First, since Ball and Brown (1968), it is evident that external financial reports, while in some senses
‘‘value relevant”, are not especially timely sources of information. This is hardly surprising, as stock
prices change in response to new information, but periodic financial statements provide a summary
of past transactions. Hence, studies that examine the responsiveness of stock prices around the release
of periodic financial statements typically display a low level of explanatory power (Kothari, 2001). In
other words, it cannot be claimed that periodic financial statements cause stock price changes, except
to a very small extent. Investors likely rely on the myriad of other information sources, including analysts
reports, media coverage and ad-hoc disclosures made by the firms themselves.
Indeed, it seems to us that the role of Ball and Brown (1968) in demonstrating what is commonly
interpreted as the value relevance of financial reports (particularly earnings) has resulted in an overstatement
of how important earnings information is as a source of timely (i.e., new) information. For
example, Francis et al. (2002) document what they characterize as an increasing usefulness of quarterly
earnings announcements for a large sample of US firm-quarters over the period 1980–1999.
While their characterization of usefulness is measured as the extent of market reaction per unit of unexpected earnings, it is worth noting that only a very small proportion of the abnormal stock return
at the time of an earnings announcement can be explained by a measure of ‘‘earnings news”
More recently, explicit recognition of this point is provided by Ball and Shivakumar (2008). They
capture the relative importance of earnings announcements in providing new information to the share
market via the r-squared from a regression of calendar year stock returns on their four quarterly
announcement date stock returns. Using a large sample of US firm-quarters (years) from 1972–
2006, Ball and Shivakumar show that around 9% of the total annual information is explained by the
four quarterly earnings releases. Hence, on a quarterly basis, around 1–2% of annual information is explained
by each quarterly release