Pairs Trading

[Yisterwald]

Quote from Hooked2000:

It is not an easy trade by any means. a lot of preparation is needed to attempt this trade. You need to know fundamentals of each stock, ex divi dates, earning dates, as well as technical levels on each side of the pair. Also must have the capital and stomach to wait out the eventual reversion to mean, unless you are playing the opposite.

Agreed. View anyone who tells you it's an easy trade as a huckster with something to sell you. It does require a lot of preparation and homework, and the strong stomach mentioned by Hooked2000 bears closer inspection. Pairs traders most often (though not always) add to their losers, adding additional units or "layers", as the PairCo guys refer to them, and accumulating a maximum position on a scaled-in basis.

This is all well and good when you have 5 units on and it's going your way, but you will sometimes find yourself with your largest position ( or 3 largest positions ) going against you. You have to be able to stomach that as a pairs trader -- your significant losses will always be with size on. It's not uncommon to take a quick .50 winner on a new trade, while at the same time suffering a .50 loss on the day in an old position ten times as large. I don't think you need me to tell you it takes fortitude to stare that in the face.

You will have a large win rate as a pairs trader -- 80% or more -- but your expectancy is all in how you manage the big, big losers that that appealing trail of small winners will lead you to.

 

[itmediaco]

Quote from Yisterwald:

Agreed. View anyone who tells you it's an easy trade as a huckster with something to sell you. It does require a lot of preparation and homework, and the strong stomach mentioned by Hooked2000 bears closer inspection. Pairs traders most often (though not always) add to their losers, adding additional units or "layers", as the PairCo guys refer to them, and accumulating a maximum position on a scaled-in basis.

This is all well and good when you have 5 units on and it's going your way, but you will sometimes find yourself with your largest position ( or 3 largest positions ) going against you. You have to be able to stomach that as a pairs trader -- your significant losses will always be with size on. It's not uncommon to take a quick .50 winner on a new trade, while at the same time suffering a .50 loss on the day in an old position ten times as large. I don't think you need me to tell you it takes fortitude to stare that in the face.

You will have a large win rate as a pairs trader -- 80% or more -- but your expectancy is all in how you manage the big, big losers that that appealing trail of small winners will lead you to.

Your pairs trades are 80% successful? What is the success rate of your long/short portfolio?

this thread is getting more and more bizarre...

 

[Yisterwald]

Quote from itmediaco:

Your pairs trades are 80% successful? What is the success rate of your long/short portfolio?

this thread is getting more and more bizarre...

Not bizarre at all. It's no real trick to get a win-rate that hovers near 80% -- give or take a % or 2 -- in pairs trading, as long as you're willing to accumulate positions. My trading partner and I did it for years. I think pairs scalpers who go home flat would have to sell their souls to the devil to get that, though.

Still, the losers are so harsh that you really want about 90% winners -- and that's pretty tough....

 

[itmediaco]

Quote from Yisterwald:

Not bizarre at all. It's no real trick to get a win-rate that hovers near 80% -- give or take a % or 2 -- in pairs trading, as long as you're willing to accumulate positions. My trading partner and I did it for years. I think pairs scalpers who go home flat would have to sell their souls to the devil to get that, though.

Still, the losers are so harsh that you really want about 90% winners -- and that's pretty tough....

What I'm trying to understand is... How much of your success in pairs trading is derived from the probabilities of individual stocks (i.e., directional and volatility view) OR a statistically determined optimal level of correlation that has positive expectancy?

So far, I haven't found any evidence that pairs trading is more effective than a hedged portfolio, and if anything, it's far less efficient than trading synthetics.

 

[spinn]

this just has to be the lamest strategy I have ever seen.

How do these fools afford a website??????????

oh...yeah...commissions

have a very merry festivus pairs traders!!!!!

 

[Yisterwald]

Quote from itmediaco:

What I'm trying to understand is... How much of your success in pairs trading is derived from the probabilities of individual stocks (i.e., directional and volatility view) OR a statistically determined optimal level of correlation that has positive expectancy?

I think you might get different answers from different traders, but for my partner and I -- as well as the small trading group we developed our strategies with -- it was all about the directional and volatility view. We didn't ever try to find optimal values for correlation, nor did we enter into trades on correlation values alone.

So far, I haven't found any evidence that pairs trading is more effective than a hedged portfolio, and if anything, it's far less efficient than trading synthetics.

I realize that anecdotal statements aren't evidence, but I can tell you that during periods of high volatility you can achieve a much better rate of return with pairs than a simple hedged long/short portfolio. I'd count that as more effective. The thing is that volatility makes it easier to make money lots of ways, not just with pairs, and, like you, I'm skeptical that pairs have been any more effective than a hedged portfolio for the last few years.

But during the bubble, and in the crash that followed, they kicked ass...

 

[FullyArticulate]

Pair trading is practically identical to spread trading. If you think spread trading is "lame", you haven't been trading very long. Options traders, Futures traders, and Equity traders all spread between strikes, deliveries, and even contracts. Hogs spread against Corn is no different from PBY spread against AZO.

Pairs eliminate the broad market movements from the equation. Events like Feb 27th are non-events to pair traders. All I have to pick is which stock will outperform another, not whether the market will go up or down.

For example, leading into the last earnings, I felt PBY would outperform AZO. AZO had been on a tear, and PBY was waffling in "can they become profitable again" indecision. I thought PBY had turned it around. Take a look at the 8PBY-1AZO spread and you'll see the kind of moves that you can get. Even if you didn't want to get involved in earnings, you still could have made a killing just by jumping on the bandwagon the day after.

More important, when you're trading spreads, you have 3 times the information to draw upon--the technicals of both members of the spread, and the technicals of the spread as a whole.

On top of all of that, you can arrange your pairs trades to be carry-positive (i.e. you earn interest just by holding them)

Intraday pair trading works fine too. I've had a number of GG vs GLD intraday spread trades, although that's not really my style.

Yes, it costs more in commissions. So, I pay .008/"share" instead of .004, it's not that big of a deal.

 

[stock_trad3r]

The problem is everyone is already doing this

A system works great...when there aren't 1000 other people copying it

 

[gkishot]

Does anyone know of a good free software for charting pair spreads?

 

[StockBagger]

Quote from sandy1:

Yes, backtesting would be nice, but we all know that anyone can optimize the data to get any results you want. Do you really believe backtesting results posted on a website??

All I know is that I have been using this system for a couple months and I am making about $300-$500 a day, trading 500 share long and 500 share short positions. I follow 30 pairs every day and close out my positions EVERY day on the close. I average 9 trades per day. I would say I make money on 4 trades, break even on 3, and get stopped out of 2 (the plug-in had a built in stop alert).

If anyone is interested, I will email the developer and ask him if he would start a thread discussing his approach.

Sandy

Sandy,

When I spoke to the guy selling this and I asked for his overall results or a blotter screenshot, he stopped responding to me. When that happens it usually isnt good. I cant imagine a guy selling a system for 50 bucks that is worth more then that. I could be wrong!