|
on Forecasting |
| By: | Andrejs Bessonovs (Bank of Latvia) |
| Abstract: | We develop and assess a suite of statistical models for forecasting Latvia's GDP. Various univariate and multivariate econometric techniques are employed to obtain short-term GDP projections and to assess the performance of the models. We also compile information contained in the GDP components and obtain short-term GDP projections from a disaggregate perspective. We propose a novel approach assessing GDP from the production side in real time, which is subject to changes in NACE classification. Forecast accuracy of all individual statistical models is assessed recursively by out-of-sample forecasting procedure. We conclude that factor-based forecasts tend to dominate in the suite. Encouraging results are also obtained using disaggregate models of factor and bridge models, which could be considered as good alternatives to aggregate ones. Furthermore, combinations of the forecasts of the statistical models allow obtaining robust and accurate forecasts which lead to a reduction of forecast errors. |
| Keywords: | out-of-sample forecasting, real-time estimation, forecast combination, disaggregate approach |
| JEL: | C32 C51 C53 |
| Date: | 2015–07–03 |
| URL: | http://d.repec.org/n?u=RePEc:ltv:wpaper:201501&r=for |
| By: | Fabio Busetti (Bank of Italy); Michele Caivano (Bank of Italy); Lisa Rodano (Bank of Italy) |
| Abstract: | The paper uses dynamic quantile regressions to estimate and forecast the conditional distribution of euro-area inflation. As in a Phillips curve relationship we assume that inflation quantiles depend on past inflation, the output gap, and other determinants, namely oil prices and the exchange rate. We find significant time variation in the shape of the distribution. Overall, the quantile regression approach describes the distribution of inflation better than a benchmark univariate trend-cycle model with stochastic volatility, which is known to perform very well in forecasting inflation. In an out-of-sample prediction exercise, the quantile regression approach provides forecasts of the conditional distribution of inflation that are superior, overall, to those produced by the benchmark model. Averaging the distribution forecasts of the different models improves robustness and in some cases results in the greatest accuracy of distributional forecasts. |
| Keywords: | quantile regression, Phillips curve, time-varying distribution |
| JEL: | C32 E31 E37 |
| Date: | 2015–07 |
| URL: | http://d.repec.org/n?u=RePEc:bdi:wptemi:td_1027_15&r=for |
| By: | Guney, Selin |
| Abstract: | The specific purpose of this paper is to investigate the potential of a time series analysis technique, namely the Time Varying Parameter Vector Autoregressive Model (TVPVAR) technique, in the development of daily forecasting models for cattle prices in the presence of structural changes. More specific objectives are to integrate smoothing techniques and stochastic volatility into TVPAR modeling framework based exclusively on time series for cash-cattle prices, and to compare the accuracy and evaluate the forecasting performance of this model with the standard VAR model based on forecast accuracy measures. |
| Keywords: | Forecasting, Cattle Prices, Structural Changes, TVPVAR, Agribusiness, Agricultural and Food Policy, Demand and Price Analysis, Farm Management, Livestock Production/Industries, E37, Q11, Q13, Q18, |
| Date: | 2014 |
| URL: | http://d.repec.org/n?u=RePEc:ags:aaea15:205109&r=for |
| By: | Buck, Steven; Soldati, Hilary; Sunding, David L. |
| Abstract: | Urban water managers rely heavily on forecasts of water consumption to determine management decisions and investment choices. Typical forecasts rely on simple models whose criteria for selection has little to do with their performance in predicting out-of-sample consumption levels. We demonstrate this issue by comparing forecast models selected on the basis of their ability to perform well in-sample versus out-of-sample. Our results highlight the benefits of developing out-of-sample evaluation criteria to ascertain model performance. Using annual data on single-family residential water consumption in Southern California we illustrate how prediction ability varies according to model evaluation method. Using a training dataset, this analysis finds that models ranking highly on in-sample performance significantly over-estimated consumption (10% − 25%) five years out from the end of the training dataset relative to observed demands five years out from the end of the training dataset. Whereas, the top models selected using our out-of-sample criteria, came within 1% of the actual total consumption. Notably, projections of future demand for the in-sample models indicate increasing aggregate water consumption over a 25-year period, which contrasts against the downward trend predicted by the out-of-sample models. |
| Keywords: | Water, Forecasting, California, Demand and Price Analysis, Environmental Economics and Policy, Research Methods/ Statistical Methods, Resource /Energy Economics and Policy, C19, L95, Q25, |
| Date: | 2015 |
| URL: | http://d.repec.org/n?u=RePEc:ags:aaea15:205737&r=for |
| By: | Pavel V. Shevchenko; Jonas Hirz; Uwe Schmock |
| Abstract: | Recently we developed a new framework in Hirz et al (2015) to model stochastic mortality using extended CreditRisk$^+$ methodology which is very different from traditional time series methods used for mortality modelling previously. In this framework, deaths are driven by common latent stochastic risk factors which may be interpreted as death causes like neoplasms, circulatory diseases or idiosyncratic components. These common factors introduce dependence between policyholders in annuity portfolios or between death events in population. This framework can be used to construct life tables based on mortality rate forecast. Moreover this framework allows stress testing and, therefore, offers insight into how certain health scenarios influence annuity payments of an insurer. Such scenarios may include improvement in health treatments or better medication. In this paper, using publicly available data for Australia, we estimate the model using Markov chain Monte Carlo method to identify leading death causes across all age groups including long term forecast for 2031 and 2051. On top of general reduced mortality, the proportion of deaths for certain certain causes has changed massively over the period 1987 to 2011. Our model forecasts suggest that if these trends persist, then the future gives a whole new picture of mortality for people aged above 40 years. Neoplasms will become the overall number-one death cause. Moreover, deaths due to mental and behavioural disorders are very likely to surge whilst deaths due to circulatory diseases will tend to decrease. This potential increase in deaths due to mental and behavioural disorders for older ages will have a massive impact on social systems as, typically, such patients need long-term geriatric care. |
