Forecasting Macroeconomic Risks

Abstract

We construct risks around consensus forecasts of real GDP growth, unemployment, and inflation. We find that risks are time-varying, asymmetric, and partly predictable. Tight financial conditions forecast downside growth risk, upside unemployment risk, and increased uncertainty around the inflation forecast. Growth vulnerability arises as the conditional mean and conditional variance of GDP growth are negatively correlated: downside risks are driven by lower mean and higher variance when financial conditions tighten. Similarly, employment vulnerability arises as the conditional mean and conditional variance of unemployment are positively correlated, with tighter financial conditions corresponding to higher forecasted unemployment and higher variance around the consensus forecast.

Capturing Macroeconomic Tail Risks with Bayesian Vector Autoregressions

“…Still other work (e.g., Loria, Matthes, and Zhang (2019)) has extended the analysis of Adrian, Boyarchenko, and Giannone (2019a) -henceforth, ABG -to better understand tail risks. 2 Earlier work of Manzan (2015) used quantile regression to assess the value of a large number of macroeconomic indicators in forecasting the complete distribution of some key variables. 3 The interest in tail risks reflects an underlying perception or assumption of asymmetries in distributions of outcomes.…”

“…Although some of this evidence of skewness in GDP growth has not established formal statistical significance, Jensen, et al (2020) develop statistically significant evidence of increasingly negative business cycle asymmetry over 1 For evidence on the euro area, see Figueres and Jarocinski (2020). 2 Adams, et al (2020) apply the ABG methodology to forecasts from the Survey of Professional Forecasters. To assess the joint distribution of economic and financial conditions, Adrian, Boyarchenko, and Giannone (2019b) develop a new approach that combines non-parametric and Monte Carlo methods.…”

“…The 5 percent quantile drops significantly, from -1.65 in period 1 to -5.29 in 4 Jensen, et al (2020) also develop a dynamic stochastic general equilibrium model consistent with their empirical findings. Adrian, et al (2020) extend an econometric formulation of a DSGE model to include endogenous risk. Salgado, Guvenen, and Bloom (2020) present evidence that skewness in firm-level panel data on growth of employment, sales, and productivity is pro-cyclical and develop a heterogenous agents with some of the same features.…”

Measuring Systemic Financial Stress and its Risks for Growth

“…In particular, systemic FSIs support macroprudential policy in its crisis management function by allowing policymakers to monitor realised systemic risk in more or less real time (Freixas et al (2015)). Moreover, in typical growth-at-risk regression frameworks (Adrian et al (2019), Adams et al (2020), Boyarchenko et al (2023), the CISS is found to predict well short-term tail risks to real GDP and other macroeconomic variables of interest. This strong short-term predictive power of the CISS has proved helpful in counterfactual policy analysis, which can serve the preventive arm of macroprudential policy (Chavleishvili et al (2021)).…”

“…The results from both approaches corroborate previous studies that find particularly strong effects of financial stress on future economic activity in the lower tails of the growth distribution. While Adams et al (2020) and Figueres and Jarociński (2020) also use our new CISS for the US and the euro area, respectively, in single-equation quantile growth-at-risk regressions, this paper is the first to use it in the QVAR framework. Furthermore, we quantify the macroeconomic relevance of systemic stress by performing conditional growth projections based on historical shocks in the CISS.…”