nep-for New Economics Papers
on Forecasting
Issue of 2007‒07‒07
thirteen papers chosen by
Rob J Hyndman
Monash University

  1. Does Age Structure Forecast Economic Growth? By David E. Bloom; David Canning; Guenther Fink; Jocelyn E. Finlay
  2. Econometric Analysis with Vector Autoregressive Models By Helmut Luetkepohl
  3. A Comparison of Estimation Methods for Vector Autoregressive Moving-Average Models By Christian Kascha
  4. Macroeconomic modelling in EMU: how relevant is the change in regime? By Javier Andrés; Fernando Restoy
  5. Quantile Forecasting for Credit Risk Management using possibly Mis-specified Hidden Markov Models By Konrad Banachewicz; André Lucas
  6. Decomposing violence: terrorist murder and attacks in New York State from 1933 to 2005 By Gómez-Sorzano, Gustavo
  7. Global Inflation By Matteo Ciccarelli; Benoît Mojon
  8. Non-negativity Conditions for the Hyperbolic GARCH Model By Christian Conrad
  9. Pathways to Disability: Predicting Health Trajectories By Florian Heiss; Axel Börsch-Supan; Michael Hurd; David Wise
  10. Random Walk Expectations and the Forward Discount Puzzle By Philippe Bacchetta; Eric van Wincoop
  11. Price setting during low and high inflation: evidence from Mexico By Etienne Gagnon
  12. Spousal Influences on Parents’ Non-Market Time Choices By Rachel Connelly; Jean Kimmel
  13. Actualización del modelo trimestral del Banco de España By Eva Ortega; Pablo Burriel; José Luis Fernández; Eva Ferraz; Samuel Hurtado

  1. By: David E. Bloom; David Canning; Guenther Fink; Jocelyn E. Finlay
    Abstract: Increases in the proportion of the working age population can yield a "demographic dividend" that enhances the rate of economic growth. We estimate the parameters of an economic growth model with a cross section of countries over the period 1960 to 1980 and investigate whether the inclusion of age structure improves the model's forecasts for the period 1980 to 2000. We find that including age structure improves the forecast, although there is evidence of parameter instability between periods with an unexplained growth slowdown in the second period. We use the model to generate growth forecasts for the period 2000 to 2020.
    JEL: C53 J1 O4
    Date: 2007–07
  2. By: Helmut Luetkepohl
    Abstract: Vector autoregressive (VAR) models for stationary and integrated variables are reviewed. Model specification and parameter estimation are discussed and various uses of these models for forecasting and economic analysis are considered. For integrated and cointegrated variables it is argued that vector error correction models offer a particularly convenient parameterization both for model specification and for using the models for economic analysis.
    Keywords: VAR, vector autoregressive models
    JEL: C32
    Date: 2007
  3. By: Christian Kascha
    Abstract: Classical Gaussian maximum likelihood estimation of mixed vector autoregressive moving-average models is plagued with various numerical problems and has been considered di±cult by many applied researchers. These disadvantages could have led to the dominant use of vector autoregressive models in macroeconomic research. Therefore, several other, simpler estimation methods have been proposed in the literature. In this paper these methods are compared by means of a Monte Carlo study. Different evaluation criteria are used to judge the relative performances of the algorithms.
    Keywords: VARMA Models, Estimation Algorithms, Forecasting
    JEL: C32 C15 C63
    Date: 2007
  4. By: Javier Andrés (Universidad de Valencia); Fernando Restoy (Banco de España)
    Abstract: We analyse the likely effects of changes in the monetary and financial regimes of EMU countries on the dynamics of output and inflation. In particular, we evaluate the impact of the regime shift on the forecasting performance of reduced-form models. Data for both the pre-EMU and the EMU regimes are generated by a relatively standard open-economy-DSGE model with sticky prices and wages and restricted access to financial markets for some individuals. We find that the effects of the shift in the monetary regime on the processes followed by macroeconomic variables depend on the nature of the shocks hitting the economy. For plausible shocks distributions the reduction in the accuracy of VAR-based inflation forecasts is relatively large and significant. The effect of the regime shift on output forecasts seem rather more modest and statistically insignificant. The impact on ouput forecasting accuracy would be comparatively much larger if the new monetary union regime is accompanied by a moderate relaxation of constraints affecting financial market access.
