nep-for New Economics Papers
on Forecasting
Issue of 2012‒05‒22
eleven papers chosen by
Rob J Hyndman
Monash University

  1. Bayesian Forecast Combination for Inflation Using Rolling Windows: An Emerging Country Case By Luis Fernando Melo; Rubén Albeiro Loaiza Maya
  2. Semi-parametric forecasting of Spikes in Electricity Prices By Adam E Clements; Joanne Fuller; Stan Hurn
  3. Inflation forecasting in Angola: a fractional approach By Carlos P. Barros; Luis A. Gil-Alana
  4. Privileged Information Exacerbates Market Volatility By Gabriel Desgranges; Stéphane Gauthier
  5. A dynamic factor model of the yield curve as a predictor of the economy By Marcelle Chauvet; Zeynep Senyuz
  6. Asymptotically Optimal Algorithm for Short-Term Trading Based on the Method of Calibration By Vladimir V'yugin; Vladimir Trunov
  7. Ein Modell für die Wirtschaftszweige der deutschen Volkswirtschaft: Das "MOGBOT" (Model of Germany's Branches of Trade) By Quaas, Georg; Köster, Robert
  8. Imposing Curvature Conditions on Flexible Functional Forms to GNP Functions By Chapda Nana, Guy; Larue, Bruno
  9. The Hog-Cycle of Law Professors By Christoph Engel; Hanjo Hamann
  10. Pseudo-Predictability in Conditional Asset Pricing Tests: Explaining Anomaly Performance with Politics, the Weather, Global Warming, Sunspots, and the Stars By Robert Novy-Marx
  11. Don't stand so close to me: the urban impact of immigration By Antonio Accetturo; Francesco Manaresi; Sauro Mocetti; Elisabetta Olivieri

  1. By: Luis Fernando Melo; Rubén Albeiro Loaiza Maya
    Abstract: Typically, when forecasting inflation rates, there are a variety of individual models and a combination of several of these models. We implement a Bayesian shrinkage combination methodology to include information that is not captured by the individual models using expert forecasts as prior information. To take into account two common characteristics in emerging countries’ economies, possible parameter instabilities and non-stationary dynamics, we use a rolling estimation windows technique for series integrated of order one. The empirical results of Colombian inflation show that the Bayesian forecast combination model outperforms the individual models and the random walk predictions for every evaluated forecast horizon. Moreover, these results outperform shrinkage forecasts that consider other priors as equal or zero weights.
    Keywords: Forecast combination, Shrinkage, Expert forecasts, Rolling window estimation, Inflation forecasts. Classification JEL: C22, C53, C11, E31.
    Date: 2012–04
  2. By: Adam E Clements (QUT); Joanne Fuller (QUT); Stan Hurn (QUT)
    Abstract: The occurrence of extreme movements in the spot price of electricity represent a significant source of risk to retailers. Electricity markets are often structured so as to allow retailers to purchase at an unregulated spot price but then sell to consumers at a heavily regulated price. As such, the ability to forecast price spikes is an important aspect of effective risk management. A range of approaches have been considered with respect to modelling electricity prices, including predicting the trajectory of spot prices, as well as more recently, focusing of the prediction of spikes specifically. These models however, have relied on time series approaches which typically use restrictive decay schemes placing greater weight on more recent observations. This paper develops an alternative, semi-parametric method for forecasting that does not rely on this convention. In this approach, a forecast is a weighted average of historical price data, with the greatest weight given to periods that exhibit similar market conditions to the time at which the forecast is being formed. Weighting is determined by comparing short-term trends in electricity price spike occurrences across time, including other relevant factors such as load, by means of a multivariate kernel scheme. It is found that the semi-parametric method produces forecasts that are more accurate than the previously identified best approach for a short forecast horizon.
    Keywords: Electricity Prices, Prices Spikes, Semi-parametric, Multivariate Kernel
    JEL: C14 C53
    Date: 2012–05–16
  3. By: Carlos P. Barros; Luis A. Gil-Alana
    Abstract: This paper forecasts inflation in Angola with an ARFIMA (AutoRegressive Fractionally Integrated Moving Average) model. It is found that inflation in Angola is a highly persistent variable with an order of integration constrained between 0 and 1. Moreover, a structural break is found in August, 1996. Using the second sub-sample for forecasting purposes, the results reveal that inflation will remain low, assuming that prudent macroeconomic policies are maintained.
