The Fed and Treasury put the Financials through the wringer on this one. There is no precedent for this. The group of 19 were given one choice in this matter; either they consent to the process and agree to the conclusions or they go out of business. The circumstances justified the heavy hand of the government. There was a very troubling question in depositors minds, “Is my bank safe?” Even more significant was the question in investors minds, “Should I risk my capital in theses entities?” As of Friday both of these questions appear to have been answered in a positive manner.
Unfortunately, the two largest financial institutions in the United States were able to avoid the rigors of the stress test. Fannie Mae and Freddie Mac should be put under the same microscope as the commercial banks. This is very much a case of, ‘What is good for the goose is also good for the gander’.
Putting Fannie and Freddie to the stress test is relatively easy. Unlike the commercial banks they do not have a diversity of assets. The Agencies only risks are in residential mortgages. The two Agencies have the same business model. They both have a large book of funded mortgages and a significant off balance sheet guarantee business.
The following chart sets out the Feds parameters for evaluating mortgage assets.
On a combined basis the Agencies have approximately $1.4 Trillion of Sub Prime and Alt-A exposure. Based on the Fed’s criteria this would require at least $200 billion of new equity just from this category of the Agencies balance sheet. An average rate to evaluate the Agencies total book of business would have been in the 6-8% range if the Fed had used the same standards to evaluate the Agencies as they did when evaluating Wells Fargo Bank. For the basis of this discussion a modest 7% will be used to stress test Fannie and Freddie.
This chart shows the current amounts outstanding in the Agencies funded and guarantee books.
Multiplying the total risk book by the 7% haircut produces an addition capital requirement of $700,000,000,000. This amount is approximately 10 times the total amount of capital required for the 19 public financial institutions that were subject to the Fed’s stress test. The average Baseline haircut of 4% suggested by the Fed produces a capital shortfall of $400 billion. Applying the Federal Reserve Board's rules to the Agencies demonstrates just how large our problem is.
The cost to the taxpayer for cleaning up the Agencies will take many years to calculate. It may take a decade to stabilize these important institutions. The guidelines established in the Fed stress test do provide some insight as to the magnitude of the losses we may face. Those losses will certainly exceed the Baseline Case of $400 billion. It is quite likely the losses will approach three quarters of a trillion dollars.
It is difficult to put these very big numbers into perspective. By way of comparison, the losses at the Agencies will probably be larger then all of the costs that will be incurred in the Iraq war.
A billion is $1,000,000,000.
ReplyDeleteA trillion is $1,000,000,000,000.
7% of 10 trillion is $700,000,000,000, not $700,000,000.
Be careful with figures.
But otherwise an excellent post.
tks anonymous.....I will watch my zeros more carefully.
ReplyDeleteI am watching the private mortgage insurers. If they fold, Fannie and Freddie will have to take very large write-downs.
ReplyDeletei have never seen an estimate of GSE assets and commitment greater thank 5 trillion. how sure are you about the $10 trillion figure?
ReplyDeletei'm afraid that you have made some analytical mistakes in calculating the GSE's net exposure.
ReplyDeletethe GSE's have two lines of business:
1. a Single Family Guarantee Business that takes mortgages originated in the primary market and transforms them into MBS (or in the case of FHLMC, PC's) that can be sold in the secondary market. as part of this economic function the GSE's assume the credit risk associated with the underlying mortgage pools (net of any credit enhancement purchased from a mortgage insurance company).
2. a portfolio business where they buy mortgage securities that are available in the secondary market. through this business line they assume interest rate and liquidity risk. they do not take explicit credit risk. that is held by the securitizer of the mortgage (usually, the other GSE's) or other investors via lower rated MBS tranches.
what you have called the "funded portfolio" is the sum of the GSE's credit exposure via the single family business and the interest rate exposure in their on-balance sheet portfolio.
thus your "funded portfolio" number represents the maximum theoretical credit risk exposure of either GSE. adding the guaranteed portfolio to this double counts the risk.
furthermore, you really should adjust the "funded portfolio" exposure by excluding any MBS held on balance sheet that for which the credit risk is owned by someone else (ie, private label MBS and other GSE MBS).
that's how you reconcile the actual $5T economic exposure to SFR mortgages the GSE's have with the $10T number posted above.
PS: you may also want to reconsider using the bank SFR portfolio cum loss assumptions for the GSE's. the GSE's own the prime conforming (ie, the safest) segment of the mortgage market and have delinquencies and charge-offs that are MUCH lower than the banks. they also have disclosed on their web sites vintage curve by origination year. it's pretty clear from the data on their web sites that the 03-05 books will top out at ~ 1% cum losses. 06 looks like it's tracking to the low single digits. 07/08 could be mid-single digits. 07/08 is only 30% of the overall exposure, so its total contribution to cum losses will probably be about 1-2%.
basically, it looks from the data to date as if the GSE's will do 2-3% cum losses on the entire book. that's a far cry from 7%, given the leverage involved.