Say I currently hold a position in a July futures contract and I'm using a 14-day ATR trailing stop. Then when it's time for me to rollover to the September contract, which price series should I use to calculate the trailing stop.
At first, I thought it should the September contract since that is the contract that I will be trading now, but the September contract may have been a different beast for the past 14-days because liquidity may have been lacking then. But now, it is the new front-month contract and it also seems reasonable to use the price series of the previous front-month contract to calculate the 'starting rollover' ATR, which is the July contract.
The latter is the standard approach if one is using a continuous data series I believe. I haven't yet checked if this is a significant issue.
Louis
which price series to use?
-
- Roundtable Knight
- Posts: 118
- Joined: Tue Apr 15, 2003 7:44 pm
- Location: Arizona
Hi Louis-
If you tested with continious contracts then the concept of continuous contracts is dictating that you use the July contract (which is adjusted from the Sept contract) for your numbers.
And depending on the length of your buffer (Maxbarsback in TS) you might need to adjust prior to the July contract to keep everything even.
This is one of those issues where you need to make adjustments in your real-time trading to match up with your historical testing.
Gordon
You are correct to note that the July and Sept will be different beasts.Say I currently hold a position in a July futures contract and I'm using a 14-day ATR trailing stop. Then when it's time for me to rollover to the September contract, which price series should I use to calculate the trailing stop.
If you tested with continious contracts then the concept of continuous contracts is dictating that you use the July contract (which is adjusted from the Sept contract) for your numbers.
And depending on the length of your buffer (Maxbarsback in TS) you might need to adjust prior to the July contract to keep everything even.
This is one of those issues where you need to make adjustments in your real-time trading to match up with your historical testing.
Gordon