Monday, 21 April 2008

Portfolio Update: Long SPX and Crude

Just updated my portfolio page to reflect trades as of today's open based on my setups built around the Commitments of Traders reports. I went long the S&P 500 and crude oil based on bullish signals in those two setups. Good luck this week!

4 comments:

KN said...

Hi Alex

could you please elaborate on your statistical robustness calculation. I do not know if you have a post on this, if you have can you just point me in the right direction.

Thank you.

Alex Roslin said...

Hi Kn,

I've posted on that from time to time all over the blog. One way to find those posts is do a search for "robust" or some similar keyword in the search bar in the right-hand column. Also, you may look at the table linked to my Latest Signals page, where the notes talk about one measure I use - confidence intervals.

I've also chosen my setups based on other measures not yet on that table - like Sharpe, Robust Sharpe and regressed annual return. I plan to put those up soon. I'm right now working on some improvements to finding top setups, which include out-of-sample testing. Stay tuned.

Regards,
Alex

MT said...

Hi Alex, I've tried to do some copper setup fine-tuning and by mistake I altered trade delay by actually speeding up the trade reversal 3 weeks before signal occurs. The results are actually stunning - profit for the whole period is 20131%, last 2 years - 446%, last year - 123%. Of course this is not possible, since we do not know COT 3 weeks from now. However it might show, how valuble could be the knowledge, whether specs will increase or decrease their position substantially.

My 'inverse' setup for copper is:
std.dev -0,16 , delay (-3), mov.avg. 8.

MT said...

Hi Alex, I've tried to do some copper setup fine-tuning and by mistake I altered trade delay by actually speeding up the trade reversal 3 weeks before signal occurs. The results are actually stunning - profit for the whole period is 20131%, last 2 years - 446%, last year - 123%. Of course this is not possible, since we do not know COT 3 weeks from now. However it might show, how valuble could be the knowledge, whether specs will increase or decrease their position substantially.

My 'inverse' setup for copper is:
std.dev -0,16 , delay (-3), mov.avg. 8.