Market Microstructure & Execution

How a limit order book actually works, the two-sided queue of resting orders, price-time priority, matching, and why your queue position is an asset with real option value.

The fundamental execution tradeoff, pay the spread for certainty with a market order, or post a limit order that is a free option you write to the market, bearing non-execution and adverse-selection risk.

The spread is compensation for three distinct costs, order processing, inventory, and adverse selection, plus the Roll model that recovers the effective spread from the autocovariance of transaction prices alone.

The market maker's winner's curse, because informed traders trade only when they have an edge, the flow that hits your quotes is systematically toxic, and the spread is the price you charge to survive it.

The canonical model of how a spread arises from pure adverse selection, competitive quotes set as Bayesian conditional expectations, updated trade by trade, with the bid and ask derived explicitly.

The strategic-informed-trader model, a single insider hides in noise-trader flow, the market maker prices linearly off aggregate order flow, and the equilibrium delivers Kyle's lambda, the price-impact coefficient λ = σ_v / (2σ_u).

Why your own trading moves the price against you, the split into permanent (information) and temporary (liquidity) impact, the propagator that ties them together, and why impact is the dominant cost for institutional size.

The strikingly universal empirical finding that a metaorder's price impact grows as the square root of its size relative to volume, its functional form, the propagator and latent-liquidity arguments for why, and how to use and abuse it.

The foundational optimal-execution framework, trade a position off to minimize expected impact cost plus a risk penalty on the leftover exposure, yielding a closed-form hyperbolic-sine trajectory and the efficient frontier of execution.

The trader's problem of converting a decision into fills at least cost, implementation shortfall as the objective, the impact-versus-timing-risk dilemma, and the extensions (adaptive, transient-impact, multi-asset) beyond Almgren-Chriss.

The three benchmark execution algorithms, time-weighted, volume-weighted, and percentage-of-volume, their scheduling math, when each is appropriate, and how each can be gamed.

The canonical high-frequency market-making model, an inventory-averse quoter derives a reservation price that skews with position, then sets optimal bid/ask offsets balancing spread capture against inventory risk, with a closed-form optimal spread.

Why microseconds are worth millions, co-location and the speed race, queue position as the prize, latency arbitrage and the sniping of stale quotes, and the Budish-Cramton-Shim argument that the continuous limit order book manufactures an arms race by design.

Slippage, spread, commission, and market impact that reduce strategy returns in practice.