Financial Markets & Market Microstructure

Trading under uncertainty requires separating the event's probability from the price paid and accepting that even a well-estimated 65% event loses 35% of the time.

A binary contract costs its market price and settles to either 1 or 0; YES and NO payoffs are complementary only when their resolution rules cover exactly the same event.

A 0-to-1 contract price can be read as a break-even probability under simple assumptions, but market frictions and trading motives can separate price from true event probability.

Realized profit and loss come from the settlement payoff minus the purchase price: one minus price on a win and negative price on a loss.

Expected value probability-weights every possible PnL; for one YES share it simplifies to personal probability minus purchase price.

Scaling one trade to 100 comparable trades turns abstract EV into total cost and payout counts, while making clear that majority accuracy can still lose money.

A binary purchase breaks even when expected payout equals execution price plus costs; before costs the required win probability is exactly the price paid.

Accuracy counts winning predictions, while profitability weights every win and loss by price, size, and cost; either metric can be high while the other is poor.

Fair value is the payoff expectation under your probability model, and edge is the difference between that value and the executable price—not a guarantee of realized profit.

Expected profit measures money per share or position, while ROI divides that edge by capital paid; fixed shares and fixed budget therefore create different exposure patterns.

YES and NO probabilities complement each other, but their executable prices need not; each side's EV must be computed against its own payoff and price.

A no-hedge strategy uses a reference market to estimate fair value and buys at a discount, but keeps the binary settlement risk instead of locking a spread.

Arbitrage closes every settlement branch with offsetting payoffs, while no-hedge trading accepts a directional outcome in exchange for potentially larger expected return.

Gross edge becomes net edge only after execution fees, spread, slippage, capital cost, and other non-overlapping expenses are measured without double counting.

A resting maker order is disproportionately filled when informed traders know its price is stale, so an advertised discount can disappear before execution.

Trading calibration asks whether execution-time probabilities match settlement frequencies on the orders the strategy actually fills, with Brier score preserving confidence quality.

Positive expected value permits losing streaks and negative finite samples; enough independent, comparable trades are needed before average performance can reveal a small edge.

Positive EV justifies consideration, not an all-in bet; fixed size, bankroll fractions, and Kelly sizing trade growth against estimation error and survival.

Many tickets can be one underlying bet when they share an event, reference, signal, or model; portfolio risk must be grouped by common failure source.

The most damaging prediction-market mistakes collapse probability, direction, price, realized PnL, and locked arbitrage into one misleading intuition.

How binary event contracts turn beliefs into prices, how order books and scoring-rule market makers supply liquidity, and how inventory, news jumps, microstructure noise, and boundary effects shape a robust quoting system.