Page 1 of 1

Poor Monte Carlo on Alt.Data

PostPosted: Thu Jun 13, 2013 5:36 am
by rjay
I've recently started checking the Monte Carlo Av.Annual Return for Alternate Data and I've found that it's often much lower than the Av.Annual Return for the Alt.Data.

For example, whereas the 5Port Av.Annual return is broadly similar to the Monte Carlo Return (as show on the 'Detailed Analysis' Performance tab), when I check the A1AvgAnnReturn and A1MCAvgAnnReturn under the Summary tab for each of my 3 Alternate Data periods, the AvgAnnReturn is often as much as 5 times higher than the MC.

Is this normal because I can't see why so many of my strategies would give such bad MC results. The Drawdown & PctProfitable for each alternate data also looks very good, which confuses me even more.

incidentally it would be nice if a future version of SS includes the Alt.Data Monte Carlo runs on the 'Detailed Analysis' Monte Carlo page.

Re: Poor Monte Carlo on Alt.Data

PostPosted: Thu Jun 13, 2013 8:20 am
by Overload
A Monte Carlo is very dependent on its time period and average days held during that time period. Is it possible that the data in your alternate data period is more bullish than your main data? That would explain it.

If you want to see detailed information for the Alternate Data Period, I'd recommend making a copy of your system and running a Detailed Analysis against just that alternate data. You'll be able to see the full MC chart, and you can verify that it's similar to the alternate data MC in the other system.

Pete

Re: Poor Monte Carlo on Alt.Data

PostPosted: Thu Jun 13, 2013 8:43 am
by rjay
I don't think the alt.data period is more bullish but It iss much shorter though (2 years vs 6 years for evaluation). So, if I'm concerned about MC, the Alt.data should be the same length as the evaluation data, yes ?

Re: Poor Monte Carlo on Alt.Data

PostPosted: Thu Jun 13, 2013 6:48 pm
by Overload
I'm not sure the length of the evaluation period matters much, as long as the alternate period is providing you with the alternate analysis you're looking for. The alternate data analysis is really nothing more than testing your system against an alternate set of prices. It's up to you how different you want that alternate data to be from your base evaluation.

Pete