Prediction markets have been taking off. Unsurprisingly, the casino-ification of everything with sports betting, election betting, etc. often gets framed as an information access democratization project with all of the fuzzy happy social good vibes that verbiage bestows. The usual framing is that markets are the best information aggregator we’ve ever built, so making them bigger, more liquid, and more accessible is an unambiguous social good. This is obviously incorrect with diminishing marginal returns in the extreme, but even more so, actually perpetuates inequality.

Information is only valuable when it’s asymmetric

The value of information is tied to its asymmetry of access: knowledge is counterfactually worth something only insofar as not everyone has access to it in an efficient market hypothesis world.

Today’s access asymmetry isn’t randomly distributed. Excutive insiders know about the acquisition first. The donor families hear about the policy shifts in real time. The 20B quant fund has the earned access from colocated serveris and proprietary data feeds. Access to information has always been a function of where you sit in the existing hierarchy of capital, power, and connection.

So “the value of information is its asymmetry” is also saying that the returns to information accrue to people who are already networked into the existing hierarchy, implying that when information becomes more valuable, by expanding markets volume and liquidity, can entrench existing power rather than democratize it. The frictions we have in markets both regulated (insider trading rules, halts, leverage caps, market cap mins) and emergent (pace of information exchange, capital access, etc.) are actually desirable as they regularize the net value extraction possible from information asymmetry.

Why I don’t like prediction markets

The idea that markets should be the primary channel for disemmination is a recent phenomena within the last fifty year. Historically, for most things that matter to people’s lives like local news, economy, foreign policy etc., the marginal value of a real-time price over ordinary reporting and news outlets is small. Fundamentally, equity markets are a growing asset class and so there’s value creation in expanding access: we’re willing to have retail pay to play (via lower information) because fundamentally the pie is growing and so over a sufficiently long time horizon of investment the EV is positive.

Prediction markets don’t allow for this. Fundamentally these assets are zero sum and the EV of retail trades is always negative, just like gambling, bringing along all of the other severe negative social effects and behaviors associated with it as well. Its hard for me to see how pursuing liquidity here is justified or worth any investment in as a society.