The interesting thing once you start delving into this whole formal modeling business is that it all starts to seem the same — after all, when it comes down to it they’re all just abstracted structures. They’re highly useful as conceptual tools to organize and understand data, and modern economics in particular couldn’t get very far without them. But in light of the current credit crisis, Roger Ehrenberg offers some insightful analysis of how models can fail when they’re placed on a pedestal:
I grew up in a time when markets were considered to be “continuous.” Portfolio insurance. Robert Merton’s 1987 treatise Continuous-Time Finance. Liquidity was presumed to be available. And while markets could and did gap due to an event, new information, etc., it could and would clear with transactions taking place at the new level. The financial markets, through price discovery in the presence of liquidity, conveyed valuable information that could be used for both security selection and asset allocation. The field of financial economics, as such, was predicated upon the existence of bids and offers and, therefore, liquidity. And this phenomenon was assumed to persist across time.
But this is not the world I observe today; quite the contrary. Price movements are not only discontinuous, but the notion of liquidity across time as traditionally assumed simply does not exist. Something has happened to rock the prevailing academic paradigm. Have the experiences of the past six months essentially blown a hole through the heart of modern financial theory?
He goes on to offer some suggestions as to what phenomena new models will need to account for:
Today we live in a world fraught with risks that we barely understand, risks that modern financial theory doesn’t have great answers for. A new model is needed that incorporates the effects of discontinuity as an outgrowth of, among other things:
- Complexity - structured securities, derivative instruments;
- Interdependency - widely disseminated holdings that can pollute portfolios globally, hundreds of trillions in counterparty exposures;
- Intermediary errors - ratings that don’t reflect the risks, financial institutions with weak control environments and poor risk management practices; and
- Bad actors - originators, underwriters, traders and managers with mis-aligned motives.
We have seen examples of each of these in the past six months, seeming “black swans” that don’t appear so unusual any more. It’s not that these risks didn’t exist before. It is that their confluence when experienced over a short period of time yielded results that were unforeseen to many. Our models and academic frameworks needs to be robust enough to handle these occurrences and to provide a model for maintaining liquidity, price discovery and information dissemination. Based upon today’s market action we’ve got a long way to go.
