As we contemplate the prospects for more government support for financial institutions, it is useful to reflect on what lessons we can now glean from the performance of the portfolio the Federal Reserve acquired in assisting the acquisition of Bear Stearns by JPMorgan Chase through its sponsored LLC, Maiden Lane.  At the hearings following the rescue before the Senate Banking Committee in April of 2008, Chairman Bernanke suggested that not only was the support of Bear Stearns’s acquisition by JPMorgan Chase necessary to avoid a potential crippling disruption to financial markets, but the Fed might even make money on the deal. 

Now, one might ask, what were the Fed’s alternatives?  Might it have struck a better deal?  How much of the performance was the unforeseen consequence of the further decline in markets following the Lehman Brothers failure? 

We can’t really answer most of these questions, because of the lack of transparency going into the deal, how Bear was initially valued, how over-collateralized the transaction was, or how much was actually known about Bear’s true condition.  All we know about the portfolio is contained in Appendix II of then-NY Fed president Timothy Geithner’s prepared remarks before the Senate Banking Committee on April 3, 2008.  On the surface, it appeared to be of reasonable quality:

“The portfolio supporting the credit extensions consists largely of mortgage related assets.  In particular, it includes cash assets as well as related hedges. 

The cash assets consist of investment grade securities (i.e. securities rated BBB- or higher by at least one of the three principal credit rating agencies and no lower than that by the others) and residential or commercial mortgage loans classified as "performing". All of the assets are current as to principal and interest (as of March 14, 2008). All securities are domiciled and issued in the U.S. and denominated in U.S. dollars.

The portfolio consists of collateralized mortgage obligations (CMOs), the majority of which are obligations of government-sponsored entities (GSEs), such as the Federal Home Loan Mortgage Corporation ("Freddie Mac"), as well as asset-backed securities, adjustable-rate mortgages, commercial mortgage-backed securities, non-GSE CMOs, collateralized bond obligations, and various other loan obligations.”

The question we can address is, how has Maiden Lane done so far?  The Fed reports on the net portfolio holdings and accrued interest and expenses for Maiden Lane each week in its H.4.1. release, “Factors Affecting Reserve Balances.”  The Fed discloses the net portfolio value based on quarterly revaluations of the assets in the portfolio, estimated as if they were sold on the valuation date in an orderly market, presumably done by the asset manager, BlackRock Financial Management.  We don’t know the specifics, though, of how the estimates were compiled.  There have been three portfolio revaluations since the initial portfolio was set up, and these were as of June 30, 2008; September 30, 2008; and December 30, 2008. 

So far the results aren’t good, and there is no prospect for a profit on the assets.  In fact, the portfolio has lost over 10% of its value, and losses are mounting.  The attached chart details the reported performance of the portfolio from its inception.  Included in the calculations are not only the value of the portfolio but also the accrued interest owed to both the Federal Reserve Bank of New York and JPMorgan Chase.  Not included are the fees paid to the asset management company. 

As part of the terms of the assistance provided, JPMorgan agreed to absorb the first $1 billion of losses.  The chart shows that JPMorgan was in the hole for almost all of the $1 billion from the outset.  The loss appeared to begin to shrink, based on the June 30 asset valuations.  But the shocker occurred in October when the portfolio was revalued using September 30 valuations.  Not only did JPMorgan lose all of the funds it had committed, but suddenly the taxpayer appeared to also be thrown into a substantial loss position of nearly $2.5 billion.  Again, over the ensuing months, the loss situation seemed to moderate; but then the portfolio was finally revalued in the most recent H.4.1 release, and losses skyrocketed again.  At present, losses now exceed $4.5 billion and the taxpayers’ share is now $3.5 billion. 

The declines in the net portfolio value probably reflect not only the deteriorating condition of the housing market, since most of the assets are housing related, but also reflects the likely optimistic valuation of the portfolio at the outset.  The chart also suggests that most of the monthly reports on the performance of Maiden Lane are worthless, since changes in the market value of the portfolio are key to assessing where the entire transaction stands and what the taxpayer losses are.

Three conclusions suggest themselves from this analysis, that have implications for the kinds of transparency Congress should demand from policy makers in future bailout situations.  First, without more detailed disclosure of the initial assets that were included in the deal, it was hard to even guess what the likely risks would be to the taxpayer under alternative economic scenarios.  Second, meaningful reporting transparency requires more frequent valuation of the portfolio than quarterly, when markets are volatile and disrupted.  Third, it clearly was the case that the transaction was not structured with adequate over-collateralization to protect the taxpayer from losses, given the downside economic risks that were perceived for housing-related assets at the time and that were also reflected in the Federal Reserve’s own forecasts for the economy as a whole.