Climate Risks of Sales Forecasts: Evidence from Satellite Readings of Soil Moisture∗

Christopher Hansman,Harrison G. Hong,F. Li,Jiangmin Xu

2019-01-01

Abstract:We show that recently-available satellite readings of soil moisture, used to study the impact of climate change on hydrological cycles, predict the output of the water-intensive food industry. Using sales and forecasts data from thirty-one countries with publicly-traded food companies from 1994-2014, we obtain causal estimates for soil moisture by using as instruments climatic measures of droughts based on temperature and precipitation. Soil moisture is particularly important for food industry sales in the last decade of historically high temperatures. Consensus sales forecasts by security analysts, widely used to guide corporate investments and adaptations, have not recognized this structural change. Yet, we find that mitigating innovations measured using patents have become more important in explaining output. ∗We thank Park Williams of Lamont-Doherty Earth Observatory at Columbia University and seminar participants at CICF, New York University, Nankai University and Imperial College of London for helpful comments. This paper subsumes earlier work on "Climate Change and Efficiency of Sales Forecasts". †Imperial College Business School ‡Columbia University and NBER §Singapore Management University, Lee Kong Chian School of Business ¶Guanghua School of Management, Peking University, China