Friday, March 20, 2009

Financial Markets and Complexity

On October 3, 2008 I arrived in Italy for an international conference on complexity science and its possible uses and implications for public policy. The program featured speakers who would talk about managing the spread of epidemics using complexity science, planning traffic-control systems with complex systems principles in mind and dealing with the very complex nature of climate change.

The program did not include a topic that was on all our minds: financial market dynamics.

This topic was on our minds, of course, because it was on everyone's minds that day. Earlier in the week investment firms had begun to fail, the stock market had been plunging and governments around the world were scrambling to determine what, if anything, they could do to stop what appeared to be a global financial catastrophe. We all wondered if our bank accounts would be empty by the time we got back to our home countries --
if we got back.

As we sat in the lecture hall, prepared to listen to discussions of epidemiology, traffic planning and climate change, speaker after speaker approached the podium, said a few words about their assigned topic, then switched to -- you guessed it -- financial market dynamics.

The speakers were in remarkable agreement: what we were observing in the global financial system had all the hallmarks of a phase transition, a sudden, discontinuous change that cannot be stopped once it is started. And the speakers all said the same thing: none of the interventions being discussed by governments and politicians was likely to have any affect whatsoever on the collapsing economy.

The phase transition that these speakers had diagnosed is like an avalanche. Before an avalanche begins, the snow field is stable. Then, one snowflake too many is placed on the side of the hill and the slightest perturbation -- a sharp noise, perhaps -- can cause the entire hillside to collapse and rumble to the valley floor.

Nothing we do after the avalanche has started will stop it.

Two days earlier, an OpEd by physicist and science writer Mark Buchanan had appeared in the New York Times saying what all those speakers also said that day at our conference: the financial markets appeared to be acting like a complex system, characterized by feedback loops that create instability.

Buchanan discussed some simulations of model, or virtual, economies using these complex systems ideas: "The instability doesn’t grow in the market gradually, but arrives suddenly. Beyond a certain threshold the virtual market abruptly loses its stability in a “phase transition” akin to the way ice abruptly melts into liquid water. Beyond this point, collective financial meltdown becomes effectively certain."

Shankar Vedantam wrote recently in the Washington Post about research using a complex systems approach which identifies a particular trigger for the financial avalanche which followed those early signs of collapse last October.

I do not know if the trigger identified by these researchers is actually the one that led to financial collapse, but this is one of the few mentions of this important topic I have seen in the press since Buchanan's early editorial. If you know of others, I would be interested to learn about them.

To read more about these ideas, you might wish to consult the Oxford Press book by Neil F. Johnson, Paul Jefferies and Pak Ming Hui entitled "Financial Market Complexity: What Physics Can Tell us About Market Behavior."

1 comment: