In Depth Analysis: CalculatedRisk Newsletter on Real Estate (Ad Free) Read it here.

Tuesday, February 12, 2008

IndyMac: We Were Not Greedy and Stupid

by Tanta on 2/12/2008 08:35:00 AM

It just looked like it to impartial observers. From IndyMac's shareholder letter:

Who is to blame for the mortgage industry's financial losses and also the record number of Americans losing their homes?

All home lenders, including Indymac, were a part of the problem, and, as Indymac's CEO, I take full responsibility for the mistakes that we made. However, objective reviewers of this mortgage crisis understand that home lenders and mortgage brokers were not the only ones responsible. Systemic problems in our secondary mortgage markets and credit markets, and our government's over-stimulation of the housing market via monetary and tax policies (the capital gains tax break on home sales encouraged speculation), were all major factors that contributed to the problem. Indymac and most home lenders were not "greedy and stupid". Most of us believed that innovative home lending served a legitimate economic and social purpose, allowing many US consumers to be able to achieve the American dream of homeownership ... and we still do.

Homeownership is the main way we Americans accumulate wealth, and, in fact, a recent Federal Reserve Bank study shows that homeowners on average have 46 times the personal wealth of renters. As innovative home lending and loan products became more widespread, the result was more people succeeding (in homeownership) and more people failing (losing their home) than ever before. But everyone, including both the government and consumer advocate groups who encouraged this lending via enforcement of CRA lending requirements, also bought into the concept that, if lenders and investors could properly price this increased risk, the higher number of failures was worth the social and economic goals of expanded homeownership. And it worked for many years; the homeownership rate, which had not moved in several decades, expanded from 64% to 69% from 1994 to 2006, allowing 4 million additional Americans the opportunity to have the American dream and build wealth.

However, in retrospect, like many innovations (e.g., the Internet, railroads, etc.), innovative home lending went too far. The housing bubble, caused primarily by the low interest rates for ARM mortgages fostered by the Fed's accommodative monetary policy and even lower rates for fixed/long-term mortgages due largely to tremendous global liquidity, combined with strong demand by institutional investors for assets with higher yields, resulted in a "systemic" underestimation of credit risk. This systemic underestimation of credit risk was not just for mortgages but for many forms of credit. By way of example, Indymac (and many other major financial institutions) has for years used one of the major credit rating agencies' models to assess and price credit risk on home loans. This model estimates expected lifetime losses on a loan level basis, and we closely monitor these average estimated lifetime losses for all of our loan production (that can be evaluated) on an ongoing basis. This particular rating agency revised its model in November 2007 (from version 6.0 to 6.1). Applying version 6.0 to our Q4-06 production (the version in place at that time) indicated an average expected lifetime loss rate of 0.88%, which we felt was a reasonable level of expected losses at which we could properly and adequately price the loans. However, now applying the updated version 6.1 to this same Q4-06 pool of loans results in an average expected lifetime loss rate of 1.88%, a 114% increase in expected losses in one year. This clearly indicates the extent to which the systemic underestimation of credit risk took place in the mortgage markets. As we began to realize this, we tightened our guidelines throughout the last year, with the result that our average expected lifetime loss rate for Q4-07 declined to 0.45% based on version 6.1, a 76% reduction in credit risk as compared to Q4-06, boding well for the future credit quality and related credit provisions/costs of our new business model.

Why didn't mortgage lenders see that things were going too far?

Lenders didn't see that things were going too far, partly because we were too close to it, but mostly because objective evidence of this credit risk did not show up in our delinquencies and financial performance until it was too late to prevent significant losses. And there were many events along the way that confirmed for those of us who believed that innovative home lending was possibly a paradigm shift (similar to widespread ownership of stocks by consumers) and definitely a legitimate marketplace: major financial institutions were offering these products and spending billions to purchase companies who specialized in these products; Wall Street firms and broker/dealers of major banks were underwriting our and others' transactions and also spending billions as recently as 2006 to buy non-GSE lenders in order to vertically integrate their home lending and securitization activities; major mortgage and bond insurers were insuring individual mortgages and pools of mortgages or bonds created from these mortgages; major credit rating agencies were providing strong ratings on our and others' transactions; and major investors around the world were purchasing these mortgage-backed bonds and even CDOs backed by these bonds (something we home lenders had no involvement in or awareness of). Very few in the private sector or in government predicted that the bursting of the housing bubble would be so severe and would result in the current wave of delinquencies, foreclosures and credit losses and the eventual collapse of the non-GSE secondary market ... even for high credit quality, full-documentation, jumbo home loans.

It is also important to understand that the rapid rise in housing prices is one of the key culprits in this current housing and mortgage crisis. In modern times, housing prices have declined in certain regions of the country but never on a nationwide basis. As a result of this fact and the important social and economic benefits that are clearly derived from homeownership, the government (first through FHA/VA programs and then through the GSEs) encouraged a USA mortgage market built upon very high leverage, with LTV ratios nearing 100% for first-time homebuyer programs. However, as home prices decline, either regionally or nationally, the leverage in a home loan, combined with the leverage of a financial institution or securitization structure, can result in significant losses for financial institutions, investors and consumers. Add to this mix a housing market that has not had a single regional market decline in over 15 years and, in fact, had a huge boom in prices from 2003 to 2006, and you can begin to understand how home lending was impacted. Automated risk-based models, on which the entire market relied, replaced portions of traditional underwriting and credit evaluation, and only in retrospect is it now clear that these models did not perform as predicted during a period of severe economic stress. As events unfolded, this proved to be particularly the case with respect to programs such as piggyback loans and high LTV cash-out refinance transactions, including home equity and second mortgages.
I have, of course, never claimed that it was merely the lenders at fault, not Wall Street or the Fed or the rating agencies or other culprits. Nonetheless, I am still capable of being amazed that the party with the most information about the loans is still willing to blame the party with probably the least information (the rating agencies) for the faulty loss estimates. Yes, the rating agencies should have demanded more information. I say this as someone, like Michael Perry, who knows damned good and well that there is much more information to be had.

I'll skip the part about how we didn't know about those CDOs because it's early in the day and I need more coffee before I can wrap my mind around the implications of that.

Otherwise, for practical matters:
In addition, to prevent consumers from making the wrong mortgage choice in the future, Indymac has decided to adopt as our policy that borrowers without $50,000 in demonstrated liquid assets or $250,000 in demonstrated net worth are not eligible for the following products(2):
1. ARM loans with initial fixed terms of less than five years.
2. Loans with negative amortization or prepayment penalties.
3. Limited documentation loans.
"To prevent consumers from making the wrong mortgage choice." God help me.