Get access



  • Note: The views expressed in this paper are those of the authors and do not necessarily represent the views of Danmarks Nationalbank. The authors are grateful to Niels Haldrup (CREATES, Aarhus University) for inspiration and practical advice regarding the choice of model specification. They thank Piet Philip Christiansen for collecting a vast amount of financial data for the purpose of the analysis. The paper also benefits from constructive suggestions from participants at the joint Eurostat/ECB FDI Workshop in November 2009. Finally, the authors highly appreciate the helpful comments received on earlier drafts of the paper from Allan Sall Tang Andersen, Bent Christiansen, Robert Evans, Niels Lynggård Hansen, Ulrich Kaiser, Nikolaj Malchow-Møller, Niki Saabye, and two anonymous referees. Any remaining errors are the responsibility of the authors alone.

Jannick Damgaard, Danmarks Nationalbank, Havnegade 5, DK-1093 Copenhagen, Denmark (


Late and significant revisions are often observed in FDI equity income in many countries, hampering the quality of preliminary balance of payments statistics. We test a range of models on Danish data and find that forecasts for FDI equity income based on a combination of past profitability and consensus data for changes in expected private consumption growth outperform forecasts solely based on historical profitability. When the refined models are applied to the Danish balance of payments, the largest improvements are observed for outward and inward FDI separately. Revisions of net FDI equity income only decrease marginally because the significant revisions in gross terms resulting from the historical models have a tendency to (partly) cancel out each other on a net basis.