On Thursday, the Federal Reserve will kick off not just U.S. action on Basel III’s capital rules, but also a long-delayed decision on the 2009 Basel II.5 rules governing market-risk capital. Three years later and what did we get? A U.S. decision on market-risk capital that comes about a month after the Basel Committee concluded in the “fundamental rewrite” analyzed in a FedFin report earlier this week that Basel II.5 needs to be junked in favor of a whole new approach to market-risk capital. After all this time and the Basel decision, the U.S. would have done well to make us all wait a bit longer for trading-book rules crafted with an eye towards all the hard lessons the fundamental rewrite aims to cure. But, never fear – next Thursday, they will finalize the Basel II.5 rules, warts and all, so that banks and regulators can figure them out just about the time they need to start all over again with the new, improved version of the global trading-book regulations. Better to tidy up current trading-book capital rules now to fix their most egregious offenses and, then, do them over in full at once so that banks and supervisors don’t tumble over themselves trying to figure out a series of cascading, complex standards that may befuddle them all so thoroughly that none can be held accountable for the urgent increases in capital needed for these vital activities.
The old Basel II.5 rules about to be loosed upon big U.S. banks do pack a punch – about triple the capital allowed for market risk under the 1996 standards that to this day define the official trading-book standards for big U.S. banks. The 2008 financial crisis showed regulators – who sometimes take a lot of convincing – that trading-book positions are subject to stress. As a result, Basel II.5 changes the ten-day horizons and other factors that drive value-at-risk (VaR) calculations, also adding stress testing, credit-valuation adjustments and other improvements better to reflect actual market risk under real stress, not the hypothetical happy talk on which the prior rules are premised.
All of this might be worthwhile for the U.S. to put into effect now if it weren’t for the basic truth underpinning the fundamental trading-book rewrite: VaR still doesn’t work. As a result, the regulators are proposing in the wholesale redesign to replace VaR with an “expected shortfall” (ES) way to model market risk. This, they hope, will do a better job even than the buffed-up VaR in Basel II.5. But, will it?
At best, the jury, including that in Basel, is out on this. ES is at least as complex as VaR and, in some ways, it’s even harder to measure because it bases its stress and liquidity assumptions on things that haven’t happened. As a result, much in ES is subjective and, perhaps, at least as vulnerable to the fun and games that so grievously damaged VaR. Even if ES is better on balance than VaR – as it may well prove – risk results from taking one longstanding, well-understood approach to measuring market risk off the table in favor of an untested one fraught with even more complexity risk than the prior Basel II.5 standards.
The solution to the complexity conundrum for the trading-book rules lies in other aspects of the fundamental rewrite. These address the simple, largely qualitative problems that led to the systemic risk in financial markets – risk, by the way, largely untouched in Basel II.5 despite the capital hike noted above. First, the fundamental rewrite puts the breaks on use of internal models, requiring a series of benchmarks and other controls so that risk can’t be made magically to disappear by dint of the latest, greatest computer model. Second, the standards put a far firmer boundary between the banking and trading books so that banks can’t move holdings between their books as favorable capital winds beckon. To be sure, the measurement methodology for at least one of the proposed boundary criteria is a complexity-risk edifice in its own right, but the concept is sound and an urgent improvement that brings long-delayed rigor to the regulatory-capital framework.
So, back to the FRB and its brother U.S. regulators. By finalizing Basel II.5, they are leaving for another day all the risk-control and arbitrage lessons laid bare by the crisis the fundamental rewrite seeks to fix. At the same time, they are forcing big banks to build a lot of new models soon to be scrapped when the ES standards send VaR to the junkyard.
Could U.S. regulators take current VaR and stress test it much as current capital standards have been stressed since 2009 to build a disciplined interim capital regime without forcing all these implementation costs? Yes. Could U.S. regulators impose now the qualitative controls included in the fundamental rewrite? Yes again. And, would doing this instead of piling one complex rule atop another make a lot more sense in terms of creating a sound capital standard to which both banks and regulators can be held accountable? Right again, but simple won’t satisfy regulators bent on keeping up with the Joneses even if the Joneses have moved to another, better neighborhood.