For example, suppose we were trading a strategy which buys one Brent contract and sells one Gasoil contract. The minimum price fluctuation for a Brent contract is \$10, and the minimum price fluctuation for Gasoil is \$25, although the contract price for each is on the same order of magnitude. I have heard that the spread should be priced to take this discrepency into consideration, but I'm not sure I see how or why.

  • $\begingroup$ I am not sure you need to account for the contract tick size, but you definitely need to account for differing variance or 'beta' of the two contracts. $\endgroup$ – experquisite Oct 29 '14 at 23:30
  • $\begingroup$ @experquisite Can you elaborate on this or give a link to something I can read? This is interesting. My general strategy has been to do a linear regression regression on two price series to find a ratio, but I haven't considered heterogeneous variances. $\endgroup$ – c12345 Oct 30 '14 at 14:04
  • $\begingroup$ If your basket only has two contracts, then a linear regression on prices is fine. Some people attempt to trade the "ratio" of two prices, though, or trade flat 1:1 contracts, which is usually a mistake. There is always some debate whether one should be regressing prices or log-prices. $\endgroup$ – experquisite Oct 30 '14 at 16:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.