Machine learning tools are well known for their success in prediction. But prediction is not causation, and causal discovery is at the core of most questions concerning economic policy. Recently, however, the literature has focused more on issues of causality. This paper gently introduces some leading work in this area, using a concrete example—assessing the impact of a hypothetical banking crisis on a country’s growth. By enabling consideration of a rich set of potential nonlinearities, and by allowing individually-tailored policy assessments, machine learning can provide an invaluable complement to the skill set of economists within the Fund and beyond.
IMF Research Perspective (formerly published as IMF Research Bulletin) is a new, redesigned online newsletter covering updates on IMF research. In the inaugural issue of the newsletter, Hites Ahir interviews Valeria Cerra; and they discuss the economic environment 10 years after the global financial crisis. Research Summaries cover the rise of populism; economic reform; labor and technology; big data; and the relationship between happiness and productivity. Sweta C. Saxena was the guest editor for this inaugural issue.
In December 2008, the IMF Executive Board discussed the Seventh Review of Data Standards Initiatives, and Directors requested staff to return to the Board within about a year with a proposal for the inclusion of selected financial indicators in the Special Data Dissemination Standard (SDDS). This paper responds to the 2008 request taking into account recent developments. The recent financial crisis has heightened the need for policymakers, financial regulators and capital market participants to put in place conditions that would help prevent the occurrence of similar crises in the future. One of the areas identified by the international community as key in crises prevention is the availability of timely and more detailed financial data that could provide early warning signals of impending risks and vulnerabilities
Mr. Paul Louis Ceriel Hilbers, Mr. Alfredo Mario Leone, Mr. Mahinder Singh Gill, and Mr. Owen Evens
Following the severe financial crises of the 1990s, identifying and assessing financial sector vulnerabilities has become a key priority of the international community. The costly disruptions in global markets underscored the need to establish a set of monitorable variables for evaluating strengths and weaknesses in financial institutions and to alert authorities of impending problems. These variables, indicators, of financial system health and stability known collectively as macroprudential indicators, are the subject of this Occasional Paper by the Monetary and Exchange Affairs Department and the Statistics Department. Macroprudential indicators take measures at both the level of aggregated financial institutions and at the macroeconomic level; financial crises often occur when weaknesses are identified in both. The authors provide a breakdown and explanations of these indicators and a review of the theoretical and empirical work done thus far. Work at other international and multilateral institutions is included as well as the experiences of several national central banks and supervisory agencies. This paper provides a valuable reference source of current knowledge about macroprudential indicators and issues related to their analysis, identification, measurement, and possible dissemination.