BAD stuff happens. But often it doesn鈥檛, because somebody somewhere saw it coming.
That job often falls to government agencies. They find variables they can measure fairly easily to predict serious events 鈥� floods, epidemics, terror attacks and the like. Is rainfall above average for the time of year? Has chatter among terror suspects ramped up? Likewise, as individuals, we do things like check the air pressure in car tyres to keep ourselves safe.
Advertisement
All of this would be senseless if we had no way to measure rainfall or tyre pressures accurately. Worse yet would be to believe our measurements were accurate when they were not. Yet this is the situation with an important set of new banking rules, crafted after the 2007 financial crisis and set to take full effect by 2019.
Known as the Basel III regulations, they require banks to calculate a number (let鈥檚 call it R) to estimate the riskiness of the assets they hold, and keep it below a certain limit.
Banks were supposed to do something similar before the financial crisis, but many found ways to game the numbers. The new rules change the maths a little in an effort to stop that, which sounds eminently sensible. Except for one thing: it appears that banks will simply never have enough data to calculate R with much accuracy at all.
As mathematicians Chen Zhou and Jon Danielsson of the London School of Economics have , making a reasonably accurate estimate of R 鈥� within 5 per cent of the true value, say 鈥� would require decades of financial data on hundreds or thousands of different assets. Banks generally have only a few months or years of data. Many of the assets they hold have only existed for a short time.
Hence, the reality is that the banks鈥� estimates of R will be wildly uncertain. Given that the point of the new regulations is to help control risk, this is a fundamental failure.
Scientists have highlighted this problem in opinions submitted to the Basel Committee on Banking Supervision, which is responsible for crafting the new rules. One was physicist and former banker Imre Kondor of the E枚tv枚s Lor谩nd University in Budapest, Hungary, who made a submission a couple of years ago. He believes the political environment now makes it unlikely that anything can be done to get better rules in place.
Most of us are not equipped to critique the mathematical technicalities of banking regulations, but we should all be concerned about them because they affect our collective future. The banks themselves have little incentive to push for more accurate measures; the illusion of safety the new regulations may create would give them cover to take on excessive risks of the kind that led to the last crisis.
This article appeared in print under the headline 鈥淩isky business as usual鈥�
