Pair Trading Strategy Journal

Hi Slave2Market,

Just looked at your post you referred to and read your interesting comments. The Flash Crash certainly did some damage to the PairTrade signals pairs, those losses on page 5 would probably have wiped me out.

But you are still pairs trading so I guess you are making it work for you even if it is a low return. The low risk does have some appeal but I think it can only be low risk if trade sizes are kept small. That's my approach anyway, lots of small trades.

The losses on my current trades are almost up to the gains I have had from my closed trades, for some reason almost all of my current trades are still diverging even though they have gone well beyond their historical spread. I must be patient though, they can suddenly snap back and bingo, a nice little profit can be taken.

The other thing I have considered is to watch the 20 day % from mean chart and enter the trade only when the chart shows that divergence has finished and convergence has started. It is sometimes a false indication but it could stop me from getting into the trades to early and having to sit through continued divergence.

Any thoughts or advice is welcome, as you have probably gathered, I am quite new to this game.
 
Quote from JohnP1000:

... Any thoughts or advice is welcome, as you have probably gathered, I am quite new to this game.
More...
In terms of advice on pairs trading, here’s my experience.

1. Don’t waste your time trying to day trade pairs. It’s extremely difficult to overcome slippage and commissions on an intraday timeframe. It might be possible to trade one side of a pair based on the action of the other side during the day but I wouldn’t call that “Pairs Trading” … I would call it a sector or sympathy play e.g. if Home Depot has great earnings and Lowes is not trading up then it might make sense to go long Lowes.

2. Don’t get hung up on the Cointegration babble … it’s not any better than Correlation or any other simple measure of whether a pair normally reverts to the mean. It is mainly useful when constructing a stationary time series out of several data series and then trying to find a method to profit from the group.

3. Trading small and trading a large basket of pairs should definitely increase your chances of survival. This should keep your losses on any single pair to a level that will not wipe out your account. Without a doubt you are going to occasionally get hammered by the following:
a. Being on the wrong side of a buyout.
b. A surprise announcement by management where Forward Earnings Guidance is dramatically lowered.
c. A major downgrade.
d. A variety of other things … think along the lines of MF Global.

4. Don’t trade Biotech, small Drug stocks or other high fliers. I think this is common sense but the risk of getting hammered is not worth the small gains.

5. Use a low-cost broker that has an extensive borrow list. Trading pairs is twice the commission in addition to Short Interest borrowing costs.

I definitely don’t consider myself to be a Pairs Trading expert and it is the LEAST profitable method that I have ever traded. Fortunately it can be scaled to a reasonable level and over the course of a year it can provide a 5% to 10% annual return with relatively low volatility … if done correctly and with a little luck. Currently as you noted, most pairs are not reverting ... this is where the drawdown fun can be enjoyed!

I’m sure plenty of imaginary traders will say, just buy a REIT, Bond Fund or MLP and get your low returns without having to trade. I would completely agree but it’s funny how that guaranteed return doesn’t look so good when you consider the big picture e.g. look at PHK, REM and many others in the last few weeks!!

Regards,

Slave2Market
 
Hi Slave2Market,

Day trading is out of the question for me, I have a day job that gets in the way. This is probably a good thing really as I don't think I am cut-out to be a day trader. I don't mind spending an hour or so in the evenings to place a few trades and with the US market being open when I get home it works out quite well.

I have taken in your points a) to d) and realise that I need to be more careful than I have been in checking for possible buy-outs etc. I have avoided any biotech stocks for the reasons you state, they are just too risky and I have the same view of the banks.

Some of my trades are starting to converge and I closed out UPS/FDX for a small profit which was boosted by a dividend payment against the UPS short so I am not too displeased with the result.

My profit from closed trades is still a bit more than that from my open trades but only to the tune of a decent meal in the pub but as I am only into my 5th week I don't think I can complain. Again, the PTF program is throwing up loads of trade exit signals, for all the trades I haven't taken, and the profit is wonderful, eighteen exit signals and only two showing a loss although they are both quite chunky.

I am in the UK and setting up spread bets through HL Markets who are linked into IGIndex. I don't know if IG is a low-cost broker compared to others but they do allow very small bets, as low as 24pence per point so it's possible to experiment a bit without risking too much money. Mind you, at 24p you don't make too much either!

If I can get half way to matching the results coming out of PTF then I will be very happy indeed, it may not be big on instant profits but it does look as if it could be a steady way to grow equity and once scaled up, could provide a bit of extra income.

I have just looked at PHK, it looks quite unwell.

Good luck with your trading.

Regards,
John
 
Quote from Slave2Market:

In terms of advice on pairs trading, here’s my experience.

1. Don’t waste your time trying to day trade pairs. It’s extremely difficult to overcome slippage and commissions on an intraday timeframe. It might be possible to trade one side of a pair based on the action of the other side during the day but I wouldn’t call that “Pairs Trading” … I would call it a sector or sympathy play e.g. if Home Depot has great earnings and Lowes is not trading up then it might make sense to go long Lowes.

2. Don’t get hung up on the Cointegration babble … it’s not any better than Correlation or any other simple measure of whether a pair normally reverts to the mean. It is mainly useful when constructing a stationary time series out of several data series and then trying to find a method to profit from the group.