    Keywords: Angola; inflation, long memory
    JEL: C22
    Date: 2012–02
  4. By: Gabriel Desgranges (THEMA); Stéphane Gauthier (CREST)
    Abstract: We study how asymmetric information affects market volatility in a linear setup where the outcome is determined by forecasts about this same outcome. The unique rational expectations equilibrium will be stable when it is the only rationalizable solution. It has been established in the literature that stability obtains when the sensitivity of the outcome to agents' forecasts is less than 1, provided that this sensitivity is common knowledge. Relaxing this common knowledge assumption, instability obtains when the proportion of agents who a priori know the sensitivity is large, and the uninformed agents believe it is possible that the sensitivity is greater than 1
    Keywords: Asymmetric Information, Common Knowledge, Eductive Learning, Rational Expectations, Rationalizability, Volatility
    JEL: C62 D82 D84
    Date: 2011–05
  5. By: Marcelle Chauvet; Zeynep Senyuz
    Abstract: In this paper, we propose an econometric model of the joint dynamic relationship between the yield curve and the economy to predict business cycles. We examine the predictive value of the yield curve to forecast future economic growth as well as the beginning and end of economic recessions at the monthly frequency. The proposed nonlinear multivariate dynamic factor model takes into account not only the popular term spread but also information extracted from the level and curvature of the yield curve and from macroeconomic variables. The nonlinear model is used to investigate the interrelationship between the phases of the bond market and of the business cycle. The results indicate a strong interrelation between these two sectors. The proposed factor model of the yield curve exhibits substantial incremental predictive value compared to several alternative specifications. This result holds in-sample and out-of-sample, using revised or real time unrevised data.
    Date: 2012
  6. By: Vladimir V'yugin; Vladimir Trunov
    Abstract: A trading strategy based on a natural learning process, which asymptotically outperforms any trading strategy from RKHS (Reproduced Kernel Hilbert Space), is presented. In this process, the trader rationally chooses his gambles using predictions made by a randomized well calibrated algorithm. Our strategy is based on Dawid's notion of calibration with more general changing checking rules and on some modification of Kakade and Foster's randomized algorithm for computing calibrated forecasts. We use also Vovk's method of defensive forecasting in RKHS.
    Date: 2012–05
  7. By: Quaas, Georg; Köster, Robert
    Abstract: Das MOGBOT ist ein ökonometrisches Modell für die Zweige der deutschen Volkswirtschaft nach der Gliederung A*10 der Klassifikation WZ 2008 (ergänzt durch das Verarbeitende Gewerbe, Abschnitt C), die den Daten der Volkswirtschaftlichen Gesamtrechnungen nach der Generalrevision 2011 zugrunde liegt. Den hier präsentierten vier Versionen des Modells liegen unterschiedliche ökonometrische Ansätze zugrunde. Es handelt sich um ein an die Zeitreihenanalyse angelehntes Modell (Version A), ein theoriegestütztes strukturelles Regressionsmodell in der Keynes/Klein-Tradition (Version B) und ein VEC-Modell ohne (C1) und mit (C2) exogenen Variablen. Da der Zweck des Modells in erster Linie in der Prognostik gesehen werden muss, wurden alle vier Versionen vor allem unter dem Gesichtspunkt getestet und bewertet, wie sie die (Vierteljahres-) Daten der letzten 18 Jahre mit Hilfe einer dynamischen Lösung des Modells erklären können. Dabei zeigt sich, dass das traditionelle Strukturmodell am besten zu den Daten passt, aber in zwei Fällen wenig plausible Prognosen produziert. Die Ergebnisse des VEC-Modells können verwendet werden, um diese Anomalien zu korrigieren. -- MOGBOT is an econometric model of Germany's branches of trade in the grouping A*10 of the classification scheme WZ 2008 (supplemented by processing trade, section C) according to which the data of the National Account System are reported after the major revision in 2011. The four versions of the model that are presented in this paper follow different econometric approaches. Version A is mainly based on time series analysis and version B is a theory-led structural regression model in the Keynes-Klein tradition. Both are supplemented by two versions of a VEC-model: one without (version C1) and the other with exogenous variables (version C2). Because the primary purpose of the model has to be seen in forecasting, all four versions are mainly tested and assessed from the point of view of how they can explain the (quarterly) data of the last 18 years with a dynamic solution of the model. It turns out that the traditionally designed structural model fits best, but produces implausible forecasts in two cases. The results of the VEC-models can be used to correct these anomalies.
