This is the second in an unintentional series of essays about what AI will do to investors and how you must prepare. (This article lays the groundwork for understanding the vicious investment environment of 2029, which we'll explain in detail later.)
As a young investor, I was totally "EMH-brained." That is, I feared the efficient market hypothesis at every turn. Factor anomalies baffled me. I could scarcely believe that the mean-variance-optimal portfolio was nowhere close to market weighting. Most of all, I would obsessively ask of any trading strategy, "Why isn't this priced in already? Why can this make any money? Where is our edge?"
At Bridgewater, I taught classes using poker, figgie, and games of gamesmanship we devised. The purposes of the classes were:
- (endorsed everywhere) expected-value thinking
- (generally popular) fun
- (relevant more at faster shops) teaching market microstructure so that our young investors might better model flow-driven displacements
- (perhaps antisocial) getting them to ask, "Why is the other guy willing to make negative-expectancy trades? Where does my structural advantage come from?" Bluntly, "How can I get one over on the other guy?" In the ever-hilarious locution of edge-centric market thinkers, "What right do I have to get one over on the other guy?"
For short-term traders, this is ≈most of the job. Advantage gambling. But longer-term investors rarely trouble themselves about this! Yes, you need to know your view is truly variant, but there isn't much hand-wringing over "Why me? Why am I so special as to be the one who gets to do this trade?" Bridgewater was somewhere in the middle–––we had more (and cleaner) economic data than anyone, incredible infra for backtesting and nurturing perceptions, fifty years of economic reasoning, and a totally unique culture of brilliant people. Much ink was spilt on what made us special, especially when it came time to scale up our excellence using AI.
It shocked me to learn that most great fundamental investors don't spend much time metacogitating about "process" and "edge"! They just think about where security prices will probably go. "Yeah, we're gonna out-research and out-think the next guy." And empirically, this has worked well! When only three serious alpha-seekers are valuing this Bolivian mining conglomerate midcap spinoff, you can just focus on doing better than your two rivals, instead of obsessing over what makes you better. Competition makes the critical difference.
At one end of the spectrum, consider (an exaggerated version of) Sequoia('s VC arm). You talk to a founder; she wants you to fund her company. There's no question of "adverse selection" here, no "why me" (you may have to win the deal, but you don't have to wonder why it came to you). You're Sequoia; you get first look at every VC deal on the planet and make good money doing it; life is good. The only question before you is: will this founder make a kickass company? You have the luxury of considering only the fundamentals here.
For the other extreme, consider (an exaggerated version of) a credit ETF market maker. The bonds you're trading are slow and off-exchange, but the ETF reprices fast. You're answering RFQs and your bid keeps getting hit (i.e., people are selling the underlying bonds), but it's only on a few names that comprise the ETF. Earlier today, you saw the ETF trade up in a way that looked like one particular AM's sloppy footprint. You're not sure if it's still going on. There's no clean create/redeem arb yet (based on what non-stale quotes you're getting to see), but it certainly looks weird. Your bond marks say the ETF is rich, and you're tempted to bring the mids back into alignment with each other. But you're not alone–––there are other market makers on this ETF, and they're not adjusting their ETF quotes down as hard as it seems to you they should. The question here isn't "How much are these bonds worth?" or even "Of the bonds in this ETF that I'm not seeing quoted, how correlated should their moves be to the bonds I am seeing quoted?" You have answers to these questions, running systematically, and so do the other market makers you're competing with, and yours probably aren't that much better than theirs. No, the question is–––what the hell is going on? What are they seeing that I'm not? This price action looks obviously wrong... why might it not look obviously wrong to someone reasonable?
You know where you fall on the spectrum. Sure, podshop TMT quarter-guessing L/S is more competitive than globally opportunistic nano-cap long-only, but most public market investors are far closer to the first end of the spectrum than the second! There's just not much competition to identify great Nigerian retailers and hold them for a year.
As something becomes systemized and infinitely scalable, the questions transition from "What's true?" to "What's my edge?" This systemization process (Ray Dalio's vision of turning great discretionary investment frameworks into computer programs, tested through time and operating in every market) has come to great fruition in short-term trading. In long-term macro–––moderate fruition. But not much in fundamental equity L/S. Too much idiosyncrasy. The only way to scale has been "hire more analysts and teach them your ways," and this is hard. Analysts take time to teach, and they sometimes leave. (As Steve Cohen says, "When you have people, you have problems." Notably, he and other pod-kings haven't scaled fundamental investment orgs–––they've built platforms inside of which smaller fundamental investment orgs can exist.)
Meanwhile, you need AUM to hire them, you need statistically strong performance to get AUM, and to get statistically strong performance, you need an incredible win rate (either through luck or skill), or you need to take a lot of uncorrelated bets. If you win 52% of the time and take 100 bets a year, only 62% of your years will see outperformance before fees and trading costs. It'll take 17 years to prove with p<0.05 that you're not just a coin-flipper.* But of course, to take more bets or get your win rate up, it really helps to have analysts! This is why most fundamental equity L/S managers, even if they're intrinsic winners, struggle to scale. As a result, there isn't that much competition. You can spend your time thinking about the stock at hand, not torturing yourself over whether you have unnoticed information or an incredible framework.
Today's essay was mostly to explain the peaceful situation that prevails today. If you disagree with this point-in-time view, sound off on LinkedIn. Next time, I'll tell you what's coming. There's good news and bad news.