Financial markets are efficient
What this means exactly is a deeper rabbit hole than we have time (or lifespan!) for. For now, just take my word for it (#971152) -- verify later.
Part of why has to do with information dispersion as per the first Money Stuff story of today (Susquehanna getting fucked on some options): rumours spread, arbitrage trades close, rents are traded/competed away. Result: trading profits, consistently beating the market, are only available for markets that are small and locked; where large players have trouble accessing or trading in volume; where shorting isn't fully available; or where unsophisticated players make errors, trade themselves out of position etc. "there are many other possibilities, and the literature of prediction market irrationality is in its earliest stages."
...ooh oh ohooohhh, you mean like PREDICTION MARKETS?!
Prediction markets are relatively small and inefficient compared with the stock market, and also fairly retail-oriented and, you know, fun for gambling. It is entirely plausible that, now, early in their development, there would be a lot of irrational investor biases that you could harvest. [...] So go find some systematic prediction-market biases, and systematically bet against them.
Betting "No" On Everything Outperforms
We have discussed, a handful of times, the simplest of these strategies, the “Nothing Ever Happens” factor: Bettors on prediction markets seem to be biased in favor of stuff happening (because that’s more fun), so betting “No” on every event seems to have positive returns.
It's true, on my Predyx bets I make more off my No bets than my positive bets: Case in point, recent F1 weekend when Antonelli miraculously won from #19 on the grid, I made ~15x my money... on that leg of the bet. It was so unlikely an outcome, though, that I only put like 78 sats on it. The profit in % was great; the profit in sats (~1,100) was rivalled by the 5,000-something No bids I'd placed on LeClerc and Lando Norris not winning (trading at 0.85-0.89).
The Nothing Ever Happens trading strategy usually carries the day. The rationale is that trading platforms select for people who already suspect/think something happen #1556891, and that the sports-gambling dominant clientele aren't sophisticated, rational traders: thus, there are plenty of SUCKERS at the poker table, plenty of FISH to financially harvest.
because prediction markets are so far mostly for sports gambling, you might consider all the standard sports-gambling biases. For instance, because betting heavy favorites is boring and betting underdogs is exciting, there might be positive returns to taking the other side and betting on the heavy favorites.
Also, the anti-Trumpian trade, from this new journal article “Political Bias in Decentralized Prediction Markets: Evidence from Trump-Related Contracts on Polymarket"
SUM:
“Polymarket traders like Trump, therefore they will systematically overestimate the probability of anything Trump-related happening, therefore you can systematically make money by betting against Trump-related things"
The dudes' strategy had a daily yield of $14 a day (I think? Not sure my econometric interpretation is right)
The anti-Trump strategy ranks at the 100th percentile of the benchmark distribution, generating a cumulative return of +$1,006 over the full 287-day period against -$1,028 for the mirrored pro-Trump strategy. An OLS trend regression yields a slope of +$14.07 per day (R 2 = 0.662, p < 0.001).
Go forth, traders, and multiply your discoveries of systematic overpricing/mispricing/overexcited one-sided supporters.
We shall arbitrage harvest all the suckers! (I'm sure in some sense this holds for the SN rewards pool, too: #1572772)
Not at all unrelated, there's a pretty cool arbitrageuers-sniff-out-profits story with stablecoins and USDT-> bank account redemption at Erebor Bank. Markets are efficient.