(This article was first published on Commodity Stat Arb, and kindly contributed to Rbloggers)
Ok, so they start with the same letter and are part of the same group of elements, but why are they so often proposed as pairs in trading strategies? As a spread strategy they look especially tempting right now:
Palladium outshone Platinum for most of 2010. Is the run over? Or will it go further?
Cointegration testing
Before attempting to answer this question, perhaps an analysis of the relationship is warranted. Do PL and PA exhibit comovement over time? A chart of the historical prices of both (backadjusted for contract roll) is as follows:
Cointegration testing in different years leads to the following results:
Year

ADF P value

Hedge ratio

Cointegration half life

2004

0.60

0.30

18

2005

0.94

0.25

33

2006

0.57

0.30

21

2007

0.35

0.27

66

2008

0.29

0.22

18

2009

0.46

0.22

27

2010

0.96

0.33

39

In other words, the ADF test [http://en.wikipedia.org/wiki/Augmented_Dickey%E2%80%93Fuller_test] for cointegration has no significant results in any of the past 7 years.
In a statistical sense there is no reason to assume that PL & PA have any sort of relationship whereby they could be part of a pairs trading strategy.
Looking at the 2010 chart one wonders how many traders have been drawn into this spread strategy waiting for the reversion to arrive…
To leave a comment for the author, please follow the link and comment on his blog: Commodity Stat Arb.
Rbloggers.com offers daily email updates about R news and tutorials on topics such as: visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...