| Date: | 2015–07 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1507.07162&r=for |
| By: | Lyocsa, Stefan |
| Abstract: | We use a simple linear regression framework to present evidence, that complex relationships between stock markets and economies may be used to predict changes in the output of 27 OECD countries. We construct new unidirectional return co-exceedance networks to account for complex relationships between stock market returns, and between real economic growths. Although there is heterogeneity between individual country level results, overall our data and analysis provides evidence that topological properties of our networks are useful for in-sample prediction of next quarter changes in the output. |
| Keywords: | harmonic centrality centralization networks co-exceedance economic growth |
| JEL: | E44 G15 O40 |
| Date: | 2015–07–27 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:65774&r=for |
| By: | Mikkel Hermansen; Oliver Röhn |
| Abstract: | The global financial crisis and the high associated costs have revived the academic and policy interest in “early warning indicators” of crises. This paper provides empirical evidence on the usefulness of a new set of vulnerability indicators, proposed in a companion paper (Röhn et al., 2015), in predicting severe recessions and crises in OECD countries. To evaluate the usefulness of the indicators the signalling approach is employed, which takes into account policy makers’ preferences between missing crises and false alarms. Our empirical evidence shows that the majority of indicators would have helped to predict severe recessions in the 34 OECD economies and Latvia between 1970 and 2014. Indicators of global risks consistently outperform domestic indicators in terms of their usefulness, highlighting the importance of taking international developments into account when assessing a country’s vulnerabilities. In the domestic areas, indicators that measure asset market imbalances (real house and equity prices, house price-to-income and house price-to-rent ratios), also perform consistently well both in and out-of sample. Domestic credit related variables appear particularly useful in signalling upcoming banking crises and in predicting the global financial crisis out-of-sample. The results are broadly robust to different definitions of costly events, different forecasting horizons and different time and country samples.<P>Résilience économique : L'utilité des indicateurs d'alerte rapide dans des pays de l'OCDE<BR>La crise financière mondiale et les coûts associés élevés ont ravivé l'intérêt pour les « indicateurs d'alerte rapide » des crises. Cette étude fournit des données statistiques sur l'utilité d'un nouvel ensemble d'indicateurs de vulnérabilité, proposé dans une étude connexe (Röhn et al., 2015), pour prédire les récessions graves et les crises dans les pays de l'OCDE. Pour évaluer l'utilité des indicateurs la méthode de signalisation est employée. Celle-ci prend en compte les préférences des décideurs politiques entre les crises manquantes et les fausses alarmes. Les résultats de l’analyse statistique montrent que la majorité des indicateurs aurait aidé à prédire les récessions sévères dans les 34 économies de l'OCDE et la Lettonie entre 1970 et 2014. Les indicateurs de risque global surclassent systématiquement les indicateurs domestiques en termes d’information utile, soulignant l'importance de prendre les développements internationaux en compte lors de l'évaluation des vulnérabilités d'un pays. Dans les champs domestiques, des indicateurs qui mesurent les déséquilibres du marché des actifs (les prix réels des logements et le cours des actions, le ratio du prix des logements au revenu disponible et le ratio du prix des logements au coût des loyers), performe bien dans et hors de l'échantillon. Les variables reliées au crédit domestique semblent particulièrement utile dans la signalisation des crises bancaires et à prédire la crise financière mondiale hors-échantillon. Les résultats sont globalement robustes pour différentes définitions d'événements onéreux, différents horizons de prévision et différents échantillons de temps et de pays. |
| Keywords: | recession, crisis, resilience, vulnerability, résilience, crise, déséquilibres, vulnérabilité |
| JEL: | E32 E44 E51 F47 |
| Date: | 2015–07–28 |
| URL: | http://d.repec.org/n?u=RePEc:oec:ecoaaa:1250-en&r=for |
| By: | Mujahid, Irfan; Kalkuhl, Matthias |
| Abstract: | Food price variations can be very costly when they abrupt and unanticipated. In the current new era of market uncertainty, monitoring food prices become highly important to foresee any potential crisis. This study proposes an alternative approach in monitoring food price movements in many different markets within a country by focusing only on the key markets. Using monthly retail rice prices from the 25 major markets in Indonesia, we identify the key markets whose price movements can help to forecast price movements in all other markets. The key markets are identified using granger causality tests conducted in the vector error correction model framework. The relevance of monitoring the key markets in detecting price crisis is tested using Probit and Poisson models. We found that albeit not all of alert phases lead to crises, monitoring the key markets can help to forecast price movements in all markets across the country. |
| Keywords: | volatility, crisis, transmission, early warning system, Indonesia, Agricultural and Food Policy, Food Security and Poverty, C22, F1, F47, Q1, |
| Date: | 2015–07 |
| URL: | http://d.repec.org/n?u=RePEc:ags:aaea15:205277&r=for |