Monday, May 23, 2011

Did the government cause the Financial Crisis?

A recent blog post by Cato Institute Senior Fellow Dan Mitchell describes a recent article by Peter Wallison, a member of the Financial Crisis Inquiry Commission (FCIC) as "superb".  In this article Mr. Wallison repeats his frequent claim that government affordable housing goals caused the financial crisis by forcing banks to lower mortgage standards.

What disturbs me most about articles of this sort is that they harm the cause of private enterprise.  By trying to excuse private failings while creating a government problem, they look like they are in the pay of big business (as the left has claimed, sorry I don't have a reference handy). Free market solutions are thus discredited as help for big business rather than being evaluated on their merits.  It is better to admit when markets fail and look at why they fail (and I mention a couple reasons near the end of this piece which can be laid at the government's doorstep).

Having read the full FCIC report, including Peter Wallson's dissent, along with other research into the financial crisis, I've become tired of reading Mr. Wallison's repeated claims of government culpability.  He's been giving essentially the same argument for the last two years regardless of the actual evidence, and even regardless of his own numbers.  In no way would I term any of his articles on this subject "superb" in terms of a factual analysis, though they are "superb" for those would would blindly take the view "government is always wrong and the private sector is flawless".

What follows is a critique of Mr. Wallison's FCIC dissent, followed by a few of my own comments on government culpability in the financial crisis.

Peter Wallison's FCIC dissent is a very one sided treatment of the crisis, excusing every possible flaw in the financial industry while magnifying every government flaw.  The short form of Mr. Wallison’s one sided treatment can be seen where he discounts “Failures in Risk Management” as a cause for the crisis, terming these a “hindsight narrative” (FCIC dissent, p. 446).  Yet blaming the government for relaxed lending standards, which at the time were not seen as risky and were adopted for purely private mortgage lending, is apparently not a “hindsight narrative”.  It appears that the government, which Mr. Wallison would most likely characterize as not qualified to set lending standards, is expected to have better foresight into the consequences of its actions than “expert” bankers, who are blameless for using subprime loans as collateral for AAA bonds.

Mr. Wallison’s selective reading of history is apparent in his repeated claim that there has been no significant deregulation of the banking industry in the last 30 years, instead stating that regulation has increased (FCIC dissent p. 445-6).  This must depend on one’s definition of regulation, I’d consider the removal of interest rate caps, forced elimination of branch banking restrictions, removing the separation of commercial and investment banking, and prohibitions on the regulation of new risky financial products traded in opaque markets to be deregulation.  I would hope that nobody denies that the financial industry of 2008 was (and still is) much different from the industry 30 years earlier.

In his latest article, Mr. Wallison again brings out Edward Pinto’s analysis of the mortgage market, claiming that the FCIC ignored this data showing government (HUD) caused the crisis by lowering lending standards.  Yet the FCIC report does discuss Mr. Pinto’s analysis (p. 219), and explains why it was not considered a valid analysis.  Mr Pinto’s analysis uses a very broad definition of “subprime” and “alt-A” mortgages, classifying many more loans in these categories than others.  Using the numbers from the Pinto analysis which Mr. Wallison presents in his own FCIC dissent it is clear why this analysis is not valid.  These figures show that the bulk of bad mortgages, both by percentage and absolute numbers, were in privately financed mortgages, not those backed by Fannie, Freddie, or government agencies like HUD.  From the FCIC dissent, Table 3 on p. 462, about 3.9 million out of 10.2 million private nonprime mortgages (38%) were delinquent as opposed to 2.4 million out of 15.5 million GSE nonprime loans (15%).  By comparison prime loan default rates were under 5%.  Thus, it appears there’s a significant difference between the GSE loans which only Mr. Pinto considers subprime and private market nonprime loans (which everybody considers subprime).

Using mortgage industry delinquency numbers (Mortgage Bankers Association 2Q 2009 delinquency report), two factors identify bad loans.  The first is subprime loans (using the MBA’s narrower definition of subprime, not Mr. Pinto’s very broad definition) and the second is adjustable rate mortgages.  Oddly, loans which are directly part of government housing policy (e.g. FHA) had relatively low default rates.  If government mandated lower lending standards caused the crisis, one would expect government held loans such as FHA loans to have the highest default rates.  From this and other data I conclude that the biggest failure in mortgage underwriting was in purely private loans, not those acquired by the government.

There is still the claim that government housing policy somehow forced these lower lending standards on the private sector and thus caused the bulk of bad mortgages.  There has certainly been pressure to increase low income mortgage lending, and at least modifying (some would say lowering) standards, both from HUD and from Community Reinvestment Act (CRA) agreements.  Yet two factors argue against this as a cause for the bulk of bad loans.

