Why Prices Near 1c and 99c Behave Differently

Most guidance about reading prediction market prices assumes you are somewhere in the middle of the range. Near the edges, several mechanical effects take over and the price stops being a clean probability estimate.

The tick is a floor

Prices move in one-cent increments on most venues. A contract genuinely worth a third of a cent cannot trade at a third of a cent; it trades at one cent, or it does not trade.

So the entire probability range from 0% to 1.5% is compressed into a single price of one cent. Reading a one-cent price as "1%" is a rounding artefact. It means "somewhere below about one and a half per cent, and we cannot be more precise than that."

The same applies at the top. A 99-cent price covers everything from about 98.5% upwards.

Capital cost dominates the return

Buying a contract at 98 cents to collect a dollar is a 2.04% gross return. If the market resolves in three months, that is roughly 8% annualised before fees — which sounds fine until you notice that you lose 100% of the capital in the case where you are wrong, and that the case where you are wrong is exactly the unmodelled tail.

This asymmetry is why the extremes stay compressed. The people who could push a 97-cent contract to 99 have better uses for 97 cents, and the people who could push a 3-cent contract to 1 have better uses for the other 97. Capital lockup is the general version of this.

Fees are proportionally enormous

A fee structure that costs a fraction of a cent per contract is negligible on a 50-cent contract and can be a significant fraction of the entire position on a 2-cent one.

Consider a fee of a quarter of a cent per contract. On a 50-cent purchase that is 0.5% of capital. On a 2-cent purchase it is 12.5% of capital, and it comes off both the entry and the exit. You now need the event to happen more than 2.5 cents' worth of the time simply to break even on a 2-cent contract.

This is the mechanical core of the favourite-longshot bias: the cost structure makes correcting tail mispricing unattractive, so it survives.

Moves at the extremes carry more information

A move from 2 to 4 cents doubles the implied odds. A move from 48 to 50 barely changes them. Measured in points they are identical; measured in evidence they are not remotely.

The practical consequence: a two-point move at the extreme deserves more attention than a two-point move in the middle, and any ranking of movement that treats them equally will systematically under-report the interesting ones. This site's movers board weights moves by how unusual they are for the market in question, which partially corrects for this, and telling a significant move from noise explains the method.

The "free money" illusion

Every so often somebody notices a contract at 98 cents on an outcome that seems certain and concludes it is free money. Three things are wrong with that.

The risk is not zero, it is just small and badly estimated. Two per cent of the time, something you did not think of happens. Resolution disputes, source failures, technicalities, genuine surprises. Over many such trades the losses arrive in a lump.

The return is small and the capital is committed. You cannot use that 98 cents for anything else until settlement, which may be weeks after the outcome is publicly obvious.

The exit is illiquid. If you change your mind, the bid may be 94. You have a 4% loss available immediately in a position whose maximum gain was 2%.

None of this means such trades are bad. It means they are a carry trade with tail risk, not free money, and they should be sized like one. Position sizing when you might be wrong is the guide.

Reading the extremes honestly

Below 5 cents: treat the number as a category, not a measurement. "Unlikely, and probably somewhat less likely than the price says."

Above 95 cents: treat as "very likely, with the remaining pennies mostly representing capital cost and tail risk rather than genuine uncertainty."

Between: the price is a reasonable probability estimate, subject to everything in what implied probability does not tell you.

For cross-venue comparison

Differences at the extremes are the least reliable signal on any gap board. A two-cent difference between a 3-cent market and a 1-cent market is enormous in relative terms and almost entirely explained by tick size and fee asymmetry.

This site therefore treats extreme-price pairs with more scepticism than mid-range ones when reporting a theoretical cross-market edge, because the arithmetic difference is real and the opportunity usually is not. Common mistakes when comparing venues lists this among the classics.

Next

When the outcomes do not add to 100% covers the multi-outcome version of the same arithmetic.