How the credit crisis spread

credit-crisis-spread

Where did the credit crunch start? According to Reginald Smith at the Bouchet-Franklin Research Institute in Rochester, it began in the property markets of California and Florida in early 2007 and is still going strong.

To help understand how the crisis has evolved, Smith has mapped the way it has spread as reflected in the stock prices of the S&P 500  and NASDAQ-100 companies. The picture above shows how the state of affairs changed between August 2007 and October 2008. Each dot represents a stock price and the colour, its return (green equals bad and red equals catastrophic).

Smith says the problems first emerged in housing stocks, soon followed by finance stocks then mainstream banks before hitting stocks across the board.

The graphic may be dramatic but it shows only how the collapse occurred, not why. That’s much more subtle and is related to the far more complex network of links that exist between the companies involved.

However, the graph does bear a remarkable resemblance to any number of other network-related catastrophies, such as the spread of disease, forest fires and fashion. That’s almost certainly because  all these events can be described terms of the physics of self-organised criticality.

Smith says it’ll take years, perhaps decades, to fully understand and analyse the credit crunch. Econophysicists could start by brushing up on their knowledge  of self-organised criticality.

Ref: arxiv.org/abs/0901.1392: The Spread of the Credit Crisis: View from a Stock Correlation Network

2 Responses to “How the credit crisis spread”

  1. cDave says:

    The forest fire model only tends to self-organised criticality when two ratios are close to zero.

    Tree Growth / Fire Spread Rate.
    Lightning Strikes / Tree Growth.

    I think the analogy just about holds.

    You’d need to modify the model to have a more complex network than the standard grid. Maybe with slow time variance. My intuition is that it would still be SOC.

    One of the predictions that this model makes is that high burn years are followed by low burn years, as there are less trees around. So propping up failing institutions, could be storing up more trouble.