International Financial Contagion: Theory and Evidence in Evolution

Abstract

This Research Foundation monograph is the result of more than three years of empirical research in the area of contagion.It all started for me in a seminar where I was discussing an empirical fact based on a very innocent graph (similar to those presented in Chapter 2).I was showing the conditional covariance between the Mexican and Argentine stock markets, and I was puzzled by the increase in comovement during the major financial crises of the 1990s.I claimed that those changes were empirical support for one of my theoretical papers.The implication of this paper was that during financial turmoil, comovement should increase.Raman Uppal of the London Business School was in the audience, however, and he told me that the evidence I was using to support my contention was wrong and that the "facts" were missing an important statistical regularity.He was right.Since then, I have repeatedly encountered in the literature mistakes similar to the one I made in that seminar.In the past three years, I have devoted part of my research (perhaps as revenge for what Raman did to me!) to highlighting what is wrong with the traditional econometric techniques used to measure contagion and to trying to find new methodologies to solve the problems.I think all of us involved in the search are getting closer.But we still have a long way to go.This monograph is a summary of what we thought we knew, what we know, and what is yet to be known.The Introduction provides an overview of the monograph, so I leave the details of the organization for later.I would like to thank Raman Uppal.Even though he has never worked in the area of contagion, and probably never will, his comments were the ones that motivated me to pursue this research.Without his insights, I am sure this monograph would not exist.My thanks go also to my colleagues Rudi Dorn-

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 33 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.