  1. According to Mr. Wallison's own numbers, about 30% of the GSE (Fannie and Freddie) backed loans were affordable before the new goals (and reduced lending standards) began.  The GSE goals were around 50% affordable loans.  Thus GSEs only had to add 20% more affordable loans.  While banks had CRA agreements they needed to meet, most of these loans could be sold to the GSEs. It is not clear how this extra 20% or so of affordable loans could "force" Mr. Wallison's claim that 50% of all loans were subprime.
  2. Mr. Wallison equates "subprime" or "alt-A" with "affordable housing loan".  Not in a specific statement, but it is quite clear throughout his dissent that in his view subprime mortgages only exist in large numbers to meet government goals.  This goes against the evidence that large numbers of these loans were for high end properties.  While apparently being forced to lower lending standards for affordable housing goals, banks also lowered standards for high end loans.  If these "subprime" loans were so risky, why did the banks make them?  If banks didn't think they were risky (as Mr. Wallison claims) then why was it so bad for the government to promote them?
  3. Lending standards have dropped across the board.  Commercial real estate loans, home equity loans, automobile loans and credit cards are all much more easily obtained today than 30 years ago, yet I’m not aware of government programs or quotas which forced banks to reduce their lending standards in these areas.  Personally, I averaged roughly one credit card offer per day in the mail in early 2008 and close to that number of mortgage refinance offers.  It was easy to maintain an overall credit card limit of over 2 times annual income and still have banks knocking down the door to lend even more.  I expected a credit crisis, though in credit cards rather than mortgages.  It was clear banks were not interested in safe, quality lending but instead encouraged high levels of consumer indebtedness.
Mr. Wallison also claims that past housing bubbles, such as 1979 and 1989, did not result in a financial crisis.  This is quite true, yet the financial markets were much different during those earlier housing bubbles.  Mr. Wallison dismisses any role for securitization in the crisis by stating that securitization has been used for decades to finance lending (FCIC dissent, p. 447).  This may be true, but as far as I know it is only in the last 10 years or so that multi-tranche loan pools have become common.  CDOs are a relatively recent invention, as are credit default swaps (CDSs) and other more complex derivatives.  Giving a AAA rating to 90% or so of the value of a pool of subprime mortgages (see the FCIC report’s sample mortgage pool) also only became common in the last few years.

It is true that the financial crisis was triggered by problems in the mortgage market, yet the reason it was so serious (shutting down the full banking system overnight) could not have been due to a few bonds defaulting or getting lower ratings.  Rather, it was the amplifying effect of relatively new financial products which caused the system to collapse like a house of cards.  This presents the dilemma of unregulated banking today.  While people and businesses might sometimes lose money, that wasn’t the problem in Sept. 2008.  The problem was that most if not all of the major sources of credit dried up, leaving businesses having nothing to do with housing on the brink of collapse.  The philosophical issue is how to deal with the collateral damage when extremely large players in the economy get into trouble.

I would argue:
  1. The Financial Crisis could not have happened in 1980 or even 1990, even if mortgage standards were lowered to the pre-crisis levels.  Mortgage debt was either kept by banks (which had to account for the risk), sold to GSEs (which had higher standards), or had to be securitized as traditional bonds, which would have garnered relatively low ratings.
  2. The Financial Crisis was inevitable.  The industry was packaging many forms of consumer debt into high yielding AAA rated bonds.  The lower rated portions were being repackaged as CDOs with a high percentage of AAA ratings.  But the fact is that consumer debt is not as safe as government or high quality corporate debt.  ANY credit reversal would have triggered a similar crisis.  That it was mortgages reflected the higher dollar total of mortgage debt and spiraling housing prices.
This is not to say that government doesn’t share in the blame.  Some areas include:

Mark to Market:  From what I’ve read, Mark to Market rules left firms open to wide swings in the price of mortgage backed securities and other bonds.  Many of these securities being thinly traded outside public exchanges, their values were largely a matter of opinion and guesswork.  While, from what I understand, Mark to Market was adopted to prevent firms from valuing worthless assets or liabilities at face value on their books for extended periods of time, the downside of Mark to Market is to leave firms open to day to day panic swings in markets, when prices reflect immediate fear rather than an honest market value.

Bank Capital Rules:  AAA rated mortgage backed securities were given preferred treatment as capital over other AAA securities.  This increased the demand for AAA mortgage backed securities and made banks highly susceptible to ratings downgrades.

The Bear Stearns, Fannie, and Freddie bailouts:  Up until the Lehman Brothers bankruptcy the government had not allowed any big firms to fail.  The Lehman failure led in rapid order to the financial system shutting down in fear of the next failure.  We could likely have gone through a much smaller crisis if Bear Stearns had not been propped up in early 2008.  Alternately the TARP bailouts, etc. might have been smaller if the government had supported Lehman, reducing or avoiding the immediate panic.

Thus, looking at the facts, I can’t see the logic in blaming government housing policy for the Financial Crisis.  As with the failure of any complex system, there are numerous factors involved in the crisis.  And as with failures of other complex systems, the way in which the various factors combined to cause the collapse of the financial markets was unexpected and difficult if not impossible to foresee.  Mortgages may have triggered the crisis, but many other factors had to be in place to cause the disaster which resulted.