Abnormal Overnight Earnings Return Factor in China
Log in to collectAcademic paper
Earnings Announcements in China: Overnight-Intraday Disparity
Junhao Liu; Ole‐Kristian Hope; Danqi Hu
- University of Melbourne
- CAUniversity of Toronto
- ?University of Melbourne - Faculty of Business and Economics
- ?University of Toronto - Rotman School of Management
- Peking University
- CAKellogg's (Canada)
- ?Guanghua School of Management, Peking University
- ?Northwestern University - Kellogg School of Management
- ?Peking University - Guanghua School of Management
Strategy in a nutshell
We study Chinese A-share stocks (Shanghai and Shenzhen), excluding firms with market cap <100m yuan and stock price <1 yuan. Using CSMAR data, we form hedge portfolios based on abnormal overnight returns on earnings announcement days (AOR_EA). Overnight return = (Openi,t−Closei,t−1)/Closei,t−1(Open_{i,t} - Close_{i,t-1}) / Close_{i,t-1}(Openi,t−Closei,t−1)/Closei,t−1. Each month, firms are ranked into quintiles (Q1–Q5) using the latest earnings announcement overnight return. We go long Q5, short Q1. Stocks are value-weighted; portfolios rebalance monthly.
Economic rationale
Chinese equity markets, dominated by retail investors, are less efficient than developed peers. Evidence (Liu, Hope, Hu 2023) shows strong overnight reactions to earnings news, largely driven by auction mechanisms that aggregate dispersed information and generate value-relevant price signals. While overnight returns lack a continuous-trading counterfactual, results suggest auction-driven signals offer exploitable predictability in a less efficient setting.