I wrote mortgage-backed securities prospectuses for Citigroup, Son.
Really, I did. You may think I should be ashamed of myself, and that I must be partly responsible for the $7 billion that Citigroup will pay under its settlement with the Justice Department announced July 14. But I’m neither ashamed nor responsible, and here’s why.
I’m not defending Citigroup for doing what the Justice Department alleged, primarily lying about the quality of the mortgages in certain residential mortgage-backed securities (RMBS) it sold. But the size of the settlement seems to be largely motivated by a set of myths about the recession, most importantly that it was caused by nefarious doings in the RMBS market.
(There may have been another motivation for that enormous settlement. As the New York Times reported, the settlement included $2.5 billion for “the financing of rental housing, mortgage modifications, down payment assistance and donations to legal aid groups, among other measures intended to provide relief to consumers.” None of this has much to do with the harm alleged—defrauding investors—but may have a lot to do with AGs getting to fund their own policy hobby-horses without the bother of going through an elected legislature. One would have thought prosecutors would be more sensitive to the ethical problems involved in using other people’s money for one’s own purposes.)
My side of Citigroup’s RMBS business seems to have come off well in the settlement. Citigroup had several RMBS programs. One (Citigroup Mortgage Loan Trust Inc., or “CMLTI”) was housed in Citigroup’s investment bank, and another (Citicorp Mortgage Securities Inc., or “CMSI”) in CitiMortgage, a Citigroup subsidiary headquartered in O’Fallon, Missouri. CMSI originated and serviced mortgages, selling off a good many as RMBS but also holding a substantial portfolio. CMLTI, in contrast, simply bought mortgages and packaged them into RMBS. I worked exclusively for CMSI, and don’t know much about what went on at CMLTI, whose prospectuses were written by Thacher Proffitt, a major law firm in the RMBS world that went under in the recession.
If there was any hanky-panky at CMSI in O’Fallon, I probably wouldn’t have seen it from my perch in New York City. But in any case, none of the five examples of Citigroup skullduggery recited in the Justice Department’s statement of facts involved CMSI; four involved CMLTI and one involved Citigroup acting purely as an underwriter. Whatever went on at CMSI, it apparently didn’t rate a marquee appearance.
Back to the Great Recession. I don’t know what caused it—I’m not sure anyone does at this point—but I’m pretty sure that blaming conniving issuers in the RMBS market is strictly from Hollywood (Inside Job, for example, which is actually a pretty good movie if you don’t take it too seriously). I’m backed in this belief by a convincing (and surprisingly readable) paper by a trio of economists at the Federal Reserve Banks in Boston and Atlanta. They wrote that
“The facts refute the popular story that the crisis resulted from financial industry insiders deceiving uninformed mortgage borrowers and investors. … Borrowers and investors made decisions that were rational and logical given their ex post overly optimistic beliefs about house prices.”
I’m not going to walk you through all of this myth-busting paper, but two of the 12 facts the authors present are particularly relevant for this unapologetic prospectus writer: “Mortgage investors had lots of information” (Fact 6), and “Investors understood the risks” (Fact 7).
A typical RMBS prospectus presented pages of information summarizing the credit scores of the borrowers and the methods used to evaluate their credit, the ratio of loans to sale prices, scenarios describing the results if various percentages of mortgages were to go into foreclosure (more about this below), and much else that the investors—all substantial institutions—wanted to know. Nor did investors stop there. Many demanded and received spreadsheets showing this information for each loan (loan-level detail), which they analyzed using their own proprietary techniques or off-the shelf analytic packages. (The spreadsheets were publicly available on the SEC’s website as “free-writing prospectuses.”) Investors who bought RMBS could also track their investments on a website showing each month’s payment information at loan level.
RMBS weren’t sold like stocks and bonds, where the characteristics of the security are determined before the buyers are found. For RMBS, the underwriter essentially bought a pool of mortgages and then devised a “structure”—that is a system of various “classes” of securities that would absorb all the principal and interest payments thrown off by the mortgage loans. For example, a pool of mortgage loans might be structured so that one class paid 5.5% interest while another paid 6%, and one class might be expected to pay off between 4 and 8 years while another paid off between 8 and12 years, in each case given assumptions stated in the prospectus about the rate at which homeowners would prepay their loans. From about 5% to 20% of the classes, depending on the quality of the loans, would be rated anywhere from AA to junk and would absorb all of the losses on the mortgages until their principal was wiped out; only then would any loss be visited on the AAA-rated classes.
