Tell me more ×
Quantitative Finance Stack Exchange is a question and answer site for finance professionals and academics. It's 100% free, no registration required.
up vote 8 down vote favorite
6

Explain pair trading to a layman. What is it, why would you want to do it, and what are the risks? Provide a real life example.

share|improve this question
3  
See wikipedia: en.wikipedia.org/wiki/Pairs_trade. Can you try to put a little more effort into this question? – Shane Feb 1 '11 at 4:21
3  
@Shane: Many of the best answers on crossvalidated/stackoverflow are ones for explaining difficult concepts in plain language. It shows a sign of deep understanding of a subject if you can explain it clearly to a layperson. Is that not something we want to show emphasize here? – Neil McGuigan Feb 1 '11 at 4:33
3  
Also, wikilinks should be discouraged. Stack answers are often 100x better than a wikipedia article, as they are voted on. – Neil McGuigan Feb 1 '11 at 4:34
4  
I don't disagree, but during the private beta we really want to have high quality questions and answers to set a precedent for the rest of the site. I think that you can just expand a little on your question and improve it in the process... – Shane Feb 1 '11 at 4:45

3 Answers

up vote 16 down vote accepted

Pair trading is a market neutral bet. Instead of saying the market in general is going higher, you say one investment under/overvalued relative to another, typically similar, investment. The bet is that the spread between the two will widen or narrow depending on how you set it up.

For instance, say I feel GM is going to outperform Ford over the next year. I will buy GM's stock and short Ford's stock. By doing this the market is taken out of the picture, and I make money if the difference between GM's stock and Ford's is greater than it was when I undertook the investment.

share|improve this answer
up vote 1 down vote

"Quantitative Trading", Ernie Chan's book is a good starting point to learn pairs trading.

share|improve this answer
1  
Lots of downvotes but no comments... please enlighten us – Pete Feb 3 '11 at 22:54
I think this book is great to start with pairs trading, it has matlab code than you can test quickly with daily close prices and see how pairs trading works. Thought it could be of your interest so I don't understand the downvotes, I'm a newbie at stackexchange sites. – Juan M. Almodóvar Feb 4 '11 at 15:45
I won't down vote, since I like Chan's book. While Chan's book is good reading, this isn't a direct answer to the question. – dkantowitz Feb 7 '11 at 23:03
up vote 7 down vote

Quantitative pair trading (as we are on the quantitative finance forum) is based on cointegration.

Two stocks are said to be cointegrated if they move together, which means that they share the same long term trend.

Precisely: It exists a linear relationship between the price of the 2 stocks so that is mean reverting. (for instance the difference between the 2 is mean reverting). But it can be another relation.

Once you have a mean reverting basket, you can study this mean reversion (average, speed to come back to the mean, etc...) And it exists optimal strategies to trade this basket.

Don't forget that past behavior is not always a good indicator of future behavior. A cointegration relationship can evolve/break. Then:

1/ Think also about the exit/stop loss strategies. 2/ Try to make all your coefficients time varying

share|improve this answer
2  
I'm not certain that all quantitative trading of pairs involves cointegration. There can be other, quantitative reasons to trade a pair. For instance, I trade the crack spread: the 3-way spread between crude oil, gasoline, and heating oil futures. Those spreads get out of whack, creating trading opportunities, and we can fully quantify the relationship. But the futures are not cointegrated. – pteetor Feb 3 '11 at 1:13
this strategy has a good realised sharpe? ;) – RockScience Apr 1 '11 at 6:10
but I agree with your point. One can create any quantitative strategy with two pairs. – RockScience Apr 1 '11 at 6:16

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.