Stress-Free Stress Tests

ProPublica’s Jesse Eisinger has an interesting piece out on the latest round of “stress tests” required by the Federal Reserve. The Fed won plaudits for its handling of stress testing of systemically important banks in 2009 — the fact that the general parameters of two alternative economic stress scenarios were disclosed helped impart confidence to a spooked market that the tests were credible.

Eisinger makes the case that, this time around, the Fed gave freer rein to the banks to define test parameters beyond a single recession scenario, and also allowed banks to make their own estimates of asset quality and losses as well as the impact of their capital management strategy. This has the faint echo of the Basel II Accord’s provision allowing some banks to calculate their capital charges based on an “internal models approach,” in which the bank’s own risk models would support the analysis. That didn’t work out so well.

Whereas the Fed’s 2009 stress tests were designed to determine how much additional capital was required to keep the banks afloat and lend confidence to the market, the current round is designed to validate bank’s decisions to step up dividend payments to soothe a restive shareholder base. Eisinger makes the critical point that shareowner interests and prudential regulatory priorities don’t necessarily align perfectly, and that regulators may be too deferential in the face of banks’ considerable political clout.

We should resist the temptation to shrug and chalk this up to politics as usual in the U.S. because, as the crisis demonstrated so vividly, the stakes are sky high. Regulators need backbones to do what they’re supposed to do.