Individual and Social Rationality

Classical asset pricing theory assumes rational, self-interested agents. In essence, it assumes greediness in a most wicked way: selfish and smart.
One of the paradoxical predictions of economic theory is that such greediness is not necessarily bad for the common good. This was first conjectured by Adam Smith, and later formally proven to be right in the form of the First Welfare Theorem, a proposition that states that, under certain conditions, competitive markets will generate allocations that cannot be improved upon. This means that there is no way to make someone or a group of agents better off without making at least one agent worse off. Economists call this Pareto Optimality. But there are conditions. For instance, in a financial markets setting, securities need to be traded that allow one to insure against any risk that could potentially materialize. Not only downturns in the economy, but also earthquakes, hurricanes, and bush fires. Another assumption is that there is no adverse selection, which as we all know is a serious issue in, e.g., health insurance.
The First Welfare Theorem is an equilibrium prediction. It obtains provided markets equilibrate. The theory is silent, however, about when and how markets equilibrate, though. We do have some interesting work on equilibration dynamics if the equilibrium is Pareto Optimal. But, surprisingly, we have also observed that competitive markets have a hard time equilibrating when the equilibrium is NOT Pareto optimal...
One member of the team, Elena Asparouhova, was the first to notice this. For her dissertation work, she studied markets for loans when banks face good and bad borrowers but don't know who is good. She found out that, when the predicted equilibrium was not Pareto optimal, markets had a hard time equilibrating. They would tend to move to the competitive equilibrium, only to veer off because banks ended up offering loan contracts that made everyone (the bank and the borrowers) better off. Only, that opened up opportunities for greedy banks to start offering contracts to "cream-skim" the good borrowers, and the market then tended to go back to the predicted (Pareto-inferior) equilibrium. This work was published in the Review of Finance and received the Goldman-Sachs Asset Management best-paper award.
In more recent work, we have observed this even more dramatically. We set up credit markets in a way that classical asset pricing theory makes a unique, sharp prediction about prices and choices. Yet the predicted equilibrium is bad. The issue is re-financing. If agents are individually rational (selfish and smart), then they would never want to re-finance and the debtor is declared bankrupt. However, if agents are willing to re-finance, the debtor has the ability to re-cover, and everyone is eventually better off. Everybody understands this, and invariably (we have run many experiments!) subjects do re-finance. The dismal equilibrium prediction never materializes... (for a while, at least).
We think that this finding has little to do with altruism, which is claimed to affect the behavior of some significant fraction of humans. Instead, it is better to think about it as "social rationality:" everyone realizes that everyone is better off by not being too greedy.
The situation is unstable though: inevitably, greed takes over, not enough people re-finance, and the debtor is declared bankrupt... We have some interesing descriptive statistics on how the credit market collapses when greed wins over social rationality. (So, one can say that classical asset pricing theory eventually does make the right prediction!)
One wonders to what extent this tension between individual and social rationality explains recent dynamics in the sovereign bond markets in the eurozone.

Bossaerts Team (aka Caltech Laboratory for Experimental Finance)

0 yorum: