By exploring global stock return patterns, the authors examine whether the earnings announcement premium that has been found for U.S. stocks holds true globally. In 16 of the 20 years in their sample period and in 9 of the 20 countries with enough observations to perform country analysis, a significantly positive premium exists during earnings announcement months compared with nonannouncement months. The size of the premium is greatest in countries with the greatest increase in idiosyncratic volatility around the time of the earnings announcement, suggesting that uncertainty over the information to be disclosed is what drives the premium.

In their sample of approximately 200,000 earnings announcements from 46 countries over the period of 1991–2010, the authors find that a portfolio of stocks expected to announce earnings within the month has an average incremental monthly return of 59.7 bps, 7.16% annualized, over stocks without announcements. This result compares favorably with such pricing anomalies as size, book-to-market ratio, and momentum and could offer opportunities for practitioners after they take transaction costs into consideration.

Evidence of an earnings announcement premium for non-U.S. stocks could lead to a better understanding of what drives this premium. By making use of cross-country variations in the magnitude of the premium, the authors are able to explore which factors seem to be most clearly driving its existence.

The authors’ results demonstrate that the premium is greatest in countries with the greatest increase in idiosyncratic volatility around the time of the earnings announcements, suggesting that uncertainty over the information to be disclosed is the main factor that leads investors to demand higher preannouncement returns, which creates the earnings announcement premium. Conversely, the authors do not find evidence that the premium is related to increased levels of systematic risk around the time of earnings announcements or increased levels of investor attention with regard to the announcements. They do find that the earnings announcement premium seems to be more pronounced for smaller stocks.

To explore returns in months around times when earnings announcements are expected and months outside the expected earnings announcement window, the authors use a regression framework, controlling for size, momentum, and book-to-market ratio. Their primary source of data is the Bloomberg database, supplemented by I/B/E/S in cases in which Bloomberg does not provide earnings announcement data for a particular firm and fiscal year. They eliminate firms for which a fiscal year-end date is not given and those for which expected earnings announcement dates are more than 150 days after the fiscal year-end date. They also remove from the sample those firms for which Datastream does not provide return data, the exchange listing code, and the listing currency, as well as firms concurrently trading American Depositary Receipts in the United States.

The final sample consists of around 200,000 annual earnings announcements issued by 28,772 firms over the period of 1991–2010 and covers 46 countries. Among those countries, 20 had enough data to allow for country analysis.

The authors’ analysis is interesting from a theoretical standpoint because they search for a pattern observed in the United States in other countries around the world and confirm the presence of a similar earnings announcement premium in other countries. The connection between increased idiosyncratic volatility and an earnings announcement premium is also interesting and seems worthy of further exploration. Aside from the obvious impact of currency exchange costs and volatility, the biggest caveat for practitioners aiming to exploit the presence of an earnings announcement premium is the substantial trading costs that seem to be involved in executing the strategy, particularly on a global basis.