    Keywords: forecasting, general equilibrium models, monetary union, inflation and output dynamics
    JEL: E17 E32 E37
    Date: 2007–06
  5. By: Konrad Banachewicz (Vrije Universiteit Amsterdam); André Lucas (Vrije Universiteit Amsterdam)
    Abstract: Recent models for credit risk management make use of Hidden Markov Models (HMMs). The HMMs are used to forecast quantiles of corporate default rates. Little research has been done on the quality of such forecasts if the underlying HMM is potentially mis-specified. In this paper, we focus on mis-specification in the dynamics and the dimension of the HMM. We consider both discrete and continuous state HMMs. The differences are substantial. Underestimating the number of discrete states has an economically significant impact on forecast quality. Generally speaking, discrete models underestimate the high-quantile default rate forecasts. Continuous state HMMs, however, vastly overestimate high quantiles if the true HMM has a discrete state space. In the reverse setting, the biases are much smaller, though still substantial in economic terms. We illustrate the empirical differences using U.S. default data.
    Keywords: defaults; Markov switching; misspecification; quantile forecast; Expectation-Maximization; simulated maximum likelihood; importance sampling
    JEL: C53 C22 G32
    Date: 2007–06–13
  6. By: Gómez-Sorzano, Gustavo
    Abstract: I apply the Beveridge-Nelson business cycle decomposition method to the time series of murder in the state of New York. (1933-2005). Separating out “permanent” from “cyclical” murder, I hypothesize that the cyclical part coincides with documented waves of organized crime, internal tensions, breakdowns in social order, crime legislation, social, and political unrest, and recently with the periodic terrorist attacks in the state. The estimated cyclical terrorist murder component warns that terrorist attacks in the state of New York from 1962 to 2005, historically occur in the estimated turning point dates, of whether a declining, or ascending cycle, and so, it must be used in future research to construct a model for explaining the causal reasons for its movement across time, and for forecasting terrorist murder and attacks for New York.
    Keywords: A model of cyclical terrorist murder in Colombia; 1950-2004. Forecasts 2005-2019; the econometrics of violence; terrorism; and scenarios for peace in Colombia from 1950 to 2019; scenarios for sustainable peace in Colombia by year 2019; decomposing violence: terrorist murder in the twentieth in the United States; using the Beveridge and Nelson decomposition of economic time series for pointing out the occurrence of terrorist attacks.
    JEL: H56 N42 H75 K14 H76 D74 C80 K42 C22
    Date: 2006–12–31
  7. By: Matteo Ciccarelli; Benoît Mojon
    Abstract: This paper shows that ination in industrialized countries is largely a global phenom- enon. First, inations of (22) OECD countries have a common factor that alone accounts for nearly 70% of their variance. This large variance share that is associated to Global Ination is not only due to the trend components of ination (up from 1960 to 1980 and down thereafter) but also to uctuations at business cycle frequencies. Second, Global In- ation is, consistently with standard models of ination, a function of real developments at short horizons and monetary developments at longer horizons. Third, there is a very robust "error correction mechanism" that brings national ination rates back to Global Ination. This model consistently beats the previous benchmarks used to forecast ination 1 to 8 quarters ahead across samples and countries.
    Keywords: Inflation, common factor, international business cycle, OECD countries
    JEL: E31 E37 F42
    Date: 2007–06
  8. By: Christian Conrad (KOF Swiss Economic Institute, ETH Zurich Switzerland)
    Abstract: In this article we derive conditions which ensure the non-negativity of the conditional variance in the Hyperbolic GARCH(p; d; q) (HYGARCH) model of Davidson (2004). The conditions are necessary and suffcient for p < 2 and suffcient for p > 2 and emerge as natural extensions of the inequality constraints derived in Nelson and Cao (1992) for the GARCH model and in Conrad and Haag (2006) for the FIGARCH model. As a by-product we obtain a representation of the ARCH(1) coeffcients which allows computationally effcient multi-step-ahead forecasting of the conditional variance of a HYGARCH process. We also relate the necessary and suffcient parameter set of the HYGARCH to the necessary and su±cient parameter sets of its GARCH and FIGARCH components. Finally, we analyze the effects of erroneously fitting a FIGARCH model to a data sample which was truly generated by a HYGARCH process. An empirical application of the HYGARCH(1; d; 1) model to daily NYSE data illustrates the importance of our results.
    Keywords: Inequality constraints, fractional integration, long memory GARCH processes
    JEL: C22 C52 C53
    Date: 2007–04
  9. By: Florian Heiss; Axel Börsch-Supan; Michael Hurd; David Wise (Mannheim Research Institute for the Economics of Aging (MEA))
    Abstract: The paper considers transitions in the health and disability status of persons as they age. In particular, we explore the relationship between health and disability at younger ages (say 50) and health and disability in future ages. We consider for example, the future health path of persons who are in good health at age 50 compared to the future health path of persons who are in poor health at age 50. To do this, we develop a model that jointly considers health and mortality. The key feature of the model is the assumption of underlying “latent” health that determines both mortality and self-reported responses to categorical health and disability questions. Latent health allows for heterogeneity among individuals and allows for correlation of health status over time, thus allowing for state dependence as well as heterogeneity. The model also allows for classification errors in self-reported response to categorical health and disability questions. All of these are important features of health and disability data, as we show with descriptive data. The model accommodates the strong relationship between self-reported health status and mortality, which is critical to an understanding of the paths of health and disability of the survivors who are observed in panel data files. Our empirical analysis is based on all four cohorts of the Health and Retirement Study (HRS) -- the HRS, AHEAD, CODA and WB cohorts). We find that self-reported health and self-reported disability correspond very closely to one another in the HRS. We find that both self-reported health and disability are strong predictors of mortality. Health and disability at younger ages are strongly related to future health and disability paths of persons as they age. There are important differences in health and disability paths by education level, race, and gender.
