|
on Forecasting |
| By: | Andrea Bastianin; Marzio Galeotti; Matteo Manera |
| Abstract: | Call centers' managers are interested in obtaining accurate point and distributional forecasts of call arrivals in order to achieve an optimal balance between service quality and operating costs. We present a strategy for selecting forecast models of call arrivals which is based on three pillars: (i) flexibility of the loss function; (ii) statistical evaluation of forecast accuracy; (iii) economic evaluation of forecast performance using money metrics. We implement fourteen time series models and seven forecast combination schemes on three series of daily call arrivals. Although we focus mainly on point forecasts, we also analyze density forecast evaluation. We show that second moments modeling is important both for point and density forecasting and that the simple Seasonal Random Walk model is always outperformed by more general specifications. Our results suggest that call center managers should invest in the use of forecast models which describe both first and second moments of call arrivals. |
| Date: | 2018–04 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1804.08315&r=for |
| By: | Jörg Döpke (Hochschule Merseburg (University of Applied Sciences Merseburg)); Ulrich Fritsche (Universität Hamburg (University of Hamburg)); Karsten Müller (Hochschule Merseburg (University of Applied Sciences Merseburg)) |
| Abstract: | Based on a panel of annual data for 17 growth and inflation forecasts from 14 institutions for Germany, we analyse forecast accuracy for the periods before and after the Great Recession, including measures of directional change accuracy based on Receiver Operating Curves (ROC). We find only small differences on forecast accuracy between both time periods. We test whether the conditions for forecast rationality hold in both time periods. We document an increased cross-sectional variance of forecasts and a changed correlation between inflation and growth forecast errors after the crisis, which might hint to changed forecaster behaviour. This is also supported by estimated loss functions before and after the crisis, which suggest a stronger incentive to avoid overestimations (growth) and underestimations (inflation) after the crisis. Estimating loss functions for a 10-year rolling window also reveals shifts in the level and direction of loss asymmetry and strengthens the impression of a changed forecaster behaviour after the Great Recession. |
| Keywords: | Macroeconomic Forecasting, Forecast Error Evaluation, Germany |
| JEL: | E32 E37 G11 |
| Date: | 2018–05 |
| URL: | http://d.repec.org/n?u=RePEc:hep:macppr:201803&r=for |
| By: | Gary Koop (Department of Economics, University of Strathclyde, UK; Rimini Centre for Economic Analysis); Dimitris Korobilis (Essex Business School, University of Essex, UK; Rimini Centre for Economic Analysis) |
| Abstract: | This paper develops methods for estimating and forecasting in Bayesian panel vector autoregressions of large dimensions with time-varying parameters and stochastic volatility. We exploit a hierarchical prior that takes into account possible pooling restrictions involving both VAR coefficients and the error covariance matrix, and propose a Bayesian dynamic learning procedure that controls for various sources of model uncertainty. We tackle computational concerns by means of a simulation-free algorithm that relies on an analytical approximation of the posterior distribution. We use our methods to forecast inflation rates in the eurozone and show that forecasts from our flexible specification are superior to alternative methods for large vector autoregressions. |
| Keywords: | Panel VAR, inflation forecasting, Bayesian, time-varying parameter model |
| Date: | 2018–05 |
| URL: | http://d.repec.org/n?u=RePEc:rim:rimwps:18-20&r=for |
| By: | Dimitris Korobilis (Essex Business School, University of Essex, UK; Rimini Centre for Economic Analysis); Davide Pettenuzzo (Sachar International Center, Brandeis University, USA) |
| Abstract: | This paper proposes a simulation-free estimation algorithm for vector autoregressions (VARs) that allows fast approximate calculation of marginal parameter posterior distributions. We apply the algorithm to derive analytical expressions for independent VAR priors that admit a hierarchical representation and which would typically require computationally intensive posterior simulation methods. The benefits of the new algorithm are explored using three quantitative exercises. First, a Monte Carlo experiment illustrates the accuracy and computational gains of the proposed estimation algorithm and priors. Second, a forecasting exercise involving VARs estimated on macroeconomic data demonstrates the ability of hierarchical shrinkage priors to find useful parsimonious representations. We also show how our approach can be used for structural analysis and that it can successfully replicate important features of news-driven business cycles predicted by a large-scale theoretical model. |
| Keywords: | Bayesian VARs, Mixture prior, Large datasets, Macroeconomic forecasting |
| JEL: | C11 C13 C32 C53 |
| Date: | 2018–05 |
| URL: | http://d.repec.org/n?u=RePEc:rim:rimwps:18-21&r=for |
| By: | Demary, Markus |
