Banking-industry comments on FASB’s new fair-value proposal seem to land somewhere between heart-rending howls and gut-wrenching screams. And, that’s before the impact of the accounting rule is combined with pending regulatory changes on reserves. All of these aim at the same, rightful goal: ending procyclical incentives – but the uncoordinated combination of all of these initiatives will turn the pending pile-up of reform proposals into a crushing heap of jumbled initiatives at cross purposes with each other. Before FASB advances its fair-value rule, it should look hard at what regulators are up to on reserves.
We won’t even try to summarize FASB’s 218-page fiesta of fancy accounting for bank loan and deposit books. Suffice it to say that it would require fair-value (aka mark to market) accounting for loans held to maturity, essentially treating them for valuation purposes the same as assets available for sale or those housed in the trading book. This isn’t easy to do – see the valuation problems that beset even nominally-marketable assets during the crisis. As a result, the proposal creates a wide range of subjectivity that could bedevil investor analytics. So far, though, FASB says it doesn’t care because it despairs of historical-cost accounting.
Bank regulators stand firm with FASB on the view that banks should plan better for credit risk. This is, of course, a big point in the pending Basel III rewrite, where regulatory capital would be far more robust for high-risk holdings. It’s also, though, a critical issue in the often-overlooked discussion of loan-loss reserves. These are allowed to only a very limited extent in regulatory capital. Basel I and II aimed at focusing capital on unexpected losses, with reserves held separate for expected ones. This is nice in theory, but creates perverse incentive for banks to under-reserve to present an overly-optimistic portrait of their capital along with unduly rosy earnings during the boom part of the credit-risk cycle. Of course, upon a bust, all is for naught because capital isn’t sufficient and reserves are non-existent.
Regulators are thus noodling over the role of reserves in regulatory capital. We’ve suggested in the past that they ramp this up and make it clear in the final Basel III rules that reserves count for far more in the regulatory-capital equation than now permitted. As noted, this isn’t theoretical perfection, but it’s a real solution to a serious problem with no adverse, risk-inducing result we know of.
Importantly, international accounting-standard setters are pursuing a path consistent with the one bank regulators should tread. A week ago, IASB proposed requiring banks to reserve against expected, not incurred, losses. This means that, just as with regulatory loan-loss reserves, banks would put aside funds at each point they think reserves might be needed. This will cost a fair amount up front because no one has put reserves aside for expected rainy days, but it will prevent the boom-bust cycle resulting when under-reserved, under-capitalized banks are forced to scramble under stress.
Why can’t FASB go for expected reserves as did IASB? Its reason to date is that expected losses are subjective, creating the potential for earnings management – a big reason the SEC likes FASB’s proposal. But, valuing illiquid assets like loans planned for long-term portfolios is arguably even harder than valuing liquid assets and, thus, even more subjective and ripe for manipulation. None of this is easy, and banks will need to bear a cost no matter what’s done in these reform efforts. But, FASB needs to find a way to work with international agencies and U.S. bank regulators to craft rules that promote sound banking for good customers instead of just punishing all the bad practices that admittedly got us to this painful point.