I have now arrived in Oslo for the 23rd Symposium of the Society for Nonlinear Dynamics & Econometrics. This is a conference that I like very much for it's always a nice mix of theory and empirical work from people in academia and Central banks the world over. The BI Norwegian Business School has an impressive architecture.
The paper I present shows an interesting source of dependence over long period. We show (with my two coauthors Alain Hecq from Maastricht University and Sébastien Laurent from Aix-Marseille University and GREQAM) that types of persistence that are commonly observed (e.g. the so called Golden Rule of realized volatilities) can actually arise from dependence across a very large number of variables (such as traded assets in financial markets).
Beyond the financial application, we believe that this new theory has a number of interesting consequences in the context of Big Data since it is the first time cross-section dependence can have such dramatic effect for the time-series dependence of each variable. It has a wide applicability to river basins, sectorial inflation, asset markets...
During the conference, Jim Stock from Harvard University gave a very insightful account of his tenure at the US Council of Economic Advisors for President Obama. He mentioned in particular the peculiar time when he was in charge of advising the president about the state of the economy during the October 2013 Governement shutdown brought about by the Republicans in Congress. By Federal Law Jim Stock didn't have the right to receive any help from anyone and had to reconstruct many datasets by himself about employment in particular (and the impact of the shutdown on job destruction).