wilmott.com

Posted: **June 3rd, 2006, 2:15 pm**

Any thoughts on this area....??? ie is there juice in this area???

Posted: **June 3rd, 2006, 2:32 pm**

Is there "juice"? Sure there is. Sometimes it's sweet. Sometimes it's like spoiled milk. Try to trade "well-known" pairs that have a lot of participation and interest. Those "pairs" should behave more predictably. If something weird happens with one leg of the pair, don't let a small loss become a big loss.

Posted: **June 3rd, 2006, 5:34 pm**

what is it?

Posted: **June 3rd, 2006, 7:23 pm**

well how mathematical does this area get....???? eg what kind of maths is used and how deep does it get...???

Posted: **June 3rd, 2006, 11:57 pm**

QuoteOriginally posted by: nazzdackIf something weird happens with one leg of the pair, don't let a small loss become a big loss.Exactly. Very tight risk controls are used in pairs trading. My approach to pairs trading therefore would be the following:Step 1. Identify pairs that are popularly traded and try to get a feel for how the pairstraders operate (especially their risk control approach). Do not trade the pairs yourself, just monitor them.Step 2. When "something weird" happens to a pair, the pairstraders will suffer losses. At some point, following nazzdacks advice they'll have to take action. Get me out, get me out, get me the f___ out! The spread will widen even further as they exit their position.Step 3. At this point, and at this point only, you come in and buy into the busted pairs trade.You might call it "being the pairs trader of last resort" or you might compare it to Buffett's manoeuver in September 1998 when he tried to buy up the arbitrage portolio of LTCM for a song when they were having problems.Something imaginative like this might work; but me-too pairs trading, forget about it.

Posted: **June 4th, 2006, 12:41 am**

QuoteOriginally posted by: playerwhat kind of maths is used and how deep does it get...???If you have (or can find) a table of critical values for the Augmented Dickey Fuller test, if you have read Chapter 1 of any book on co-integration, and if you know how to do a two-variable orthogonal least squares regression, you have the maths that it takes to get started in pairs trading.Unfortunately so do 1.0E06 other people.

Posted: **June 4th, 2006, 9:13 pm**

can someone explain what volatility pairs trading is???

Posted: **June 4th, 2006, 9:45 pm**

http://en.wikipedia.org/wiki/Pairs_trade(article could be improved, but gets the basic idea across)

Posted: **June 5th, 2006, 5:59 am**

That defines pais trading.....what about vol pairs trading??? Is it just the same thing??

Posted: **June 9th, 2006, 11:59 am**

You trade implied vol spreads on markets whose vols are correlated. Its not good enough to pair-up markets whose RETURNS are correlated....their GARCH parameters need to be correlated.Once you've idenrified a pair, and the spread is out, you can trade be either:1. just long vol outright on the relatively low vol2. long/short straddles both atm3. long/short straddles otm and use cash to buy a put for some d/side protection...any bonus skew helps here.4. var swaps

Posted: **June 9th, 2006, 4:06 pm**

you have any papers on this??

Posted: **June 12th, 2006, 10:47 am**

http://papers.ssrn.com/sol3/papers.cfm? ... _id=881550

Posted: **June 13th, 2006, 12:35 am**

Is there any way to trade the correlation skew there?Just a thought.

Posted: **June 13th, 2006, 11:21 am**

by corr skew are talking about CDO's??

Posted: **June 14th, 2006, 12:14 am**

no. I mean by locking in the implied vol spread wrt moneyness (or strike), you are actually trading the corr skew. Maybe that is overkill.