Any good book on Statistical Arbitrage?

I haven’t noticed any discernable changes in the stat arb space this year. I have heard that more hedge funds are returning to this type of trading, but due to my smaller scale and diversification across many pairs and baskets I can be a lot more nimble than them and hence still extract an edge.

Intra-day is where I have a noticeable edge so I don’t want to give too much away suffice to say anything I do is really an extension of prices getting out of whack. This will be caused by volume and news events. Plot some cointegrated intra-day pair ticks and notice what happens. Also stat arb doesn’t mean you always have to go long/short – you can just do one to “catch up” to your index or basket. Finally, some stocks have wider day ranges compared to their overnight ranges – Gummy has something on that which was a great insight for me.

I’m not sure of the protocol for linking to another discussion forum (as opposed to a file website) but have a look at Wilmott. Search for “pairs trading” in the 2005/06 posts, there’s some good intra-day ideas there. That’s part of where I got my start. Also search there for a PhD thesis by a guy called Burgess. It’s no longer on his personal site but it is still posted to the forum itself. Its 370 pages but worth the read.

As for gummy taking down his site. It says it on the front page of his site that he wants to get back to writing and painting. This all happened suddenly when he came back from his recent Caribbean cruise. It’s a real pity but you have to respect his decision – he did it for 10 years without any pay. I for one owe him a great debt of gratitude.
 
Quote from Matt1234au:

...You can get a good Excel spreadsheet and explanation by hunting around http://www.gummy-stuff.org/. He has a Google search add in at the site - look for “spearman”. Also try “pairs trading”....

I found this fellow's web site years ago when looking for how to automate yahoo stock scans. It's a treasure trove of excel and VBA, as well as general trading and math ideas.

Looks like I'll have to archive the whole web site now LOL.
 
Quote from Matt1234au:

I see an ad in efinancialcareers.com today where they say - Minimum Sharpe ratio of 4, Sharpe 10-15+ preferred.

Crazy. If someone has such an outstanding performance why get a job at a hedge fund? How much does a proven track record with a Sharpe ratio of 15+ worth?
 
Great posts Matt, thanks.

Does anybody here know of a source to get ETF constituents in a machine-readable format for all the ETFs at once? Or all the ETFs in a family maybe?
 
Quote from black diamond:

Great posts Matt, thanks.

Does anybody here know of a source to get ETF constituents in a machine-readable format for all the ETFs at once? Or all the ETFs in a family maybe?


TC200 has the components for most of the heavily traded ones. You can D/L to .txt file
 
Anyone know any papers about risk-management in non-market/sector neutral pairs trading strategies?

e.g. PGE-LLY vs XOM-BAC (not saying they are a valid pair, but just as an example). PGE-LLY may be a sector neutral trade, but XOM/BAC has two separate sectors involved ... which makes risk management a bit more complex. Any thoughts?
 
Gummy? A woodie plogic wannabe with a convoluted website and mental process. What a mess! No wonder people are confused
 
Yeah, I agree that gummy is not the most browsing friendly website. It takes an effort just to find things.

I am also interested in risk management for pair trading strategies. Are you supposed to figure out the correlation of returns among different pairs and use some kind of model(like mean-variance) to form the portfolio?
 
Most of the stuff I'm reading on stat arb (super late to the game haha) talks about the importance of cointegration...seems to be more important than =correl
 
Back
Top