Original posted on FT Alphaville by Tracy Alloway:
What is it with Lehman Brothers and 105-titled operations? Buried in Volume II of the 2,200-page Court Examiner’s report is the story of Cap * 105 — a commercial property valuation system used by Lehman’s Principal Transactions Group (PTG) and its real estate servicer, TriMont. While no where near as controversial as the machinations of Repo 105, it’s worth a look since it provides some insight into the failed bank’s valuation practices.
Here’s how it worked:
Lehman’s primary method for valuing the collateral underlying its PTG positions was the Cap * 105 method. Cap * 105 calculated the current capitalization of the underlying property (i.e., outstanding debt plus equity invested to date), and then multiplied this number by 105% to estimate the value of the collateral as of the specific valuation date. The additional 5% represented the presumed appreciation of the collateral.
The technique was meant to rein-in valuations in the midst of rising real estate prices, in other words keep the presumed appreciation or property, or value inflation, under control. It basically had a tendency to limit or undervalue Lehman’s collateral, and appears to have had zero market-based inputs.
In early 2007, and to its credit, Lehman Bros decided Cap *105 was no longer a suitable valuation tool given real estate prices were falling — rapidly. Cap * 105 was no longer tending to undervalue collateral, instead it was doing the exact opposite; overvaluing it sans-market signals.
So Lehman began switching to an Internal Rate of Return, or IRR, model.
Unlike the Cap * 105 models, the IRR models were meant to use some market-based inputs, in the form of market-based yields. Here’s how that worked:
Under the discounted cash flow method, an IRR model calculated the current value of collateral by determining the Net Present Value (“NPV”) of all monthly discounted Net Cash Flows (“NCF”). As a first step, the IRR model calculated the NCF produced by the asset by taking monthly expected revenue and subtracting monthly expected expenses. To this result, the IRR model applied a discount rate to produce the NPV of the NCF. In order to reflect fair value, the discount rate should reflect, for both equity and debt investments, the yield an investor would require to purchase the property.
The problem was though, that the switch from Cap * 105 to IRR was slow to happen. According to the Examiner’s report, at least 228 PTG positions still relied on the Cap * 105 method, or a variant of it, in May 2008. That’s about a third of the total PTG portfolio in the second quarter of 2008.
The Examiner also found evidence that (only) 153 positions migrated from Cap to IRR between May and July 2008. When they did switch, collateral values for the positions dropped by 20 per cent.
Furthermore, even when IRR models were used, the discount rate employed was generally based on Lehman’s expected rate of return, or on the interest rate associated with the underlying loans at origination — not, crucially, necessarily reflecting yield investors would want to buy the property.
Par example, from the report:
That’s the discount rate used by TriMont on three PTG properties, versus a discount rate used by a third-party valuation co., Cushman & Wakefield (C&W), for a property called Heritage Fields. You can see that the varying rates result in a difference in collateral valuations of about 20 per cent.
And so, to the Examiner’s conclusion:
The Examiner finds sufficient evidence to support a determination that Lehman did not appropriately consider market-based yield when valuing PTG assets in the second and third quarters of 2008. While the Examiner recognizes that the valuation of illiquid assets requires judgment and that there is a wide range of reasonable valuations for any particular asset, Lehman’s systemic failure to incorporate a market‐based yield generally resulted in an overvaluation of PTG assets. Accordingly, the Examiner finds that there is sufficient evidence to support a finding, for purposes of a solvency analysis, that the values Lehman determined for certain of these assets were unreasonable.
Note, however, that perhaps unlike Repo 105, the Examiner’s report found no evidence that Lehman workers charged with valuing PTG positions breached their fiduciary duties.Related links:
Lack of consistency in Lehman’s asset valuations – FT