More: Can science reinvent the economy?
Bubbles are nothing new (see âThe madness of crowdsâ). Now, though, a bubble in just one country can cause the whole worldâs economy to collapse. Globalisation would be hard to roll back â so how do we contain its dangers?
Economists have been worrying about this kind of âsystemic riskâ since the bank failures of the Great Depression in the 1930s. Regulations introduced soon afterwards required banks to hold adequate reserves, and governments guaranteed many bank deposits. For 70 years, that avoided a repeat.
Advertisement
The present crisis owes much to the emergence of a globally interlinked âshadow banking systemâ â investment banks, hedge funds, mutual funds, insurance companies and the like â not subject to these safeguards. Producing policy prescriptions this time will mean penetrating that complex web to find out how it was woven â and what made it break.
The essential tool will be network theory â a branch of applied mathematics used to tease out complex relationships in areas from computing to biology. Last year, for example, economists Domenico Delli Gatti of the Catholic University of Milan, Italy, and his colleagues, including Nobel prizewinner of Columbia University in New York, used the approach to show how a few banks in a network generally with an outsized share of links.
âFinancially robust lenders can supply credit at better conditions and therefore tend to increase their market share, attracting a higher number of links,â says Delli Gatti. Thatâs good in the short term for people looking for money, but it suggests systemic risk grows naturally as a system becomes dependent on a few pivotal institutions. Their failure can trigger avalanches of further trouble.
Financial food webs
That conclusion is borne out by research from of Yale University and his colleagues. They studied how hedge funds compete to attract investors by seeking high leverage â borrowing heavily to amplify the funds they can invest, as well as their potential profits. In simulations of market activity, the researchers show how this creates enormous credit dependencies between banks and hedge funds that can push markets through an abrupt âphase transitionâ from stability to instability, much as solid ice abruptly melts into liquid water.
So how might we engineer our financial networks more robustly? Some scientists think that answers lie in areas where similar risks emerge. Ecologists, for instance, have spent decades getting to grips with the ability of ecosystems to adapt and thrive in the face of all sorts of challenges, including everyday processes like the movement of species, but also continental drift and climate change.
Many stable food webs have a âdisassortativeâ structure that is very different from what seems to emerge in financial networks (Nature, vol 451, p 893). Species with many links to others tend to be linked to species that have very few: an insect may pollinate many different plants, while each of those plants may be pollinated by only one or a few insects. Food webs also tend to be built up out of a , meaning that disruption in one area does not necessarily spread to other parts.
Translating insights from stable ecological networks into policy prescriptions in the very different world of finance will need a great deal of further study, however. For a start, more compartmentalisation might increase a systemâs robustness, but it might also slow the free flow of cash â something that could actually increase the fragility of a financial network in a crisis.