Strategic Economics

Price Discovery and How Information Gets Into a Price

Nobody announces what an asset is worth. A price is assembled, one trade at a time, out of what individual people are willing to back with money — and you can measure how much of that assembly happened, where it happened, and how much of a given move was learning rather than noise.

  • Advanced
  • 14 min total
  • 14 chapters

What decision this helps you make: Where to get a defensible price for something that does not have one: which venue, which instrument, and which mechanism is actually doing the discovering — and when the honest answer is to build a mechanism rather than to look one up.

What this topic is

Price discovery is the process by which scattered private information — what individual people know, believe, and are willing to bet on — gets compressed into a single public number. It is not the act of finding a price that already existed somewhere. There is no such price. The number is manufactured out of disagreement: someone thinks the asset is cheap and buys, someone thinks it is dear and sells, and the level at which they transact becomes the market's current summary of everything anyone was willing to act on. Every subsequent trade revises it.

Why it matters

Almost every important number in business is downstream of somebody's price discovery. The value of your company, the interest rate on your debt, the price of your input, the exchange rate on your revenue, the level of your index-linked pension — all of them are the output of a mechanism that aggregated beliefs, and every one of them can be broken by using a mechanism that had no informed participants in it. Understanding how information gets into a price tells you which prices to trust, how much of a move was genuine learning, and — most usefully — how to build a mechanism when the price you need does not yet exist.

Who should learn it

Founders and boards pricing a secondary or a tender, anyone who has to defend a valuation, investors trying to separate signal from noise in a price move, operators forecasting demand or cost, and readers who want the working version of the claim that markets aggregate information.

What you will understand

  • How a trade moves a price mechanically, and why an informed trader gives up half their edge to price impact in the standard model
  • Why a market cannot be perfectly efficient, and why that is a structural result rather than a complaint
  • How to separate the permanent component of a price move — the learning — from the temporary component that reverts
  • How to build a price-discovery mechanism when no existing price is trustworthy, and what makes one credible

Prerequisites

Common misconception

"The market price reflects all available information, so a move without news is irrational." Both halves are wrong in an interesting way. Grossman and Stiglitz proved that if prices revealed everything, nobody would pay to gather information, so nobody would — which means prices cannot be fully revealing in equilibrium. And French and Roll showed that price variance during trading hours is far larger than during closed hours of the same length, which no plausible schedule of public news announcements can explain. Prices move because people trade, and people trade on things nobody has published. A move without news is usually the most informative kind there is.