Choosing Futures Roll Rules and Continuous-Price Adjustments
Summary
The document considers how to build continuous futures price histories for a multi-market backtest. The questioner describes using backward Panama adjustment, rolling when open interest switches, and selecting the first listed contract across a diverse set of futures. The replies do not establish one universally correct method; they emphasize that roll timing and adjustment should fit the strategy and that the mechanics of the chosen series must be understood.
One reply favors weighted rolls on a predetermined schedule, such as near the last trading day or at the start of a month, because the timing is consistent and spreading the transition over several days can smooth gaps between contract months. It contrasts this with unadjusted prices, which preserve rollover gaps and suit the view that such gaps behave like ordinary price gaps. A separate reply suspects the reported missing bars reflect a data problem, observing that the cited yen contract is liquid. The exchange does not diagnose the gaps or compare methods empirically, so backtest results may depend on data quality and roll conventions.
Key ideas
- Futures roll timing and price adjustment should match the purpose of the backtest.
- Open-interest switching is one possible roll rule, while fixed-date weighted rolls provide predetermined timing.
- Weighted rolls can smooth price differences between adjacent contract months.
- Unadjusted series preserve rollover gaps, which some traders treat like ordinary market gaps.
- Missing bars in a liquid futures contract may indicate a data-quality issue, but the document gives no diagnosis.
Tags
Full text
# how to choose a price adjustment, a roll date and a data center for my trading strategy? # how to choose a price adjustment, a roll date and a data center for my trading strategy? I have many doubts about Which roll date and price adjustment should I use. I need to backtest like 50 diferents futures. 6 index(mini sp500, Nikkei 225…), 10 Agriculture (soybean, Oat, Corn….),3 Meats (live Cattle, Lean Hog, Feeder Cattle), 8 Currencies (yen , Australian Dollar, Pound, Swiss Franc…), 5 Metals (Silver, Gold, Palladium…), Treasury Notes (10 years, 5 years…), Us Bond 30 year and some more… My backtest is for 15 years from 2000 to 2015. I have choosen the backward Panama canal method, rolling with the open interest switch and with a depth #1 in all of then. My question is…Is that correct? Or I should use differents kind of methods for the differents kinds of futures(agricultures, metals, currencies…) Another question is that the SCF FUTURES of some futures have gaps in the graphics. There are severals with this gaps between 2009 until 2012 (the mayority of the currencies and the agricultures futures) . The example below is the yen future. I don’t know why produce this gaps or undiscontinuous bars. Thank you very much for your time . ## Answer by Jacob Amos (score 0) https://quant.stackexchange.com/a/24677 I would first take a look at Quandl SCF Roll Methodology (if you haven't done so already). In my own personal backtests I have used weighted rolls based on either the last trading day or the first day of the month, simply because I prefer the determinism of when the roll takes place and how it is to take place. The other advantage of this weighted roll is that it can smooth the price gaps between months, just because if the spread between first and second month is big, this approach eases into the second month over a couple of days. On the other hand, the guy that wrote this post prefers no adjustment, because the "rollover gaps are treated just like traditional gaps and have a very high probability of being filled at some point in the future." That's a common belief in the world of technical analysis, and if you agree then you might also prefer an unadjusted roll methodology. TLDR: At the end of the day, it depends on a combination of what you're trying to do and your personal preference. Just make sure you know what's going on with your chosen roll methodology. ## Answer by user42108 (score 0) https://quant.stackexchange.com/a/59835 "Another question is that the SCF FUTURES of some futures have gaps in the graphics. There are severals with this gaps between 2009 until 2012 (the mayority of the currencies and the agricultures futures) . The example below is the yen future." Looks like there's something wrong with your data. 6J is liquid and trades 23 hours per day. EDIT: what are "SCF FUTURES"?
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.