I haven't done much this year, so this will be a post that is quite brief in terms of new stuff, and will be a bit of a copy / paste from last years update. Performance wise this was a year of two halves, if the first half is three and a bit months long. I made out like a bandit initially and hit a HWM in late April, and then went straight down until October for a 20% drawdown. In cash terms this was quite painful, as I'd put more money in my trading account for the first time ever in early April (I've normally been withdrawing profits to invest elsewhere); my largest ever drawdown in live trading. I've made a modest recovery since then and my current d/d is around 17%. For the year then I am down around 1%. This is a little better than the AHL fund I compare myself against (-3% on a lower vol target), and a little worse than than the benchmarks (SG CTA +1%, BTOP50 +5%). There's roughly a 70% chance then that 2024 will end up being my 3rd losing calendar year (after 2018 and 2021). If was a CTA marketing person, I'd point out that both years were followed by +20% years. So let's see how this *very* small sample plays out. In terms of achievements from last year, there isn't much. I said I was busy with a non profit project, and I was, and will still be until around Q2 next year; plus I've had some additional unplanned family commitments. I've added a collaborator to pysystemtrade as otherwise that would be stagnating without me to approve pull requests. I don't plan to do very much personally with this now. I've done a small number of blog posts on new research; after a big gap I have recently restarted my monthly cycle of posts. Progress this year / Goals for next year: - teaching - despite the massive cuts in the UK university sector I've been retained; although perhaps I am relatively cheap compared to getting a tenured professor to teach something as part of a larger workload of research and admin - usual round of conferences - not sure why, but a bit of a lighter load in 2024 here; maybe I'm too annoying to the organisers when I diss the sponsors / other spreakers... or perhaps because I didn't have a book coming out. - consulting (picked up a couple more regular[ish] clients to do a couple of hours or so a month for, so getting a nice bit of income from this stream now) - exercise (was very poor in 2024 with my lowest number of Strava hours logged since 2019 and about 30% down from average- must do better next year) - my huge pile of research ideas. I'm currently reading up on stat arb. Stat arb across futures sounds like it might be fun, don't you think? - books: my book to do list is unchanged from last year so I hope to start writing "a book on systematic trading and investing for dummies; covering 'delta one' instruments stocks, ETF and crypto (sorry). Basically 'Leveraged Trading' but without the leverage" .... in Q2 next year. Best of luck to everyone. Rob
Familiar story. Stonking start and a demoralising drawdown into -ve territory for the rest of the year. The fact even the most hardened believers (us) have this nagging question in the back of our minds, "does it still work?", is I suppose why there is alpha and why it can never be mainstream. There also is another costly lesson for me. I defied the sage advice in ST and made the discretionary call to discard cocoa from my system at the beginning of the year on the basis it hadn't made money for decades in the backtests. If I hadn't I'd be up. I have therefore promised to myself to revisit ST and recommit to it as my religion of choice in 2025. Good luck and happy holidays to all.
LOL. Everyone in the systematic world I know does. I do this for a living, some of my alphas have pretty high sharpes (like 3ish), have all sorts of processes to evaluate alpha decay etc. And yet the question “is alpha X still working?” creeps up in every drawdown or even a flat period. Especially if a bunch of them draw down at the same time
I wish everyone a merry christmas! I am currently hunting down some smaller discrepancies between live trading and the backtest. I narrowed the effect down to the calculation of EWMA and the impact of roll dates and am interested in how you deal with this effect. My system did a roll in platinum on christmas. As EWMA is calculated on backadjusted prices the time series of platinum now has different close prices when compared to the day before christmas. This change in close prices results in different EWMAs when comparing the EWMA as of 23rd of December. This ultimately resulted in different portfolio positions in other instruments after doing dynamic optimisation. I observe this behavior quite often (potentially every time some instrument rolled). This sort of difference pops up regularly in my system because I always compare my portfolio as of prior EOD with the portfolio that I did want to have backtest-wise as shown in the attached image. I wonder how to deal with this effect. Perhaps higher shadow costs to encourage slower adjustments?
A level shift in prices as done in Panama back adjustment should not affect the calculation of ema, or vol scaling if done in price difference space. Rob
Thank you! We work with differences between EMAs and not absolute level of EMAs - I got it now. Have to dig deeper why this behaviour exists in my implementation
9 Jan 2024 - US National Day of Mourning Trading Schedule for Globex, BrokerTec, EBS and Trading Floor https://www.cmegroup.com/content/dam/cmegroup/trading-hours/files/day-of-mourning-january-9-2024.pdf
Hi Rob, Have the V2X and VIX positions been for Trend? Or for a different strategy? Not sure if you've ever broken down stats per market, but are you able to say whether trading these volatility indices has been profitable for trend, and if only on short side or both? I keep hearing that VIX etc cannot be traded on technical price based rules, it must be treated differently and traded on volatility rules like measuring the VRP, or on term structure rules, and other similar stories about why not to include them in trend following. Would love to hear your experience on this, especially if they have been in trend positions. Thanks!
1 more question - which instrument is COAL in your list on IB? I was looking at ICE - Rotterdam Coal (ATW code in IB), but watched several times where there was no bid at all only offers, in the most liquid contract. Makes me nervous seeing that..