Destructive Oscillation

We’ve all seen the crystal goblets shatter when the right pitch of sound hits them long enough and loud enough.  The classic “Galloping Gerty” video shows how even steel and concrete is no match for the right kind of oscillation, even from something as ephemeral as the wind.

[Video of Tacoma Narrows Bridge]

A system consisting of yield spreads and leverage can oscillate, too.  Occasionally it oscillates out of control.  It’s the nature of the beast.

What some of those who worship at the altar of “free market capitalism” just can’t imagine is that the system allows and encourages destructive oscillation.  Thus, when the catastrophic oscillation pushes too far and the system breaks down, they desperately look for some other cause for the collapse.

As I watch the attempts to re-write history so that the CRA (Community Reinvestment Act) or Fannie and Freddie become the cause of the subprime mortgage meltdown, I can’t help but wonder why such easily disproved assertions have so many supporters.

By the way, I’ve got some good news for them.  I can actually see a way (with somewhat contortionist logic) to partially blame CRA and the Agencies.  So hang in there, free marketeers, you can still spread the blame where you want it to be, just not for the made-up reasons you cite.

In other words, social engineering did help push the mortgage crisis farther, but not the way you say it did.

Nobody likes being proved a liar.  I would think that goes double for those who sell their opinions in exchange for their paychecks.  So why would commentators, columnists and Fellows of the American Enterprise Institute or the Hoover Institute persist in blatantly lying about the subprime mortgage crisis?

For that select group of intelligent, generally well-informed 6-figure and 7-figure people, I’m sure they know they are not telling the truth, but they do it anyway.

I conclude that they must think it terribly important to put out the misinformation they’re peddling, since they are willing to risk their reputations, and possibly even their paychecks, to sell the falsehoods.

I think they rationalize their actions the way a detective can rationalize lying to a suspect or a prosecuting attorney can set up a “sting” operation and have his agents commit crimes in the process of snaring a corrupt target.  What they are doing is wrong on some scale, and they know it, but they are serving a “greater good” by protecting capitalism from stifling regulations imposed by those they view as anti-business, or even worse, socialists.

Their first mistake is that they went looking for political reasons for the bubble and subsequent meltdown, rather than looking at the economic incentives that drive economic behavior.

I know from first hand that even the most ardent anti-welfare people over the age of 65 take and spend their Social Security checks long after they’ve been paid back their initial investment, with interest.  They collect welfare from the current working class, but the economic incentive is too powerful to take the money, whether they need it to survive or not.

In housing finance, the big incentive is deductibilty of mortgage interest.  That incentive (not available in most of “socialist” Europe) pushed a greater percentage of our population into owning their homes rather than renting, clearly a social goal.

Almost unnoticed by the meltdown reconstruction commentators was another powerful incentive added in the late 90’s, an incentive that drove people to “flip” houses rather than live in them.  The capital gains tax on houses, formerly exempted once in lifetime (and then for only $125K), was changed to once every two years, and to $500K for a married couple filing jointly.

Using a top combined income tax rate of 50% for couples earning in the mid-six figures per year, that exemption gave a couple that could afford to finance and improve a high-end house cosmetically in a “hot” area the potential to make the equivalent of a half million dollars a year in ordinary earned income.   Not only that, they could do it over and over again, and pay no taxes whatsoever.

The house-flipping phenomenon wasn’t quite as powerful in the lower end housing stock, since the typical house in that price range sold for only $150K to $250K, hardly the stuff of half million dollar profits after two years’ ownership.  Still, if you asked a typical working class family in a $200K house how long it would take them to save up $50 or $80 thousand after tax dollars, chances are they would say “forever.”

Still, this did give a broad swath of the population an incentive to make their living by selling at a profit and not pay taxes on the income rather than working and paying payroll tax or self-employment tax on the first dollar of their income and regular income tax above certain mimima.

What else was going on?

Let’s look at the Community Reinvestment Act since that is such a popular whipping boy for some dishonest commentators. Primary among the facts that say the CRA isn’t a root cause of the bubble and subsequent meltdown is that it existed for 30 years before there was a problem.  Second is that the problem loans weren’t CRA loans.

The bad subprime mortgage loans may have looked like CRA loans to those with a heavy demagogic filter over their eyes, since subprime loans were made to people with poor credit scores on less-expensive housing, just as most CRA loans were made to people with poor credit histories in less desirable neighborhoods.  Both subprime and CRA mortgage loans charged the borrowers a premium rate to adjust for additional credit risk.  At that point, the similarity breaks down completely.

