STIRs, single contract vs back-adjusted price series

General discussions about futures.
Post Reply
AceofAce
Senior Member
Senior Member
Posts: 49
Joined: Thu Sep 30, 2010 5:06 am

STIRs, single contract vs back-adjusted price series

Post by AceofAce » Thu Mar 02, 2017 8:59 am

Typically quant futures strategies trade on or analyse back-adjusted price series data to issue signals or carry out research.
Does anybody have a view whether it makes sense to make an exception for Short Term Interest Rate contracts and use single contract time series? They are fairly liquid over long periods spanning more than 2 years, sometimes, and more importantly may give different signals. Take 3M Euribor as an example. The back-adjusted (on Open Interest) time series has been trending upwards since July 2016. However, the single contract price series of the most liquid month (IMH8) has been trending downwards over the same period.

Tim Arnold
Site Admin
Site Admin
Posts: 8523
Joined: Tue Apr 06, 2004 1:41 pm
Location: Boston, MA
Contact:

Re: STIRs, single contract vs back-adjusted price series

Post by Tim Arnold » Thu Mar 02, 2017 9:24 am

Yes it makes sense. Every market has the potential to be back adjusted differently, using different roll criteria and selective months. In this case you might want to roll once per year, to get a good 20 year back test window. Basically, trade what you test, and test what you trade -- particularly when it comes to roll criteria. Rolling early, or late, or not at all, is a huge area of performance adjustment. The action of a single month over a long period can be much different than the 'always front' month. The psychology of the traders in these markets is quite different.

You can also trade many different roll strategies at the same time, and/or test spreads against these, to diversify.

AceofAce
Senior Member
Senior Member
Posts: 49
Joined: Thu Sep 30, 2010 5:06 am

Re: STIRs, single contract vs back-adjusted price series

Post by AceofAce » Thu Mar 02, 2017 12:34 pm

Many thanks Tim. Indeed you are right. I am doing a research on rolls timing and I noticed for example that in the back adjusted time series, if you change the roll point from market on open to market on close, the backtest gives you hugely different results. Equally important the results vastly differ when you roll on OI on the first trigger day or the second.
I am not sure how to approach this for optimum results in real trading but I feel the answer is not to pick the one that gives me the best backtested sharpe ratio.

Post Reply