexy than the examiner's finding that Lehman shunted $50bn of assets off its published balance sheet, to exaggerate its financial strength, using the highly questionable Repo 105 technique (see my earlier note on this).
And, to be clear, the examiner does not believe that Lehman deliberately understated losses on its loans and investments in a way that could lead to substantial damages claims by creditors.
But his report tells a disturbing story of a bank with $700bn of assets and 900,000 derivative positions woefully ill-equipped to assess whether the values that its traders were putting on their deals were the correct values.
And before I quote one or two choice passages from the report, I will state the bloomin' obvious - which is that trusting the valuations of traders, whose enormous bonuses depend on whether their investment and dealing positions are showing a loss or profit, is as sensible as trusting a bunch of five-year-olds not to eat the sweeties in a chocolate factory.
Now Lehman did have a so-called Product Control Group whose job was to assess the valuations or "marks" put on assets by the assorted business desks. This is what the examiner says about the capability of the Product Control Group in respect of its checks on the prices claimed for collateralised debt obligations, those toxic bonds made out of home loans:
"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...
"The effectiveness of the Product Control Group was also limited because it did not have the technical sophistication to develop complex models for pricing CDOs, as did certain of the desk personnel (commonly referred to as 'quants') they were charged with monitoring."
Or to put it another way, in the absence of reliable market prices the Product Control Group lacked the intellectual tools to challenge the prices put on CDOs by those who created them.
This is profoundly shocking, and not just for what it says about woeful risk controls at Lehman.
It calls into question the assurances given by those who run all the world's big investment banks that they have reliable techniques to control the risks taken by their employees.
The point is that the bosses of Barclays, Goldman, Morgan Stanley, JP Morgan and so on have never claimed that they personally understand each and every one of the millions of investments that are on their respective balance sheets. Nor could they ever do so. The size and complexity of their businesses would baffle an X-Men style mutant superhero with a brain the size of a planet.
But they do claim that they have highly skilled risk controllers who vet their traders' and bankers' valuations on their behalf. So the really important question raised by the Lehman report is whether these extant banks' respective risk controllers and product control groups are a cut above Lehmans'.
What was the practical consequence of the physical and intellectual under-resourcing of the team that was supposed to keep Lehman's bankers and traders on the straight and narrow?
Well, on one measure some half of Lehman's CDO portfolio was unreviewed in May 2008. Bizarre mistakes were made, such as using a lower discount rate to value tranches of CDO that were intrinsically more risky. And on one securitisation called CEAGO, the court examiner valued one tranche of bonds at 3% of the price put on them by Lehman's Product Control Group.
Here's the important point. We pay money to be passengers in planes not because we have a detailed understanding of all those complex computer and engineering systems that keep planes in the air, but because we are confident that the airlines and manufacturers have that understanding.
The corollary for banks like Lehman is that they are given licences to trade because they are trusted to keep a firm grip on the high complicated risks they are running. The examiner's report should make us ponder whether we've been a bit too trusting, not just in Lehman's case but for all those global mega investment banks.