After those similarities, the first difference  to note about CRA loans is that they were a really tiny piece of the overall market, on the order of one tenth of one percent, or possibly less.  Second, they were not sold by the originating banks, so they couldn’t become part of the subprime mortgage bonds that went bad.  Third, 90% of the subprime MBS were issued by non-banks, institutions that had absolutely nothing to do with CRA, which only regulates banks who take deposits in formerly “red-lined” neighborhoods.

But I did promise to toss a bone to the CRA-haters, albeit with tortured logic.

Here it is:

After 25 years of experience with CRA loans, banks and rating agencies had enough history on them to expect that they performed more poorly than “prime” loans, but  even with 10% to 15% of the borrowers eventually losing their homes to foreclosure, losses only averaged 3% to 4% of the principal of pool of loans.

Since a typical “bad” credit getting a CRA loan was paying about 2% per annum more than a prime mortgage loan, over the average five year life of the loans, that amounted to 10% in extra interest collected and only 3% to 4% in eventual credit losses.  Done right, lending to poor quality credits was actually more profitable than lending to good credits.

So what did the Wall Street (non-bank) finance companies do?  They took this history, and let the mortgage brokers know that they would buy loans that looked, on paper, the same as CRA loans, and that they would pay a hefty premium.

By 2003, some subprime mortgage loans were being sold at prices as high as 104% of principal.  Add in a point or two of borrower-paid (but financed) origination costs, and a $200K subprime loan might have as much as $10,000 profit in it for the broker.

Soon the mortgage brokers were steering even decent credit borrowers into subprime loans to collect that huge premium.  House prices were going up because the 1997 change in the tax law was giving the entire housing market a boost, and subprime borrowers were able to refinance and take out additional cash at the loan closing, along with the brokers, finance companies, rating agencies, and everyone else in the game.

So, indirectly, the fully documented, retained for life neighborhood CRA loans to weak credit borrowers established a track record of performance that encouraged pure capitalists to copy, even though the new loans were no longer held to maturity, but sold immediately into bond deals.

Pretty cool, eh?  You can still blame the CRA program for leading to the subprime explosion, but you have to change the story line, and say that the better-than-expected profits in CRA loans led to private issuers copying the terms of those loans across all types of neighborhoods.

But the devil is in the details.  While most CRA loans were fixed-rate loans (and until 2002, most subprime loans were, too), the subprime market “borrowed” an innovation from the prime mortgage market to make subprime loans more affordable.

They started putting out “hybrid” ARMs that carried a fixed rate for two years, and then adjusted to a very fat margin over LIBOR.  The typical borrower was encouraged to refinance at the reset date, and everybody got rich.

Now comes part two of the “blame the socialists” logic — Fannie and Freddie.

When Fannie and Freddie were “privatized” they kept some of their special status that they had when owned by the taxpayers.  They had irrevocable credit lines with the US Treasury.  They had never drawn on those lines, but the mere existence gave Fannie and Freddie “super-AAA” status in the bond markets.  We called it an implied guarantee, and it let them raise capital at lower cost than any private competitor.

The quid pro quo for this special status was that Fannie and Freddie were beholden to Congress, and subject to laws that Congress passed to achieve social goals.

One of those requirements was that Fannie and Freddie support affordable housing.  For that purpose, the private-label subprime MBS market seemed to be a perfect solution.

Subprime loans typically averaged $150 – $160K, made against $200K houses (20% – 25% down payment).  In most subprime deals, well over half the loans were fully documented as to borrower income, appraisals, etc., with the loans themselves even made on Fannie/Freddie standard forms.

The beauty of the private-label subprime MBS is that Fannie and Freddie could invest in only the very top of the “credit stack” without taking very much of the actual credit risk of the borrowers.

In a typical deal, that meant the “agency qualifying” loans in the deal supported a bond that Fannie or Freddie purchased, a bond that had 20% or so in subordination protecting it.  At the historic loss ratio of about a third of the principal of a foreclosed loan, it looked like Fannie and Freddie were safe from losses until about 60% of the borrowers defaulted.  Pretty darn safe, in other words.

Another (non-Agency qualifying) mortgage group usually supported AAA bonds that were cut into sequential series (called A1, A2, A3 and A4) and sold to other institutional investors with a little more yield.  It was actually the longest maturity bonds in these series that became the “reference bonds” for the ABX credit default swap index that played such a crucial role in accelerating the mortgage meltdown.

Coming back to the loans themselves and the bonds Fannie and Freddie bought, decades of experience suggested that 5% in losses was quite high and 3% might be normal, so a bond that didn’t get touched until losses hit 20% seemed very safe.

So now you know how to blame Fannie and Freddie – they were buyers of a particular senior class of subprime bonds.

The loans backing the Freddie and Fannie investments were fully documented as to home value and borrower income, and the LTV’s did not exceed 80%.  Aside from claiming that they were 100% LTV, or that they were “liar loans,” or that the housing Agencies guaranteed their credit, or that the CRA was used by ACORN or the unions to force banks to make the subprime loans, the TV “experts” blaming Fannie and Freddie are telling you the truth.

