The author survey the statistical evidence on average stock return and economic theories that try to explain it. The stastical evidence suggests a period of low returns, followed by a slow reversion to a high long- term average. However, that evidence is quite uncertain. Standard economics predicts tiny stock returns. The author surveys new economic models that predict high returns, but by fundamentally changing the description of stock market risk. He warns that a low forecast for stock returns does not mean one should sell.