All this should have set off alarms among the nation’s financial watchdogs. Some regulators were clearly uncomfortable, particularly those at the Federal Deposit Insurance Corporation. 3 But few took action to rein in the runaway lending.
The silence was most deafening at the principal regulator of the financial system, the Federal Reserve. Fed officials believed in letting the market police itself; investors would make sure the bonds they bought were sound, or at least priced appropriately. If some homeowners were liable to default, their loans would carry higher interest rates, compensating investors for the added risk. The Fed, beginning with Chairman Greenspan, was philosophically predisposed to believe that markets worked efficiently if left alone; there was no need for heavy-handed oversight.
After the crisis, some argued that regulators lacked the legal authority to intervene more aggressively in the mortgage market; this is untrue. Congress had given the Fed authority to act against unfair or deceptive mortgage lending as far back as the early 1990s. 4 But regulators did not begin to use this power until late 2006, too late to stop the housing bubble from imperiling the economy. 5
Regulators were hampered to some extent by the complexity of their own system—an alphabet soup of agencies with partial and overlapping responsibilities. In addition to the Fed and FDIC, these included the OCC, the OTS, and the NCUA. The Federal Trade Commission, Securities and Exchange Commission, and FHFA also had a say. 6 With so many voices, it was hard to reach consensus on any action that would cut across all financial institutions. Many financial institutions, and arguably the ones that did the most egregious lending, were the most lightly regulated. These were finance companies with names like New Century Financial, Ameriquest, and Novastar.
That the credit rating agencies weren’t on high alert over mortgage-backed securities was also in a sense a regulatory failure. The agencies had significant power over bond markets because of capital and liquidity standards that required banks to use ratings in their risk management. Wall Street banks needed ratings to sell any bonds, and particularly mortgage-backed securities. Few investors were equipped to evaluate these extraordinarily complex debt instruments; gathering the data needed to understand their component parts and untangle their structures would have been an overwhelming task.
The agencies issued tens of thousands of ratings on mortgage-related securities during the housing boom, and in hindsight it’s clear that many of those ratings were too positive. In many cases, agencies began downgrading their opinions not long after they were issued; it had quickly become clear that the securities were much riskier than their original ratings implied.
Several types of errors skewed the agencies’ opinions, starting with the quality of the underlying data they were given to evaluate. Rating agencies typically assume the information they receive from issuers is correct, and historically it mostly has been. So when the agencies received data from mortgage lenders—about homebuyers’ debt levels, income, purchase prices, and so on—the agencies took it as true. The agencies made no secret of this; they didn’t consider it their responsibility to verify such data, and thus couldn’t tell when homebuyers were stretching the facts or lying outright. As a result, ratings on mortgage securities worth trillions of dollars were based on falsified data.
The rating agencies missed the mark badly enough on many mortgage securities to stoke some long-smoldering criticism of their basic business model. For years critics had said the agencies were subject to a conflict of interest, because they are paid by people who issue bonds (as opposed to the investors who buy them). The agencies thus allegedly have an incentive to produce favorable opinions, regardless of their accuracy. The higher the rating, the higher the price issuers can obtain for their securities. Some critics charged that bond issuers routinely shop for good ratings, steering business to the agency offering the sunniest opinion.
There is little evidence for this. Ratings are based on statistical models, which constrains analysts’ ability to rate bonds for spurious reasons. Still, because issuers pay the agencies to rate their securities, there is at least an appearance of a conflict.