Presumably these are the critics who don't understand statistics or the mechanics of portfolio construction, but have to write meaningless articles about end of year performance? GAT
This is good news for us in ‘the community’, and I look very much forward to updates or anything you’re willing to share. I’m especially curious how accurate a backtest is for intraday trading. I believe that many intraday backtests turn from firmly positive after TC to very negative once you assume that you cross the bid-ask as opposed to getting filled by posting liquidity. Anyway, looking forward to this believe that we can all learn something, if nothing else. Merry Christmas.
I tend not do do detailed annual performance reviews, since I work on a UK fiscal year (7th to 6th April) basis. Nevertheless, if you're interested, I was up ~26% in 2019 in my IB trading account, which broke down to -3% in my 'equity neutral' bucket, and 29% in my futures trading. That's exactly the same as the fund I benchmark myself against on a vol adjusted basis. Clearly a good year for this sort of thing. Best of luck for 2020, GAT
You benchmark against AHL Diversity Alternative with 34 markets your strategy follows with fixed capital allocation over all 34 markets? What´s your correlation matrix across all of these "markets"? I was counting 18 financial futures markets out of the 34 at the start of your journal. On the "true commodity" side of things, I counted only 6 markets. How did this work out in the period Oct- Dec 2018? Why 34 markets? Quote: ....Clearly a good year for this sort of thing.... 2020 will be even better for "this sort of thing"......
If I can read between the lines, you're saying that my performance was good because there were strong trends in the financial markets in 2019, and my portfolio is heavilyl biased torwards financials. That is a reasonable thing to say, but isn't wholly accurate. To begin with, I'm not sure what you mean by 'fixed capital allocation'. I don't adjust my risk capital allocation to each trading strategy, but obviously position size will vary according to signal strength, and I don't have an equal risk weight across all 36 markets*. If I did then I'd have 31% in the 11 non financial markets I trade. Let me explain. * Used to be more. Since I started trading I've stopped trading a few markets for various reasons. I also use Eurostoxx purely for hedging. Not sure where 34 come from: maybe it was a list of positions, as I don't always have positions in every market. The sector structure at AHL historically was: fixed income, equities, FX, commodities, and later volatility. Now you could argue that 'commodities' has much more heterogeneity than fixed income, consisting as it does of several distinct sub sectors. That's partly more out in the non equal weight across sectors with a higher weight to commodities than say FX, but it's true that AHL's portfolio at least when I was there was quite heavy on the financials, which was a lot to do with the greater liquidity in those markets, although there were some other reasons. From memory it was about 35% non financials. For my own portfolio I broke commodities down further into metals, energies and ags. Within that I basically went for the most liquid markets that I could get cheap data for, with some attempt to get geographical diversification. My portfolio optimisation technique will allocate more to more diversifying markets, however to a degree it will be pulled back towards equal weight, so if the starting market list is unequal (i.e. with too many financials) it will probably still have unequal weights. My current weights are 'hacked' using my manual handcrafting method and so are more diverse: Here's my current list: Non financial (44%): Ags 20% Energy 8% Metals 16% Financial (56%): Bonds 15% STIR 4% Equities 13% FX 20% Vol 4% Not sure that 56% is 'heavily biased' towards financials, but of course you are free to disagree. Nevertheless, one thing I am hoping to achieve (probably not this year) is an expansion of my list of markets, which will result in probably more commodities. However that will need to come after I've modified my system so it is better at taking selective positions rather than trying to hold positions in all markets at the same time. GAT PS Oct-Dec 2018 (not that 3 months proves anything) I was down ~6%. In fact for the whole of 2018 (not that 1 year proves anything) I was down 1.6%.
This is an awesome result GAT, congrats!. I went live middle of last year and achieved +12% gross (10% net - not annualised - I have expensive markets which require ICE data, perhaps I should review whether they are strictly necessary and can be replaced, for example I chose IPE gasoil instead of NYMEX heating oil, etc). I am wondering if you have the headline return - gross and net for the second half of the year or is it a pain to get? I would love to benchmark, although I know you would have kicked my butt on returns, as would only be the right thing .
My net return was about 6% in the second half of the year, so you beat me . I spend around 0.6% a year on commissions so my gross return was probably around 6.3%. GAT
OK well I think I have the same vol target (25%) but I won't gloat as you would be the first to point out that 6 months proves nothing and half a year doesn't make a track record. Lets check back on this in a year I guess! Any more on this thread willing to share?
Hi Elder, Did you use Gat's pysystemtrade code for your trading? I used it last year and just broke even for the year, so I'm surprised others have had significantly different performance from mine.
Hey no, despite it being an excellent option, I didn't. I am very poor at reading other people's code so I wrote my own which actually took me much longer than I had anticipated (hence why I only went live middle of last year). I doubt that your returns would be code related. What's your vol target and how many markets (and asset classes) are you trading?