Research projects

Project CEMAPRE internal

TitleSystemic Tail Risk: Dynamic Factors, Clustering, Breaks, and Forecasting.
ParticipantsJoão Nicolau (Principal Investigator)
SummaryModeling the tail of a distribution (the extreme left and right outcomes) is crucial in finance and
macroeconomics because a large share of real-world damage (and sometimes profit) comes from rare but
very large moves. Models that fit only the “middle” of the distribution (means, variances,
near-normal behavior) can still be seriously misleading precisely about the events that determine
resilience and survival.
My coauthors and I have studied several aspects of tail estimation. In this project, we aim to
extend this research agenda by advancing work on (i) clustering of extreme values, (ii) a tail index
dynamic factor model, (iii) tail index estimation for discrete data, (iv) forecasting banks’
risk-sentiment tail behavior, and (v) structural breaks in conditional tail risk.