29 April 2008

Financial Sleight of Hand

Are securitizated mortgages at fault for the subprime collapse?

Yes and no. As the inventor of junk bonds, who was also an innovator in the securitization market, who later went to prison for securities fraud explained, the real problem in the subprime collapse was bad underwriting, and bad security pricing.

But this doesn't get collateralization off the hook. Magicians do their sleight of hand by distracting the vast majority of the audience from what really matters, despite the fact that trick is conducted in plain sight. Collateralized mortgages work the same way. If you make a loan directly to a consumer, you focus like a laser beam on the quality of the underwriting decision in the loan, and are more prone to price the risk appropriately. In a collateralized mortgage transaction, the risk of bad underwriting is still central to the value of the security, but there are many distractions from this risk.

A collateralized mortgage pool does eliminate the risk that you will have the bad luck to get the loan that goes bad, rather than the loan that does not default, by random chance. The average default risk is shared, and interest rates can be priced to exceed that risk, seemingly eliminating or virtually eliminating risk in the deal.

Only, it isn't that simple.

When you have one loan, you simply have to make a good guess about the risk of default. When you have many loans in a pool, you have to know how much of the risk of default is independent from one loan to another (and hence can appropriately be averaged out) and how much of the risk of default is cyclic and will impact all loans.

This distinction is particularly important for adjustable rate loans where rising interest rates foreseeably increase the risk of default across the portfolio.

It is also an important distinction for loans in real estate bubble markets where a widespread decline in property prices is foreseeable, because a decline in real estate prices creates an incentive for homeowners with mortgages to walk away from their homes (particularly in California where a large share of residential mortgagages are non-recourse, and with first time, low down payment home buyers who have no assets from which a deficiency judgment after a foreclosure can be pursued).

As a result, the risk in a collateralized mortgage pool is greater than a mere averaging of default rates and losses in current years and the recent past would suggest. To know the risk that pools of mortgages faced you needed to know the likely future range of possible interest rates, the impact that this would have on default risk, and the likely impact of future real estate prices on default risk, which is hard to estimate because real estate price bubbles are fundamentally localized phenomena.

Investors know there there is some cyclic risk, of course, and so they insisted that the originating subprime lenders make good upon excessive defaults in the pool. This practice also hedged against insider manipulation of the pools by lumping bad loans in one pool and good loans in another without disclosing that fact.

The investors were wise when they imposed these terms. Cherry picking of loans in pools compounded by insider trading incentives didn't happen on a widespread basis. And, as a result of the excess default terms, much of the shareholder equity in companies in the subprime lending industry has been appropriated by collateralized mortgage investors -- buffering their losses and putting reckless subprime lending companies out of business.

But it turns out to be much harder to estimate cyclic risk than it is to estimate total risk in isolated loans. The available data are ill suited separating out cyclic risk from aggregate risk, so it was hard for even very smart people to make the distinction practically. As a result, investors were in a poor position to compare subprime lending company capital that was insuring them against excessly risky underwriting, to the amount of cyclic risk that they faced in their investments. They underestimated the cyclic risk, and as a result, didn't realize that subprime company capital was dwarfted by the cyclic risk in the loan pools.

By removing the underwriting decision several steps from the investment decision, understanding of the underwriting risk declined, so it was easier to misvalue the collaterized securities.

Collateralized mortgage pools also actively encouraged bad underwriting. By offering a purportedly much lower risk way for investors to participate in what had traditionally been a high risk, high transaction cost investment opportunity, the money available to make subprime loans increased dramatically. Since loan originators make profits more or less in proportion to the number of loans originated, the incentive for the subprime lenders was to make as many loans as possible, even if that meant making loans to people unlikely to be able to pay or with poor documentation.

This incentive certainly applied at the lowest level of subprime lending organizations which have pretty simple commission structures. Why, however, didn't senior management insist on taking a more cautious course that would protect shareholder equity from the risk of loans going bad? Surely, subprime lending executives who had to approve the underwriting standards used by their companies and had long experience in the field in most cases, understood that risk better than their collateralized mortgage pool investors.

The disconnect between shareholders and management at subprime lending companies was acute in this case, and it isn't the only case where this disconnection is important in explaining bad corporate decision making. High revenues boosted management compensation at subprime lending companies, even as shareholder equity was put at risk. Subprime lending company boards of directors, like most boards of directors, were rubber stamps for management.

A key terms that collateralized mortgage pool investors failed to bargain for was some mechanism that would align subprime lending company management with subprime lending company shareholders. The false assumption that the two were aligned was another reason that collateralized mortgage pool investors failed to appropriately recognize the risk they faced and as a result mispriced the securities. If shareholder and management interests had been aligned at subprime lending companies, it is likely that management would have insisted on tighter underwriting to control the risk to their shareholders from cyclic risk in the pools whose defaults they partially guaranteed. Management had no strong incentive not to take big risks with someone else's money.

The total dollar amount spent on the excessive part of executive compensation at subprime lenders was small compared to the losses. We could and probably should make subprime lending executives cough up their big bonuses. But the problem wasn't so much that the executive compensation was large in dollars as it was that the executive compensation packages gave executives insufficient reason to manage their companies in the interest of the shareholders.

Junk bonds, discussed in the link above, and the incredibly high levels of leverage found at investment banks like the collapsed Bear Sterns, go to the same idea. Leverage increases profits for shareholders and makes more money available for senior management in exchange for higher risk to shareholders.

