The Economics of the Bid-Ask Spread

The bid-ask spread looks like a rounding error — a quarter tick here, a cent there. Compounded across every trade a strategy makes, it is often the difference between a profitable system and an unprofitable one. But the spread is more interesting than a cost line: it is a price, set competitively, for three specific risks. Understanding what the spread charges for tells you when it will widen, when it will vanish, and who is on the other side of your fills.

A market maker's dilemma

Picture the participant quoting both sides of a futures contract. She posts a bid at 5000.00 and an offer at 5000.25, hoping to buy at the bid, sell at the offer, and pocket the quarter point. Her business has three costs, and the spread must cover all of them.

1. Adverse selection: the cost of trading with people who know more

The deepest insight in spread economics comes from Glosten and Milgrom (1985): even a market maker with zero operating costs and zero risk aversion must quote a positive spread, because some fraction of incoming orders comes from traders with better information. When an informed trader lifts her offer, the price is about to rise — she has sold exactly when she shouldn't have. She cannot identify informed orders individually, so she charges everyone a spread wide enough that profits from uninformed flow cover losses to informed flow.

This is why the spread is best understood as an information gauge. It widens around scheduled economic releases, around earnings, and in the minutes after unexpected news — not because market makers become greedy, but because the probability that the next order is informed has jumped. The spread is the market pricing its own ignorance in real time.

2. Inventory risk: the cost of holding what you didn't want

Every fill pushes the market maker away from her preferred (usually flat) position. Long 200 contracts after a wave of selling, she is exposed to exactly the kind of directional risk her business model tries to avoid. Inventory models (Ho and Stoll, 1981) show how quotes shift to manage this: an inventory-heavy market maker lowers both her bid and offer, leaning her prices to attract flow that unwinds the position. Volatility amplifies this cost directly — the longer and wilder the ride back to flat, the wider the spread must be. This is the mechanical link between volatility and spreads that every trader observes but few can explain.

3. Order-processing costs: the mundane remainder

Exchange fees, clearing fees, technology, connectivity, personnel. In modern electronic futures markets this component is small per contract, but it sets a hard floor: no one intermediates for free.

Reading the spread like an instrument

Once you see the spread as a three-part price, it becomes diagnostic:

• A tight, stable spread with deep queues signals a market where uninformed flow dominates and intermediation is competitive — the cheapest environment for execution.
• A spread that widens while depth thins signals rising adverse-selection risk: intermediaries suspect informed flow. Executing aggressively into this state pays a premium precisely when the premium is highest.
• A spread pinned at one tick with massive queues (common in the most liquid futures, like ES) shifts the entire competitive game from price to queue position — the spread can't compress further, so time priority becomes the scarce asset.

The spread you pay vs. the spread you see

Practitioners should measure two related but distinct quantities. The quoted spread is the displayed ask minus bid. The effective spread is twice the distance between your actual fill price and the mid-price at the moment you traded — what you truly paid for immediacy, including any price movement your own order caused. For any order larger than the inside size, the effective spread exceeds the quoted spread, sometimes dramatically.

Measuring effective spread requires timestamped records of your own order lifecycle against market state — one of the quieter arguments for treating execution logging as core infrastructure rather than compliance overhead. A firm that retains every signal, order, and fill with precise timestamps can compute its true cost of liquidity per strategy, per session, per market regime. A firm that doesn't is guessing.

Why systematic traders should care most

Discretionary traders pay the spread occasionally. Systematic strategies pay it structurally — every entry, every exit, every rebalance. A strategy that trades 20 times a day in a one-tick-spread market pays roughly 10 ticks daily in half-spreads before any edge is counted. Whether that toll is survivable depends entirely on the strategy's gross edge per trade, which is why spread and impact modeling belongs in the backtest, not the post-mortem.

The spread is not friction on top of the market. It is the market's honest quote for immediacy, information risk, and inventory risk — and it is the first price every strategy pays.

References

• Glosten, L. & Milgrom, P. (1985). "Bid, Ask and Transaction Prices in a Specialist Market with Heterogeneously Informed Traders." Journal of Financial Economics, 14(1).
• Ho, T. & Stoll, H. (1981). "Optimal Dealer Pricing Under Transactions and Return Uncertainty." Journal of Financial Economics, 9(1).
• Kyle, A. S. (1985). "Continuous Auctions and Insider Trading." Econometrica, 53(6).
• Harris, L. (2003). Trading and Exchanges. Oxford University Press.