Why feedback loops could tie regulators in knots

A hard thing for risk professionals to explain about the market turbulence since the start of the year is that so little has changed that might be causing it. A slowdown in China, falls in the price of oil, European monetary policy, the unclear US outlook – all of these were in investors' minds three or six months ago. At the same time, implied volatility has remained relatively low, which suggests uncertainty about the future is not entirely to blame either.

It is a situation that exposes the shortcomings of the conventional models risk managers typically use, which treat markets as efficiently reflecting how investors expect risk factors to behave in future. The industry has known for a long time that models overlook feedback effects inside the market – how others will react based on price changes, and how price changes will be shaped by how others react. But in the past the models have been seen as good enough to do the job required.

That no longer holds true, partly because of concerns about diminished liquidity and partly because of worries that investors have crowded into unfamiliar assets seeking yield. As a result, the buy side is focusing on how to augment traditional models. Much of what they are doing comes down to tracking the positions of participants in the market more closely, aiming better to understand and predict what others will do in a stress scenario.

Such activities are not limited to the buy side, though. Regulators too are thinking about the risks that arise within markets rather than from outside. The Bank of England is running a ‘desk-based exercise' for asset managers, which it has been careful to differentiate from the type of stress tests it runs for banks. Part of the difference is that for asset managers the exercise aims to simulate feedback loops and identify whether a shock could trigger a self-perpetuating sell-off among a group of buy-side firms.

As they pursue these lines of thinking, though, regulators face a clash of ideas. The thrust of regulation since the financial crisis has been to make individual institutions safer by forcing them to hold more capital.

However, such an approach runs against the views of people like Jon Danielsson, director of the Systemic Risk Centre at the London School of Economics, who first used the term ‘endogenous risk' to describe the risks inside markets. Danielsson thinks the stability of single firms is less critical than whether the system as a whole is unbalanced. As he sees it, capital requirements could act as a transmission mechanism for risk in a crisis by turning organisations into forced sellers. From this perspective, rules that force firms to act in unison can add to systemic risk even if they aim to do the opposite.

Buy-side risk managers might be embracing the concept of endogeneity. But if regulators are to do the same they will have to consider whether some actions taken in response to the financial crisis could add to the problem.