Those classes didn’t just happen; the underwriter structured them to satisfy its clients’ demands for securities with certain payment, maturity, and risk characteristics. To imagine RMBS being sold to a bunch of passive ignoramuses is sheer fantasy.
What went wrong, of course, is that the worst-case scenarios in the prospectuses and the investors’ models, while extreme by post WW II historical standards, were nowhere near as bad as the situation that actually overwhelmed us in 2007-8. But even using the prospectus scenarios, anyone could project that if housing prices went much lower than the prospectus assumptions, things could get very dicey indeed.
Actually, even for subprime RMBS, the tsunami left most AAA investors high and dry: 90% of the AAA-rated classes, often 80% of the offering, rode out the storm without loss. Floyd Norris in one of the Times articles linked above tracked a CMLTI series and found that its AAA classes (89% of the total) had cumulative losses of about 4%. Holders had been reaping interest income of around 6% a year in a very low-interest environment, so their pain was somewhat mitigated. But holders of lower-rated classes suffered a good bit more: The AA classes were wiped out in less than six years, the BBBs in less than three.
The real devastation, however, was in a different form of asset-backed security, the now-infamous collateralized debt obligations (CDOs), whose AAA classes were savaged. (The Fed trio offer an interesting explanation of why the RMBS analysts got it mainly right while the CDO analysts, often working in the same building for the same employer, got it so wrong.)
The Great Recession originated in a huge bubble in housing prices. That’s an unsatisfying explanation because we don’t have much understanding of bubbles. For unexplained reasons, sophisticated observers come to believe that the price of an asset will continue to rise. In that climate much that would normally not be reasonable—like lending to homeowners without checking their income or assets—becomes perfectly sensible. (You can learn all you need to know about the frenzy by checking out the documentary The Queen of Versailles, where a rich couple’s chauffeur mentions that he owns six houses, all in foreclosure.) While many saw a bubble, few thought its bursting would be anywhere near as cataclysmic as it was.
(In 2005, I inserted the following paragraph in the “Risk Factors” section of CMSI’s prospectuses
"Housing price cycle
"A number of commentators have recently suggested that home prices in the United States are at a cyclical high, and likely to fall substantially in the near future. A substantial fall in housing prices could cause an increase in defaults on the mortgage loans, and would reduce the amount that could be realized on foreclosure."
In late 2006 I added that “There have been reports that housing prices in some areas are currently declining.”
I wasn’t psychic; I’d simply read a cover story in The Economist on the asset bubble.)
When things go terribly wrong we hunt for the persons responsible. But often there is no one responsible, and our demand for heads to put on pikes simply reflects something primitive in our psychological makeup—an inchoate feeling that all events are the result of human passion and calculation, a leftover perhaps from ancient systems that saw all natural phenomena as imbued with spirits having human wants and desires. ‘Taint so, but it takes constant vigilance to keep from falling for this sort of emotional logic.
Even as wary an observer as Floyd Norris can get caught up in the passion of the time. While realizing that the Citigroup settlement did nothing for the ostensible victims—the investors—and that the prosecutors never established a link between Citigroup’s dodgy practices and the investors’ losses, Norris still finds “a sense in which Citigroup’s punishment seems reasonable”:
“Many of those most responsible for the worst abuses are long gone or simply untraceable. That includes the mortgage companies that failed early in the crisis and the loan officers and borrowers who lied to get loans approved.
“The big banks are the survivors of a crisis that they helped to create, and they survived in part because the government chose to bail them out. The direct cost to the government of those bailouts may have been recovered, but the indirect costs—in the form of lost tax revenue and increased spending on the social safety net—are still enormous. The big banks are now doing reasonably well, and the effect of settlements like this one can be seen as providing partial reimbursement of those indirect costs.”
But unless Citigroup’s illegal actions to some extent caused the crisis, there can be no argument for “reimbursement.” And here, as Norris understands full well, the verdict seems to be “unproved.”