Stats › Individual stocks › Earnings drift
US earnings drift stopped paying outside microcaps in 2006
| Study | Sample | Finding |
|---|---|---|
| Chordia, Goyal, Sadka, Sadka and Shivakumar, Financial Analysts Journal 2009 | 1972 to 2005 | The long short strategy returned 0.04% a month value weighted in the most liquid stocks and 2.43% a month in the most illiquid. Transaction costs account for 70% to 100% of the paper profits |
| Ng, Rusticus and Verdi, Journal of Accounting Research 2008 | US listed | Higher transaction cost firms show weaker immediate reaction and larger drift. PEAD strategy profits are significantly reduced by transaction costs |
| Chordia, Subrahmanyam and Tong, Journal of Accounting and Economics 2014 | 1976 to 2011 | A one standard deviation increase in SUE was worth 0.14% of monthly return in 1976 to 1993 and 0.01% in 1994 to 2011. Returns to a basket of prominent anomalies roughly halved after decimalisation in 2001 |
| Hou, Xue and Zhang, Review of Financial Studies 2020 | Jan 1967 to Dec 2016 | With NYSE breakpoints and value weighted returns, the six month earnings surprise premium is 0.16% a month with a t statistic of 1.44. It does not clear the bar. 65% of the 452 anomalies they tested do not |
| Martineau, Critical Finance Review 2022 | 1 Jan 1984 to 31 Dec 2019 | Analyst surprises stop predicting day +2 to +60 returns from 2006 for everything but microcaps, and from 2016 for microcaps too. Meanwhile the announcement day reaction grew from about 20 basis points in 1984 to 1990 to about 120 basis points in 2016 to 2019 |
| Kettell, McInnis and Zhao, working paper 2022 | 1974 to 2020 | Hedge PEAD falls by about 2 basis points a quarter, t of -3.06. Roughly 5% in the 1980s and 1990s, roughly 4% in the 2000s, 3% or lower in the late 2010s, indistinguishable from zero after 2017. Controlling for falling earnings persistence removes the trend |
| Subrahmanyam, Journal of Investment and Management 2026 | Feb 2001 to Dec 2024 | Reconciles the recent disagreement. All stocks: t of 2.18. Excluding microcaps: t of 1.43. His conclusion is that earnings drift does not exist in anything but microcaps |
Sources listed in full at the bottom. Note that these studies do not all measure the same thing, so the rows are not directly comparable to each other. They are comparable in direction, and the direction is the same in every one.
After a US company beats or misses expectations, its share price used to keep drifting the same way. The drift was worth about 2% each way over 60 trading days on 1974 to 1986 data. It has not been significant outside microcaps since 2006.
Members only
The rest of this page is for members
Below this point there are 6 sections, 1 chart, 1 table and 8 named sources, roughly 1100 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.