Lehman Brothers used an accounting gimmick to increase a key percentage in its balance sheet, thereby appearing to be better off that it actually was. This accounting gimmick was used to hit certain target numbers in an attempt to (1) keeping stock prices from dropping, and (2) forestall regulators from stepping in and forcing bankruptcy at an earlier date.
In the end, it was not enough to save the company.
Lehman capitalized on a “bad” accounting rule that allows what is called off-balance sheet financing. Specifically, Lehman was able to remove liabilities from the balance sheet, thereby reducing leverage (the relative percentage of financial liabilities) and increasing the relative percentage of capital (or stockholders equity).
Was it an attempt to deceive? Obviously, but activities that fudge some number happen every day in every company. Was it material, or large enough to deceive? At first glance the numbers involved are pretty small. I’ll walk you through the analysis, but you have to go through a lot to find any percentages boosted a material amount. But materiality is a judgment call, and Ernst & Young might not have reasoned the Repo 105 use was material (I’ll explain how to justify E&Y’s position).
The accounting rule. Many financial institutions have short term needs for cash. The business practice is to borrow money and transfer ownership of high-grade securities as collateral. A few days later, the loan is repaid (with a small fee) and the ownership of the collateral reverts. If the collateral’s value does not fall within 98-102% of the amount borrowed, it should be accounted for as a sale of securities. But if the value of the securities is within that range, should be accounted for as a loan. Don’t let your eyes glaze over, I’ll explain it all.
Impact of loan accounting. I’m going to start with a hypothetical company with assets of $100, liabilities of $70, and stockholders equity of $30. The company wants to pay off a current debt of $10, and doesn’t mind borrowing to do so. Therefore, it borrows $10, increasing the asset cash and increasing a liability such as notes payable. This loan takes place with (or without) collateral, but there’s no entry for the collateral because the company still owns and possesses the asset. Then, it uses the cash to pay off the other liability. The net impact is that the company ends up where it began, with assets of $100, liabilities of $70, and stockholders equity of $30.
Notice the number shaded in yellow, SHE as % of assets. As they say, keep your eye on the bouncing ball and that number is about to bounce.
Impact of sale accounting. Start with the same balance sheet of 100-70-30. Now, the $10 of collateral is “sold” to another party for $10. There’s an exchange of assets, as the collateral goes off the balance sheet when the cash goes on. The cash is then used to pay off the other liability. The net impact of the “sale” is that both liabilities and assets are decreased.
What we really care about is that liabilities go down when SHE doesn’t. That increases SHE relatively (and therefore SHE as a percentage of assets). The key percentage (SHE divided by assets) increases by 11.1%, [(33.3% - 30.0%)÷30.0% = 11.1%]. [ An alternative metric is debt to equity, the liquidity ratio. This ratio goes from 2.33 (70/30) down to 2.00 (60/30). It improves by 1/7th, or 14.3%.]
If the company borrows money in the next accounting period, for the purpose of repurchasing the asset, then the balance sheet is restored to 100-70-30.
Explaining what Lehman did. Lehman, as do all other financial institutions, used repurchase agreements to satisfy some its short term needs for cash. It is unknown how many of these institutions account for these as loans, and how many as sales. They should for them as loans, it is the treatment closest to the economics of the transaction. Cash comes in to the company when the loan is finalized (with high-grade securities serving as collateral), a few days later the company pays off the loan balance with interest (and lien to the collateral is released).
Lehman, though, “sold” the high-grade securities (instead of considering them collateral), then promised to repurchase them a few days later with a fee equivalent to the amount of interest. It is just a matter of what words you choose to characterize the transaction, there is no economic difference between loan or sale.
During Lehman’s financially distressed period of late 2007 and 2008, it rapidly increased the use of repurchase agreements by 2 or 3 times right before the end of an accounting period, and accounted for these as sales. It would transfer ownership to securities (assets serving as collateral) valued at either 105% (for fixed term) or 108% (for equities) of the amount of cash received. The securities used were very high grade stuff for the most part. Actually, it works out to very high grade stuff for about 100% of the cash transfer, and very very low-grade stuff for the 5% cushion. Upon receipt of cash, Lehman would use it to pay down current liabilities. The entire period for these end of period repurchase transactions was 8-10 days, a few days before period end, and a few days after. Not all the securities transferred were fixed term, but right before bankruptcy most were. The timing of these transactions pretty much proves the intention was to reshape the balance sheet by decreasing liabilities relative to SHE.
