Publications & Working Papers

We detect cyclical variation in the predictive information of economic fundamentals, which can be used to substantially improve and simplify out-of-sample equity premium prediction. Economic fundamentals based on stock-specific information (notably the dividend yield) deliver better predictions in expansions. Economic fundamentals based on aggregate information (notably the short rate) deliver better predictions in recessions. Accordingly, a simple forecast combination of one predictor that generates cyclical forecasts and one predictor that generates countercyclical forecasts can deliver statistically significant and economically valuable equity premium predictions in both expansions and recessions. A prominent two-predictor forecast combination that performs well is the dividend yield and the short rate. Strategies designed for ex-ante timing of the business cycle can provide additional economic gains in equity premium prediction.


This paper investigates whether business cycles cause financial cycles or vice versa. We also assess whether the US plays a leading role in causing the domestic business and financial cycles of other countries. The literature has established that business and financial cycles are linked through several channels such as credit constraints, the real effects of financial information and the reversal of overoptimistic expectations. Our analysis evaluates the direction of Granger causality using a novel approach based on the mixed-frequency vector autoregression model for the G7 countries. Our approach exploits the fact that real economic activity measured by industrial production is observed at a higher frequency than aggregate credit. We find strong evidence of bidirectional causality between the business and financial cycles, especially in recessions. Furthermore, the US is a global leader since the US business cycle significantly affects other countries’ business cycles, especially in terms of expansions.


We examine the short-run and long-run dynamics of the correlation between exchange rate and commodity returns, and assess the extent to which the long-run correlation is determined by economic fundamentals. Our empirical analysis is based on the dynamic conditional correlation model with mixed data sampling (DCC-MIDAS) of Colacito, Engle and Ghysels (2011). This model provides a framework that captures the high-frequency relation between exchange rate and commodity returns as well as the low-frequency relation of volatility and correlation to economic fundamentals. Using both economic and statistical criteria, we find that the DCC-MIDAS model augmented with economic fundamentals performs better than competing models in sample and out of sample.


Standard vector autoregression (VAR) models suffer from the curse of dimensionality. This is because VAR estimation requires an unrealistically large sample size when more than two lagged variables are involved in the model. This paper resolves the estimation issues associated with high dimension by estimating three alternative VAR models: single index additive vector autoregression (SIAVAR), nonlinear additive vector autoregression (AVAR), and linear vector autoregression (VAR). The three models are used to produce impulse response functions and one-step-ahead forecasts. We perform simulation experiments to compare the models in various settings. The methodology is also illustrated on economic and financial time series variables.