A Measure of Financial Statement Similarity

Abstract

We propose a measure of financial statement peer-to-peer similarity. The measure captures multiple dimensions of financial statements, considers both the presence and amount of a financial statement item, and therefore provides a holistic view to users who often need to compare a firm with its peers. The measure could be useful for screening firms for further scrutiny. One of such settings is regulatory enforcement of financial reporting. We find that firms with lower financial statement peer-to-peer similarity are more likely to receive comment letters from the SEC. This is even more the case after the SEC launched RoboCop in 2013. Overall, our study highlights the multidimensionality of financial statements and introduces a measure to capture such multidimensionality.