Hello Rob. Late November wasn't a very good time for most of us. One thing that stands out to me was the increase in the account's annualised volatility, the current level which stands out well above the annualised volatility level at any time since your automated system has been operating. When you eventually get a chance to look at what happened in late November, I would also be interested in a comparison between the old system vs Dynamic Optimisation system as far as the annualised volatility level goes. If you are able to model this, of course. Would annualised volatility have increased as much under the old system? My system (still) runs at a much more conservative level than your non-DO system, but even so, annualised volatility jumped only by about 60% on that particular day. Thanks, KH
Your graph is interesting in the sense that it shows a gradual increase in volatility since May 2021. Have you gradually changed some parameter settings? The system I'm running is close to what you would call the "old system", how it was described in Rob's book "Systematic Trading". I don't do any of the dynamic optimizations. My system also had a sudden step up in volatility in November, caused by the concerns about the Omicron variant of the Covid virus.
Hi @HobbyTrading The graph is a bit misleading. This is the IB account. I started using the framework with CFDs (not in this IB account) in November 2020, and added IB Sector ETFs before deciding to go fully futures in this IB account. There was very little trading in April / May, and until the Python system was ready to run in mid July 2021. For the first 3 months (until September rollovers) settings were stable, but were gradually increased as I became more comfortable with the framework and my attempt at Python coding. The settings are still very conservative (intentionally), with maintenance margin still only about 1/6 of account balance. I'll continue to gradually increase the account multiplier, usually at contract rollover time, and hopefully add another 2 or 3 instruments, over the next 6 months or so until margin increases to about 1/3 of account balance. I am still not comfortable with DO (i.e. I don't fully understand its operation), so was just curious as to whether the DO method behaved differently to the original method during the late November event. KH
I think it would be fair to say that the increase in Annualised Volatility of only 60% is far less than many longer term trend followers would have experienced in late November. KH
I run my account on an elevated volatility. Until the spike in November the maintenance margin on my positions was almost 70% of net liquidation value (NLV, i.e. account value). Once the volatility hit in November (and my account got a beating), all position sizes got reduced. Since then is the maintenance margin roughly one-third of NLV. By the way, once you increase your volatility setting you may want to start tracking the "initial margin requirement", not only the maintenance margin. If your initial margin gets too high IB won't let you increase your position sizes, or add a new position to your portfolio.
Hi Rob, Merry Xmas. Just getting around catching up with your latest blog posts (https://qoppac.blogspot.com/2021/12/my-trading-system.html). I see that in 2021 you have added quite a few new features to pysystemtrade. I have been using a late 2020 version of the code throughout 2021, and look forward to merge the latest changes and try out especially the dynamic instrument selection from a larger set of universe. Some random questions, in case you or anyone else following could answer. 1- How did you decide to use Barchart to backfill historical data. Have you considered/looked into CSI? 2- You seem to be running a high number of different strategies. Does it make sense to run so many with limited capital < 500k? (I have been running only 3 momentum plus carry) Can't wait for the new book. Best regards, Emre