Abstract
The value-relevance of earnings components is an important issue for accounting regulators because it can help them assess the benefit of mandating disclosures with respect to value-relevant earnings components (see, e.g., Barth et al. [2001]). Hope and Kang (hereafter HK) examine the value-relevance of earnings components—specifically, foreign and domestic earnings—by building on Bodnar and Weintrop (1997; hereafter BW). BW show that while both domestic and foreign earnings are value- relevant, the foreign earnings response coefficient is higher than the domestic earnings response coefficient. They also find that the foreign earnings response coefficient is higher than the domestic earnings response coefficient because of higher growth opportunities in foreign operations relative to domestic operations. The objective of HK is to provide an alternative explanation for BW's finding. Specifically, they argue that BW's result could be a statistical artifact due to the omission of an explanatory variable that is correlated with the foreign earnings in BW's research design. HK make the case for the omission of “other information” that is correlated with foreign earnings in BW's research design. They include a variable for “other information” based on analysts' future earnings forecasts in BW's research design and find that the foreign and domestic earnings response coefficients are similar in magnitude.
My discussion centers, first, on certain limitations of the research design and proxy for other information employed by HK; second, I will then interpret the results in light of these limitations. For highlighting the limitations of the research design and proxies employed, I begin by providing a simple framework of stock valuation that provides insights into the drivers of earnings response coefficients. Finally, I will state my conclusions and suggest future research directions.