More: Can science reinvent the economy?
The current financial crisis started with an enormous housing bubble in the US, UK and elsewhere. People borrowed ever more against the value of their homes and lenders sold the risk on, in an impenetrable web of financial transactions. Common sense might have pointed to a brewing catastrophe. Yet amazingly, some financial institutions were assessing their risks with models that ignored the possibility that housing prices might fall.
The reason was a mixture of shortsightedness 鈥� a decade or more of continuous growth had obscured the longer-term reality of market fluctuations 鈥� and rampant greed: if there are megabucks to be made today, why worry too much about tomorrow?
Advertisement
That鈥檚 why we need objective signs of trouble building, says physicist Didier Sornette of the Swiss Federal Institute of Technology (ETH) in Zurich. He heads a new lab at ETH called the , which aims to detect developing market bubbles from the mathematical patterns they throw up.
His work indicates that can be done by looking for any quantity, such as housing or stock values, that is rising over a period of time. That is an indication, he suggests, of potentially dangerous positive feedback, in which confidence engendered by upward moves on a market encourages ever more people to invest 鈥� without any objective increase in the worth of what is being traded.
That sort of behaviour can be identified by plotting the logarithm of prices in a given market over time. Anything greater than a straight-line increase is a danger signal. Despite its simplicity, this is an approach that hasn鈥檛 been applied before 鈥� and there is evidence it works. In 2005, Sornette and his colleague Wei-Xing Zhou found the signature of a fast-growing bubble in US housing data over the previous two years, especially in the north-east and the west of the country, and predicted a bursting point of mid-2006 that turned out to be fairly accurate (Physica A, vol 361, p 297).
鈥淲ith the right monitoring, it would have been clear that the last 15 years of excesses were leading to an unsustainable regime that could only blow up,鈥� says Sornette. His group is now developing similar techniques for monitoring, for example, the stock values of the 500 largest US companies.