Shareholders suffer a lot from downside risk. Management does not. Management can usually walk away with the compensation they received at anything but the eve of bankruptcy (although the 2005 bankruptcy law may change this situation) and usually can find future employment -- perhaps as a manager in their own company if it is acquired, perhaps with some other firm in the industry or a related industry.

Why should the public care if shareholders (predominantly savy institutional investors and sophisticated individuals in the top 5% of wealth with expert advisors)are harmed? The people who own publicly traded bonds that lose value in publicly traded bankruptcies are members of the same class of people who own publicly traded stock that loses all of its value in corporate bankrupcies.

There are several reasons why the public should worry about excessive risk taking by public companies that lead to bankruptcies:

1. Bankruptcies result in abrupt jobs losses to large numbers of people.

2. Bankruptcies delay or prevent payment to trade creditors and employees who are in short term relationships with the company and are not in position to evaluate the credit risks they take in offering short term trade credit. The amounts involved, moreover, are usually of fairly minor importance to the corporation's payout to its long term investors (bondholders and bank lenders for the most part), but are often important to the employees and trade creditors who a stiffed or have to wait to receive what is due to them.

3. Bankruptcies, as a result, undermine the ability of commerce to flow smoothly by undermining the likelihood that undisputed legitimate debts will be paid in full and on time, even with big businesses that are expected to be reliable in paying their debts.

4. Corporate bankruptcies always involve what amounts to a mutiny of management against shareholders to whom the management owes a fiduciary duty. Management keeps getting paid even after a bankruptcy, for a while at least. Shareholders lose everything immediately.

5. Bankruptcies often lead to partial or complete government bailouts harming the taxpayer. Bear Sterns is a blatant example of a government bailout. But any time a company sheds its defined benefit pension plan in bankruptcy, the Pension Benefit Guarantee Company has to step in and this risks government funds to make good on the PBGC obligation. Bank collapses put the FDIC on the hook to make good on private debts. Unemployment payments are a government obligation. And so on.


michaelmalak said...

Three more aspects of the subprime crisis:

1. The "look the borrower in the eye" was affected not only by securitization, but also by the lethal combination of non-discrimination laws and automated underwriting. The Option ARMs and Interest Only mortgages that were intended for salespeople and entrepreneurs who have high but irregular incomes were extended to low-income people who falsely claimed alternative source of income. NINA loans became known as "liar loans". Automated underwriting systems have no means of verifying the probability of future income, and going by gut feeling can lead to illegal racial profiling.

2. Foreclosures are acting as an accelerant. The lenders who made the bad loans are the ones who hold the deeds, and they are absentee landlords, driving down the property values of neighbors. Buffalo and Cleveland are among the municipalities suing lenders over their increased costs for crime and blight control. States should consider outlawing out-of-state lenders from extending mortgages to homes in their state. This would increase stability through vested/shared interest, at the expense of slightly more costly and less liquid mortgages.

3. Securitization can't be blamed for the dot-com bubble. In the cases of both the dot-com bubble and the housing bubble, the underlying cause is, as Ron Paul says, the Federal Reserve printing money.

Andrew Oh-Willeke said...

Mostly, I disagree.

Re: Point 1

Option ARMs and Interest only loans are mostly being driven by the California real estate bubble. No doc loans were a problem, but a separate one.

There also was almost certainly illegal racial profiling going on, which is why such a large percentage of minorities in minority neighborhoods who got subprime loans had credit good enough to get FHA loans instead, compared to members of other ethnicities. Anti-discrimination laws did nothing to increase subprime marketing and the "look the borrower in the eye" approach doesn't actually work nearly as accurately as automated systems.

Re: Point 2

This proposal is unconstitutional. Discrimination against out of state economic players who meet all in state regulatory requirements is not allowed (even for lawyers).

Re: Point 3

Ron Paul is wrong about the importance of fiat money, among other reasons, because crashes similar to the tech bust and the subprime bust were much more common before the gold standard was abandoned than it was afterwards. Markets have been producing crashes since the Tulip bust several hundred years ago.

The dot.com bust was a simpler affair in a lot of ways, with "irrational exuberance" as Greenspan described it as the main cause. Also, one of the main lessons of the economics of "bustology" is that abstaining isn't necessary in the economic best interests of smart active participants, even if they know precisely what is happening. The tech bust is more like historical busts like the tulip bust and the gold bust and some of the early waves of investments associated with industrialization.

In contrast, the subprime bust is more like the panics that resulted in periodic waves of runs on banks prior to the FDIC.

michaelmalak said...

Yes, we have been having financial panics for several centuries -- ever since the Bank of England invented fractional reserve banking. Thus we've never had a "true" gold standard in the U.S., i.e., one without fractional reserve. Fractional reserve should be prosecuted as fraud rather than expected as normal by law.

Regarding banning out-of-state lenders -- you're right, U.S. Congress would first have to pass a bill to allow states to do it.

Andrew Oh-Willeke said...

The currency supply affects of fractional reserve banking are inherent in any economic system that allows people to make unsecured loans (or, at least, loans in excess of a person's assets exclusive of debt instruments). It arises naturally in any non-barter economy with widespread trade credit and an informal lending sector.

Under your proposal, banks would be reduced to depository institutions with safe deposit boxes, similar to street name security ownership firms and the American Depositary Receipt people. Think Gringots.

The argument that such draconian measures are necessary it weak, as evidenced by the stablity of the conventional commecial banking sector in the midst of this crisis.