RSS Feeds

Email a Friend Printer Friendly

Little Crashes Could Protect Stock Market

August 07, 2002

EMBARGOED UNTIL WEDNESDAY 7 AUGUST 2002 19:00 BST
UK CONTACT - Claire Bowles, New Scientist Press Office, London:
Tel: +44(0)20 7331 2751 or email claire.bowles@rbi.co.uk

WALL STREET`S sneezing fit continues, and the rest of the world can`t shake off its cold. Could the answer be to "immunise" the stock market by injecting it with a dose of minor downturns? If it works, a few jabs at the right time and place could make devastating crashes a thing of the past.
Michael Hart and his colleagues at Oxford University have made a computer simulation of the stock market, mimicking a group of traders who try to profit by buying when the majority wants to sell and selling when the majority wants to buy. Each time the model is run, the virtual traders make different decisions and the stock market index traces a different path.
To discover the most likely future for the stock market in the model, Hart calculated the possible future scenarios over a certain period of time. Some paths just wobbled up and down with no major fluctuations, while others eventually dived into a catastrophic crash. "All the model runs can be thought of as different parallel world lines running into the future," explains Hart.
The researchers found that paths that eventually crashed tended to have an above-average number of upward movements in their history before the crash, while paths that didn`t crash had experienced more downward movements over a similar period. It seems that small drops take tension out of the system and allow it to ride through the period of market instability, says Hart.
To spot when a crash was looming, the researchers ran their parallel world lines into the future. When many of the possible lines ended in disaster, they found that a crash was almost inevitable. "The fact that all the lines are following the same path means that the event is highly probable," explains Hart.
The team hopes the model could be applied to the real world to spot major dips in the stock market and smooth them out before they happen. They envisage a regulator who would monitor the future stock market based on the parallel-worlds analysis. "If they see the warning signal of converging paths, then they step in and `immunise` the market," says Hart.
To do this, the regulator could sell small amounts of stock- enough to force the market down but not make it crash. Another, less likely control would be laws to force major market movers to adjust their positions in the market by buying or selling stock.
The team has discussed its ideas with the Bank of England, but the model is far from ready to simulate the real stock market. The bank declined to comment on how seriously it`s taking the approach.
One concern is how to gather data about the real world. Ton Coolen, an applied mathematician from King`s College London, says applying the model to the stock market might be going too far. "Computing all the possible future world lines for the stock market would require getting inside the heads of the traders to see how they think. As far as I can see, this would be impossible," he says.
The logistics would also be daunting. "A regulator would need stacks of money and a large buffer stock to change the direction of the stock market," says a spokesperson for the Financial Services Authority. "It sounds crackers to me."
Walter Kemmsies, head of global sector strategy at UBS Warburg, thinks gentle intervention in the stock market to prevent crashes would be a great idea, if it could be made to work. "Buying shares would become safer and people could be confident that the share price was not based on dreams."
But one city trader still sees a downside to the idea. "It would make the markets terribly boring," he says.
Author: Kate Ravilious, Oxford

New Scientist