Repo 105 Let Lehman Hide Leverage on Reporting Dates
An accounting technique moved tens of billions of dollars off the balance sheet at quarter end and brought it back days later. It was disclosed to nobody and legally supported by an opinion obtained abroad.
The Ordinary Transaction
A repurchase agreement involves selling a security and agreeing to buy it back shortly afterward at a slightly higher price. Economically it is a secured loan, and it is normally accounted for that way. The security stays on the balance sheet and a liability is recorded.
That treatment reflects substance. The firm never really parted with the asset, since it committed to reacquire it.
The Variation
Accounting rules permitted treating such a transaction as a genuine sale in narrow circumstances, specifically where the collateral delivered was sufficiently in excess of the cash received that the firm was deemed to have surrendered control.
Lehman used transactions where it delivered collateral valued at around 105 percent of the cash received, which is the source of the internal name. Under the applicable interpretation, that overcollateralisation supported sale treatment.
Sale treatment removes the asset from the balance sheet, and the cash received can be used to pay down other liabilities. The result is that both assets and debt shrink, and reported leverage falls.
The transactions were executed shortly before quarter end and reversed shortly after. They existed to change a number on a reporting date and nothing else.
The Legal Opinion
A revealing detail is that the firm reportedly could not obtain a supporting legal opinion under American law and instead relied on an opinion obtained under English law, executing the transactions through its London entity.
Seeking a jurisdiction whose rules permit a treatment your home jurisdiction does not is not automatically improper. It becomes damning when combined with the timing and the absence of disclosure, because it indicates awareness that the treatment was not straightforwardly available.
Why It Mattered
Leverage was the central question about investment banks during that period. Investors, counterparties, and rating agencies were assessing whether these firms held enough equity against their assets.
Reducing reported leverage at exactly the moments when it was measured directly affected that assessment. Counterparties deciding whether to continue lending were reading numbers that had been temporarily improved.
The practice was described in detail in the bankruptcy examiner's report, which concluded there were grounds to consider claims relating to the failure to disclose it.
The General Pattern
This is window dressing, and the structural insight is identical to other period end manipulations. Any metric measured at a point in time invites management of that point.
The analytical responses are to seek average balances where disclosed, to compare period end figures against intra period data where available, and to treat large reversing transactions clustered around reporting dates as requiring explanation.
Regulators subsequently required more disclosure of intra period borrowing precisely because point in time leverage had proven manipulable.
The Bottom Line
Repo 105 moved assets off the balance sheet for a few days spanning each reporting date. When a firm goes to considerable trouble to improve a number only on the days it is measured, the number was the objective.