Pair Trading Strategy Journal

Quote from thetrendfollowe:

...
I was reading an academic paper on Pairs Trading by a couple of guys from Yale the other day. Interestingly, according to them, stocks from different sectors didn't make much of a difference to returns. They tested their strategy from 1960s to 2002 and it showed significant outperformance of the S&P500, about double the excess return with third to half of the risk.

In addition, their strategy outperformed in bear and sideways markets (like the 70s) but underperformed during the recent 1990s bullmarket. Therefore, it would compliment my trendfollowing system rather nicely.

Nizar.
More...
I did not read the paper in all but I can say be very carefull. Things evolved over time. As you can see in their sugar coated equity curve, the early days of pairs trading where very good but people catched on the deal and opportunities are smaller theses days. I suggest you read the book Statistical Arbitrage before getting wet. It say that Merrill was doing that all automated in the 80's with good results but now with all the hedge funds and access to technology more people are doing it and the easy strategy is arbed away.
Whenever you see an academic paper think twice. Academics have their own schedule. They are in the business of "knowledge", we are in the business of making money.
 
Quote from total_keops:

I did not read the paper in all but I can say be very carefull. Things evolved over time. As you can see in their sugar coated equity curve, the early days of pairs trading where very good but people catched on the deal and opportunities are smaller theses days. I suggest you read the book Statistical Arbitrage before getting wet. It say that Merrill was doing that all automated in the 80's with good results but now with all the hedge funds and access to technology more people are doing it and the easy strategy is arbed away.
Whenever you see an academic paper think twice. Academics have their own schedule. They are in the business of "knowledge", we are in the business of making money.
More...

Yeh I have read that book just last week.
Wasn't the best.
Thanks for your thoughts.
 
Quote from jonnysharp:

2 new trades;

Long BAS @ 10.80
Short HAL @ 17.93
More...

Hey Jonny, are you still in the BAS/HAL pair (1/12/09), or did I miss the close of the pair?

Walt B
 
Hi Jonny.

Gotta question for you.

What do you do when you enter a pair but then a few days after entry they start diverging and actually don't converge back due to fundamental reasons (eg. one reported profit, the other a huge loss). Would you exit this pair on a discretionary basis because there's a good reason?

Or does your system have rules to prevent this from becoming too damaging?

Eg. a time-based stop (exit after X days if the pair does not revert to the mean) or a maximum stop loss (at like 4sd) ?

I'd be keen to hear your thoughts.

Nizar.
 
Quote from thetrendfollowe:

Hi Jonny.

Gotta question for you.

What do you do when you enter a pair but then a few days after entry they start diverging and actually don't converge back due to fundamental reasons (eg. one reported profit, the other a huge loss). Would you exit this pair on a discretionary basis because there's a good reason?

Or does your system have rules to prevent this from becoming too damaging?

Eg. a time-based stop (exit after X days if the pair does not revert to the mean) or a maximum stop loss (at like 4sd) ?

I'd be keen to hear your thoughts.

Nizar.
More...

I strictly stick to the system and wait for an exit signal, which is when the pair comes back to its mean.
 
Quote from jonnysharp:

I strictly stick to the system and wait for an exit signal, which is when the pair comes back to its mean.
More...

Hi Jonny.

Can you read my question again brother?
My question was relating to what you would do when the pair DOESN'T revert back to its mean within your expected timeframe.

So you'd just keep waiting?
Potentially forever?

Thanks.
 
Quote from thetrendfollowe:

Hi Jonny.

Can you read my question again brother?
My question was relating to what you would do when the pair DOESN'T revert back to its mean within your expected timeframe.

So you'd just keep waiting?
Potentially forever?

Thanks.
More...

Yes I wait as it will always come back eventually.
 
You must log in or register to reply here.
Back
Top