A basic view of Econophysics

Econophysics is a relatively new field in which ideas from Physics are used to analyze the economy. The principle motivation for such an application is that economic theories make rather broad assumptions about the homogeneity of various ‘explanatory’ variables in its models. In order to make more accurate models it is important to allow for non-equilibrium states in the economy. So, one can therefore allow the input variables for a model to be a function in itself whereby there is a preferred state but there are other states in which it can exist. This idea is taken from statistical mechanics, in which the basic theme is to determine what energy state a particle would want to be in based on its surroundings. The formulae are all statistically driven, so essentially you can figure out in which state the particle will have the highest probability of existing in. Similarly, the hope is to figure in which state a certain economic variable will exist in.

While the idea that one can allow the inputs to take on various values based on sound economic assumptions is seemingly promising, it appears to me that Econophysics has many drawbacks. The most fundamental problem is that Economics tries to explain how rational humans would make transactions, and the number of rational humans who understand statistical mechanics is minute. So while it seems that making broad assumptions about inputs to models does not account for everything, it is in fact the best way to think about human behavior. Further, macroeconomic issues are always riddled with political problems as well; so rather than include the complication of statistical mechanics, a far more important variable is that of politics. Hence, the scope of use for tools in econophysics would be limited to trying to predict more micro issues such as stock market fluctuations.

Now, if you are like me, by now the biggest problem that you would see with this whole idea is that it is based on probabilities. It is fair to assume that until the mortgage crisis, the economy was behaving in a way in which one would say that the ‘probable’ events were occurring. However, given that it was inconceivable that such a huge crash would occur, it is highly unlikely that any model based on statistical mechanics would have given the right predictions during the mortgage crisis. Hence, while there may be some use to these models in day trading during normal times, it is definitely not the most reliable tool to predict large crashes or high impact events. Evidence of this is the underperformance of various quantitative hedge funds. These funds, while they do not necessarily use statistical mechanics, do use ideas from statistics to predict various stock market fluctuations. Even large, prestigious funds such as Goldman Sachs Quantitative Resources Group saw very tough times during the market crash.

Nevertheless, Econophysics is a cool new field and deserves more attention. Perhaps there is a way to really use these tools effectively.