Maximizing Diversification Across Systems & Markets
The Trend Following Trading Think Tank
A farmer throws many seeds, a tennis player hits more than one ball, and a trend following trader places many probing bets in the markets, using a consistent risk budgeting framework for statistical equality.
Where we, that is Jerry Parker, Richard Brennan, and Simon M, all agree, is this: Diversification benefits are gratis. There’s no cost other than having the computer run a few more lines of code.
In my experience, MORE is better with respect to the number of:
systems we trade, each with a clear design and edge.
markets we trade using the systems mentioned under 1.
trading speeds we use across these systems and markets.
All the points 1-3 increase our total sample size and, as a result, improve the statistical body of evidence, which is our compass. It reduces the role of luck, improves robustness, and increases the consistency with which we can extract our trading edge in the markets.
On the other hand, LESS is better when it comes to the number of parameters on which a system depends. More parameters increase the risk of over-optimization, reduce sample size, and demand an almost exact replay of historical price sequences, which is very unlikely. Less is also better with respect to overlays that aren't rooted in the core system logic. If your edge is in following and exploiting price trends, you may want to keep things pure and stay focused on the objective rather than use overlays that add mean-reverting trades (—> early profit taking) to a method which is meant to do the opposite. Or: Trade the overlay on a standalone basis, i.e., decouple it like a satellite from the core, to see what its true and pure edge is.
These are some of the topics we discussed here:
Yesterday, on 13 May 2024, cocoa July futures dropped about 20% in a single session. Boom. Our trend system is long cocoa, so that move went against us. However, we trade about 100 different markets, i.e., cocoa is one position out of many. Overall, we lost 13 basis points yesterday, no big deal.
Sometimes, we hear that trading many markets results in performance dilution, and that this sort of dilution is bad. And yes, trading many markets can dilute the impact of big outlier trades relative to a portfolio that trades a lesser number of markets - depending on how positions are sized initially and also depending on how, or whether, positions are dynamically adjusted over time, e.g., in response to changing volatility of correlation patterns.
However, the opposite of dilution is concentration. Viewed from this PRESENT point of time, and looking back over the past couple of months, you would have liked to be massively long cocoa and for cocoa to be the dominant position in your portfolio. This view however isn’t the correct one as far as all possible FUTURE points in time are concerned, as yesterday's cocoa price move has shown. Personally, I like it when the impact of our many losing trades get diluted.
#happytrading
Your great narrative, mate, evoked feelings. If you're a long-term trend follower, why are you writing about something Cocoa did the other day? Who cares about Cocoa's one-day move? I'm perplexed that you offer a systematic trend program, but "options are used to manage/replace the delta of open futures positions and control risk on a discretionary basis." If I average the results of your three programs this year through the end of April, I get 7%. The replicator is up 16%; you know, the one who trades a few contracts. I'm trying to figure out how systematic you are with your discretionary options. Perhaps you are a competent discretionary trader. However, being discretionary does not make you systematic.
Stan Druckenmiller: “I strongly believe the only way to make long-term returns in our business that are superior is by being a pig. I think diversification and all the stuff they’re teaching at business school today is probably the most misguided concept anywhere. And if you look at all the great investors...they tend to [make] very, very concentrated bets. They see something, they bet it, and they bet the ranch on it. And that’s kind of the way my philosophy evolved...if you really see it, put all your eggs in one basket and then watch the basket very carefully.”
I suppose it could be argued that textbook MPT diversification (stocks + bonds + commodities long only) is a different kind of diversification from TF diversification (more markets = more potential of finding one outlier to pay for all the stop losses) and that going all in on a diversified trend system is a concentrated bet. But then you also have Dalio calling diversification (more MPT style) the Holy Grail.
Isn't it fascinating how views on diversification are so diametrically opposed and yet both camps have good long-term outcomes? Choose your poison I guess.