nep-for New Economics Papers
on Forecasting
Issue of 2014‒11‒28
seven papers chosen by
Rob J Hyndman
Monash University

  1. Forecasting the Brent oil price: addressing time-variation in forecast performance By Manescu, Cristiana; Van Robays, Ine
  2. Dynamic Prediction Pools: An Investigation of Financial Frictions and Forecasting Performance By Marco Del Negro; Raiden B. Hasegawa; Frank Schorfheide
  3. Balance Sheets of Financial Intermediaries: Do They Forecast Economic Activity? By Rodrigo Sekkel
  4. Relative Performance of Semi-Parametric Nonlinear Models in Forecasting Basis By Onel, Gulcan; Karali, Berna
  5. Conditional forecasts and scenario analysis with vector autoregressions for large cross-sections By Bańbura, Marta; Giannone, Domenico; Lenza, Michele
  6. Analysing and forecasting price dynamics across euro area countries and sectors: a panel VAR approach By Dées, Stéphane; Güntner, Jochen
  7. Determination of long-run and short-run dynamics in EC-VARMA models via canonical correlations By George Athanasopoulos; D.S. Poskitt; Farshid Vahid; Wenying Yao

  1. By: Manescu, Cristiana; Van Robays, Ine
    Abstract: This paper demonstrates how the real-time forecasting accuracy of different Brent oil price forecast models changes over time. We find considerable instability in the performance of all models evaluated and argue that relying on average forecasting statistics might hide important information on a model`s forecasting properties. To address this instability, we propose a forecast combination approach to predict quarterly real Brent oil prices. A four-model combination (consisting of futures, risk-adjusted futures, a Bayesian VAR and a DGSE model of the oil market) predicts Brent oil prices more accurately than the futures and the random walk up to 11 quarters ahead, on average, and generates a forecast whose performance is remarkably robust over time. In addition, the model combination reduces the forecast bias and predicts the direction of the oil price changes more accurately than both benchmarks. JEL Classification: Q43, C43, E32
    Keywords: Brent oil prices, central banks, forecast combination, real-time, time-variation
    Date: 2014–09
  2. By: Marco Del Negro; Raiden B. Hasegawa; Frank Schorfheide
    Abstract: We provide a novel methodology for estimating time-varying weights in linear prediction pools, which we call Dynamic Pools, and use it to investigate the relative forecasting performance of DSGE models with and without financial frictions for output growth and inflation from 1992 to 2011. We find strong evidence of time variation in the pool's weights, reflecting the fact that the DSGE model with financial frictions produces superior forecasts in periods of financial distress but does not perform as well in tranquil periods. The dynamic pool's weights react in a timely fashion to changes in the environment, leading to real-time forecast improvements relative to other methods of density forecast combination, such as Bayesian Model Averaging, optimal (static) pools, and equal weights. We show how a policymaker dealing with model uncertainty could have used a dynamic pools to perform a counterfactual exercise (responding to the gap in labor market conditions) in the immediate aftermath of the Lehman crisis.
    JEL: C53 E31 E32 E37
    Date: 2014–10
  3. By: Rodrigo Sekkel
    Abstract: This paper conducts a real-time, out-of-sample analysis of the forecasting power of various aggregate financial intermediaries’ balance sheets to a wide range of economic activity measures in the United States. I find evidence that the balance sheets of leveraged financial institutions do have out-of-sample predictive power for future economic activity, and this predictability arises mainly through the housing sector. Nevertheless, I show that these variables have very little predictive power during periods of economic expansions and that predictability arises mainly during the financial crisis period.
    Keywords: Econometric and statistical methods
    JEL: C53
    Date: 2014
  4. By: Onel, Gulcan; Karali, Berna
    Abstract: Many risk management strategies, including hedging the price risk using forward or futures contracts require accurate forecasts of basis, i.e., spot price minus the futures price. Recent literature in this area has applied nonlinear time-series models, which are refinements of the linear autoregressive models that allow the parameters to transition from one regime to another. These parametric nonlinear models, however, involve complex estimation problems, and may diminish forecasting accuracy, especially in longer horizons. We propose using a semi-parametric, generalized additive model (GAM) that may improve the forecasting performance with its simplicity and flexibility while still accounting for nonlinearities in local prices and basis. Empirical results based on weekly futures and spot prices for North Carolina soybean and corn markets support evidence of nonlinear effects in basis. In general, generalized additive models seem to yield better forecasts of basis.
    Keywords: basis, futures markets, forecasting, generalized additive models, nonlinear models, Agricultural Finance, Demand and Price Analysis,
    Date: 2014
  5. By: Bańbura, Marta; Giannone, Domenico; Lenza, Michele
    Abstract: This paper describes an algorithm to compute the distribution of conditional forecasts, i.e. projections of a set of variables of interest on future paths of some other variables, in dynamic systems. The algorithm is based on Kalman filtering methods and is computationally viable for large models that can be cast in a linear state space representation. We build large vector autoregressions (VARs) and a large dynamic factor model (DFM) for a quarterly data set of 26 euro area macroeconomic and financial indicators. Both approaches deliver similar forecasts and scenario assessments. In addition, conditional forecasts shed light on the stability of the dynamic relationships in the euro area during the recent episodes of financial turmoil and indicate that only a small number of sources drive the bulk of the fluctuations in the euro area economy. JEL Classification: C11, C13, C33, C53
    Keywords: Bayesian shrinkage, conditional forecast, dynamic factor model, large cross-sections, vector autoregression
    Date: 2014–09
  6. By: Dées, Stéphane; Güntner, Jochen
    Abstract: This paper uses a panel VAR (PVAR) approach to estimating, analysing, and forecasting price dynamics in four different sectors – industry, services, construction, and agriculture – across the four largest euro area economies – Germany, France, Italy and Spain – and the euro area as a whole. By modelling prices together with real activity, employment and wages, we can disentangle the role of unit labour costs and profit margins as the factors affecting price pressures on the supply side. In out-of-sample forecast exercises, the PVAR model fares comparatively well against common alternatives, although short-horizon forecast errors tend to be large when we consider only the period of the recent financial crisis. The second part of the paper focuses on Spain, for which prediction errors during the crisis are particularly large. Given that its economy faced dramatic sectoral changes due to the burst of a housing bubble, we use the PVAR model for studying the transmission of shocks originating from the Spanish construction sector to other sectors. In a multi-country extension of the model, we also allow for spillovers to the other euro area countries in our sample. JEL Classification: C33, C53, E31, E37
    Keywords: cost pressures, forecasting, impulse response analysis, panel VAR models
    Date: 2014–08
  7. By: George Athanasopoulos; D.S. Poskitt; Farshid Vahid; Wenying Yao
    Abstract: This article studies a simple, coherent approach for identifying and estimating error correcting vector autoregressive moving average (EC-VARMA) models. Canonical correlation analysis is implemented for both determining the cointegrating rank, using a strongly consistent method, and identifying the short-run VARMA dynamics, using the scalar component methodology. Finite sample performances are evaluated via Monte-Carlo simulations and the approach is applied to model and forecast US interest rates. The results reveal that EC-VARMA models generate significantly more accurate out-of-sample forecasts than vector error correction models (VECMs), especially for short horizons.
    Keywords: Cointegration, Error correction, Scalar Component Model, Multivariate Time Series.
    Date: 2014

This nep-for issue is ©2014 by Rob J Hyndman. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.