You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Pair Trading Strategy Journal
- Thread starter jonnysharp
- Start date
Quote from Vespasian:
Here's the main problem I see with trading pairs.
Your looking for something to STOP happening and go back to so called "normal" or mean revision.
In my experience the market can stay abnormal far longer then my bank account in most cases.
How do you deal with this issue ?
PTF uses a ''moving'' mean, not a static. In addition to that I use a 10 day time stop. I dont see that problem, Vespasian.
Quote from henderson:
agreed....the 2 ways I used to look at it:
Time stop = exit date - trade date
when Time stop = 0; get flat
Stop Price (Cost basis) =< Price + transactions costs + debit interest - ECN rebates - lending rebates; get flat....debit interest can get high so need to include it.
Thanks,
Can you give a trade example?
What has the greatest volume, the open or the close?
Quote from Vespasian:
May I ask your worst draw down with this approach and why specifically 10 days?
Thanks
I started trading live on april 7 and never went negative on my account. Unrealized about 7%. 10 days by backtesting and experience of traders who use PTF for a much longer time than me.
Quote from cipherscribe:
What has the greatest volume, the open or the close?
As a general rule, the avg. volume is about the same on both ends.
The last half hour almost always trades much more than the 15:00-15:30 half hour whilst the 10-10:30 can trade with the volume of the 9:30-10 period, moreso on the days that reports come out at 10.
Quote from CBuster:
This occurs because standard deviation in this case is not constant. In fact, the measure of standard deviation being used by PTF, or any use of bollinger bands etc (i.e. standard deviation from a moving average) exhibits serial correlation.
What S.D basically measures is how far from the average the ratio "usually" is. This means that as the ratio moves away from the average, and the longer it stays away, the more SD rises. This ultimately means that on any given day, SD can rise or fall by more than the ratio itself which in turn means that the % from mean can get bigger but num SDs can get smaller.
So mathematically it is understandable. But I totally agree that interpretting this situation can be a little tricky. If you are looking to add layers to a trade, should you do it as % from mean increases or num SDs goes up...
Yes interpreting can be tricky, for layers, I would really prefer to see both increase, % and SDs.
I can tell you this though. I have gotten in a few bad trades that SDs were decreasing while % increased. So now I really dont enter unless % is near extremes.
Overall, I started trading pairs june 26 and I'm now up 3.5 % ... I keep positions small. I was underwater when I started by -1.8 % for about ten days.
I love how I can avoid the market volatility.
Just sold fdx/bought hubg yesterday..