Remember the Lehman Examiner's Report? The 4000+ page report by the court-appointed examiner was lauded for a couple of weeks after it was released, and then largely forgotten. The media and blogosphere quickly moved on to the next outrage-du-jour (among others, the SEC's suit against Goldman), and never looked back. Well, I did not forget about it, and thanks to the uptick in flights — and thus reading time — in the last few months, I can now credibly claim to have read....well, not every single word in the Examiner's Report (some appendices are just pages of CUSIPs), but all of the substantive sections, including all of the substantive appendices, and many of the underlying documents.
First of all, let me just say that Anton Valukas and the lawyers at Jenner & Block who wrote the Examiner's Report did a masterful job. I was, and continue to be, in awe of the quality and comprehensiveness of the report. The fact that they were able to do all the research, conduct all the interviews, and write the 4,000+ page report in a little over one year is amazing, and makes me incredibly jealous of their productivity.
Now, since I read the whole damn thing, I think I have a pretty good handle on what went wrong at Lehman, and why it failed. Obviously, there isn't just one reason why Lehman failed, despite what most commentators would have you believe. But there's one issue that stands out to me as the biggest problem at Lehman — in short, they were misrepresenting their liquidity pool. In a huge way.
It's disappointing that this issue has been almost completely overlooked, because the brazenness of their misrepresentation was shocking. I think the best way to think about it is this: on Friday, September 12, Lehman claimed that it had a $32.5bn liquidity pool, and on Monday, September 15, Lehman needed $16bn to finance non-central bank eligible collateral. So why, if their liquidity pool was twice the size of their funding requirement, did they have to file for bankruptcy? The answer is that Lehman didn't actually have a $32.5bn liquidity pool; they had, at most, a $2.5bn liquidity pool. That is not a typo.
Earlier in 2008, Lehman's two main clearing banks, JPMorgan and Citi, started requiring Lehman to collateralize its intraday exposures. (Previously, the clearing banks would repay Lehman's tri-party repo lenders at the beginning of the day, and wouldn't require Lehman to pay back this advance until the end of the day.) Lehman reluctantly agreed, but requested that the banks release the collateral at the end of each day. Why did they care if the banks released the collateral every night if it just had to be posted again the next morning? Because Lehman calculated its reportable liquidity at the end of each day, and if the clearing-bank collateral was released at the end of each day, Lehman considered it part of the "liquidity pool." By the end, roughly $19bn of the $32.5bn liquidity pool consisted of clearing-bank collateral.
In no functional sense was the clearing-bank collateral "unencumbered" — if Lehman requested the collateral back, JPMorgan and Citi would have at the very least required them to pre-fund their trades (which Lehman didn't have the cash to do), and more likely would have just stopped clearing their trades. People at Lehman admitted as much to the Examiner. And once a broker-dealer's clearing bank stops clearing its trades, the broker-dealer is finished. Including the clearing-bank collateral in its liquidity pool was not only inappropriate, but also aggressively deceptive.
But wait, there's more. Lehman was also including in its liquidity pool non-central bank eligible CLOs and CDOs. And they had the audacity to mark these CLOs and CDOs at 100 (par) for purposes of the liquidity pool, even though JPMorgan's third-party pricing vendor marked them at 50–60.
Unfortunately, I can't do much in the way of legal analysis of the propriety of Lehman's decision to include clearing-bank collateral and junk CLOs/CDOs in its liquidity pool, because as the Examiner rightly points out, "No law, SEC regulation or GAAP-style rule governed the definition of a 'liquid' asset in the context of a CSE’s liquidity pool." The SEC considered an asset to be "liquid" if it could be monetized in less than 24 hours (though the SEC never enforced this definition, or wrote it into an actual rule). The "24 hours" standard is, in my opinion, the consensus in the industry — that is, I think if you asked most market participants, they would agree that at least most of the assets in a reportable "liquidity pool" should be able to be monetized in 24 hours.
Lehman's Global Treasurer, Paolo Tonucci (who really comes off as a Bad Guy in this episode), claimed that "Lehman's internal definition of a 'liquid' asset, appropriate for inclusion in the liquidity pool, was one that could be monetized within five days." (Examiner's Report, pg. 1412) Tellingly, in a footnote, the Examiner notes that "Tonucci could not cite a particular Lehman document that established this five-day definition." Lehman's International Treasurer, Carlos Pellerani, said he had never heard of this alleged "five-day" standard. (Examiner's Report, Appendix 20, pg. 10)
Of course, even by this "five-day" standard, Lehman was still misrepresenting its liquidity pool. Lehman broke down its liquidity pool into "ability to monetize" categories: assets with a "high" ability to monetize could be liquidated in one day, assets assigned a "mid" rating could be liquidated within five days, and assets assigned a "low" rating could be monetized within one to two weeks. (Examiner's Report, Appendix 20, pg. 10) By September 12, $30bn of Lehman's $32.5bn liquidity pool had a "low" ability to monetize:

This is a freaking liquidity pool, and you're telling me that 92% of it can only be monetized in a week or two? Unreal. By contrast, here's how Goldman defines its liquidity pool in its 10-Qs (emphasis added):
The U.S. dollar-denominated [Global Core] Excess [i.e., liquidity pool] is comprised of only unencumbered U.S. government securities, U.S. agency securities and highly liquid U.S. agency mortgage-backed securities, all of which are eligible as collateral in Federal Reserve open market operations, as well as certain overnight cash deposits. Our non-U.S. dollar-denominated excess is comprised of only unencumbered French, German, United Kingdom and Japanese government bonds and certain overnight cash deposits in highly liquid currencies. We strictly limit our Global Core Excess to this narrowly defined list of securities and cash because we believe they are highly liquid, even in a difficult funding environment. We do not believe that other potential sources of excess liquidity, such as lower-quality unencumbered securities or committed credit facilities, are as reliable in a liquidity crisis.Using this definition (which is standard), Lehman had a liquidity pool of maybe $1.9bn on Friday, September 12.
Now, Lehman wasn't always misrepresenting its liquidity pool. By the end of May, at least, it had a legitimate liquidity pool of over $40bn. But over the Summer of 2008, its liquidity pool steadily deteriorated, and Lehman simply refused to recognize this deterioration in its reported liquidity. JPMorgan and Citi demanded around $14bn more in collateral, and large asset managers — most importantly, Fidelity — pulled Lehman's repo lines and stopped rolling their paper. As of June 30, 2008, Fidelity had $14.1bn of repo lines in place with Lehman; by Thursday, September 11, Fidelity had pulled the entire $14.1bn balance.
As you can see from the chart above, Tuesday, September 9th was the real Day of Days for Lehman in terms of funding. On Monday night, news broke that Korea Development Bank had decided not to invest in Lehman. On Tuesday, repo counterparties pulled roughly $14bn of funding. Here's the breakdown of counterparties who pulled repo lines that day:

Notice that these are all large institutional asset managers — PIMCO, Morgan Stanley, Barclays, etc. — rather than those supposedly skittish retail investors. This was, essentially, the end for Lehman. There was no way they could replace $14bn of secured funding that quickly, especially after reporting such terrible earnings the next day.
All of this highlights the importance of establishing liquidity requirements with a clear definition of a "liquid" asset — which, fortunately, is something Basel III does, and does rather conservatively. I'll save that for another post, though.