Friday, October 10, 2008

Whither PMI?

The favorite talking point on right-wing radio has been that this whole financial mess is the fault of Democrats. It is true that Democrats have been staunch defenders of Freddie Mac and Fannie Mae compared to Republicans. It is also true that the Clinton Administration's passage of the Community Reinvestment Act in 1992 forced mortgage lenders to issue loans to higher-risk, lower-income borrowers than they would otherwise prefer. Both of these set the economy up for a run of defaulted loans.

If this were the only side of the story worth telling, then the current crisis would have manifested solely as an insurance collapse, not an investment banking collapse. The other side of the story worth telling is how this event disrupted the investment banking market so quickly and effectively.

(There may still be an insurance collapse, but it will take longer to become visible than the investment banking collapse, because of the reinsurance shell game -- insurance companies taking out policies from other insurance companies to cover their losses in the event that some external circumstance forces them to pay out more claims against their own policies than their actuaries predicted would occur. AIG was the first domino to fall, but I'm sure its shockwaves are invisibly propagating through the reinsurance community as we speak.)

Why all this focus on insurance? Because I have yet to hear a single pundit, politician, or self-proclaimed economic expert mention PMI. It is standard banking practice to attach mortgage insurance to any loan for which the borrower hasn't put down 20% or more of the purchase price. The borrower pays the premium (factored into his amortized payment schedule) for a policy that indemnifies the lender in the event that the borrower defaults on the loan. The Community Reinvestment Act may have artificially expanded the mortgage lending market into brackets of lower income and higher risk, but it did not preclude those lenders from continuing their established practice of insuring those risky loans with PMI.

A spike in loan defaults from low-income, high-risk borrowers shouldn't hurt the lenders -- because by definition, those are the loans for which they would have required PMI from the getgo. It would just translate into a spike in mortgage insurance claims. But barely anybody is talking about insurance in this mess; it's all about the investment banks.

I view this as proof that CDOs really are the problem. If a CDO is based on 10 underlying mortgages, all of which go bad, but all of which are covered with PMI, then the underlying lenders will recoup their loss, but none of that insurance payout will go to the holders of the CDO. Banks are failing not because their mortgages went bad, but because they invested en masse in CDOs that were not entitled to the PMI payouts of those bad mortgages.

The legislation that allowed CDOs to be created was the Gramm-Leach-Bliley Act, a Republican-championed bill that was passed seven years after Clinton's Community Reinvestment Act. (For those irrational Clinton-haters among you, yes, Clinton did sign both bills -- ironically, a decision he continues to defend even in light of the subprime mess.)

So here's what I want to know. I want some statistics:

(a) percentage of defaulted loans that would not have been made had the Community Reinvestment Act never been passed
(b) percentage of defaulted loans that had PMI
(c) percentage of defaulted loans that had been repackaged into CDOs
(d) percentage of overlap between groups (a), (b), and (c)

If it turns out that the overlap between (b) and (c) is low, that would disprove my conjecture, as would (c) being any lower than 90%. If (b) is anything less than 90%, then the banks have fucked up on an even more fundamental level than we've been led to believe up to this point.

Everybody has been making the general prediction that the economy is going to continue getting worse over the next year or two. I'm going to make a specific sub-prediction: that, just when it seems like the investment banks are finally getting their shit together, we'll see the same thing happen with the insurance business. The reinsurance market is the other shell game that will prove unsustainable under these conditions.


Anonymous said...

PMI is not a silver bullet, though. According to Wikipedia (, PMI fills the gap between the lender's costs and the proceeds from selling the foreclosed property. Consequently, PMI does not just pay off the loan. The lender has to first sell the foreclosed property. If credit is not readily available for investors or average homeowners, the lender will not be able to sell the property. As a result, there may be a spike in PMI claims, but, that statistic will probably be a lagging economic indicator.

Owen T. Cunningham said...

Wow, that's a pretty glaring oversight on my part. Thank you for clueing me in.

In thinking about foreclosure sales, it makes me wonder -- why aren't hordes of foreign real estate speculators scooping up all these properties? A handful of sufficiently motivated Russian billionaires could easily take over a hefty chunk of our future housing market.