Trading and Exchanges Chapter 14: Why Bid-Ask Spreads Exist (Part 1)
Harris calls chapter 14 the most important chapter in the book. Bold claim for page 297, but he backs it up. The lesson is simple and painful: uninformed traders lose money no matter what they do. Not because they pick the wrong side. Because they trade at all.
The Spread in One Sentence
The bid-ask spread is the price you pay for trading right now instead of waiting. You want to buy? You pay the ask. You want to sell? You get the bid. The gap goes to whoever was standing there offering you that immediacy.
Think of it like concert tickets. The venue sells them for $50. A scalper outside sells them for $80. That $30 gap is the spread. You are paying for speed and convenience.
When the spread is wide, trading is expensive and limit orders look attractive. When the spread is narrow, just use a market order and trade immediately.
Two Pieces of the Spread
Economists break the spread into two components, each existing for a different reason.
The transaction cost component covers the dealer’s business expenses. Rent, staff, technology, exchange fees, capital tied up in inventory. If everyone knew exactly what things were worth, this would be the entire spread. Prices would just bounce between bid and ask. Economists call this “bid-ask bounce,” and it is a minor form of volatility caused by impatient traders demanding immediacy.
The adverse selection component is the interesting one. Some traders know more than the dealer. When an informed trader buys, the price is about to go up and the dealer loses. The dealer does not know which clients are informed. So the dealer widens the spread on everyone, recovering from uninformed traders what they lose to informed ones.
This component is called “permanent” because the price changes do not reverse. They reflect genuine new information about value.
Here is the cool part. You can think about adverse selection from two angles and get the same number. Information angle: the dealer adjusts their value estimate after each trade based on the probability the trader was informed. Accounting angle: the dealer charges everyone extra to cover expected losses to informed traders. Same result either way. This is the Glosten-Milgrom theorem, one of the cleanest results in microstructure theory.
The Most Important Lesson in the Book
Harris literally titles section 14.3 “the most important lesson in this book for most readers.”
Uninformed traders lose no matter how they trade. Use limit orders? Informed traders eagerly trade against you when your price is wrong. You regret trading. Or informed traders compete on your side and grab the good fills, leaving you with nothing. You regret not trading.
Use market orders? The spread is wider because of informed traders, so you pay extra just for immediacy.
Flip a coin to decide? You will be right about direction half the time, but the spread eats your profits every time. Uninformed traders lose simply because they trade. The only escape is not trading.
Equilibrium Spreads in Auction Markets
In order-driven markets with no designated dealer, regular traders choose between limit orders and market orders. The spread adjusts until both strategies are equally attractive.
Harris starts with a silly simplified model. All traders identical, no information advantage, no time preference, free order cancellation. Result: equilibrium spread is zero. Trading is zero-sum with no friction, so expected costs for both strategies must be zero.
Obviously not reality. But it gives a clean baseline to add friction back.
Commissions. If limit orders cost more in fees, the spread widens until both sides share the cost equally.
Order management costs. Canceling and resubmitting limit orders takes time and money. Worse, when limit order traders cannot reprice fast enough, market order traders get a free timing option. They watch prices move and pick off stale limit orders before anyone can cancel them. The slower your cancellation, the bigger this option is for the other side.
Volatility. For volatile instruments, prices can move a lot before a limit order trader reacts. The timing option becomes very valuable. Spreads must be wider to compensate.
Time and risk. Impatient traders prefer market orders. Risk-averse traders prefer market orders because the outcome is predictable. To attract anyone to the limit order side, spreads must widen. In the real world, the most patient and risk-tolerant traders post limit orders. The most impatient use market orders. The spread is set by whichever trader sits right at the boundary of indifference.
The Takeaway
The bid-ask spread is not just a number on your screen. It is the price of immediacy, the cost of adverse selection, and compensation for providing liquidity. If you are uninformed, the spread is the mechanism through which you subsidize informed traders. Every time you trade, a piece of your money flows to someone who knows more than you. Harris does not sugarcoat it.
Part 2 will cover what makes spreads wide or narrow across different securities, and how volatility, competition, and trading activity shape the numbers you see on your screen.
Previous: Chapter 13: Dealers