Last week I wrote post reconciling the existence of skilled traders with the efficient market hypothesis (EMH). By far the most skilled trader is someone going by “Fausterion18”, who in his most recent Reddit post (archive.is) “make $1.8M in a single trade and $18 million for the year”. But for the past two years he has turned a couple-hundred-thousand-dollar account into tens of millions of dollars, by my estimate and his post history. I think this is the clearest evidence yet of actual skill in the market, with a verifiable track record using ordinary financial instruments, like options, that are available to retail traders, not proprietary methods or exotic contracts arranged between investment banks and hedge funds.
Yes, it’s possible Fausterion18 is not showing his loses, but his overall P&L is in excess of tens of millions of dollars of profit. [I have corroborated the screenshots of his trades with the actual timestamped trades on the exchange, so it is real and not a Photoshop.] Some may argue he’s engaging in deception, but again, his trades are real.
To try to reconcile his demonstrable skill with the EMH, let us assume that outcomes or phenomena can be split into two categories: outcomes which are determinate and outcomes which are indeterminate (random or unknowable). So in the latter, some examples would be the exact position of water droplets from a sprinkler, winning lottery numbers, or who will win the Super Bowl (at best we can narrow it down to maybe a dozen teams based on historical wins, but this is not useful in the context of betting, as those odds are already priced in).
The second category is not that useful–there are seemingly infinite things which are indeterminate, and we cannot make use of them to profit. At best we can derive a probability distribution function, like of the radius of the position of droplets relative to the sprinkler head, but this precludes the possibility of profit because the behavior of the system is public and known, governed by the laws of physics, so there is no possible information asymmetry or advantage. A hypothetical betting market would price the payoffs of the locations of the droplets in such a way as to prevent anyone from realizing a long-term excess return, as to be expected in a perfectly efficient market.
The first category is more interesting, that being phenomenon which can be predicted or understood in some way with greater certainty or fidelity than if it were random. This includes things which are more subtle such as a defective roulette wheel [0] or a slightly biased coin. So regarding skill, it is possible that market direction and fundamental trends, like GPU usage/demand or the tendency of stock prices to exhibit certain patterns based on hedge fund activity or other external perturbances, falls into the first category. There is an underlying system or process that can be understood, and hence exploited in some way. But because the behavior of this system is subtle or non-obvious, the market cannot price it in, unlike Super Bowl odds.
Additionally, in theory, it could be possible for a sufficiently high-IQ person (like a modern day Von Neumann or William James Sidis…they are famous, but it’s reasonable to assume there are other people also smart as them but who are unknown) or AI to understand everything that is determinate with such high precision to predict the future by deducing the underlying system or process, so as to appear to have ESP to outsiders (and also to profit). Yet this form of ESP is still consistent with the known laws of physics. I think this explains the success of some traders which cannot merely be dismissed as survivorship bias. They can effectively ‘read’ the future. I am sure such individuals exist, as demonstrated by the skill exhibited by Fausterion18 and other exceptional traders.
[0] There are famous examples of people who made a lot of money in Las Vegas by predicting where a roulette ball would land, due to imperfections of the wheel design. Thus, the outcome was not completely random and it was possible to realize a long-term excess return as those individuals had done. To ~100% of the population and the casino operators, the position of the ball was random, but upon closer inspection it was not. Same for the stock market, in which it’s possible for someone who is sufficiently smart or has sufficient computational power to decipher order from the noise.