Revisiting Accounting Entropy: An Internal Summary Measure of Accounting Numbers Classification
Abstract
Numerical accounting measures such as the standard classified balance sheet provide multi-faceted information about a reporting entity. We decompose the numerical accounting measures into two summary statistics which are designed to capture two facets of the underlying information about the entity. One facet is about the overall scale of the entity and the other about the distribution of various different activities of a given scale. While a summary statistics for scale can be empirically natural such as reported Total Assets, a summary statistics for the distribution of various different activities is not as easy. In this paper, we derive theoretically, and empirically validate and apply, a new such measure based on the entropy concept from information theory of Shannon in the 1940s and pioneered by Theil in the 1960s. We empirically calculate
entropy-based summary statistics using Balance Sheet information for a large sample of US publicly traded firms. In our main empirical analysis, we first validate our new measure by demonstrating its association with many existing empirical proxies that indirectly measure information content about the traded firms. Second, we use the setting of financial analyst attention allocation to demonstrate that our measure is useful when studying behavior which may respond differently to scale and classification/composition.
For enquiries, please contact Ms. Heidi Lam at heidilam@cuhk.edu.hk.