Using a demographics-based proxy for smartness, we show that the portfolio distortions of “smart” investors reflect an informational advantage, while the distortions of “dumb” investors reflect psychological biases. Specifically, smart investors outperform dumb investors by about 3% annually on a risk-adjusted basis. Furthermore, among investors with high portfolio distortions, smart investors outperform passive benchmarks by 2%, and the smart-dumb performance differential is 5%. At the stock level, a portfolio of stocks with smart investor clientele outperforms the dumb clientele portfolio by 3.50% annually. These findings suggest that behavioral and information-based explanations for portfolio distortions apply to distinct subsets of investors.