3. Trading small and trading a large basket of pairs should definitely increase your chances of survival. This should keep your losses on any single pair to a level that will not wipe out your account. Without a doubt you are going to occasionally get hammered by the following:
a. Being on the wrong side of a buyout.
b. A surprise announcement by management where Forward Earnings Guidance is dramatically lowered.
c. A major downgrade.
d. A variety of other things … think along the lines of MF Global.

4. Don’t trade Biotech, small Drug stocks or other high fliers. I think this is common sense but the risk of getting hammered is not worth the small gains.

5. Use a low-cost broker that has an extensive borrow list. Trading pairs is twice the commission in addition to Short Interest borrowing costs.

I definitely don’t consider myself to be a Pairs Trading expert and it is the LEAST profitable method that I have ever traded. Fortunately it can be scaled to a reasonable level and over the course of a year it can provide a 5% to 10% annual return with relatively low volatility … if done correctly and with a little luck. Currently as you noted, most pairs are not reverting ... this is where the drawdown fun can be enjoyed!

I’m sure plenty of imaginary traders will say, just buy a REIT, Bond Fund or MLP and get your low returns without having to trade. I would completely agree but it’s funny how that guaranteed return doesn’t look so good when you consider the big picture e.g. look at PHK, REM and many others in the last few weeks!!

Regards,

Slave2Market
More...

Good post, this sums up my experience, on shorter time frames you're fighting commissions, on longer time frames you're fighting news, overnight margins, dividends etc etc. Time is risk.
 
Came across an idea involving pairs in the book Hedge fund market wizard.
For e.g, consumer discretionary ETF eventually converges to follow Dow uptrend even it takes a few days. But when the market falls, Dow decreases much faster causing a widening of profitable spread if you held long ETF short Dow.
Is this considered pairs trading?
What are the points to note if implementing this strategy?
 
Quote from virtualmoney:

Came across an idea involving pairs in the book Hedge fund market wizard.
For e.g, consumer discretionary ETF eventually converges to follow Dow uptrend even it takes a few days. But when the market falls, Dow decreases much faster causing a widening of profitable spread if you held long ETF short Dow.
Is this considered pairs trading?
What are the points to note if implementing this strategy?
More...
The DIA/XLY spread is low-volatility with a longer term trending behavior. I don't see any compelling tradable characteristics unless it is done on the micro-second time scale.

If you analyze ETF spreads you will find that most exhibit extremely high levels of correlation on an intraday time frame. It’s very difficult to overcome slippage and commission in order to make a profit on an intraday ETF spread trade. This is primarily the domain of the HFT firms.
 
I would deeply appreciate your response, as short as it may be, to the three take-aways (listed below) about pairs trading. Also, what is your experience and depth into pairs trading ?

Exactly 2 weeks ago, one poster (tony501) lamented (me, too) about this thread virtually going dormant. Yet it was at that time a new, important theme developed: the merits of pairs trading. The pace of postings picked up ! Conclusion ? ....... The first one out of the chute for me is that there is an innate attraction to this style trading, regardless of what I am picking up as being a bit downbeat about its worthiness to allowing time and effort being devoted to it. I make some general observations and ask you to comment on their validity based on your experience. It is extremely important to me that you do so, and hopefully you will find the results significant for you, too.

When I discovered this thread Pairs Trading Journal, four years ago, I knew nothing about the subject except for my own, 100% organic ,out- of- my- own- brain, trial and error trading experiences. I was like a savage in the wilderness stumbling on to civilization when I found this thread. I learned there were resources: posters willing to contribute, software to assist, and a history of writings relating to the characteristics of pairs. After catching up with a year of postings and continuing for a long time after reading posts--never posting myself--I had three major take-aways about what this population of posters were about within the world of pairs trading. (1) There is indeed enthusiasm akin to the same fascination as scientists trying to find life in outer space. There is a core belief in the righteousness of this style trading. (2) This is a generalization: poster / traders in this style-- not unlike many if not most other styles of trading--expect, when doing things with optimum correctness, the trade will be gotten into and closed in fairly short order, somewhere between momentum trading and swing trading. I found one active source on the web that a trader can subscribe to that distinctly impressed me as being at the momentum end of that spectrum, approaching day trading. (I hope I am not being unfair.) (3) With few exceptions, all pairs trading is done with applied analytics incorporating software that identifies pairs and their optimum entry and exit from a trade.

Would I be correct in believing that those three conclusions would describe pairs traders in a distribution, highly peaked, bell shaped curve with very tight, close-in standard deviations ? In other words all three of those conclusions fit virtually all pairs traders ?

The reason I ask is because I wonder how many suitably satisfied pairs traders like me fit only the first take-away ? I either have not evolved out of savage-hood or am believing there is a meritorious alternative to using sophisticated analytics. (My suspicion is that a few selected metrics will be helpful.) With respect to quickness in and out of the trade, I am agnostic to that, driven by the direction and characeristics of each individual trade's finger print. I find myself frequently believing in a trade needing to "mature" to realize its full potential, yet scalping on the way.

I recognize some posters as having long and deep experience with pairs trading while I don't recognize others currently posting. Everyone's comments would be greatly appreciated to this narrow, but I think very important subject. Please ask others you know that are pairs traders to jump into this. Focus on those three take-away's I have gleened and described. Also describe yourself for experience and depth in pairs trading on a scale of 1 to 5. (Fractions between are accepted !)
 
You must log in or register to reply here.
Back
Top