Price as probability
A contract that pays £1 if an event happens and nothing if it does not will trade at roughly the market's estimate of the probability. At 34p, the market is saying about a 34% chance.
That translation is the entire appeal. Instead of surveying opinions, you observe what people will risk money on, which filters out cheap talk. The theoretical case is strong: participants with genuine information have an incentive to trade on it, and the price aggregates what the crowd collectively knows.
Where they work and where they do not
Evidence suggests prediction markets do reasonably well on well-defined, short-horizon, heavily-traded questions — election outcomes, policy decisions, corporate events. They are meaningfully better than naive polling averages at some things and no better at others.
They are weak where it matters most: thin markets with few participants produce noisy prices, low-probability events are systematically overpriced (the same favourite-longshot bias seen in bookmaking), and resolution ambiguity can make a contract unfair regardless of what happened. The question "who will win the election" is clean; "will there be a recession" requires a definition that someone must adjudicate.
Manipulation
Because prices are read as forecasts, there is an incentive to move them for reasons unconnected to belief. Buying a candidate's contract to generate a headline about momentum can be cheap in a thin market, and documented attempts exist. Deep liquidity is the only real defence, and most of these markets do not have it.
The UK position
This is genuinely awkward regulatory territory. A binary contract on a financial underlying is a financial instrument, and the FCA banned the sale of binary options to retail consumers in 2019 after finding they caused substantial consumer harm.
A contract on a political or sporting outcome is generally a bet, regulated by the Gambling Commission, and betting exchanges have offered exactly this for years — a peer-to-peer order book in event outcomes, which is a prediction market in all but name.
US-based event exchanges that have grown rapidly are generally not open to UK residents. Anyone offering these to UK consumers needs either FCA or Gambling Commission authorisation, and firms operating without either are a well-established fraud pattern.
How you would actually trade it
For UK residents, the accessible form is a licensed betting exchange, where you can back and lay outcomes against other users at odds you set, with the operator taking commission on net winnings.
Winnings from gambling are not subject to UK tax, and losses are not deductible — the same trade-off as spread betting.
Check the resolution criteria before committing to anything. Most disputes in these markets are not about what happened but about whether what happened meets a definition written months earlier.
What goes wrong
- Binary outcomes. You are usually right or wiped out, with nothing in between.
- Thin liquidity makes prices unreliable and exit expensive.
- Resolution disputes where the criteria are ambiguous.
- Manipulation is cheap in small markets.
- Unlicensed operators targeting UK users are common; verify FCA or Gambling Commission authorisation.
- Favourite-longshot bias means unlikely outcomes are systematically overpriced.
This entry describes how a market works. It is not a recommendation, a price forecast, or advice about whether this belongs anywhere near your money. How entries are written.