Insider trading, the buying and selling of financial instruments based on privileged information which is not available to a wider public is rightly forbidden. However, there is an other tool, often used by companies and investors to hedge their real-life exposure: Prediction markets.
Prediction markets such as Kalshi and Polymarket allow you to place bets on a number of topics in politics, sports, the economy etc. But more than being gambling sites, watching the score on different bets can give you a good feel for how a wider community feels about a topic. And while this isn’t necessarily always the case, people with inside knowledge may have staked their money here, which will no doubt give some added credibility to the odds.
Take for example the open market for the ‘Best Picture’ award at this year’s Oscars: One Battle After Another is clear favourite on Kalshi with a chance of 74% at the time of writing. In second place is Sinners with a paltry 15%. Or who will win the English football premier league? According to Polymarket at the time of writing this, Arsenal are favourites with 59%, followed by Manchester City at 40%.
There is something irresistibly neat about prediction markets. If democracy is government by counting heads, then prediction markets are forecasting by counting pounds. Or dollars. Or the jittery conviction of someone convinced they’ve spotted the future before everyone else.
The premise is disarmingly simple: Create a contract that pays £1 if an event happens — say, a candidate wins an election or inflation exceeds 4%. If that contract trades at 63p, the market is implying a 63% probability. No hand-waving. No “sources suggest”. Just a number, backed by money. It is tidy and quantified, it feels modern and faintly superior. And like most tidy things in public life, it deserves both admiration and suspicion.
The strongest case for prediction markets is that they aggregate information brutally efficiently. Instead of asking people what they think will happen, you ask what they are prepared to bet will happen. That small shift — from opinion to financial commitment — matters. People become marginally less reckless when their wallet is involved.
Markets compress dispersed knowledge into a price. Analysts, obsessives, insiders (legal ones, I hope), and contrarians all feed information into the system. The result is a constantly updated probability. Compared to a panel show populated by professional guessers, this is progress.
Another virtue is that markets speak in probabilities rather than certainties. A 70% chance is not a prophecy. It is an admission of uncertainty. In a culture addicted to binary thinking — win or lose, boom or bust — probabilistic language is a civilising force. It reminds us that the world is not scripted.
They also update in real time. When new information arrives, prices adjust. No need to wait for the next opinion poll or tomorrow’s newspaper. Markets behave like nervous systems: twitchy, reactive, always processing.
But prediction markets do not as such ‚predict‘ the future, they price expectations. And that is the entire point. Prices reflect incentives and beliefs. If participants are biased, ill-informed, or simply following a fashionable narrative, the price will embody that collective bias. Markets are only as wise as their participants — and participants are human, which means herding behaviour, overconfidence and overreaction to headlines. But it also means underreaction to structural change, narrative capture and the comforting delusion that this time it’s different.
There is a persistent myth that markets are uniquely rational. In truth, they are often merely efficient at aggregating shared delusions.
Liquidity adds another wrinkle. For a prediction market to function properly, it needs depth — enough participants and volume that prices genuinely reflect diverse information. Many prediction markets, however, are comparatively thin. It does not take vast sums to shift prices. That makes them vulnerable to manipulation, or at least to distortion.
In theory, anyone attempting to push prices away from “true” probabilities will lose money as others trade against them. In practice, that presumes there are enough others, sufficiently motivated and capitalised, to do so. Sometimes there aren’t.
Then there is the ethical discomfort. Create markets around elections, corporate collapses, pandemics or geopolitical crises, and you have attached financial incentives to unpleasant outcomes. The sums may be small, and the causal link tenuous, but the optics are awkward. One cannot entirely dismiss the queasy feeling that certain events should not double as tradable entertainment.
Even in more mundane contexts, incentives can blur. Employees betting against their own firm’s product launch. Traders taking positions aligned with their policy preferences. The line between forecasting and influencing can grow fuzzy.
The Deeper Risk: False Objectivity. Prediction markets produce a number. The number moves. It looks empirical and scientific. It feels objective. But the number is not a measurement of reality. It is a measurement of belief — specifically, the weighted beliefs of those who chose to participate. If the crowd shares the same blind spots, the aggregated belief will simply reflect a beautifully averaged error.
History offers no shortage of examples where markets mispriced risk for years. Financial crises were not caused by a lack of prices; they were caused by collective misjudgement embedded in prices. There is no reason to assume prediction markets are immune to the same pathology.
When the market says there is a 90% chance of something happening, it is also quietly admitting that one time in ten it will be wrong. The trouble is that humans are very bad at internalising what “one time in ten” really means. When the unlikely occurs, we treat it as a failure of foresight rather than the statistical inevitability it always was.
Prediction markets tend to perform best when the question is clearly defined, objectively verifiable, and reasonably near-term. They shine when participants possess genuinely diverse information and the market is sufficiently liquid.
They struggle with long-term structural questions, low-liquidity environments, and situations where the market itself may influence the outcome. And they are perpetually vulnerable to the fact that crowds, however aggregated, can be collectively wrong.
The sensible position, therefore, is neither reverence nor dismissal. Prediction markets are useful tools. Often better than pundits. Frequently better than casual polling. Certainly better than the confident man on television who has a “strong feeling”. But they are not crystal balls.
Treat them as one input among many. A probabilistic signal in a noisy world. Respect their ability to synthesise information, but resist the temptation to outsource judgement to a fluctuating price. Admire their elegance. Exploit their utility. Distrust their aura of inevitability.
But in the meantime, prediction markets can be a very useful tool to gauge public sentiment and how this possibly shifts over time. But remember: when the market looks supremely confident, that is precisely when humility is most required.
The grumpy older man 