• Accountants and analysts as financial intermediaries Volume I, Part I • Measuring corporate earnings and profitability Volume I, Part II • Financial ratios, the risk of failure, and stock returns Volume I, Part III • Equity valuation Volume I, Part IV • Price-earnings ratios, market-to-book ratios, and stock returns Volume II, Part I • Earnings and stock returns Volume II, Part II • Fundamental analysis and stock returns Volume II, Part III

We use confidential, U.S. Census Bureau, plant-level data to investigate aggregation in external reporting. We compare firms’ plant-level data to their published segment reports by grouping a firm’s plants that share the same four-digit SIC code into a “pseudo-segment.” We then determine whether each pseudo-segment is disclosed as an external segment, or whether it is subsumed into a different business unit for external reporting purposes. We show that a pseudo-segment is more likely to be aggregated when the agency and proprietary costs of separately reporting the pseudo-segment are higher and when firm and pseudo-segment characteristics allow for more discretion in the application of segment reporting rules. For firms reporting multiple external segments, aggregation of pseudo-segments is driven by both agency and proprietary costs. For firms reporting a single external segment, we find no evidence of an agency cost motive for aggregation.

We use a simple k-nearest neighbors algorithm (hereafter, k-NN*) to forecast earnings. k-NN* forecasts of one-, two-, and three-year-ahead earnings are more accurate than those generated by popular extant forecasting approaches. k-NN* forecasts of two- and three-year (one-year)-ahead EPS and aggregate three-year EPS are more (less) accurate than those generated by analysts. The association between the unexpected earnings implied by k-NN* and the contemporaneous market-adjusted return (i.e., the earnings association coefficient (EAC)) is positive and exceeds the EAC on unexpected earnings implied by alternate approaches. A trading strategy that is long (short) firms for which k-NN* predicts positive (negative) earnings growth earns positive risk-adjusted returns that exceed those earned by similar trading strategies that are based on alternate forecasts. The k-NN* algorithm generates an empirically reliable ex ante indicator of forecast accuracy that identifies situations when the k-NN* EAC is larger and the k-NN* trading strategy is more profitable. Data Availability: Data are available from the public sources described in the text.