Do Asset Prices Help Predict Inflation? Evidence from Individual Stock Prices
利用个股价格数据，结合机器学习方法，发现个股价格能显著提升通胀预测的准确性，尤其在中期和长期以及高通胀或通缩时期，比综合指数和行业指数等提供更丰富的信息。
This paper revisits the predictive power of asset prices for inflation, focusing on individual stock prices rather than aggregate indices. Using a large panel data of firm-level stock prices and applying machine learning techniques, we demonstrate that individual stock prices significantly enhance the accuracy of inflation forecasts, particularly over medium- to long-term horizons and during periods of high inflationary and deflationary pressure. Compared to composite and industry-level stock indices, other aggregate asset prices, and Fama-French factors, individual stock prices contain valuable heterogeneous information, offering richer insights for inflation forecasting. These findings provide new empirical support for macro-finance theory, affirming the predictive value of asset prices from a micro-level perspective.