Lehman Japan scandal – More Repo 105 (Part 3)

Picking up from the day before yesterday:

Was Lehman Japan’s fixed income division (“FID”, the bond trading desk) involved in the Repo 105 activity? You bet.

It’s right on Page 795 (Volume 3) of the Valukas report:

In addition, documentary evidence suggests that as Global Head of Accounting Policy, Marie Stewart, was consulted on new applications of the Repo 105 mechanism [footnote 3048]. Stewart said that it was very “typical of people in the Front Office” to try to apply the Repo 105 mechanism to new situations such as when FID Asia tried to include Australian securities in a Repo 105 transaction. [footnote 3049]

Emphasis added.

The FID Asia desk was in Roppongi Hills, Tokyo.

Footnote 3049, leaving out the cross reference codes to the exhibits:

Examiner’s Interview of Marie Stewart, Sept. 2, 2009, at p. 10, e-mail from Annie Lin, Lehman, to Todd Wiener, Lehman, et al. (April 16, 2007) [cite omitted – and could be 2008, based on quotes below] (asking whether Lehman Brothers Australia Ltd. (“LBAU”) could “do a back to back position transfer between LBAU and LBIE in order to benefit from repo 105 rule”); e-mail from Marie Stewart, Lehman, to Annie Lin, Lehman, et al. (Apr. 17, 2008) [cite omitted] (“You cannot do Repo 105 in Australia. We will not approve it even if it technically works. I will explain further when I am back from vacation . . .”); e-mail from Thomas Siegmund, Lehman, to Kaushik Amin, Lehman (Apr. 17, 2008) [cite omitted] (“As B/S will be super tight, I need to make sure we make the best use of 105; I want to investigate whether 105-mechanics can apply to Aussie paper. An answer like below, ‘no . . . will explain when back from vac’ is unacceptable in any circumstances and really out of line in the current situation!”).

Emphasis added.

The writing clearly indicates that Lehman’s FID Asia was aware of some significance to a “Repo 105” transaction, although they may not have been aware of–what later everyone found out to be–the scam with the manual adjustments and the Linklaters’ opinion.

It almost goes without comment that the people involved with Repo 105 in Australia gave an emphatic “no”. My guess as a professional is that the other side of the repo might have been an Australian bank whose own audit or internal control system would have red-flagged the transaction. Like everything in the world of double-entry accounting, the transaction has to posted on another set of ledgers, the other banks’ ones!

So if Lehman Japan was keeping GAAP books and recording Repo 105, it meant they were showing the liability to pay back, at least on the internal systems like I-STAR. But what it would have also meant, from the Japanese perspective, is that Lehman was running on thinner capital every quarter end. A smart regulator or auditor would have been concerned about this. And they probably were. If. If.

In footnotes in Volume 8, Lehman Japan CFO Enrico Corsalini is mentioned twice. Once as a sender of an e-mail, and the other time as a recipient. (Footnotes 158 and 351).

This footnote is part of a discussion about SEC capital requirements (looks like the 15c-3 “minimum net capital rule”), listing that rule, two others, and a “Cash Capital Model”. This is not part of government regulations, but rather a model that “measures long-term funding sources against requirements”, and defined internally within a company. So the company can count whatever it thinks it can rely on as a funding source as part of its capital. Without knowing the details, it sounds like a made-up internal capital measurement that Lehman relied on. [Update: Reading further on SEC Rule 15c-3, most of the large brokers were allowed to use something called “Alternative Method”, by which they could rely on modeling for some of the minimum net capital calculation.]

Remember what I said about why Lehman was doing Repo 105? Because they wanted to borrow money and not have it show as a liability on the balance sheet. The machinations, according to Valukas, were done manually within LBIE (Lehman Brothers International – Europe or “Libby”). But the ingredients had to go in at a desk like Lehman Japan, or Lehman in New York.

At minimum, this footnote shows that having adequate capital was a concern within Lehman in early 2008. If the Lehman Fixed Income Division – Asia was trying to do “Repo 105’s” using Australian debt, and the home office treasurer was getting e-mails from the Lehman Japan CFO about capital models, how likely is it that people at a second-tier senior level did not know the importance of “Repo 105” to managing liabilities and keeping up a capital level?

So the problem in Japan would have been that running Repo 105 through their books (both asset and liability side) would have made the Lehman Japan unit more thinly capitalized than whatever it would have been without. Thin capitalization can shut you down in some jurisdictions. So this must have been a concern in Japan, which is where the second footnote mentioning Enrico Corsalini comes in:

(Footnote 351, on page 61 of Volume 8):

The main text on page 61 says:

H. Standard Chartered Bank

In mid‐April 2008, Lehman’s Chief Risk Officer, Christopher M. O’Meara, considered the possibility of a combination with Standard Chartered, but he concluded that Lehman’s Executive Committee would not support such a deal. 351

Now, this was a communication between the former CFO of it all (O’Meara), who by April 2008 was Chief Risk Officer, and the Asia CFO (Corsalini). “Consider the possibility of a combination” likely means that O’Meara wanted Lehman to merge with Standard Chartered, as this would supply the needed capital to keep Lehman from being too thinly capitalized. This was April; the bankruptcy occurred in September, and the belief of some commentators is that Lehman had been insolvent weeks before then.

Why share potential merger information with a subsidiary CFO? Just to chat? Or to say, ‘look it, we are trying to get around the problem of thin capitalization’.

Although the Valukas report is very thorough, it still and all isn’t everything. If you look at Volume 9, there is a large list of manual journal entries, some done involving Tokyo, that the Examiner’s subcontractor said would take even longer to analyze.

But what is already there, if you go through it, says a lot even about the units that were not at the center of action in New York or London.

[Update: Two additional items.

One, what makes this whole Repo 105 story strange is, as I mentioned before, there has to be a counterparty to a repo transaction. So the bank(s) on the other end of the Lehman 105 were being given this unusual deal at quarter end: they could loan out $1 and hold $1.05 of collateral. (Not sure if the bank gets to re-hypothecate the $1.05 security for $1.05 to a third party.) Since the talk is that this transaction was unusual, the bank repo desks involved in it probably had to wonder what was up with Lehman.

Two, where were the repo auditing experts at Ernst & Young? The minute the junior or senior auditors came across the 105’s, you’d think they’d flag the unusual transaction? What the general public isn’t aware of is that there are almost always separate “notes to the financial statements” where things about the balance sheet, or items not in the balance sheet, are listed with a brief explanation.

Could it be that even after the great Sarbanes-Oxley Act of 2002, auditors were still looking the other way to keep the audit client?

What these characters were doing set off a chain reaction that many observers feel almost brought the whole financial system down. Up to now, that is, post crisis, we’ve been told that it was the “toxic” mortgage-backed securities that were the trigger to this. And that Goldman Sachs was setting up financial fires to profit off the resulting declines. But now, it looks like pure fraud at Lehman, and one that two outside groups should have been suspicious about: the counterparty banks and the auditor.]