The Impact of SEC Interventions and Information-Complexity on Analyst Forecast Accuracy

Abstract

Regulation G, adopted in March 2003, regulates the reporting of non-GAAP accounting performance measures by US publicly traded firms. In this study, we examine the impact of Regulation G on the understanding of GAAP and non-GAAP financial performance measures by financial analysts. We expect analyst forecast errors to decrease following the adoption of regulation G. Examining the REIT industry setting for which the analysts provide both the forecasts of the non-GAAP, Funds from Operations (FFO), and GAAP, earnings per share (EPS), our results show that analyst forecast accuracy is a function of the complexity of computing a performance measure, and that Regulation G has had an impact on analyst activities. As EPS is computationally more complex than FFO, we find that analyst forecasts for EPS are less accurate than FFO forecasts. The magnitude of the difference is a function of the level of complexity of computing EPS compared to the complexity of computing FFO. We find that the adoption of Regulation G has decreased forecast error differences. We find that the publication of reconciliation of GAAP vs. Non-GAAP performance measures has improved the EPS forecast ability of analysts under Regulation G. This is consistent with the conjecture that the disclosure of the reconciliation items is providing analysts with information about net income components that was not previously available, thereby increasing their ability to forecast EPS. Since the computation of FFO has been normalized by the industry for years, and consistent with the less complex nature of FFO computation and long experience of analysts in forecasting FFO, we do not find an impact of regulation G on FFO forecast.