What moves a contract price

This is the mechanism everybody has in mind and it does work. Something becomes known, participants update, and the price moves to where the new marginal buyer and seller meet. On a liquid market with a genuinely surprising release, this can happen in seconds and the move is usually the most informative thing on the screen.

Two refinements matter. First, what moves a price is news relative to expectation, not news in absolute terms: an outcome everybody anticipated arriving on schedule moves nothing, because it was already in the price. Second, the size of the move is a statement about how much the release changed the probability, not about how important the release was.

This is also the only category where a large move genuinely deserves to be read as a signal about the world. The rest of this article is about the other categories, which are collectively more common and are routinely mistaken for this one.

As a resolution date approaches, prices tend to drift toward 0 or 1. Nothing needs to happen for this: there is simply less remaining opportunity for the outcome to change, so the probability of the current leader holding rises mechanically.

On a chart this looks like momentum and it is not. A contract climbing from 70¢ to 85¢ over two months while nothing much happened has not been vindicated by events; it has run out of time to be wrong. Traders who read that drift as confirmation of their thesis are reading a calendar.

The reverse also holds. A contract that has failed to move toward certainty as its deadline approaches is telling you something — usually that the market thinks the outcome is genuinely still open, or that there is a resolution question in the way.

Buying a contract commits capital until settlement, and that capital could have been earning something else. On a market resolving next week the effect is negligible. On one resolving in two years it dominates the last few cents of the price.

Work it: paying 97¢ for a contract that pays $1.00 in two years is a total return of 1.00 ÷ 0.97 = 1.0309, so 3.09% over the whole two years — about 1.53% a year. If cash yields more than that, nobody rational pays 97¢ for the contract however certain the outcome, and the price settles somewhere lower.

This explains a pattern that otherwise looks like market stupidity: long-dated contracts on near-certain outcomes trading at 88¢ or 90¢ rather than 98¢. Most of that gap is not doubt. It is the price of tying money up, and it closes on its own as the settlement date approaches.

Somebody closing a position to free capital for something else moves the price without knowing anything new. So does somebody entering because they just heard about the market, somebody rebalancing, and somebody whose account is being closed for unrelated reasons. None of it is information about the event and all of it looks identical on a chart.

In a thin book this is the majority of all movement. A market with $200 of depth can move six points on a single ordinary-sized order, and it will move back when someone posts on the other side. Reading that as a probability revision is reading noise.

The practical filter is depth rather than price change. A three-point move that consumed the whole book is much weaker evidence than a three-point move that traded through real size — and the second one is the one that would have been hard to do without conviction.

A price can stall well short of certainty for reasons that have nothing to do with the event. If the resolution criteria are ambiguous about the case that actually occurred, participants price the chance that the market resolves against the obvious reading, and that shows up as a contract stuck at 92¢ when the outcome looks settled.

The eight points there are a mixture: some genuine doubt about the world, some doubt about the wording, and some cost of locked capital. Only the first is what most readers assume they are looking at. Separating the three is most of what experienced participants are doing when they look at a market that seems obviously mispriced.

A market under active dispute is the extreme case. Its price is a view on a governance or administrative process, not on the underlying event, and the two should not be conflated.

A single large order tells you that one participant was willing to commit money at that price. It does not tell you they are right, and on a small market it does not even tell you the price moved for a reason — one order can be the entire market.

Large participants are wrong regularly, and their public trades are sometimes hedges against an exposure you cannot see, in which case the position says nothing about their view on the event at all. Following flow is a strategy that requires knowing why the flow exists, and usually you do not.

None of which is a recommendation either way. The point is narrower: “a whale bought” is an observation about order flow, not evidence about the outcome, and the two get conflated constantly.