Disagreement about the past: An empirical assessment of bank analysts' GAAP and non‐GAAP earnings measures

Abstract

Prior research documents that analysts' non-Generally Accepted Accounting Principles (non-GAAP) or "street earnings" exclude transitory items in order to facilitate security valuation. We study sell-side analysts' reports for large European banks and find significant variation in analysts' non-GAAP actual earnings measures. These measures are not always easily reconcilable to firms' reported non-GAAP earnings, to GAAP earnings or to non-GAAP earnings reported by Institutional Brokers Estimate System (I/B/E/S). By contrast, reported measures of GAAP earnings in analysts' reports differ from one another (or from firms' reported GAAP earnings) far less often. When evaluated against analysts' own actual non-GAAP earnings measures, forecasts appear more accurate than those based on I/B/E/S actuals. Results for GAAP earnings are less conclusive. Our results indicate that as well as disagreeing about future earnings, analysts also disagree significantly about what earnings were in the past.

The Field of Investment Advice: The Social Forces That Govern Equity Analysts

“…EPS). Consensus numbers represent an aggregate of considered individual analyst opinions, but also allow a variety of individual reactions to the numbers (Clatworthy, Ho and Zhu 2022) and enable a comparison between the consensus and individual analysts' forecasts (Michaelli, Rubin, Segal and Vedrashko 2023). However, as they combine in order to focus minds on one representative number, consensus numbers also become a somewhat homogenizing, normalizing force that market players are forced to pay close attention to (Macintosh, Shearer, Thornton and Welker 2000).…”