Financial Speculators' Underperformance : Learning, Self-Selection, and Provision of Liquidity

Why do so many speculators enter financial markets when most lose money and cease trading after a short period of time? In contrast to existing behavioral explanations, we offer an equilibrium model of learning by rational agents that explains these and related phenomena. Agents do not know their speculative skills, but trading profits can provide them information. Even though it is common knowledge that most traders are unskilled, some agents enter financial markets. They initially experiment by trading on a small scale, and use the information uncovered by their trading profits to decide whether to continue. Competitive market makers post price quotes, recognizing the entry process of speculators, so that the rewards from speculation are endogenous. The model reconciles several empirical regularities: (1) In a given cross-section, most individual speculators lose money. (2) Large (more active) speculators out-perform small (less active) speculators. (3) Performance positively affects subsequent trade intensity. Most new traders lose money and quickly cease speculation. (4) Performance shows persistence. We also show how learning from trade produces endogenous liquidity in financial markets. This endogenous liquidity reduces bid-ask spreads and reduces the effect of exogenous liquidity shocks on asset prices, but amplifies the effect of real shocks in the economy (e.g., labor market and learning costs) on prices. Finally, if some traders are slightly overconfident about their speculative abilities, we show that this increases bid-ask spreads, thereby hurting all traders