Size Always Matters

One size for all or scale up for special trades? That is a never-ending debate in our business and I was reminded of it this week through an exchange of a tweets that really brought this issue to light.

On Thursday @FemaleTrader_A wrote, “Most of p&l in my life came from aggressively varying size and not from picking set ups. When RR in my favour+conviction, going really big, whilst being flat/small on the rest. Not cutting a bit, but cutting the whole thing.

$$ is at edge of ‘unpleasant’ usually…”

To which @Trader_Dante replied,”Totally get what you are saying although I had many times when my conviction was strongest and my bet size largest that I was wrong and ended up undoing a lot of work.
Now I only vary size based on stats personally.”

So who is right?

Well, the answer as is so often the case in trading is that it depends.

If you listen to most quant and algo driven traders the answer is obvious. Every trade should be the same size. That’s because the underlying worldview of algo trading is that every single trade is the same as the other. Algo traders view each trade with all the romance of Stalinist industrial planners -- it’s just a widget like any other so just stamp it out of your machine and go on to the next trade.

There are good reasons for following this protocol. If your trading strategy is based on some statistical sample set, especially one that’s large than varying size per trade will greatly skew your results from the expected value. If your system is based on high reward to risk profiles than you may get lucky and actually beat your expectancy, but more often than not if you are day trading the risk/reward skew is negative sometimes significantly so and in that case, size increase could be catastrophic to your outcome.

Yet history is full of examples of traders doing just the opposite. In fact, the greatest trader of all time -- George Soros -- is famous for stating that when you are right on something you can’t be big enough. His exact words were, “It takes courage to be a pig.”

The Soros GBPUSD trade when his Quantum fund broke the Bank of England and made $1 Billion in one day was the classic example of “have a hunch, bet a bunch” style that made him so profitable. Yet when you deconstruct what happened you realize that Soros wasn’t nearly as cavalier as he would have you believe. No doubt that the GBPUSD trade was massive, but its total risk was essentially limited to Quantum’s profits for the year. When Soros executed that trade his hedge fund was already up for the year by almost a billion dollars so the size of his bet on the pound was limited to profits, not capital. Still, it was a remarkable feat. Few of us would be willing to gamble a year’s worth of winnings on one single trade. And Soros did this many times throughout his career which is of course what makes him so great.

The history of Soros’s trading also shows that all trades are NOT the same. That there are points in market history -- inflection points when being a pig is absolutely the right thing to do. This, of course, requires nuance, understanding, context -- all things that are totally foreign to an algo trading robot.

Where does it leave us regular trading folk? Back to our own devices. You have to make peace with the approach you choose. I day trade FX majors every day. There are times during the session when I am very convinced that a combo of news and price action will move the pair in a certain way. Usually, I am right. But I never scale up. Like a grinder at the poker table, I keep all my trades the same size and focus on pumping out a few pips per day. It’s not glorious but I have a bigger goal in mind. I am focused on developing long-term low vol returns that I can compound for years to come. Furthermore, I know that when I am wrong BIG, I lose my emotional balance and it can take months for me to recover my equity as I spiral into a vortex of self-destructive behavior.

As always, trading is psychological rather than logical so you need to choose your path. Much as I would like to be like George Soros, I am a pip grinder instead.

Boris Schlossberg

Leave a Comment

Your email address will not be published. Required fields are marked *