From Fixed-Event to Fixed-Horizon Density Forecasts: Obtaining Measures of Multi-Horizon Uncertainty From Survey Density Forecasts

Abstract

Surveys of professional forecasters produce precise and timely point forecasts for key macroeconomic variables. However, the accompanying density forecasts are not as widely utilized, and there is no consensus about their quality. This is partly because such surveys are often conducted for "fixed events". For example, in each quarter panelists are asked to forecast output growth and inflation for the current calendar year and the next, implying that the forecast horizon changes with each survey round. The fixed-event nature limits the usefulness of survey density predictions for policymakers and market participants, who often wish to characterize uncertainty a fixed number of periods ahead ("fixed-horizon"). Is it possible to obtain fixed-horizon density forecasts using the available fixed-event ones? We propose a density combination approach that weights fixed-event density forecasts according to a uniformity of the probability integral transform criterion, aiming at obtaining a correctly calibrated fixed-horizon density forecast. Using data from the US Survey of Professional Forecasters, we show that our combination method produces competitive density forecasts relative to widely used alternatives based on historical forecast errors or Bayesian VARs. Thus, our proposed fixed-horizon predictive densities are a new and useful tool for researchers and policy makers.

Modeling and Forecasting Macroeconomic Downside Risk

“…Our framework nests their specification and provides a richer dynamic, able to capture smoother and sharper variations of the scale and shape parameters. In particular, we introduce persistence in the skewness of the distribution of GDP growth, in line with the term structure of the growth-at-risk displaying stronger asymmetry for the short-than for the mediumrun (Adrian et al, 2018), and consistent with the pronounced skewness displayed by the Survey of Professional Forecasters' short-term predictions (Ganics et al, 2020). Finally, unlike alternative approaches, we put forward a two-component specification for the time-varying parameters of the model, in the spirit of Engle and Lee (1999), to track both secular and cyclical changes in the underlying distribution of GDP growth.…”

Forecasting Macroeconomic Risk in Real Time: Great and Covid-19 Recessions

“…Ganics et al (2020) confirm that the more flexible skewed t-distribution appears to fit better during the Great Recession.ECB Working Paper Series No 2436 / July 2020…”

Commodity prices and inflation risk

“…Including this more up-to-date information does not qualitatively change the results presented in this section.28 The probabilistic questions they ask when constructing histograms, are for headline inflation only and are with respect to current and next year's annual inflation for each quarterly survey in the year, and as such are fixed event-based forecasts. Hence, using these would involve converting from fixed event to fixed horizon densities, a complication we opt to avoid in this instance Ganics et al (2020). suggest methods to address this issue.…”