Small Enough To Fail: A Systems Approach To Financial Systems Reform

Introduction The “Credit Scrunch” of 2007 was a systemic failure. Interactions between elements of the system (banks, rating agencies, regulators, governments, financial instruments, etc.) mattered more than the specific behaviour of a particular actor. If you believe the crisis was an apocalypse or foreshadows an apocalypse, then you should be considering fundamental reform. You want to redesign, and design principles would be handy. What might they be? Systems theory, touching on if not encompassing chaos theory and complexity theory, gives us a rich background to some simple design parameters, namely input process output, governance, monitoring, feed back and feed forward. Systems theory goes some way to explaining why financial systems, due to their large amount of feed forward (positive feed back), tend to exhibit “fat tail” outcome distributions and more instability than physical systems. Bob Giffords, the technology analyst, groups together feed back, monitoring, feed forward and governance components as “feed through”, highlighting the effect of people’s perceptions on the probability of future events (Mainelli and Giffords, 2009). If people change their perception of a risk, e.g. terrorism, that perception feeds through to alter future behaviour, such as passenger levels on public transport. Systems with feed through – and human systems are marked by this – typically have non-normal event distributions. Systems theory suggests that, for the sake of resilience and robustness, larger systems should be broken into smaller, more discrete and redundant systems where possible. Feed-through suggests a need to slow processes down and introduce variety, leading to recommendations for countercyclical mechanisms, but oddly perhaps considering slowing down or breaking up information flows.