Pair Trading Strategy Journal
- Thread starter jonnysharp
- Start date
Im interested to find out what the entry and exit process was with people and Pairtradefinder - "PTF".
Im using delayed end of day data from Yahoo. You can get a live data - 7day trial with IQ data feed but I am reluctant to use live data not for the cost but for the fact that I've been a live data trader in the past and personality wise it doesnt fit with me
I end up getting glued to the screen!
That said, can i ask the PTF users about
A. ENTRY & EXIT RULES people use with PTF.
1. When do you enter? - the last hour of trade on the same day of the signal? or in the morning of the next days trade after the opening has settled down (following a previous days confirmation of signal based on the closing price of the previous day).
2. When do you exit? When you can see the profit and run regardless of PTF to own set of percentage rules? adhoc rules? or Do you exit on the opening of the next day following a confirmed close price (and confirmed signal on the pairtrade finder program?).
3. What are your basic criteria. I read Johnnysharp (Thanks JS for your posts some time ago!) used:
Correlations > 80%
Divergence of Ratio and RSI
Ideally a flat volatility.
charts displayed were: correlation, ratio, rsi and volatility.
and using divergence entry setups.
Is everyone strict on the covariance? i mean .. what if the correlation was high, the same sector pair etc and the covariance was low. Would you guys still take the trade?
Im using delayed end of day data from Yahoo. You can get a live data - 7day trial with IQ data feed but I am reluctant to use live data not for the cost but for the fact that I've been a live data trader in the past and personality wise it doesnt fit with me
I end up getting glued to the screen!That said, can i ask the PTF users about
A. ENTRY & EXIT RULES people use with PTF.
1. When do you enter? - the last hour of trade on the same day of the signal? or in the morning of the next days trade after the opening has settled down (following a previous days confirmation of signal based on the closing price of the previous day).
2. When do you exit? When you can see the profit and run regardless of PTF to own set of percentage rules? adhoc rules? or Do you exit on the opening of the next day following a confirmed close price (and confirmed signal on the pairtrade finder program?).
3. What are your basic criteria. I read Johnnysharp (Thanks JS for your posts some time ago!) used:
Correlations > 80%
Divergence of Ratio and RSI
Ideally a flat volatility.
charts displayed were: correlation, ratio, rsi and volatility.
and using divergence entry setups.
Is everyone strict on the covariance? i mean .. what if the correlation was high, the same sector pair etc and the covariance was low. Would you guys still take the trade?
Hello,
I would like to publish follow-up to my previous posts with some update on my new experience with pair trading. After few month of simulation, I have been trading pairs live on my Interactive Brokers account since March 2011. Pairs trading is really just a part of my trading approach and I have it mostly automated thanks to my friends. Thus I really do not spent more than about an hour a month with the strategy.
I have done more than 50 live trades since March 2011 and you can see my results. They are positive, but not too much. However, I learned a lot of interesting lessons and I feel my pairs trading getting improved. The major lesson is selection of pairs. I was trading pairs within the same sector as well as pairs that were from different sectors â these where picked basically based on good âbacktestâ and the stocks do not have fundamental relationship. These âcustom pairsâ generated net loss -1838 USD in my portfolio. On the opposite â pairs within the same sector generated net gain 2589 USDâ¦. For the future, I will stick with just this group of pairs.
All my results include commissions and slippage. So far I have paid 281 USD on commissions but also 519 USD on slippage. I definitely recommend to keep tracking slippage in the journal as some pairs have much bigger average slippage than others and I start to remove pairs with higher average slippage from my portfolio.
I did find cointegration helpful factor for selecting my pairs. I do not know too much how cointegration is calculated, but just looking at cointegration graphs from pairslog I can see major patterns between loosing and winning pairs. Thus I started to redefine my portfolio using cointegration criteria as well.
M
I would like to publish follow-up to my previous posts with some update on my new experience with pair trading. After few month of simulation, I have been trading pairs live on my Interactive Brokers account since March 2011. Pairs trading is really just a part of my trading approach and I have it mostly automated thanks to my friends. Thus I really do not spent more than about an hour a month with the strategy.
I have done more than 50 live trades since March 2011 and you can see my results. They are positive, but not too much. However, I learned a lot of interesting lessons and I feel my pairs trading getting improved. The major lesson is selection of pairs. I was trading pairs within the same sector as well as pairs that were from different sectors â these where picked basically based on good âbacktestâ and the stocks do not have fundamental relationship. These âcustom pairsâ generated net loss -1838 USD in my portfolio. On the opposite â pairs within the same sector generated net gain 2589 USDâ¦. For the future, I will stick with just this group of pairs.
All my results include commissions and slippage. So far I have paid 281 USD on commissions but also 519 USD on slippage. I definitely recommend to keep tracking slippage in the journal as some pairs have much bigger average slippage than others and I start to remove pairs with higher average slippage from my portfolio.
I did find cointegration helpful factor for selecting my pairs. I do not know too much how cointegration is calculated, but just looking at cointegration graphs from pairslog I can see major patterns between loosing and winning pairs. Thus I started to redefine my portfolio using cointegration criteria as well.
M
Quote from Magnas:
Screenshots of cointegration that makes difference for my pair selection:
thankyou for posting that. I had never seen the website http://www.pairslog.com/ before too ..
Cheers mate.
How, exactly, are you measuring slippage?
Im hoping someone will be able to clarify this...
I have an issue with Pairtradefinder.
Data feed: EOD *yahoo
Market: London
Trade: Today is Monday, On the backtesting pair screen: Thursday has the coloured entry line. Next day, Friday is marked orange (1st day into the trade line), today is unmarked for obvious reasons given the close of trading day is not over (its 3 hours into the London session).
Today being monday i did a paper trade test. 3 hours into the trading session uggh.. the trade on the back tester on PTF disappeared!!
I've gone back to paper trading because I found when i used PTF last time and I had live trades using REAL money and then the trade dropped off .. never to be seen then I was left with a trade hanging with no signals at all on PTF meaning i had to manually trade the pair out.
Please Check the file attached below and you will see what i mean.
the image, shows the home screen and clearly the pairs.
the lower image shows now the pair testing screen and ZERO entry/trade signals.
Can anyone clarify?
Cheers
I have an issue with Pairtradefinder.
Data feed: EOD *yahoo
Market: London
Trade: Today is Monday, On the backtesting pair screen: Thursday has the coloured entry line. Next day, Friday is marked orange (1st day into the trade line), today is unmarked for obvious reasons given the close of trading day is not over (its 3 hours into the London session).
Today being monday i did a paper trade test. 3 hours into the trading session uggh.. the trade on the back tester on PTF disappeared!!
I've gone back to paper trading because I found when i used PTF last time and I had live trades using REAL money and then the trade dropped off .. never to be seen then I was left with a trade hanging with no signals at all on PTF meaning i had to manually trade the pair out.
Please Check the file attached below and you will see what i mean.
the image, shows the home screen and clearly the pairs.
the lower image shows now the pair testing screen and ZERO entry/trade signals.
Can anyone clarify?
Cheers
Hey .. i couldnt post a followup yesterday ..as this site doesnt allow another post in short succession .. i managed to get into pairlog. I think the accounts are mannually processed so theres a bit of patience to wait and got confirm by email.
The cointegration is between 0-1.0 and therefore 0.64 is correct...