Slightly over two years since the collapse of Lehman Brothers, the financial sector reforms in developed economies are still evolving. One definitive piece of legislation has been the Dodd-Frank Act, passed earlier this year by the US Congress, to restore US financial stability. How effective would the Act’s provisions have been, starting in 2003-2004 (years during which the housing credit boom took hold) and until the fall of 2008 (when the financial system had to be rescued)? Would the Act have prevented the enormous build-up of leverage on financial balance sheets all betting against a material correction in the US housing market?
Let’s go back to 2003. One of the most staggering statistics of the credit boom was that the balance sheet size of the 10 largest global banks more than doubled, from about €7 trillion to €15 trillion during this period. And, during the same period, the regulatory assessment of the risk on their balance sheets (assessed for computing the banks’ Tier-1 capital), moved far more gradually from €3.5 trillion to under €5 trillion. The system was deemed to be very well capitalized in the second quarter of 2007—indeed, better capitalized by this standard than in 2003. Something was clearly amiss.
The apparent safety of the financial sector’s collective balance sheet was attributable to the fact that the biggest global banks had amassed vast quantities of AAA-rated (“safe”) tranches backed by residential mortgages. These assets had historically been safer than similarly rated corporate loans. This was the principal reason behind their lower risk charge (by a factor of five) under the Basel capital requirements that were in place for European banks, for allowing the US commercial banks to park these in off-balance sheet vehicles with little capital, and letting investment banks use internal models for risk management that largely ignored the tail risk of a secular housing collapse.
Even accepting that the AAA-rated mortgage-backed securities (MBS) were indeed safer than corporate loans at the time—in itself a strong assumption for the period ahead—regulatory capital requirements ignored the fact that the entire system was at risk to a common shock, the risk that mortgage defaults could reach levels at which AAA-rated tranches could take some losses. Such financial fragility—the extraordinarily high level of exposure of the system to a common asset shock—is not being sufficiently discouraged by the Dodd-Frank Act.
True, the Dodd-Frank Act will require systemically important financial institutions (SIFIs) to be identified and to be subjected to higher capital and liquidity requirements. But these requirements are unlikely to be raised in the near future, given the weak state of global economic recovery. Nevertheless, suppose a new 8% Tier-1 capital requirement had existed in place of the actual 4% in 2003. Would that have effectively curbed mortgage lending?
The problem in the build-up to the credit crisis was not the level of the capital requirement but its form. Suppose the level of the capital requirement is raised but there is no change in the Basel risk weights. The AAA-rated MBS would continue to enjoy a one-fifth risk-weight charge, compared with AAA-rated corporate loans. Consequently, the basic distortion favouring mortgage finance in the economy would remain. Worse, by raising the capital requirement, bankers face a lower return on equity (ROE). So to restore their ROE, bankers would tilt their portfolios even more towards MBS, leveraging more in an economic sense, yet remaining safer in a Basel risk-weighted sense.
There are several things that could be done differently in the Dodd-Frank Act to avoid such a correlated build-up of mortgage exposures starting in 2003.
First, rather than taking an a priori stance that one asset will always remain safer than other assets, the regulators could reconfirm their assumption by applying an annual stress test of the financial sector based on the composition of assets in different banks’ portfolios. If all of them were concentrated in mortgages, they would hardly represent a “safer” asset class from the standpoint of stabilizing the system at large. Indeed, the Reserve Bank of India (RBI), recognizing the excess concentration to retail housing risk in Indian banks, raised the risk weights on this asset class in the year before the crisis, and averted a potential hard landing in 2008. There is an important lesson in RBI’s macro-prudential move for regulators abroad.
Second, the systemic risk itself could be assessed using simple methods that investigate whether banks’ equity returns imply greater systemic risk—for example, if they are more correlated with the overall market or the financial sector as a whole. If applied during the pre-2007 period, our research shows that such measures would have identified that the most systemically risky institutions were the investment banks (also most highly leveraged) and then Fannie Mae and Freddie Mac. Charging these institutions with a higher capital requirement would be far better than simply raising the level of capital requirement uniformly for all players, as Basel 3 reforms are proposing.
My “back to the future” exercise has its limitations to be sure. We do not want legislation that will help us to win the last war, or only the next one, but it is equally dangerous to think the next one will be different altogether.
Viral Acharya is professor of finance at the Stern Business School at New York University
Comments are welcome at firstname.lastname@example.org