THE BIRDS AND THE BEES, AND THE BANKS / THE FINANCIAL TIMES COMMENTARY & ANALYSIS ( VERY HIGHLY RECOMMENDED READING )

The birds and the bees, and the big banks

By Andrew Haldane and Robert May

Published: February 20 2011 19:59

Regulators want big, complex banks to hold larger buffers of capital to protect the financial system. Big banks argue this is unnecessary because risk is diversified across their larger balance sheets. Who is right? Natural sciences – especially epidemiology, ecology and genetics – provide clues.


Are big banks less prone to failure? The traditional economics of diversification suggest so. By scaling up balance sheets across different classes of asset, risks to portfolios will tend, on average, to cancel each other out. Aggregate balance sheet risk is dampened the bigger the balance sheet. Big banks thus benefit from a law of large numbers.




Complex systems – those found in nature, but also in finance – tell a different tale. Here, scaling up risks may cause them to cascade rather than cancel out. The bigger and more complex the structure, the greater this risk.


Why? Because size and complexity increase the chances of cross-contamination of the whole barrel, even if there is only one bad apple. Errors do not cancel; they cascade. There is a flaw of large numbers.


Complex systems theory has evolution on its side. Error cascades can explain why complex ecosystems (like rainforests) tend to be more fragile than simple ones (like savannahs). They explain why there appear to be natural limits to the scalability of computer programs. And they provide insight into physical disasters, such as Three Mile Island. In banking, they help to explain financial disasters at institutions such as RBS, AIG and Citibank.


In complex systems, diversity matters more than diversification. For example, genetic diversity improves disease resistance. Diversification and diversity might be thought to pull in the same direction. But in finance, they are often engaged in a tug-of-war.


In the run-up to financial crisis, the pursuit of diversification saw banks move into the same business lines at around the same time. Correlations among banks’ equity prices converged on one. Diversification by individual banks generated a lack of diversity for the system as a whole. That lack of diversity in turn generated system-wide fragility and, ultimately, collapse.


Historical evidence broadly bears out these findings. There is no evidence that failure probabilities are lower among big, complex banks than smaller ones. But even if there were, the case for big banks holding higher levels of loss-absorbing capital would not be weakened. That case rests not on the probability of large banks failing, but on their system-wide impact. What matters is not a bank’s closeness to the edge of the cliff; it is the extent of the fall. And this will depend on a bank’s size, complexity and numbers of market counterparties.


These basic principles have long been known in the study of infectious diseases. Optimal strategies for preventing disease spread focus on “super spreaders”: not those most likely to die, but those with the greatest capacity to infect counterparties. The same calculus applies to big, complex banks. These super-spreaders of the financial world have huge balance sheets and often comprise thousands of distinct legal entities. Their numbers of counterparties are often mind-boggling. When Lehman Brothers failed, it had more than 1m such relationships. These spread financial infection on a global scale.


Fortunately, epidemiologists provide us with a simple preventative solution: target the super-spreader. For banks, this means the largest, most complex and most interconnected banks should hold higher amounts of loss-absorbing capital. This lowers the chances of them contracting disease, thus heading off its contagious consequences. Rather than seeking to equalise the probability of failure across institutions (irrespective of size), regulation would seek to equalise each bank’s contribution to systemic risk.


The present situation in banking is in many respects perverse. The magic of diversification, when assumed into banks’ risk models, means that large, complex banks often hold less capital than their smaller, simpler brethren. The rocket-scientists building models tell us this makes sense. But the rocket-scientists building rockets tell us it is nonsense. This error has cost the world dear. Through this year, the Financial Stability Board is leading the charge to boost loss-absorbing capital for the largest, systemically important institutions to correct this error. It is right to do so.


The writers are executive director for financial stability at the Bank of England, and a professor of zoology at Oxford University and former British chief scientific adviser


Copyright The Financial Times Limited 2011.