Pair Trading Strategy Journal

I also note from many pair trading discussions, there is too much emphasis purely on correlation. While this is integral to any pair trading opportunity, there should be equal value placed on technicals & fundamentals of a stock.

My basic strategy which integrates all of the above can be summarized in 4 steps...

1: I start by looking for technically [oversold] and/or fundamentally [undervalued] stocks and make a short-list of potential candidates. Call each of these stock 'X'.

2: I then look for a stock to trade against X which is correlated - a good place to find these is at the free site http://www.sectorspdr.com/correlation/ where you can simply enter stock 'X' symbol and it will come up with a stocks which typically provide 85%+ correlation. Once I find a stock to trade against 'X' I make a note. Call this stock 'Y'.

3: Using %comparison charts, both X & Y are plotted (line charts, 1 year/daily) where I can quickly see of 'X' has diverged against 'Y' to offer any potential pair trading opportunity. This is all part of my field research essentially combining technicals, fundamentals and correlation, with several 'X' stocks to match up with relative 'Y' stocks. It is at this point, that a trading decision can be made.

4: Once I am happy with 'X' versus 'Y', I put on the trade, ensuring that the 'total value' of my purchase of 'X' is almost exactly the same as my short-sale of 'Y'. For instance, if I am going long 1,000 shares 'X' at $20 (totalling $20,000), I make sure the total value of 'Y' is also $20,000, maintaining dollar-neutral. I exit the trade when I am up 3% net profit on a trade, and exit if I am negative 3%.

Hope this helps, and builds a clear picture of what is essentially a simple, logical method, but emphasizes the important factoring in of technicals/fundamentals as opposed to pure correlation overkill...
 
Sold FHN @ 13.18
Bought FFIN @ 55.11

What I like about trade:

No earnings imminent.
At Layer 2 (2.5 SDs)
% -> Mean at near term historic level
Ratio chart non-trending
Spread is at 50d low(Long the spread, so like to see it at or making new lows)
Corr chart trending upwards

What I don't like:

Nothing about this trade I don't like.

If you keep making trades like this and keep your execution costs ruthlessly low you will come out ahead.
The process is simple, but don't confuse simplicity with easiness.
It is hard to do it day in and day out, and that's what separates the winners from the losers.
 

Attachments

Bought RYN @ 40.67
Sold DRH @ 8.26

What I like about Trade:

Non-Trending ratio chart
Corr trending upwards
Spread near 50d low

What I don't:

% -> Mean reading is not that high


Having said that % -> Mean and % profitability are not necessarily correlated. i.e. if the % deviation is 8%, don't assume profitability potential of trade is 8%. It doesn't work like that.
 

Attachments

Quote from coreed:

Bought RYN @ 40.67
Sold DRH @ 8.26

What I like about Trade:

Non-Trending ratio chart
Corr trending upwards
Spread near 50d low

What I don't:

% -> Mean reading is not that high


Having said that % -> Mean and % profitability are not necessarily correlated. i.e. if the % deviation is 8%, don't assume profitability potential of trade is 8%. It doesn't work like that.
More...

Hello coreed,

Thanks for sharing your trades. Do you have any concern that DRH and RYN are in different industries?
 
Well BD1 its funny you shld point that out.


When the signal was generated, I ran RYN through BBG/Reuters to filter for news and I wondered why I had entered it under NYSE-REITs in PTF, as BBG has it listed as a Paper & Related Products company.

It looked like an error on my part.


Confused, I even went to Rayonier's site to double check. They are not a foresty company in the sense of an IP/WPP or a REIT like SPG/TCO. Somewhere in between.

I went back to NYSE.com to check the sector listing and that is how they list it - as a REIT, and that is how I chose to classify it.
 
Another trade in Regional Banks.

Bought FITB @ 11.518
Sold WBS @ 15.41

What I like about trade:

Non-Trending ratio chart
V.High corr (93%)
Spread chart at 50d high(Short the spread- so like to see it at high)
Backtests well
No earning pending

What I don't like:
Would prefer that recent action of mean line of ratio chart was not pointing down.
Best signals seem to come when mean ratio line is fairly flat.


Regional Financials are highly correlated, so easy to find good pairs. Diversification is not so important in pairs trading as you are

1: Market Neutral (By default)
2: You should be sector/industry neutral

Only risk is the relationship btwn stocks in pair.
 

Attachments

Hi Guys, just letting you all know I'm still around here every now and then. Ive been busy with my business and don't have time to keep an up to date journal. Thought i touch base and give a quick update, as i receive PMs all the time requesting the same. I'm still pair trading and have the guys at pairtrade finder managing an account for me which is going well. 2009 wasn't as good a year as 2008 but still profitable with low portfolio volatility none the less. I still think pair trading is the best strategy to manage my portfolio going forward and 2010 is shaping up as a good year as the markets look to correct and volatility is rising again, uncertain sentiment is much better conditions for pair trading than an melt up in prices. Jared and his brokers have some interesting things going on atm as some of you may know, and i think they are targeting 10m in funds by year end or something, i'm in with them, i think these guys will eventually be a 100m+ hedge fund, they seem to do well with everything. I haven't had time to read through all the pages of the journal since my last visit but it looks like some of you are still doing well, so congrats, and thanks for keeping the journal alive, ive learnt heaps from all of you and will be back to check in again soon,
Jonny.
 
Nice to see you back Jonny- You have been missed

OK new trade - Asset Man.

Sold AB @ 27.44
Bought OZM @ 13.23

What I like about trade:

At Layer 2( =2.5 SDs as per my PTF parameters)
Non-trending Ratio chart
% -> Mean at near term historic levels
Pair has good history of profitability.

What I don't like:

Corr. chart reading is low (56%)
Corr chart recent action pointing down.

***
Reason I include +/- for trade is as a rule there will always be something about a trade/chart you don't like. It is rare that all your ducks will be in a row, and they don't need to be to come up with profitable trades.
 

Attachments

Hello, I would like to get some input on what you do with analyst upgrades/downgrades.

Here is what I am doing but always looking to improve:

1. If a trade is triggered and the only news is an upgrade/downgrade I typically do not trade but I am wondering if I should be trading these triggers? If I do trade I will wait till the end of the day because the upgrades/downgraded stock tends to trend up(down) based on the upgrade. Then if everything else with the pair looks very good and it seems like the market overreacted then I will place this trade. But this is probably less than 20% to the time.

2. If the pair trade is open I get an upgrade/downgrade that is unfavorable to my long/short stock then it depends:

- If the trade is open for 8+ days and it is trending away from the ratio then I will probably just close it out right away and take my loss before it grows. I have a 10-12 trading day stop loss on my pairs.

- If the reason for the upgrade/downgrade is a big news event and it is going to negative to my position then I will likely close it out.

- If it is a top shelf analyst like Goldman then I pay more attention to the upgrade/downgrade

- If it is early in the trade and a more 'normal' upgrade/downgrade then I will just ride it out.

Reactions? What do other do? I would like to be as mechanical as possible with my trading rules but this area seems to require more subjective analysis.
 
You must log in or register to reply here.
Back
Top