    Date: 2007–07–03
  10. By: Philippe Bacchetta; Eric van Wincoop
    Abstract: Two well-known, but seemingly contradictory, features of exchange rates are that they are close to a random walk while at the same time exchange rate changes are predictable by interest rate differentials. In this paper we investigate whether these two features of the data may in fact be related. In particular, we ask whether the predictability of exchange rates by interest differentials naturally results when participants in the FX market adopt random walk expectations. We find that random walk expectations can explain the forward discount puzzle, but only if FX portfolio positions are revised infrequently. In contrast, with frequent portfolio adjustment and random walk expectations, we find that high interest rate currencies depreciate much more than what UIP would predict.
    JEL: F3 F31 F41
    Date: 2007–06
  11. By: Etienne Gagnon
    Abstract: This paper provides new insight into the relationship between inflation and consumer price setting by examining a large data set of Mexican consumer prices covering episodes of both low and high inflation, as well as the transition between the two. Overall, the economy shares several characteristics with time-dependent models when the annual inflation rate is low (below 10-15%), while displaying strong state dependence when inflation is high (above 10-15%). At low inflation levels, the aggregate frequency of price changes responds little to movements in inflation because movements in the frequency of price decreases partly offset movements in the frequency of price increases. When the annual inflation rate rises beyond 10-15 percent, however, there are no longer enough price decreases to counterbalance the rising occurrence of price increases, making the frequency of price changes more responsive to inflation. It is shown that a simple menu-cost model with idiosyncratic technology shocks predicts remarkably well the level of the average frequency and magnitude of price changes over a wide range of inflation.
    Date: 2007
  12. By: Rachel Connelly (Bowdoin College); Jean Kimmel (Western Michigan University and IZA)
    Abstract: This paper considers the effect of spouse’s characteristics on three aggregated non-paid time uses, active leisure time; child caregiving time; and home production time, using the American Time Use Survey (ATUS). The time diary of each married individual with children under the age of 13 (mothers and fathers) is analyzed, both in terms of the level of non-paid time and the wife’s share of the total level of the daily activity for the couple. Three spousal variables: the relative wage of the wife compared to her husband, spouses’ weekly hours of employment; and, in the level equations only, the spouses’ time in the same activity are considered. Each of these spousal variables needs to be estimated in order to address issues of both endogeneity and missing data. Three alternative strategies to address these problems are explored: predictions within the sample, predictions from outside the sample and propensity matching which "marries" mothers with time diaries to fathers with time diaries who have propensity scores similar to the women’s husband. The results show very little effect of one spouse on the level of other spouse’s unpaid time use. This absence of spousal effects is similar to the reduction of spousal effects in employment time described in Blau and Kahn (2005). In terms of the share of wife’s time in the activity, we find higher relative wages of the mother compared to her husband leads to a greater share of child care done by the mother on both weekdays and weekends. No consistent effect of relative wages is found on the mother’s share of leisure or home production.
    Keywords: time use, non-market time choices, family decision-making
    JEL: J13 J22
    Date: 2007–06
  13. By: Eva Ortega (Banco de España); Pablo Burriel (Banco de España); José Luis Fernández (Banco de España); Eva Ferraz (Banco de España); Samuel Hurtado (Banco de España)
    Abstract: This paper presents the update of the macroeconometric model used at the Bank of Spain for medium term macroeconomic forecasting, as well as for performing policy simulations. The many changes that the Spanish economy has experimented in the last years, and the new system of national accounts published by the national statistical office, suggested that a reestimation of the model was due. This paper presents such reestimation with newer data (up to the end of 2005), and includes some modifications that were deemed necessary in certain equations. The quarterly model of the Bank of Spain keeps a similar structure to its previous version; it still is basically a demand-driven model. It is found that the Spanish economy shows, in general, higher sensitivity than in previous periods to changes in exogenous variables, especially in financial conditions. The new model reflects, too, changes in demographic trends, and presents an external sector less sentitive to changes in price-competitiveness.
    Keywords: economía española, Spanish economy, modelo macroeconómico, macroeconometric model
    JEL: E10 E17 E20 E60
    Date: 2007–06

This nep-for issue is ©2007 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.