Every inch of ground in financial regulation is hard fought. Are capital ratios a percentage point too large or too small? How can you decide what proprietary trading is? What is and isn’t a large exposure, who decides and how do you appeal?
Rarely do regulators or their political masters step back and ask whether the foundations of the regulatory architecture are in the right place.
Debates about the right approach to regulation are often clichéd and unworkable — calls for state ownership from the left and calls to tear up the rule book from the right.
But all modern regulation, from Basel I onwards, is premised on a world governed by known, calculable risks. The models generated by banks and their regulators appreciate that tail events could occur, and the principle purpose of regulation is to mitigate the damage of these, or try, in some cases, to eliminate them entirely.
It is trying to understand and capture all of the tail risk in model-based regulatory frameworks that has led to the explosion of complexity in the rules.
But why not go beyond the black swan and start by presuming that capital markets are not based on known, calculable risks? We cannot infer the probability distribution and thus cannot assign probabilities to black swans. Uncertainty trumps risk.
Economist Frank Knight first made the distinction between economic risk and uncertainty in his seminal 1921 book Risk, Uncertainty, and Profit. In the literature that followed, the term ‘Knightian uncertainty’ was coined to describe a risk that is impossible to measure.
One way of approaching economic decisions in a state of uncertainty is by using heuristics — an approach to problem solving where you not only do not, but cannot, have all the information.
In a paper published last Friday, the Bank of England used heuristics, simple rules that ignore part of the information, to (theoretically) propose a complete rewrite of global capital markets regulation. They have not torn up the rule book, they have thrown book up in the air and said “we can’t know”.
One of the consequences is that rule book becomes radically simplified, as it no longer has to capture all of the tail risk.
So the 300-year-old institution has come up with four simple tests that use these principles.
A leverage ratio where the same amount of capital is applied to all assets regardless of their riskiness is one rule.
The others are a market-based capital ratio, a simple measure of wholesale funding and the loan-to-deposit ratio.
These rules, it turns out, perform better in predicting institutional failure than existing, modelled capital and risk-based capital rules.
Splendid! Someone call Basel and let’s get cracking on the rest of (shortened) rule book.