The astute reader is about to say that selling 105 (or 108) for 100 results in a loss. That’s true. However, the contractual agreement to repurchase creates a futures contract, and an equal-sized gain from the futures contract offsets the loss on sale of securities. Nifty. There are a few minor details left to consider, and I’m still analyzing those.
Here’s the impact of the repurchase transactions on Lehman’s balance sheets for its last annual report of November 30, 2007, and the first two quarters in 2008.
Note the huge increase (7.32%) in SHE as a % of total assets. Bounce! When using the debt to equity ratio (aka liquidity ratio), the boost is magnified a bit. When using regulatory assets instead of just total assets the boost is magnified even more (up to 15%).
Was Lehman entitled to use of Repo 105/108? Absolutely. Even if it intended to influence (or deceitfully change) the numbers reported? Yes, intent doesn’t matter. It found a rule it could utilize to its advantage, and followed it. That’s why I call it a “bad” rule. When used as enacted it could alter a company’s financial results. It would be better if all short-term collateralized financings (aka repurchase agreements) were required to be treated as loans.
Was the change in financial statement numbers material? This is really the crux of the matter. If the change is immaterial, then detailed disclosure is not needed (I still need to research this, but I think it is not needed). This is the question that has Ernst & Young holding its breath, waiting for the judicial answer. It depends on your perspective, and judgment. It can be argued that a change in total liabilities from 96.917% (November 30, 2007) to 96.744% isn’t all that large. The change in the second quarter, 2008 is a bit larger, but interim financial statements aren’t audited. However, we are talking about billions of dollars, here. A billion of liabilities is a large number, and 38.6 or even 50.4 is much bigger, still. It could be argued that $50 billion is material on the sheer magnitude of the number.
Another factor is the junk that was thrown in to make the 5% cushion. If that stuff was nearly worthless, then does that reduce the asset transfer down from 105 to within the 102 boundary?
Right now, I just don’t see what the big fuss is all about. The number differentials are just too small. Although a repugnant practice, Lehman didn’t accomplish much of anything with Repo 105 use. Every little bit helps, though. I wonder what else they were doing.
Although I’m an accounting professor, I’m neither a licensed CPA nor a certified expert for legal purposes. I’m just a blogger, and my opinion doesn’t doesn’t hold much value.
When this goes to trial, the judge (and possibly jury) will decide. They are the ultimate umpire. Who has the best lawyer is going to be a huge factor in deciding whether or not deceit and fraud occurred on the part of Lehman and whether or not Ernst & Young is negligent.
On Tuesday, April 20, Lehman ex-CEO Richard Fuld made a statement before the U.S. House of Representatives Committee on Financial Services. He said:
There has been a lot of misinformation about Repo 105. Among the worst were the completely erroneous reports on the front pages of major newspapers claiming that Lehman used Repo 105 transactions to remove toxic assets from its balance sheet. That simply was not true.
As I showed above, Fuld is correct. For the most part the assets were investment grade.
Another piece of misinformation was that Repo 105 transactions were used to hide Lehman’s assets. That also was not true.
Again, he is correct. They were used to hide liabilities. It’s a subtle distinction.
Repo 105 transactions were sales, as mandated by the accounting rule, FAS 140.
Correct, and Lehman was entitled to use it, no matter how bad the stench.
Another misperception was that the Repo 105 transactions contributed to Lehman’s bankruptcy. That was not true either. Lehman was forced into bankruptcy amid one of the most turbulent periods in our economic history, which culminated in a catastrophic crisis of confidence and a run on the bank.
Again, Fuld is correct. The transactions didn’t hasten bankruptcy. Possibly they delayed it, but not by much.
Did I get this one right? Leave a comment.
Debit and credit – - David Albrecht