Expected Change in Liquidity Forecasts Stocks’ Returns

Liquidity Forecasts and Stock Returns

AuthorsClaus Schmitt; Philipp Schuster

Paper linkhttps://papers.ssrn.com/sol3/papers.cfm?abstract_id=4389864

This paper develops a forecasting model for individual stock liquidity using two key measures: the Amihud illiquidity ratio and the effective spread from TAQ data. The model incorporates past liquidity realizations, trading volume, firm size, book-to-market ratio, and stock returns, applying a rolling five-year window for calibration and generating out-of-sample monthly forecasts. Results show the model significantly improves liquidity prediction accuracy over naïve benchmarks, reducing forecast error and identifying predicted changes in illiquidity that have meaningful implications for portfolio construction and asset pricing.

The study addresses the gap in finance literature on liquidity forecasting, recognizing liquidity’s central role in asset pricing and its link to volatility. By adapting methods from volatility prediction and employing elastic net predictor selection, the model captures time-series patterns such as flight-to-liquidity and macroeconomic effects. Findings reveal that while current liquidity levels are not strongly priced, predicted changes in illiquidity carry significant explanatory power for future stock returns, suggesting that investors demand compensation for expected liquidity shifts. This highlights liquidity expectations as a priced risk factor in financial markets.