Pair Trading Strategy Journal

Quote from bears21:

hey jonnysharp quick question with all of these backtests done do they account for splits and dividends. if not those backtests can be very misleading. thanks

yes splits and dividends are taken into account.
 
Quote from sleepy:

Hi jonny,

Do you trade ETFs in identical fashion to stocks with respect to entries and exits .. etc.
And are you currently used a time-based stop on both?

sleepy :

Yes pretty much, although I use different signal settings for ETFs, they tend to work better on lower look back periods, and yes time stop on both, max 40 days usually like to exit within 10 days though.
 
Hi I am new here , I have been using PTF for about 3 months with some good results. I trade only ETF and take profit early worrying about large drawdown as posted on this journal. I am unable to do back test longer than 1.5 year with Version 2.99. I tried to test back 3 years 5 years or 10 years and the number of tests are the same and the equity charts were not different on all periods. Do you have any of this problem ?

Thanks

PA
 
Check that both ETFs have been trading for more than 1.5 years, a lot of ETFs are relatively new. I can backtest up to 10 years worth of data on many stocks, futures, forex fine with v2.99
 
Thank you Jonny I tested on QQQQ and SPY and both have more than ten years of data and it occured on both of my computers which are HP 32 bit widows XP 2 G RAM enough hard disk and good enough speed. Are you using IQ feed data. I use Yahoo delay data.

PA
 

Attachments

Hi jonny I did specified 10 years once on one of my computer it did not work then. I tried it again as you suggested this time it was working fine, returned 37 transactions. I might have done something wrong when creating a stock group then.

Thank you very much for your suggestion,

PA
 
Quote from jonnysharp:

did you specify 10yrs when first adding SPY and QQQQ into a group? by default its only 1yr, you have to change it to 10yrs.

About backtesting: during last year and a half qqqq has significantly outperformed spy. Beginning 2009 volatility neutral ratio was approximately +1spy-3qqqq (neutral I mean if the market plunges, the spread won't move much). Nowadays this ratio is around +5spy-12qqqq. If you trade the old ratio now you'll end up essentially trading qqqq alone (as it overweights spy now).
 
Jonny I do not see my previous response to you I just want to say that as your suggestion I can test the pairs up to 10 years now.

Thanks for your help,

PA
 
Back
Top