When Is a Market Not a Market?

In a Wall Street Journal op-ed titled When Is a Sports Bet Not a Sports Bet?, Kalshi CEO Tarek Mansour makes a compelling case for distinguishing between betting platforms and prediction markets. Beyond the immediate legal consequences that follow from the ascribed ‘betting’ or ‘market’ labels lies an interesting and more foundational question: what really is a market?

By way of continuing that conversation, I want to explore the philosophical conception of a market and its basic components. What follows are some preliminary reflections on some of the points raised in the piece, thinking out loud about the question of what constitutes a market.

The definition of a market presented in Mansour’s article is structural, teleological, and epistemic: a prediction market often behaves in the same way as a betting platform. They appear the same, but they are structurally different, have a different goal, and operate on different knowledge and transparency levels.

According to Mansour, a prediction market facilitates a transaction between the users instead of pitting itself against them. Speculative behavior certainly exists, but the end goal of prediction markets is to mitigate risk and hedge against uncertainty, just like the example of the bar owner hedging against the local team not advancing through to the next playoff round.

Prediction markets are also a source of information that is considered a public good. They offer companies, public institutions, and individuals alike aggregated intelligence reflected in a transparent and clear signal or datum: market prices. This, in theory, solves Hayek’s dispersed knowledge problem.

Given their structural similarity to derivatives, prediction markets should therefore be subject to federal rather than state regulation, acknowledging also that there are some unique challenges specific to this domain in sports and election contexts that ought to be addressed and properly regulated.

Compelling as that case seems, some issues need to be examined further, keeping the legal and regulatory implications aside.

The underlying assumption of prediction markets is that more information leads to more accurate prices and forecasting. We tend to fall under the illusion that this assumption is correct, until things go haywire. As Soros points out in The Alchemy of Finance, financial markets and financial modeling are not scientific in the same way as physics or biology are, even though both rely on the scientific method. This is because financial investing operates on a desired outcome basis (hence the alchemy of finance) rather than the discovery of ‘truth’.

Prediction markets also operate in the reflexive realm, with human agents who perceive market signals in different ways, and respond to this information unpredictably, causing markets to behave differently, which creates feedback loops that can distort the very outcomes they aim to forecast.

Prediction markets can reasonably be viewed as betting platforms despite the structural difference for the following reason. Financial markets, including futures, may look like they offer a space for speculators to bet on an event with a binary outcome: will the price increase or decrease at a certain time in the future compared to the buying price? In both cases speculators risk losing their money, too. However, in theory, derivatives are the very tools that act as a hedge against future risk, based on existing financial instruments, operating within broader financial systems that offer a spectrum of scenarios and results that don’t automatically lead to one losing all the money they invested.

The underlying nature of the commodity held, and the potential scenarios, therefore, are quite distinct. Abstract as derivative contracts can be, they are anchored in a concrete, tangible commodity out there. While, theoretically speaking, companies use them as a hedge to lock in prices in the future to facilitate their operations in the present, and make their balance sheets easier to tackle, they remain a zero-sum game for speculators. What separates them from prediction markets is that futures markets essentially exist for price discovery. This process enables hedging against future risk. As such, in principle, they carry economic value and public utility.

Prediction markets, by contrast, are more focused on events either happening or not, decoupled from any underlying commodity. While a case like the bar owner’s can exist, it is mostly an abstract, ideal case scenario: the bar owner’s local club can still make it to the next round, yet customers might still not show up because the bar next door has a better promotion, prices, food, or ambiance.

Prediction markets thus risk becoming speculative platforms rather than hedging tools against any underlying exposure. Those who really want to hedge against future risk can do so by buying insurance, diversifying their income streams, or building a diversified portfolio aligned with their risk appetite. If the volume is driven by 80% speculators, then fundamentally the goal is not so much hedging as wagering over whether a specific event will take place or not.

In addition, commodifying and trading things such as values, ideas, and democratic processes like elections poses questions that need to be explored. If prediction markets are institutionalized as forecasting mechanisms and hedging platforms, values and elections become anchored in financial interest and speculation rather than civic responsibility.

This legitimizes certain practices, including market manipulation. While these may already take place every round of elections, sometimes inconspicuously, other times less so, blurring the boundaries between what is legal or not, ethical or not, institutionalizing them requires that we think about what a public good means, what sovereignty looks like, and what would constitute a better political system. All this brings forward notable ethical and legal implications.

By analogy, and this is not to compare the two domains, but rather just to give a more concrete example of the legal and ethical dimensions that need to be examined, consider how democratic societies place a ban on the buying and selling of votes, even between consenting agents, because pricing a civic duty risks fundamentally corrupting it.

So, does the claim that structure defines the difference between betting platforms and prediction markets make sense? Perhaps, but while the underlying structure may be necessary, it isn’t sufficient. Markets have far more foundational elements than just order books or matching mechanisms. Legal arbitrage may grant them a legitimate exchange status rather than betting status. Still, the pressing issues that need to be addressed are more substantially philosophical.

For further preliminary reading, check out the SEP article about markets here.