Binary Conditional Forecasts

Conditional Forecasts in Large Bayesian VARs with Multiple Soft and Hard Constraints

“…Finally, we conducted an empirical exercise where we estimated a Bayesian VAR featuring 31 quarterly macroeconomic and financial series, and we used our approach to investigate the effect of simultaneously imposing a number of inequality and equality constraints on the trajectories of CPI inflation, the unemployment rate, and the 10-year Treasury rate over the 2020-2022 period. In the future, we believe it would be useful to generalize the current approach to also work with Bayesian VARs with time-varying parameters and stochastic volatility, and it would also be interesting to extend the precision-based sampler to produce conditional forecasts for binary variables, along the lines of McCracken, McGillicuddy, and Owyang (2021).…”

Group selection and shrinkage with application to sparse semiparametric modeling

“…Unlike the products in the supermarket dataset, the identifiers of the macroeconomic series are available, allowing us to inspect the selected series. In the interest of space, we focus on a subset of six series used in McCracken et al (2021) as determinants of recessions. For this subset, the type of function fit by each estimator as applied to all n = 739 observations is reported in minus federal funds rate) and FEDFUNDS (federal funds rate) are treated nonlinearly by every estimator.…”