Pair Trading Strategy Journal

Fantastic results!!

Quote from jonnysharp:

Thanks yobo, you have been the most regular contributor to this journal and I appreciate that. Yes several weeks ago I was slighty down, however now Im up 21% so yes I am profitable. I'm not using pair trading to make a living atm, Im using it to grow my capital, although I expect next year or maybe 2010 I can make a living doing this, I would just have to put on more trades per month to increase the probability of being profitable every month, so if I take 15-20trades per month I should be profitable 10 out of 12 months, however if I take 100 trades per month I should be profitable 49 out of 50 months and could use it to make a living plus grow capital.
 
Quote from yayt:

That is amazing and impressive, especially considering the recent market turmoil! Congratulations!

Could you provide some more statistical info, such as drawdowns, et. al? I am thinking of, after completing my current project, working on quantitatively and automatically trading pairs both intraday and longer term, as you are.

I've been following this journal with great interest, keep up the great work,and again, congratulations!

thanks yayt, yeah the method is robust espically through uncertain times.

I don't keep detailed statisical info, I can say from memory I don't think ive been down more than 8% since starting and that was during that crazy week of volatility.
 
Quote from gobar:

do u sell/cover or buy/short stock @ the same time or wait for some indicators before jumping?

thanks

yes usually I enter both sides at the same time, sometimes if the market is moving strongly in one direction which can happen around the close il put on the side in that direction and wait several minutes before putting on the opposite side, most of time I usually lower my cost average doing this, gotta be careful not to be too clever though as it can be dangerous if a stock moves quickly. Also im looking at the size on the bid/ask and try and get a fill inside the spread. But yeah execution is very important.
 
Quote from cipherscribe:

Jonny,

Great Journal, very inspiring. Congratulations on getting green!

Can you elaborate a bit on how you identify pairs? What is your filtering technique that helps you decide to add a ticker to the pairtrader db?

Thanks for that.

Not sure I understand your question correctly, so il try my best. I use pairtrade finder to find suitable pairs through backtests and then add pairs based on high correlation and confirm with fundamental correlation(same industry/sector), then just before the market closes each day I open the program and check for signals and choose one's based on my own personal criteria and then wait for the program to give me a exit signal. I don't like to hold too many pairs in the same sector either.
 
Quote from Appleseed:

Great work jonnysharp
but as others have noted here winning percentages are misleading. there are systems that win 70% and lose money while others with only 40% winners or less go on to be winning strtergies. A good start mind you but it would be interesting
to seeing an update P/L side.
Isn't this inportant to you?
cheers
john

thankyou, yes PnL is important to me, I not only trade for the intellectual challenge and pleasure but also I want to make some good $$$. My win % is 73% and win/loss is 0.71 so that answers your question as too whether im profitable. I don't like posting my PnL because it changes daily plus I feel thats personal and I don't want to attract heat from some of the ''pro'' traders on these boards. I can say atm its well into 4 figures. thanks again.
 
Quote from jonnysharp:

Thanks for that.

Not sure I understand your question correctly, so il try my best. I use pairtrade finder to find suitable pairs through backtests and then add pairs based on high correlation and confirm with fundamental correlation(same industry/sector), then just before the market closes each day I open the program and check for signals and choose one's based on my own personal criteria and then wait for the program to give me a exit signal. I don't like to hold too many pairs in the same sector either.

Sorry for my ambiguity.

I have Pairtrader, and i'm following your trades within the application. My question is based around your identification of what tickers to add to the application.

I'm very new to the idea of pair trading, and whilst I understand the concepts behind pair trading, I'd like to know the step prior to that, for example, what makes HES or APC worthy of being added to the Pairtrader app?
 
Quote from cipherscribe:

Sorry for my ambiguity.

I have Pairtrader, and i'm following your trades within the application. My question is based around your identification of what tickers to add to the application.

I'm very new to the idea of pair trading, and whilst I understand the concepts behind pair trading, I'd like to know the step prior to that, for example, what makes HES or APC worthy of being added to the Pairtrader app?

No problem, I used Yahoo Finance stock screener and added every stock above 5billion mkt cap from each sector into a new stock group, then I ran backtests and added every pair above 70% correlation to my watchlist. Then I filter out pairs based on fundamentals, you don't want to pair trade a gold company against a oil&gas company.
 
Quote from jonnysharp:

Entered one new trade today;

Long TV @ 16.35
Short OMC @ 33.39

Hi Jonnysharp,

very interesting thread!

Do you take an equal amount of shares on both stocks. Or do you trade dollar neutral?

I think I read a post somewhere, where you said you try to look for pairs that have nearly equal prices?

But what would you do on pairs like TV-AMC or HES-APC?

-Screenlocal
 
OMC / TV

TV is an ADR. In my experience ADRs suck to pairs trade. The overnight gaps are terrible because of the different markets involved. Its not uncommon to have an ADR pairs trade go many std dev from the mean in a couple days for no real reason and never come back.
 
Back
Top