What Do Analyst Financial Modelling Mistakes Tell Us? Evidence from the Properties of Accounting

This study develops a novel measure of sell‐side analyst financial modelling mistakes, based on whether an analyst's simultaneous forecasts of earnings, book value of equity, and dividends are internally consistent with the accounting clean surplus relation. I find that the discrepancy of an analyst's forecasts from clean surplus (a proxy for financial modelling mistakes) is positively associated with the analyst's contemporaneous forecast errors and price target errors. This is consistent with financial modelling playing an important role in analysts’ forecasting and valuation process. More experienced analysts, analysts who work for larger brokerage firms, and analysts who are less busy produce forecasts that are less discrepant from clean surplus. This is consistent with analysts learning to avoid modelling mistakes over time and making fewer mistakes when they have better access to resources and sufficient time to prepare a careful analysis. The market reaction to earnings forecast revisions is significantly weaker when the contemporaneous analyst forecasts are discrepant from clean surplus, consistent with the market viewing these forecasts as lower quality. In sum, the results reveal the importance of financial modelling to analyst information processing.