Working Paper: CEPR ID: DP18032
Authors: Kim Christensen; Allan Timmermann; Bezirgen Veliyev
Abstract: Corporate earnings announcements unpack large bundles of information that should, if markets are efficient, almost surely trigger jumps in stock prices immediately after the news release. Testing this implication is difficult in practice because most earnings announcements occur in the after-hours market where prices are contaminated by high levels of microstructure noise. We develop a new noise-robust jump test statistic and demonstrate that stock prices almost always jump immediately after earnings announcements. Finally, we develop a trading-based approach that allows us to estimate exactly how long it takes for markets to incorporate earnings news and quantify the importance of transaction costs.
Keywords: afterhours trading; earnings announcements; jump testing; high-frequency data; market efficiency; price discovery
JEL Codes: C10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Earnings announcements (G14) | Stock price jumps (G19) |
| Earnings surprises (G14) | Stock price jumps (G19) |
| Number of analysts covering the stock (G24) | Stock price jumps (G19) |
| Earnings announcements (G14) | Price adjustments (L11) |
| Microstructure noise (C58) | Price adjustments (L11) |
| Speed of price discovery improvements (G13) | Stock price jumps (G19) |