My conversation with Simon on The Algorithmic Advantage, and a map of where to jump in
Simon from The Algorithmic Advantage had me on the show, and the episode went up last week. We hadn’t prepared much, which turned out to be the right call, because what came out is about the most complete walk-through of how I think about trading that exists in one place. Seventy-two minutes, from the meetup where a prop trader asked me what my edge was (I didn’t have one) to what my book looks like today, with a long stretch in the middle on how a solo trader should get started.
If you’re new here, start with this. If you’ve been reading for years, there are a couple of things in it I don’t think I’ve written down anywhere before, noted below.
The question
It was a traders’ meetup in Sydney, back when Quantopian was still running workshops and the local prop firms would send people along. I was telling a bloke from one of those firms about a machine learning framework I’d built for backtesting. I thought I was impressing him.
He said, “Look mate, that’s really cool, but what’s your edge?”
I didn’t have a bad answer. I didn’t have an answer at all, and I’d never even thought about the question. Everything I’ve done since, including ending up at that firm as a researcher and partner, has been an attempt to answer it properly. Simon wanted to know how much of the firm’s approach I brought home with me when I left to trade on my own, and the answer is bits and pieces, and probably not the bits people expect. The culture travels: the weekly meeting where your work gets pulled to pieces, and the paranoia that drives you to kill every idea as fast as you can so the good ones survive. The actual trading doesn’t travel at all. Low-latency stat arb on global futures needs co-location, exchange memberships, the right clearing arrangements and a team of people working on execution, and none of that fits in a home office.
So you ask a different question. If I can’t play that game, what games can I play?
Keep the day job
This is the first of the two things I haven’t written down before, and I expect to get emails about it.
Simon asked when you should quit your job to trade. My answer was that most people shouldn’t, at least for a long while. Trading income is lumpy and you don’t know what next month pays until next month has happened. If you need that P&L to cover the mortgage, the pressure forces every mistake in the book. You oversize, and you take trades you’d never have touched with a salary coming in. I’ve watched it make bad traders out of very smart people.
In addition, living off your account means pulling money out of it every month, and a permanent stream of withdrawals behaves like a strategy with an infinitely negative Sharpe ratio sitting in your portfolio. Compounding is about the closest thing a solo trader gets to a superpower, and it only works on capital you leave alone. More than a decade into this, I still run other income streams, partly for that reason (and partly because I like doing things other than staring at markets).
Of all the people in RW Pro, I reckon about one in ten trade full-time, and most of them are exited founders who don’t need the monthly income. I’d call that the correct structure for most of us. The day job isn’t a consolation prize or an indication of failure.
Breadth before polish
Simon brought up the fundamental law of active management, which says (roughly) that the quality of your portfolio scales with your skill times the square root of the number of independent bets you’re making. Square roots grow fastest at the start. So when you’ve got two or three strategies running, the gain from adding a fourth is big, and the gain from going back and polishing the first one is small.
The mistake we tend to make is trying to polish that first trade rather than adding a second and a third. You get your first trade working and then you spend your weekends bolting on overlays, because it’s the thing you understand best. The maths says to spend those weekends finding trade number two, three, four instead. Polishing becomes the right job later, once more breadth stops adding as much.
This is also why I think systematic trading is the only realistic way for one person to run a serious operation. Try running seven or eight edges that depend on you staring at the tape rather than following simple rules and see how long you last.
Where to jump in
The whole thing is worth the 72 minutes, but if you’re short on time:
- One strategy, hoping it works forever. Start at 17:58 (getting started) and then 49:30 (breadth before polish).
- Engineer who likes building things. 52:52 for what my stack actually looks like, which is less than you’d think, and 1:02:03 for the vanity projects I built before I needed them.
- Came out of a firm, or curious what the professionals do that you can copy. 8:30.
- Just want the trades. 36:03 covers the rebalance flows, including the SPY/TLT month-end trade I still put on by hand.
The full chapter list, courtesy of Simon:
- 8:30 The question that changed everything
- 17:58 How solo traders should get started
- 27:00 Building a diversified trading portfolio (the three layers)
- 36:03 Exploiting predictable rebalancing flows
- 45:17 How to find real trading edges
- 52:52 The systematic trader’s tech stack
- 59:12 R vs Python
- 1:04:16 Why machine learning isn’t an edge
- 1:07:10 The four hats of a solo trader
If you want more
A few of the ideas in the conversation have longer write-ups already. The backstory, including the meetup and the years of mistakes before it, is in the case study. The four hats got their own article, Everything Everywhere All at Once. The two ways a solo trader gets paid are in Building Trading Strategies with a Reason to Make Money over on Edge Alchemy, and the bit about what’s left on the table for us being noisy and a little mediocre, and why that’s fine, is Awkward, Noisy Edges are an Indie Trader’s Dream. Over the next few months I’ll be writing up the rest of the main ideas as a series on Edge Alchemy, starting with the two above.
Thanks to Simon for having me on. He’s running his own series on part-time trading over at The Algorithmic Advantage, and it’s worth your time.
