Yesterday, the Basel Committee released its comparison of risk weightings at big banks under the capital rules governing trading assets. How incomparable they are surprised even risk-weighting skeptics – variations reaching orders of magnitude were found among the sampled banks. With these results, calls will intensify for a replacement capital regime like Andrew Haldane’s “Frisbee,” a leverage ratio of at least ten percent. But, while many will be quick to blame banks for all the risk-weighting shenanigans, the study shows it’s supervisors at least as much as banks who lie at the heart of disparate risk assumptions. When I last year analyzed what I call complexity risk – that is, rules no one understands with consequences no one anticipates – I tried to detail how much supervisors, not just seemingly loophole-bound big banks, exacerbate this challenge. Fixing Basel means putting supervisors sometimes to shame – a task the Basel Committee has to undertake right quick if it wants to save the global capital construct.
Basel’s study found that supervisors were the main cause of risk-weighted asset (RWA) variations, accounting for 25 percent of the swings uncovered by the cross-bank comparison. And, swings there were – the one between five and 1,000 percent of median RWAs in one asset book was hardly unique. Some of the supervisory policies Basel cites err on the prudential side – for example, the U.K. takes a very different course than Basel on sovereign risk. Others are driven by different supervisory visions of trading risk – e.g., the U.S. approach to correlation — than may be warranted. Others, though, result from supervisory demands or, at the least, concessions to RWAs that meet national policy or financial-market goals with little regard for safety and soundness. Sovereign RWAs are a large part of the problem here – Greece looks pretty good to all too many supervisors – but so too are any number of “optimizing” strategies supervisors have pushed to promote market confidence in home-country banks under acute stress.
The Basel trading-book study is only part of ongoing work to judge RWAs across banks and nations. Up next is a similar study of banking-book RWAs. If it shows anything like the trading-books wild swings – which I think it will and then some – it will intensify the siren song of “simple” capital. Anticipating this, the Basel Committee is already arguing that complex banks need complicated RWAs, but that investor confusion would be mitigated by lots more disclosure.
This cure to RWA fiddles is, I think, little better than letting banks and supervisors continue to go their own way. RWA disclosures to illuminate internal models and flush out variations will be so complex that few investors, if any, will be able to pull them apart to highlight high-risk practice. Worse still, supervisors will get to hide in the bushes, because disclosure aimed at investors will take them off the hook for doing much of anything about the fundamental problem the Basel trading-book study found: themselves.
Is the answer here benchmarked portfolios? Maybe not, but a decision here needs to await another Basel study. Standardized floors against which banks need to disclose results? Quite possibly but not until the market-correlation and procyclical effects of a “herd” approach to RWAs is better understood. Broader adoption of leverage is to some extent a solution, but not if the EU and Japan still say no.
These are all hard questions that require careful analysis to craft a capital rule that balances complexity against clarity. But, as more study continues, the political heat on Basel will only grow. Thus, Basel can study on, but supervisors and banks that want a meaningful global capital accord need now to reckon honestly with Basel’s findings and construct a credible response. If not, we’re back to very high leverage standards in those countries that muster meaningful rules and the Wild West pretty much everywhere else.