Keynesian Beauty Contest

In The General Theory (1936), John Maynard Keynes described a newspaper contest: readers pick the six prettiest faces from a hundred photos. The prize goes to whoever matches the most popular choice. So you're not choosing who you find prettiest — you're choosing who you think others will find prettiest. And since others are doing the same, you're really guessing who you think others think others will pick. Keynes wrote that some players "practise the fourth, fifth and higher degrees" of reasoning. In markets, he argued, professional investors don't pick assets by fundamental value; they try to anticipate what the crowd will value. Prices become a bet on collective psychology, not a simple reflection of cash flows. The contest is a metaphor for second- and third-order thinking: what do I think? What do I think they think? What do I think they think I think?

The strategic implication is that in many settings — elections, fashion, asset prices, hiring — the "correct" move depends on what others will do. If everyone is one step ahead, the edge goes to those who think one step further — until the process becomes unstable or everyone converges on a focal point. The model explains momentum, bubbles, and the persistence of conventions that have no fundamental anchor. It also warns: when the game is "guess the average," the average can drift far from intrinsic value. The practitioner's question is always: how many steps of "what do they think?" matter here, and where do I sit in that chain?

Experimental "guess 2/3 of the average" games confirm that depth varies: many people pick a number that reflects one or two steps of reasoning; few go to the Nash equilibrium of zero. The winning guess is often in the 20–30 range — the average of a mix of naive and sophisticated players. In markets, the equivalent is that price can sit at a level that is neither fundamental value nor pure chaos but an unstable compromise between contest-players and fundamentalists. When the share of contest-players rises, price drifts from value; when they exit or get wiped out, price reverts. The model doesn't say markets are always wrong; it says the mapping from fundamentals to price runs through expectations of others, and that mapping can be noisy and regime-dependent.