Pair Trading Strategy Journal

Quote from jonnysharp:

Exited 2 trades;

Sold SE @ 13.09
Covered EP @ 7.52

Sold ESV @ 25.2
Covered ATN @ 14.39

Opened 1 new trade;

Long AXA @ 10.28
Short PUK @ 8.18
More...

I saw your trade above, long AXA short PUK. Recently I have had to adjust my pair system which has been grounded by RSI. Given the majority of all my stocks within my pairs now trade at RSI lower than 40 I wasn't generating any trade alerts since I like my stocks to have an RSI between 40 and 60. For those interested see my earlier posts regarding how I do pairs.

There are many ways to skin the cat, but recently I began analyzing the spread between the upper 2nd deviation and the lower 2nd deviation of the pair ratio versus the average spread.

In essence, if you were to graph what I am talking about it would look like a bollinger band.

If the pair's spread trades above or below 2 standard deviaitions, I enter a trade based on the spread reverting back to the mean.

I mention this because it creates a different result for pairs. For example where Johnny went long axa and short puk, my system yesterday said the opposite. Yesterday the pair ratio was around 1.29, today it is 1.25, or roughly 3% profit.

I am basing everything on the 50 day average.

Just a new method for folks to consider if you get impatient waiting for stocks to bounce back.
 
Quote from jonnysharp:

Same for both sides, half a cent per share.
More...

As far as I know, IB charges interest when shorting stocks.
Depending on the number of shares available to short, these fees can be higher.

I suppose you also pay these interest fees, this is not relevant?

Thanks
 
ok...finally have my act together...going to try to start trading pairs also...

Johnny:

one i found tonight that has a great ratio chart is CNQ/SM. The software issued a trade 5 or 6 days ago but they have even diverged more since then.

only question is the SM had a significant loss reported in last few days which has created the larger diversion. is that news reason not to take this trade?

also, i noticed you made some profit on the HP/UNT trade but then a day or two later the reverse trade was signaled and you didn't take that trade...and that too has gotten ever more divergent. is there a reason you would pass on this leg of the trade even though you took the other side and watch the pair?

Thanks

Mike
 
Quote from yobo:


I am basing everything on the 50 day average.

[/B]
More...

Besides spread and RSI, whatelse are you looking at?
You don't compare different time-frames?

I am analyzing, differential, ratio, spread, correlation and density curve, on different time frames, 15/30/90 and 180 days.

To be honest, I think I need to simplify my method, i am getting lost with so many data and charts.
 
Quote from tatankas:

Besides spread and RSI, whatelse are you looking at?
You don't compare different time-frames?

I am analyzing, differential, ratio, spread, correlation and density curve, on different time frames, 15/30/90 and 180 days.

To be honest, I think I need to simplify my method, i am getting lost with so many data and charts.
More...

I know what you mean by getting lost in data points...Been there done that. Been through many pair trading strategies.

My staple strategy examines the pair ratio and looks for abnormal deviations...2 or more, high correlation...70% or higher, RSI trending higher above 40 for the long stock, and for stock 2 trending lower below 60 utilizing a 90 day time frame for calculations.

The new method I am using now which seems to be working and averages 2-5 days per trade is the one I just described. Data points are based on the 50 day average...50 day correlation, 50 day average for the pair ratio and 50 day standard deviations. Than I calculate the spread between the upper and lower 2std deviations of the pair ratio. Once I know that I determine the average 50 day spread and again calculate the upper and lower band of the spread using 2 std deviations. When the spread deviates above the upper limit, I go long stock 2 and short stock 1 and vice versa for the other.
 
Quote from mlsignups:

ok...finally have my act together...going to try to start trading pairs also...

Johnny:

one i found tonight that has a great ratio chart is CNQ/SM. The software issued a trade 5 or 6 days ago but they have even diverged more since then.

only question is the SM had a significant loss reported in last few days which has created the larger diversion. is that news reason not to take this trade?

also, i noticed you made some profit on the HP/UNT trade but then a day or two later the reverse trade was signaled and you didn't take that trade...and that too has gotten ever more divergent. is there a reason you would pass on this leg of the trade even though you took the other side and watch the pair?

Thanks

Mike
More...

Mike, its up to you whether you filter out stocks affected by news items, although some of the best trades can come from stocks overreacting to news, for example news comes out monday and stock gets hammered into the week, sometimes thursday or friday that same week can be a good time to place a trade. Yes I don't take every signal on every pair I have, I treat each signal on a case by case basis and even though it may meet my criteria I will choose the best signal that I see.
 
Quote from jonnysharp:

New trade

Long PQ @ 3.37
Short ROSE @ 5.39
More...

Hi Johnny,

for me I decided to pass signals on penny stocks for several reasons. First of all, those stocks are a subject for many gamblers in market. They can be moved very easyly because of their low price. Even more when that stock is a low volume stock. Just see the examble with the pair you just posted. If you would have taken the signal on feb. 19, your position would have seen a massive drawdown. Even its a dollar neutral trade. Those stocks are that low for a reason and the selling pressure is usually higher than in higher priced stocks. I think there are a lot of other signals out there on pairs that include solid stocks. The risk is not worth it to take these ones imo. Just my 2 cents, Johnny.

Saico.
 
You must log in or register to reply here.
Back
Top