Projects and impact
The cluster is developing and maintaining a ‘data warehouse’, containing a variety of micro data sets on financial and operational information of firms and workers, scanner data of products, high frequency stock market data, etc.. In addition, the cluster develops and applies machine learning methods and algorithms to explore, describe, visualize and synthesize complex relationships commonly observed in a variety of problems in the various subject fields in the school.
A number of colleagues have been applying such ‘big data’ to a number of problems with high relevance and impact for academic research, economic policy and business practices. To illustrate the variety of the type of work this cluster has been engaged in, three applications (amongst many) are briefly described below.
A first application deals with the question of why large movements in exchange rates have small effects on international goods prices? This empirical regularity of low exchange rate pass-through is a central puzzle in international macroeconomics. Professor Jozef Konings together with Professor Oleg Itskhoki of Princeton University and Dr. Mary Amiti of the New York Fed have been working with granular data on exports and imports to explore this question. They show that taking into account that large exporters are also large importers can help explain differences in exchange rate pass-through across firms as well as low aggregate pass-through. These results have important implications for international macroeconomics as price sensitivity to exchange rates is central for the expenditure-switching mechanism at the core of international adjustment and re-balancing. Understanding why there is incomplete pass-through is important from a welfare perspective as the implications differ if incomplete pass-through is due to different distributions of markups across firms or to complex global sourcing patterns, which directly affect marginal costs.
Another example is the work launched by Dr. Tena, who received a grant from the Spanish football Federation in association with the UEFA. He uses big data on football results and tournaments to investigate how the performance of football clubs in European and domestic tournaments are interrelated. This research on managerial decisions in football and tournament design has an important impact on how to organise tournaments and how to manage efficiently competition in sports. Not surprisingly, this work has attracted a lot of interest from the media covered in the Guardian, BBC and various national TVs, to name a few.
A third application is the work of Dr. Ruijun Bu and his collaborators in which they use cutting-edge econometric tools, such as nonparametric filtering, continuous-time regression with time change, and mixed-frequency sampling, for modelling the dynamics of key economic indicators. These include the dynamics of interest rates, oil and equity prices and volatilities, liquidity and uncertainty indices. The methodological and empirical contributions of their work has a large impact on the practice of economic forecasting, relevant to financial institutions to set out a strategy on financial investment and risk management.