Stress Test (This Is Not a Book Review)

Oh, the Fed. For about four years now, our esteemed Board of Governors has tried to figure out how to apply the latest international bank regulatory standards, better known as Basel III. Just weeks ago, it seemed they had it all figured out, finally wrapping up the standards for the eight biggest banks. Yet now it seems they’re back to the drawing board, with one prominent member suggesting ditching one piece of Basel III and replacing it with their own system. Naturally, this raised banks’ hackles—and concerns about regulatory uncertainty—but for investors, it’s all just noise. Nothing much would change—the Fed would just formalize the status quo.

It started last month, when Fed Governor and regulator-in-chief Daniel Tarullo gave a speech on the evolution of regulations. Toward the end, he mused on the shortcomings of Basel III’s system for judging banks’ asset quality. As he sees it, the self-assessment approach used since 2004 gives banks too much leeway and incentives to game the system. His thesis isn’t new. Banks’ risk assessment models came under heavy scrutiny as 2008 unfolded and remain one of regulators’ scapegoats for the crisis. Some said the math and assumptions banks used to gauge the likelihood of an asset’s default, their potential losses and exposure in that event and how swaps contracts or other derivatives would impact total risk was just plain flawed, ultimately setting the system up for failure when the world woke up. Others think banks had a perverse incentive to make assets look “safer” than they otherwise would, so they could hold less capital against them, freeing up more cash for lending and trading, ultimately weakening their balance sheets and leaving them extra vulnerable when the crisis hit.

As an alternative, he suggests swapping self-assessments for an official Fed exam—essentially judging banks’ Basel compliance through stress tests. In his view, the Fed’s risk assessments are more accurate, objective and complete—instead of relying on the banks’ estimates of loss, with standards and estimates varying from bank to bank, they apply one consistent model for all (or so they say). To boot, they actually measure how an asset would perform during crisis. (Albeit, a hypothetical crisis with arbitrary parameters that don’t match reality.) Never mind that the Fed’s modeling wasn’t sufficient to catch an error in one big bank’s accounting methods. Or that the Fed, too, faces criticism of perverse incentives—make tests too hard, and too many banks fail, potentially killing confidence in Financials. Make them too easy, and they’re a toothless softie. (Another elephant in the room: For all the talk of banks’ incentives to keep thin capital buffers, they also have a big incentive to keep them flush. It’s called staying alive.)

The media greeted Tarullo’s comments less than enthusiastically, warning of uncertainty for banks (Fed tests are opaque!) and the impact of yet another rule change. However, this seems a stretch. There isn’t really anything different between what Tarullo suggested and what the Fed already does! Banks are already beholden to both Basel and the Fed. Every year come test time, if the Fed deems—based on its own models—that banks’ balance sheets will breech minimum capital requirements (as determined by Basel III) in the event of whichever hypothetical crisis they cook up, then the banks have to raise capital. Even if the banks’ own models hold up just fine. Heck, the Fed can already order banks to raise capital even if their capital ratios survive the test but the Fed just doesn’t like something! In practice, the Fed is already forcing banks to raise capital based on its own assessments. The banks’ internal tests don’t matter. Any formal rule change to this effect amounts to triple-stamping a double stamp.

Some say ditching the self-assessment rule would put the Fed out of step with the rest of the world, defeating the purpose of Basel III—a uniform international standard—and instead creating a fragmented system. But this isn’t quite true. The ECB and UK, for example, already do exactly what the Fed does. Officially, their banks are beholden to Basel III’s internal assessment standards. But in reality, the stress tests trump all. Last year, the UK forced banks to raise £25 billion in fresh capital after stress tests—complete with the BoE’s own risk modelling approach—suggested loan loss reserves couldn’t handle a crisis. RBS alone was £9 billion in the hole. The ECB is conducting stress tests as we type, complete with their own “Asset Quality Review.” Officials (and the new regulatory code’s fine print) have intimated banks that don’t measure up could face Cyprus-style depositor bail-ins or be shut altogether. All of this supersedes Basel III and its internal ratings-based risk assessments. No one bats an eye.

Whether or not the BoE, ECB, the Fed (or other countries) make their models and stress tests the law of the land is beside the point—what matters more is what actually happens, and it’s already abundantly clear stress tests hold more sway.

Which means not much changes for banks if the Fed rewrites the rules as Tarullo suggested. Some bank execs are grousing, but we suspect this has more to do with their frustration over the billions spent on developing more sophisticated modeling and beefing up their compliance departments. If the system changes, those expenses become sunk costs. But future regulatory compliance shouldn’t become materially more costly or difficult. Banks already beefed up for stress tests, and their solvency was already at the mercy of the Fed’s opaque assumptions. It’s a headache, but one markets have long been well aware of.