    Keywords: Prognose und Simulation,Prognosemodelle,makroökonomische Analyse der ökonomischen Entwicklung,ökonometrische Modelle und ihre Anwendung,Simulationsmethoden,Wirtschaftszweige,deutsche Volkswirtschaft,WZ 2008,Forecasting and Simulation,Forecasting Models,Macroeconomic Analyses of Economic Development,Econometric Models and Applications,Simulation Methods,Branches of Trade,German Economy
    JEL: C53 E27 O11
    Date: 2012
  8. By: Chapda Nana, Guy; Larue, Bruno
    Abstract: This paper empirically investigates the implications of the imposition of convexity in output prices and concavity in factor endowments on flexible functional forms for the GNP function. Using macroeconomic data for Switzerland, we estimate the Translog and the Symmetric Normalized Quadratic forms to investigate the manner with which curvature restrictions are imposed, the extend of curvature violations and the robustness of estimated elasticities. We also compare the predictive accuracy of the aforementioned flexible functional forms. Our result show that concavity in factor endowments is violated much more often than convexity in output prices. For the Translog, the date at which local restrictions are imposed matters a great deal in terms of remaining curvature violations in the sample, but far less for estimated elasticities. In contrast, we found that the size and sign of elasticities vary across functional forms. In-sample forecasting analysis demonstrates that the Translog model significantly dominates the Symmetric Normalized Quadratic.
    Keywords: GNP function, flexible functional forms, curvature violations, elasticities, International Relations/Trade, Research Methods/ Statistical Methods, D24, C30,
    Date: 2012–05
  9. By: Christoph Engel (Max Planck Institute for Research on Collective Goods, Bonn); Hanjo Hamann (Max Planck Institute for Research on Collective Goods, Bonn)
    Abstract: The market for law professors fulfils the conditions for a hog cycle: in the short run, supply cannot be extended or limited; future law professors must be hired soon after they first present themselves, or leave the market; demand is inelastic. Using a comprehensive German dataset, we show that the number of market entries today is significantly negatively correlated with the number of market entries 8 years ago. This is quite precisely the time young scholars on average take to prepare for the market. To get this estimate, we detrend the data, and we control for the size of student cohorts when these candidates enter university. This control variable mediates the effect of birth cohorts when candidates are born, which themselves exhibit negative autocorrelation, with a lag of some 20 years. Using our statistical model, we make out of sample predictions for the German academic market in law until 2020.
    Keywords: market for law professors, hog-cycle, time series, out of sample prediction
    JEL: K23 K00 D84 D92 J22 D22 C22 J45
    Date: 2012–04
  10. By: Robert Novy-Marx
    Abstract: Ferson, Sarkissian and Simin (2003) warn that persistence in expected returns generates spurious regression bias in predictive regressions of stock returns, even though stock returns are themselves only weakly autocorrelated. Despite this fact a growing literature attempts to explain the performance of stock market anomalies with highly persistent investor sentiment. The data suggest, however, that the potential misspecification bias may be large. Predictive regressions of real returns on simulated regressors are too likely to reject the null of independence, and it is far too easy to find real variables that have “significant power” predicting returns. Standard OLS predictive regressions find that the party of the U.S. President, cold weather in Manhattan, global warming, the El Niño phenomenon, atmospheric pressure in the Arctic, the conjunctions of the planets, and sunspots, all have “significant power” predicting the performance of anomalies. These issues appear particularly acute for anomalies prominent in the sentiment literature, including those formed on the basis of size, distress, asset growth, investment, profitability, and idiosyncratic volatility.
    JEL: C53 G0 G12
    Date: 2012–05
  11. By: Antonio Accetturo (Bank of Italy); Francesco Manaresi (Bank of Italy); Sauro Mocetti (Bank of Italy); Elisabetta Olivieri (Bank of Italy)
    Abstract: We examine the impact of immigration on the residential market within urban areas. We develop a spatial equilibrium model that shows how the effect of an immigrant inflow in a district affects local housing prices through changes in how natives perceive the quality of their local amenities and how this influences their mobility. Predictions of the model are tested using a novel dataset on housing prices and population variables at the district level for a sample of 20 large Italian cities. To address endogeneity problems we adopt an instrumental variable strategy which uses historical enclaves of immigrants across districts to predict current settlements. We find that immigration raises average housing prices at the city level; however it reduces price growth in a district affected by an inflow vis-à-vis the rest of the city. This pattern is driven by the natives’ flight from immigrant-dense districts towards other areas of the city. These findings are consistent with native preferences to live in predominantly native areas.
    Keywords: migration, housing, spatial segregation
    JEL: R23 J15 R21 F22
    Date: 2012–04

This nep-for issue is ©2012 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.