UC Berkeley

How can accounting be useful to investors and facilitate equity valuation? In the dissertation at hand, I provide three essays that add to the literature about earnings quality and financial analysis.

The first chapter provides a framework for the three studies that comprise chapter 2 to 4. I discuss the normative principles that underlie the valuation of an equity stake in a company as advocated by Penman (2010) and as taught in most MBA courses. I proceed to derive a proposition that useful accounting minimizes investors' forecast errors of abnormal earnings over the life of the firm. This demand forms the basis on which I will judge and test the usefulness of several accounting methods for valuation throughout the dissertation.

The second chapter is titled "The Matching Principle: Timely Information to Investors?" In it, I empirically test whether and how the matching principle (matching expenses to revenues) provides investors with more timely information. I hypothesize and find empirical evidence that matching provides investors with an accounting margin and rate of return that is a more informative and less noisy anchor for forecasting a firm's future business performance. Changes in accounting margins based

on adequately matched expenses are a timelier signal of changes in underlying business performance.

In the third chapter, "Does Matching Expenses to Revenues Increase the Usefulness of Fundamental Signals?", I examine whether fundamental accounting signals are more insightful the more expenses are matched to revenues. This chapter adds to the literature by comprehensively examining how matching of expenses to revenues improves or weakens the predictive ability of financial signals for future earnings. Second, I investigate whether market participants react differently to financials signals from firms with a high degree of matched expenses compared to signals fromfirms with a low degree of matched expenses. The results indicate that the degree of matched expenses significantly and positively affects the predictive power of accounting based signals (signals that are based on accounting mechanics such as abnormal inventory growth). On the other hand, the predictive ability of "non-accounting" signals such as abnormal capex growth diminishes. Furthermore, the relation between matching and the predictive power of the signals is driven at least as much

by accounting choices as the underlying business factors. At the same time, there is only weak evidence that market participants such as analysts reacts differently to fundamental signals for high-matching firms. Analyst forecast revisions only show a weakly increasing relation between signals and revisions. In contrast, there is a robust increasing relation between forecast errors and signals.

The fourth chapter is titled "Do Analysts Understand the Relation Between Investment Intensity and Earnings Growth?" The aim of this chapter is to test whether the accounting for investments is understood by valuation experts such as analysts. I identify a transitory pattern in earnings growth that is caused by changes in a firm's investment activity and whose magnitude depends on the degree of conservative accounting used for these investments. Since, the pattern is mechanical and can be predicted, I test whether analysts do so. I provide evidence that they do not fully anticipate transitory changes in earnings growth and that these cases result in abnormal returns around subsequent earnings announcements.