About This Blog

This blog on commodities, financial modelling and related topics is written by Will Smith.  I’m a PhD student at Birkbeck, University of London studying Commodities Modelling (formally, I’m studying Mathematical Finance).

3 Responses to About This Blog

  1. Imonike Mohammed says:

    Hello Will,
    I was browsing and I came across your blog. I was wondering whether you could give me a few pointers to a problem I am having. Here is the problem:

    I am trying to do cointegration analysis of commodity futures data downloaded from the Johannesburg Stock Exchange. I am specifically testing for the presence of a cointegration relationship between the white maize and yellow maize futures. I have commodity futures data starting from November 2006 to mid-April 2010 in my database. When testing I pick a specific contract say “White Maize May 2007” and its yellow maize equivalent, create a time series of each and test. I was wondering if there was a way I could use all the data I have for the 2 commodities to create each commodity’s time series without introducing any biases. I suspect this would give more accurate results if this were possible.
    Any ideas will be greatly appreciated.

    Thank you for your time.


    • Why not make a ‘front month’ contract, but avoid roll-over effects by rolling what you consider to be ‘front month’ a few days before expiry. It’s the standard way to do it. Or, more advanced, construct a weighted average of the 1st and 2nd month contracts, with the weights recalculated each day, to build a ‘constant maturity’ contract with some duration, say 45 days.

      Then do the same for both commodities and run the cointegration over the full time period.

      To remove biases, you will have to correctly account for the rebalancing trade as you sell front month and buy 2nd month. Put yourself in the position of a fund, sell some 1 month (at say the closing price), work out how much money you get, and buy 2nd month on the same day, again at the closing price.

      • Imonike says:

        Hi Will, thanks very much for your response. I think I might say a bit about myself. I am a software engineer working in Ghana. I have been teaching myself statistics,econometrics and R in my spare time. I haven’t traded commodities before, I had just studied to the point where I felt confident enough to try something like this. When you say that I should create a ‘front month’ contract, are you saying that for example if I start with the March 2007 contract for white maize as my white maize front month contract, that I should take all the data I have for this contract to start building my time series for white maize, drop the last few days to avoid roll-over effects, continue to build my time series with the next contract for example April 2007 contract for white maize, ignoring all data that exists for the April 2007 contract before the roll over point and proceed in that fashion, doing like wise for yellow maize?

        Thank you for your time.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: