Short term bump in Crude and the Fractal nature of patterns

“The techniques I developed for studying turbulence, like weather, also apply to the stock market.”

– Benoit Mandelbrot, The (MIS)Behavior of Markets

If you are not familiar with the work of Benoit Mandelbrot, he made huge contributions to the field of science and mathematics, and his research is starting to have greater impact in Finance.  Mandelbrot’s work dispels that markets are efficient, and identifies that return distributions are not normal but rather volatile and extreme, and that price changes are not independent.  See 2008.

Mandelbrot also coined the term “Fractals”.  Fractals are never-ending patterns that are self-similar across different scales.  For example, if you take a tree and then zoom in on the larger limbs, and then zoom in further on the branches and then again on the twigs, the similarities of the smaller scales continue to resemble the larger.  He observed this phenomenon of Fractals in the prices of Cotton.  He looked at Cotton prices over a 60 year period on varying time series scales, i.e.  hourly, daily and monthly and found that the changes followed a symmetrical pattern in terms of scaling and the curves were identical.  Patterns within patterns.

This week, using lessons from Mandelbrot, we decided to take a closer look at another commodity market:  Crude Oil.  Current drivers on the bullish side appear to be increased production in the U.S. and a potential lifting of the ban on U.S. oil exports which has been in existence for almost 40 years, as well as a recent drop in Crude inventories.  On the bearish side, distillate (refined Crude) inventories increased last week which seems to have pushed the price down in the short term.  Where does Crude go from here?

In the spirit of the fractal nature of patterns, we’ve taken a look at the current price pattern in Crude Oil (CL) over two time scales to see what traditionally happens in the next 10 days.  First, we looked at the price pattern from its 2013 high of $110.53 through to Friday’s close of $92.72 (approximately 4 months).  Then, we analyzed a shorter term pattern from its more recent high of $100.32 on December 27^th until Friday’s close (two weeks).  We took each of these patterns, and looked back over 20 years in Crude to see when we’ve seen similar instances of each of these current price patterns, and to see what happened in the next 10 days historically.

1. 4 month pattern:        We found 17 similar instances, and the average return in the next 10 days historically was 4.16%
2. 2 week pattern:          We found 55 similar instances, and the average return in the next 10 days historically was 2.44%

Please see the charts below.  You’ll also notice how similar the current 4 month and 2 week price patterns are.  Fractals!

Watch for a short term bump in Crude Oil.

Price of Crude Oil since its 2013 high of $110.53 on September 6, 2013 (appr. 4 months)

17 highest correlated historical instances of this price pattern in Crude Oil, and the average return

in the next 10 days of 4.16%

Price of Crude Oil since its recent high of $100.32 on December 27, 2013 (two weeks)

55 highest correlated historical instances of this price pattern in Crude Oil, and the average return

in the next 10 days of 2.44%