Intraday Stock Returns from Noise Reversals

Can investors reliably capture illiquidity-driven stock price noise, short-term deviations in price from some measurable fair value? In their February 2024 paper entitled “Intraday Residual Reversal in the U.S. Stock Market”, Jonathan Brogaard, Jaehee Han and Hanjun Kim investigate returns to a strategy that exploits reversals of short-lived noise in stock prices by buying (selling) stocks with positive (negative) price noise. Specifically, at intervals of 30 minutes, they:

• Regress stock returns cross-sectionally versus 15 standardized/normalized stock return anomalies to predict next-interval return for each stock, with the difference between predicted and actual returns designated as noise (residual). The first daily noise measurement is at 10:00AM and the last (for overnight) at 4:00PM.
• Sort stocks into tenths (deciles) based on noise from most negative to most positive.
• Reform a hedge portfolio, either value-weighted or equal-weighted, that is long (short) the stocks in the bottom (top) noise decile.

They limit their stock universe to the S&P 500 for depth and liquidity, so the hedge portfolio has positions in about 100 stocks. They consider impacts of trading frictions ranging from 0.03% to 0.07%. Using daily returns for each of the 15 anomalies and 30-minute bid-ask midpoints for S&P 500 index stocks during July 1996 through December 2022, they find that:

(more…)