Kemp: Oil Market Takes A Walk On The Wild Side (Again)
John Kemp is a Reuters market analyst. The views expressed are his own
LONDON, Sept 2 (Reuters) - Crude oil prices have been on a rollercoaster over the last four trading sessions that has seen some of the highest volatility in a quarter of a century.
The market is providing a brutal reminder of the extreme side of commodity pricing, leaving many analysts and traders struggling to identify a safe strategy.
Front-month Brent crude futures rose by more than 10 percent on Thursday, 5 percent on Friday and 8 percent on Monday, before plunging by more than 8 percent on Tuesday.
To put that in context, the percentage daily price movements were 4.6 standard deviations away from the mean on Thursday, 2.4 standard deviations on Friday, 3.7 on Monday and 3.8 on Tuesday.
If price changes followed a normal distribution, a move of 3.5 standard deviations should occur only once every eight years and a move of greater than 4.5 standard deviations should happen once every six centuries.
The occurrence of three exceptional moves in the space of just four trading sessions underscores changes in oil prices do not follow a normal or Gaussian distribution, named after the 19th century mathematical genius Carl Friedrich Gauss.
Even though they are taught that it is not an accurate description of how financial markets work, the normal distribution exerts a strong unconscious influence on how traders, analysts and investors think about the risk of extreme price moves.
But it significantly understates the probability of very large price movements and is little use in understanding how commodity markets behave. Commodity markets are much more dangerous than the comforting world of the Gaussian distribution allows.
The extreme movement of commodity prices was first observed in the early 20th century by Wesley Mitchell ("The Making and Using of Index Numbers" 1915) and Frederick Mills ("The Behaviour of Prices" 1927).
But it was rediscovered and made famous by Benoit Mandelbrot in the 1960s, who studied cotton prices and discovered far more large moves than expected if they followed a normal distribution ("The variation of certain speculative prices" 1963).
As Mandelbrot explained, the normal distribution "does not account for the abundant data (on price changes) accumulated since 1900 by empirical economists."
He went on to observe "empirical distributions of price changes are usually too peaked to be relative to samples from Gaussian populations."
There are "so many outliers" and "the tails of the distributions of price changes are in fact so extraordinarily long" they could not be modelled by a normal distribution.
View Full Article
WHAT DO YOU THINK?
Generated by readers, the comments included herein do not reflect the views and opinions of Rigzone. All comments are subject to editorial review. Off-topic, inappropriate or insulting comments will be removed.