Stat Arb - Indices Futures Trading

[MichaelPurcell]

Hi All,

I have been trading for 2 months using my method of Statistical Arbitrage.

Michael Purcell's Statistical Arbitrage Hedge Fund

I wanted to share my trades and performance as a means of accountability and also for the trading community to contribute their knowledge whilst being able to critique me.

My methodology is essentially finding highly correlated Stock Indices and trading on instances of divergence at calculated levels. Please feel free after reading the blog to make any comments or just follow it at your own leisure.

Good Luck fellow traders!

Michael Purcell's Statistical Arbitrage Hedge Fund

 

[MichaelPurcell]

My blogs trades for today :

Long emini S&P500 @ 1413.39 - Short SPI200 @ 4339

Long DIJA($10) @ 13205.3 - Short SPI200 @ 4339

Long NASDAQ100 @ 2782.6 - Short SPI200 @ 4339

SPI200 diverged away from the Americas by outperforming them by over 1% today. IMO the Americas have opportunity to correct this small divergence today and particularly during the open trading hours of the physical Australian market ~ New York time 10:00pm tonight. Reassess later today.

 

[Martinghoul]

How do you define stat arb?

 

[MichaelPurcell]

I trade short term (Hedged - Long/Short positions) between indices that are cointegrated over different time periods. If one of the pair of Indices diverges away by a x% given the specific relationship then on average it will converge back z% over a particular time frame according to my modelling. The relationship may trend overall as Indices will outperform each other in their relationship but it is particular short term divergences I look out for, given certain market/trading conditions, that on average after the event will trigger some sort of convergence and hence an arbitrage opportunity.

Please by all means critique. It is the very essence of why I joined this so I can be educated by those who work in the profession such as yourself.

Thanks

 

[ChocolateDigestive]

Please by all means critique. It is the very essence of why I joined this so I can be educated by those who work in the profession such as yourself.

Thanks

Dont worry mhoul will critique lol. I like mean reverting strategies as they are forgiving it always pays a man to be on the rightside.

 

[Martinghoul]

Please by all means critique. It is the very essence of why I joined this so I can be educated by those who work in the profession such as yourself.

Thanks

My intention was not to critique at all (I don't really know the specifics of your method sufficiently well for that). I was only curious to establish the terminology correctly. I remember that I have looked at doing this type of stuff a couple of years back. I think it's quite promising if you do it right and carefully.

 

[MichaelPurcell]

My methodology to date (almost 3 months) has yielded 31 winners and 11 losers with a return of 12.51%.

Is their a publication on the web that shows Hedge fund quarterly performance?

 

[Martinghoul]

My methodology to date (almost 3 months) has yielded 31 winners and 11 losers with a return of 12.51%.

Is their a publication on the web that shows Hedge fund quarterly performance?

There's a weekly HSBC publication that is occasionally made publicly available on the web.

 

[NVP]

Hi MP

Interesting......I'm a Forex man really but am naturally interested in any correlation Related trading ......do you study Joe Ross at all and his Arb work ?

N

 

[scose-no-doubt]

Please give me a link, NVP.

 

[RVLS]

Hello Michael.

Nice journal idea - I have been looking for threads on cointegration/hedging. Are you including fx and commissions in your returns data? Also, how are you modelling the cointegration?

 

[MichaelPurcell]

NVP and Martinghoul ~ Thanks for the reference to the information

Hi RVLS, I am including fx and the spread into the returns.
Using a VECM for the cointegration and some bands derived from a historical VaR which give me a rule of thumb probability. I also watch the volatility at certain x days as we all know these relationships get strained during volatile times

 

[MichaelPurcell]

Update on performance (since 23rd January 2012) has yielded 44 winners and 16 losers with a return of 17.84%.

 

[c0l$il]

I assume this is paper trading, right?

 

[kareem40]

Hi Michael,
Are you still trading the strategy? I trade similar strategy with Forex, but I do see and accept large draws. So I trade very small.
Thanks

 

[holmes16]

i am by no means a stat arb expert, although i've read a book or two on the subject. personally, i found some of the stat arb texts to be difficult to get your head around. what helped me was the following text which doesn't go into stat arb, but explains surrounding econometric theory:

Ender's Applied Time Series Econometrics.

here's a link.
Enders: Applied Econometric Time Series, 2nd Edition - Student Companion Site

if this book is intimidating, i would recommend starting with any introduction econometrics text.