But let’s keep trying to understand the incentives that pushed the system into destructive oscillation.

First among these forces is the incentive for upfront profits to the mortgage brokers and securitizers.  With Fannie and Freddie bidding at yield spreads as tight as 10 basis points over LIBOR for 40%, or even 50% of each deal, the remaining spread over LIBOR, which ranged from 400 to more than 600 basis points, was incredibly juicy.  All the issuer had to do was monetize that ongoing spread, and they were home free with a large profit and all their capital back in a very short time.

They did this by issuing NIMS – Net Interest Margin Securities.  Even assuming very fast prepayments and high defaults, NIMs sometimes sold for 6% to 9% of the face amount of the mortgages.  A Wall Street non-bank finance company that put up 5% capital to buy loans at 103% or 104% of the face amount of the loans could turn around and issue MBS and NIMs debt that sold for as much as 106% or 107%, after expenses, usually in two to three months.

The most aggressive firms (eg Lehman) were nearly doubling their money four to six times a year.  No wonder they all wanted to do it, and no wonder they all did everything they could to crank up the leverage as high as possible.  Even the “less efficient” subprime issuers who “only” got 106% in proceeds from loan that cost 104% and deal expenses of 1% found themselves making 100% per annum on their capital.

Look at it this way:

If a produce seller gets paid by weight, he will do what he can to maximize the weight the customer pays for.

I remember when a friend got back from China in the 1980’s, she warned me not to buy melons in the farmer’s markets if I went because they were often injected with untreated water that would make a Westerner like me sick for days.

Small-scale capitalists like those Chinese farmers aren’t the only ones who game the system.

I’m sure it’s perfectly legal what the big produce companies like United Fruit and Del Monte have done with their hybridizing and genetic selection.  They can even claim that the much thicker peels on today’s bananas protect the fruit from bruising on its way to the market.  Still, the other day I bought a bunch of organic bananas that didn’t carry one of the big brand names, and I was struck by how much thinner the peel was than the bananas I’ve gotten used to.  It was like banana peels were when I was a kid (and yes, the organic bananas did bruise more easily).  Still, when I think about the percentage of edible fruit I bought, those organic bananas that were probably an “old” strain had a much higher fruit-to-peel ratio.  They were tastier, too.

It’s all in the incentives.

If we were to summarize the incentives that led to the enormous bubble in subprime mortgage debt, I think the first among them is the lack of regulation on the lenders and brokers, so the sale price of the loan into a securitization drove the design of the product offered.

Combine that with the “infinite” leverage of being able to issue long-term debt more than equal to the cost of the raw material, and you have the perfect set of conditions to encourage lenders to get every single loan that makes sense, along with hundreds of billions of dollars in loans that were doomed when the bubble burst.

From the borrower perspective, the deductibility of mortgage interest and the tax exemption for gains from selling a primary or secondary home led millions of borrowers to “stretch” their sworn and notarized statements of income to qualify for the biggest loans they could get.

Wholesale loan purchasers that paid 101% or at most 102% for prime loans but 103% or 104% for subprime gave mortgage brokers enormous incentive to originate subprime loans.

CRA was, and is, a gnat on the elephant.

Fannie and Freddie were huge buyers with enormous leverage who could afford to buy AAA subprime bonds at L+10 (their cost of funds being L-10 or L-15).  Other investors were forced to look elsewhere for decent returns, and to play the game of “off balance sheet” vehicles like SIVs.

Wall Street created a new class of investors when it virtually sponsored CDO managers and encouraged them to buy the lower-rated classes of MBS from subprime deals.

I still think that mortgage lending and securitization was one of the best ways to lower the cost of housing for borrowers that actually live in the homes they are financing.  I think that kind of securitization, which I call “primary securitization,” deserves to be saved.

I still think home ownership stabilizes neighborhoods, and provides the best single way for a family to save for retirement and weather the ravages of inflationary periods with their savings more or less intact.

I think that the knee-jerk anti-regulation crowd has it all wrong. Imagine a world without weights and measures insured by government to see what they are striving for.  Every time you paid for something by weight, you’d have to negotiate what the weight really is.

In the end, the biggest single financial obligation taken on by most people in their lifetimes can’t be left to a wild west laissez-faire “free market,” but needs prudent regulation that applies to the activity of mortgage lending, no matter what kind of company the lender has chosen to call itself.

If we don’t do this, we are doomed to having the most aggressive risk-takers once again set the terms for every investor and every borrower.  After all, whoever pays the highest price sets the terms for everyone else unless they want to go out of business.  We tried that.  It was a disaster.

hh