Forecasting papers 2011-06-03

In this issue we have Forecasting recessions using stall speeds, Bayesian VARs: specification choices and forecast accuracy, WALS estimation and forecasting in factor-based dynamic models, Forecasting Value-at-Risk Using Nonlinear Regression Quantiles and the Intra-day Range, and more.


  1. Forecasting recessions using stall speeds
    Date: 2011
    By: Jeremy J. Nalewaik
    This paper presents evidence that the economic stall speed concept has some empirical content, and can be moderately useful in forecasting recessions. Specifically, output tends to transition to a slow-growth phase at the end of expansions before falling into a recession, and the paper designs Markov-switching models that behave in that way. While the switching models using output growth alone produce a considerable number of false positive recession signals, adding the slope of the yield curve, the percent change in housing starts, and the change in the unemployment rate to the model reduces false positives and improves recession forecasting. The switching model is particularly good at forecasting at long horizons, outperforming Blue Chip consensus forecasts.
  2. Bayesian VARs: specification choices and forecast accuracy
    Date: 2011
    By: Andrea Carriero
    Todd Clark
    Massimiliano Marcellino
    In this paper we examine how the forecasting performance of Bayesian VARs is affected by a number of specification choices. In the baseline case, we use a Normal-Inverted Wishart prior that, when combined with a (pseudo-) iterated approach, makes the analytical computation of multi-step forecasts feasible and simple, in particular when using standard and fixed values for the tightness and the lag length. We then assess the role of the optimal choice of the tightness, of the lag length and of both; compare alternative approaches to multi-step forecasting (direct, iterated, and pseudo-iterated); discuss the treatment of the error variance and of cross-variable shrinkage; and address a set of additional issues, including the size of the VAR, modeling in levels or growth rates, and the extent of forecast bias induced by shrinkage. We obtain a large set of empirical results, but we can summarize them by saying that we find very small losses (and sometimes even gains) from the adoption of specification choices that make BVAR modeling quick and easy. This finding could therefore further enhance the diffusion of the BVAR as an econometric tool for a vast range of applications.
    Keywords: Bayesian statistical decision theory ; Forecasting ; Vector autoregression
  3. WALS estimation and forecasting in factor-based dynamic models with an application to Armenia
    Date: 2011
    By: Poghosyan, K.
    Magnus, J.R. (Tilburg University, Center for Economic Research)
    Two model averaging approaches are used and compared in estimating and forecasting dynamic factor models, the well-known BMA and the recently developed WALS. Both methods propose to combine frequentist estimators using Bayesian weights. We apply our framework to the Armenian economy using quarterly data from 2000–2010, and we estimate and forecast real GDP and inflation dynamics.
    Keywords: Dynamic models;Factor analysis;Model averaging;Monte Carlo;Armenia.
    JEL: C11
  4. Forecasting Value-at-Risk Using Nonlinear Regression Quantiles and the Intra-day Range
    Date: 2011-05-18
    By: Cathy W. S. Chen
    Richard Gerlach
    Bruce B. K. Hwang
    Michael McAleer (University of Canterbury)
    Value-at-Risk (VaR) is commonly used for financial risk measurement. It has recently become even more important, especially during the 2008-09 global financial crisis. We propose some novel nonlinear threshold conditional autoregressive VaR (CAViaR) models that incorporate intra-day price ranges. Model estimation and inference are performed using the Bayesian approach via the link with the Skewed-Laplace distribution. We examine how a range of risk models perform during the 2008-09 financial crisis, and evaluate how the crisis affects the performance of risk models via forecasting VaR. Empirical analysis is conducted on five Asia-Pacific Economic Cooperation stock market indices as well as two exchange rate series. We examine violation rates, back-testing criteria, market risk charges and quantile loss function values to measure and assess the forecasting performance of a variety of risk models. The proposed threshold CAViaR model, incorporating range information, is shown to forecast VaR more efficiently than other models, across the series considered, which should be useful for financial practitioners.
    Keywords: Value-at-Risk; CAViaR model; Skewed-Laplace distribution; intra-day range; backtesting; Markov chain Monte Carlo
  5. Uncertainty and Disagreement in Forecasting Inflation: Evidence from the Laboratory
    Date: 2011
    By: Pfajfar, D.
    Zakelj, B. (Tilburg University, Center for Economic Research)
    We establish several stylized facts about the behavior of individual uncertainty and disagreement between individuals when forecasting inflation in the laboratory. Subjects correctly perceive the underlying inflation uncertainty in only 60% of cases, which can be interpreted as the overconfidence bias. Determinants of individual uncertainty, dis- agreement among forecasters and properties of aggregate distribution are analyzed. We find that the interquartile range of the aggregate distribution has the highest correlation with inflation variability; however the average confidence interval performs best in a forecasting exercise. Allowing subjects to insert asymmetric confidence intervals results in wider upper intervals than lower intervals on average, thus perceiving higher uncertainty with respect to inflation increases. In different treatments we study the influence of different monetary policy designs on the formation of confidence bounds. Inflation targeting produces lower uncertainty and higher accuracy of intervals than inflation forecast targeting.
    Keywords: Laboratory Experiments;Confidence Bounds;New Keynesian Model;Inflation Expectations
    JEL: C91
  6. FOMC communication policy and the accuracy of Fed Funds futures
    Date: 2011
    By: Menno Middeldorp
    Over the last two decades, the Federal Open Market Committee (FOMC), the rate-setting body of the United States Federal Reserve System, has become increasingly communicative and transparent. According to policymakers, one of the goals of this shift has been to improve monetary policy predictability. Previous academic research has found that the FOMC has indeed become more predictable. Here, I contribute to the literature in two ways. First, instead of simply looking at predictability before and after the Fed’s communication reforms in the 1990s, I identify three distinct periods of reform and measure their separate contributions. Second, I correct the interest rate forecasts embedded in fed funds futures contracts for risk premiums, in order to obtain a less biased measure of predictability. My results suggest that the communication reforms of the early 1990s and the “guidance” provided from 2003 significantly improved predictability, while the release of the FOMC’s policy bias in 1999 had no measurable impact. Finally, I find that FOMC speeches and testimonies significantly lower short-term forecasting errors.
    Keywords: Federal Open Market Committee ; Disclosure of information ; Interest rates ; Forecasting ; Monetary policy ; Federal funds
  7. Forecasting Corporate Distress in the Asian and Pacific Region
    Date: 2011-05
    By: Russ Moro
    Wolfgang Härdle
    Saeideh Aliakbari
    Linda Hoffmann
    This study analyses credit default risk for firms in the Asian and Pacific region by applying two methodologies: a Support Vector Machine (SVM) and a logistic regression (Logit). Among different financial ratios suggested as predictors of default, leverage ratios and the company size display a higher discriminating power compared to others. An analysis of the dependencies between PD and financial ratios is provided along with a comparison with Europe (Germany). With respect to forecasting accuracy the SVM has a lower model risk than the Logit on average and displays a more robust performance. This result holds true across different years.
    Keywords: Credit risk, Bankruptcy, Asian companies, SVM
    JEL: C14

Taken from the NEP-FOR mailing list edited by Rob Hyndman.