| Abstract: | In the IW Financial Expert Survey for the second quarter of 2018 the surveyed experts predict, on average, a steeper yield curve, i.e. a larger increase in the long-term than in the short-term interest rate. Moreover, the average forecasts indicate higher stock market indices, a depreciation of the Euro vis-à-vis the US Dollar, and lower oil prices by the end of the third quarter of 2018. However, despite the expectation of higher interest rates, the short-term interest rate is predicted to remain in negative territory. The 3-month Euribor is, on average, expected to reach -0.31 percent at the end of the third quarter of 2018, while the yield on German government bonds with 10-year maturity is expected to reach 0.81 percent by then. However, the experts do not expect the European Central Bank (ECB) to change the forward guidance of its monetary policy significantly. Stock markets are, on average, expected to increase by 9.2 percent (Stoxx 50) and 8.4 percent (DAX 30) until the end of the third quarter of 2018. During that same period, the experts predict the Euro to depreciate by 2.5 percent vis-à-vis the US Dollar, while oil prices are expected to fall by 7.8 percent. The expectation of an increase in the long rate and a slight increase in the short rate, together with the expected delayed monetary tightening of the ECB, hints at a financial market outlook characterised by a cautious approach to monetary policy normalisation. In this cautious approach, the ECB lets the market determine the first increases in long-term interest rates before it stops intervening at the long end of the yield curve, while keeping the short end of the yield curve lower. This cautious approach to monetary policy normalization is reflected in the projections of the yield curve. Moreover, the experts expect that the development of the Euro and the development of oil prices as well as the development of the stock market will support the ECB's cautious approach to monetary normalization instead of forcing a faster exit from low interest rates. The experts do not expect the ECB to change its forwards guidance in the forthcoming Governing Council meeting. The evaluation of the forecasting performance of the latest forecasts yields the result that Commerzbank and DZ Bank performed best in predicting trends within the long-term ranking, which covers all forecasts from March 2015 to March 2018. DekaBank and Deutsche Bank performed best in the short-term ranking, which covers the surveys for the third and the fourth quarter of 2017 for the 3-months ahead prediction and the survey for the third quarter of 2017 for the 6-month forecasts. When it comes to point prediction, in the long-term evaluation of the period running from March 2015 to March 2018, the experts of National-Bank performed best in predicting all indicators, while the Postbank experts produced the most precise point forecasts for all indicators for the short-term evaluation period. |
| JEL: | G12 G17 |
| Date: | 2018 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:iwkrep:132018&r=for |
| By: | Jackson, Emerson Abraham |
| Abstract: | This empirical study has provided interpretive outcome from a univariate forecast using Box-Jenkins ARIMA methodology. The HCPI_SA seasonally adjusted data for Sierra Leone shows a robust model outcome with three months ahead prediction based on the STATIC method result. Test results like Autocorrelation and also comparative values for MAPE and the Inverted Root values have indicated that the model is a good fit. Despite better choice of outcome from the STATIC result in comparison to DYNAMIC forecast, the conclusion a cautious means of advice when using results for policy outcomes and with comparative forecasts highly recommended a way forward in guiding policy makers’ decision. |
| Keywords: | ARIMA, Forecast, Headline Consumer Price Index [HCPI], Sierra Leone |
| JEL: | C52 C53 |
| Date: | 2018–01–11 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:86180&r=for |
| By: | Gary Koop (Department of Economics, University of Strathclyde, UK; The Rimini Centre for Economic Analysis); Dimitris Korobilis (Essex Business School, University of Essex, UK; The Rimini Centre for Economic Analysis); Davide Pettenuzzo (Sachar International Center, Brandeis University, USA) |
| Abstract: | Macroeconomists are increasingly working with large Vector Autoregressions (VARs) where the number of parameters vastly exceeds the number of observations. Existing approaches either involve prior shrinkage or the use of factor methods. In this paper, we develop an alternative based on ideas from the compressed regression literature. It involves randomly compressing the explanatory variables prior to analysis. A huge dimensional problem is thus turned into a much smaller, more computationally tractable one. Bayesian model averaging can be done over various compressions, attaching greater weight to compressions which forecast well. In a macroeconomic application involving up to 129 variables, we find compressed VAR methods to forecast as well or better than either factor methods or large VAR methods involving prior shrinkage. |
| Keywords: | multivariate time series, random projection, forecasting |
| JEL: | C11 C32 C53 |
| Date: | 2017–12 |
| URL: | http://d.repec.org/n?u=RePEc:rim:rimwps:17-32&r=for |