Noise traders incarnate: Describing a realistic noise trading process

We estimate a realistic process for noise trading to help theorists derive predictions from noisy rational expectations models. We characterize the trades of individual investors, who are natural candidates for the role of noise traders because their trades are weakly correlated with fundamentals, in line with how such models define noise trading. Data from a retail brokerage house, small and price-improved trades in TAQ, and flows to retail mutual funds yield consistent estimates. The properties of noise trading are highly sensitive to the frequency considered, with the common assumption of i.i.d.-normal noise appropriate only at monthly and lower frequencies. •We use retail trades to estimate a realistic process for noise trading.•Data from a retail broker, small and price-improved trades in TAQ, and flows to retail mutual funds yield consistent results.•The properties of noise trading are highly sensitive to the trading frequency considered.•The common assumption of i.i.d.-normal noise appropriate only at monthly and lower frequencies.