Research

Publications

On Testing for Bubbles During Hyperinflations

Studies in Nonlinear Dynamics & Econometrics

With: Zacharias Psaradakis, Martin Sola, and Patricio Yunis

Abstract: We consider testing for the presence of rational bubbles during hyperinflations via an analysis of the non-stationarity properties of relevant observable time series. The test procedure is based on a Markov regime-switching model with independent stochastic changes in its intercept, error variance and autoregressive coefficients. This model formulation allow us to disentangle fundamentals-driven changes in the drift, bubble-driven explosiveness, and volatility changes that may be fundamentals-driven and/or bubble-driven. The testing methodology is illustrated by applying it to data from hyperinflations in Argentina, Brazil, Germany and Poland. 

Current Research

European Sovereign Bond and Stock Market Granger Causality Dynamics  - Preparing

With: Pedro Gomes and Zeynep Ozde Kurter

Abstract: We investigate the lead-lag relationship using weekly sovereign (government) bond yield changes and stock returns for seven European countries and how it changed during the period 2008-2018. Such lead-lag analysis are evaluated through vector autoregressions (VARs) on weekly data and with a method, based on Markov-switching Granger causality, that determines the direction of the Granger causality endogenously. We find that there are significant changes in the direction of the Granger causality between stock returns and changes in sovereign bond yields that vary across the countries. Additionally, global and specific economic events are linked to changes in Granger causality. Our results indicate that stock returns take a leading role in changes of sovereign bond yields in Germany, France, Spain, Italy and the United Kingdom, while changes of sovereign bond yields significantly lead stock returns in Portugal and Greece, where these countries faced very high instability during the European sovereign debt crisis period. No significant lead-lag relationship is found between stock and sovereign bond markets in some sub-periods for all countries.

Granger Causality Regimes and Volatility Dynamics: A Unified Markov-Switching Framework 

Draft Soon

Abstract: I propose a Markov-switching vector autoregressive (MS-VAR) framework for analysing time-varying Granger causality in the presence of conditional heteroskedasticity. Existing approaches in the tradition of Psaradakis et al. (2005) embed both mean dynamics and covariance structure within the same regime indicators, creating a fundamental identification problem: volatility clusters and genuine causality regime changes are observationally similar from the perspective of the Hamilton filter, so the former can be spuriously attributed to the latter. I resolve this by separating the two channels, allowing regime switching only in the conditional mean while modelling second-moment dynamics through a state-invariant GARCH-DCC structure. I formalise the misspecification failure as a Kullback-Leibler proximity inversion and prove, under standard regularity conditions, that the fixed-covariance filter systematically misattributes GARCH volatility episodes to predictive-regime switches with positive stationary probability. Monte Carlo evidence with T = 300 confirms that the fixed-covariance estimator produces substantially worse regime classification across all four predictive states, with mean classification error roughly twice as large as the proposed estimator; differences are highly significant across all regimes. An empirical application to the trade-weighted U.S. dollar index and the Trade Policy Uncertainty (TPU) index of Baker et al. (2016), using monthly data from January 1985 to April 2026, illustrates the practical importance of the approach: ARCH-LM tests strongly reject homoskedasticity in the VAR residuals, and the MS-VAR with GARCH-DCC uncovers four distinct predictive regimes whose timing maps cleanly onto the major episodes of U.S. trade policy history. TPU-driven dollar dynamics dominate during both Trump administration trade-war episodes of 2017-2020 and 2025-2026, Dollar-to-TPU causality characterises the Plaza Accord era and the China shock of the early 2000s, and a persistent no-causality regime prevails during intervals of broadly stable multilateral trade rules. 

Markov-Switching Local Projections with Time-Varying Transition Probabilities: Oil Prices and Political Polarisation in the United States

Draft Soon

Abstract: This paper develops a Markov-Switching Local Projection (MS-LP) framework in which the dynamic response of political polarization to oil price shocks depends on a latent regime, with regime changes determined endogenously by a first-order Markov process rather than an a priori break date. Estimation combines EM initialization with direct maximum likelihood, multi-seed exploration of the likelihood surface, and parametric bootstrap inference. Applied to monthly U.S. data from 1986 to 2025, the model identifies a single, sharply dated structural break in 2008 to 2009, with near-unit regime persistence on both sides. Before the break, oil shocks have no detectable effect on the Partisan Conflict Index; in the post-2008 regime, a one standard deviation oil shock raises partisan conflict significantly, peaking at about 6.6 log points within one month. A time-varying transition probability extension, introduced as a diagnostic, yields suggestive but statistically insignificant evidence (p = 0.072) that financial stress is associated with exit from the pre-crisis regime, a test that is underpowered by construction because the sample contains a single transition. The findings indicate that the oil-to-polarization channel is a recent phenomenon: consistent with the transformation of the U.S. oil sector during the shale revolution, oil prices appear to have shifted from operating as an aggregate macroeconomic shock toward functioning as a partisan-distributional issue. 

Testing Time-Varying Granger Causality: A Markov-Switching VAR with TVTP and GARCH–DCC   - Draft Soon

Working In Progress

• The Liquidity Effect Shifts.

•  DSGE models and predictability in the real economy, with Stephen Wright and Rory Macqueen.