Uncertainty and Forecasts of U.S. Recessions

Abstract

We estimate Boosted Regression Trees (BRT) on a sample of monthly data that extends back to 1889 to recover the predictive value of disaggregated news-based uncertainty indexes for U.S recessions. We control for widely-studied standard predictors and use out-of-sample metrics to assess forecast performance. We find that war-related uncertainty is among the top five predictors of recessions at three different forecast horizons (3, 6, and 12 months). The predictive value of war-related uncertainty has fallen in the second half of the 20th century. Uncertainty regarding the state of securities markets has gained in relative importance. The probability of a recession is a nonlinear function of war-related and securities-markets uncertainty. Receiver-operating-characteristic curves show that uncertainty improves out-of-sample forecast performance at the longer forecast horizons. A dynamic version of the BRT approach sheds light on the importance of various lags of government-related uncertainty for recession forecasting at the long forecast horizon.

The effectiveness of monetary and fiscal policy shocks on U.S. inequality: the role of uncertainty

“…Similarly, on the consumption side both consumption expenditure and total expenditure inequality are considered. 1 For uncertainty, we use the news-based US economic policy uncertainty data, 2 as developed by Baker et al, (2016), though there are of course other measures of uncertainty available, but those do not necessarily relate to policymaking (see, Pierdzioch and Gupta (2017) for a detailed discussion in terms of the alternative measures of uncertainty, which in any case do tend to be very highly correlated irrespective of the approach undertaken to construct the latent variable of uncertainty). To construct the index, Baker et al, (2016) perform month-by-month searches of several newspapers 3 , for terms related to economic and policy uncertainty.…”

Forecasting economic policy uncertainty of BRIC countries using Bayesian VARs

“…But, Baker et al (2016) has solved this problem by constructing normalized indexes of the volume of newspaper articles discussing economic policy uncertainty (EPU) for a large number of developed and developing economies. Using these news-based measures of uncertainty, empirical validation of the theoretical prediction that heightened uncertainty leads to recessions for both advanced and emerging countries can be found in the recent works of Karnizova and Li (2014), Balcilar et al (2016), Kurasawa (2017), Junttila and Vataja (2018), Aye et al (2019a,b), Pierdzioch and Gupta (2017). Despite the well-established importance of uncertainty in macroeconomic developments, both theoretically and now empirically, there is no systematic effort to forecast uncertainty that will allow policymakers to act upon such forecasts while making their decisions in terms of designing appropriate policies ahead of time to deal with future business cycle downturns.…”

Gold, platinum and the predictability of bond risk premia

“…At the same time, the yields on short and long-term Treasuries are shown to capture valuable information regarding the current and future states of the economy and inflation (e.g. Hamilton and Kim, 2002;Dewachter et al, 2014;Gogas et al, 2015a,b;Plakandaras et al, 2017a,b, forthcoming;Pierdzioch and Gupta, 2019).…”