nep-for New Economics Papers
on Forecasting
Issue of 2011‒07‒13
ten papers chosen by
Rob J Hyndman
Monash University

  1. The Role of Asset Prices in Forecasting Inflation and Output in South Africa By Rangan Gupta; Faaiqa Hartley
  2. Beating the Random Walk in Central and Eastern Europe by Survey Forecasts By Anna Naszódi
  3. Ranking Multivariate GARCH Models by Problem Dimension: An Empirical Evaluation By Caporin, M.; McAleer, M.J.
  4. Testing the asset pricing model of exchange rates with survey data By Anna Naszódi
  5. Growth and Change in the Vietnamese Labour Market: A decomposition of forecast trends in employment over 2010-2020 By J. A. Giesecke; N. H. Tran; G.A. Meagher; F. Pang
  6. Forecasting Value-at-Risk Using Nonlinear Regression Quantiles and the Intraday Range By Chen, C.W.S.; Gerlach, R.; Hwang, B.B.K.; McAleer, M.J.
  7. Error Reduction strategies for the 1998-2005 USAGE Forecast By Peter Mavromatis; Marnie Griffith
  8. Neural Networks, Ordered Probit Models and Multiple Discriminants. Evaluating Risk Rating Forecasts of Local Governments in Mexico. By Alfonso Mendoza-Velázquez; Pilar Gómez-Gil
  9. Partial Least Square Discriminant Analysis (PLS-DA) for bankruptcy prediction By Carlos Serrano-Cinca; Begoña Gutiérrez-Nieto
  10. Destined for (Un)Happiness: Does Childhood Predict Adult Life Satisfaction? By Frijters, Paul; Johnston, David W.; Shields, Michael A.

  1. By: Rangan Gupta (Department of Economics, University of Pretoria); Faaiqa Hartley (Department of Economics, University of Pretoria)
    Abstract: This paper assesses the predictive ability of asset prices relative to other variables in forecasting inflation and real GDP growth in South Africa. A total of 42 asset and non-asset predictor variables are considered. Forecasts of inflation and real GDP growth are computed using both individual predictor autoregressive distributed lag (ARDL) models, forecast combination approaches, as well as, large scale models. The large scale data models considered include Bayesian vector autoregressive models, and classical and Bayesian univariate and multivariate factor augmented vector autoregressive models. The models are estimated for an in-sample of 1980:Q2 to 1999:Q4, and then one- to eight step-ahead forecasts for inflation and real GDP growth are evaluated over the 2000:Q1 to 2010:Q2 out-of-sample period. Principle Component forecast combination models are found to produce the most accurate out-of-sample forecasts of inflation and real GDP growth relative to the other combination and more sophisticated models considered. Asset prices are found to contain particularly useful information for forecasting inflation and real GDP growth at certain horizons. Asset prices are however found to be stronger predictors of inflation, particularly in the long run.
    Keywords: Asset Prices, Combination Forecasts, BVAR, FAVAR
    JEL: C32 R31
    Date: 2011–07
  2. By: Anna Naszódi (Magyar Nemzeti Bank (central bank of Hungary))
    Abstract: This paper investigates the forecasting ability of survey data on exchange rate expectations with multiple forecast horizons. The survey forecasts are on the exchange rates of five Central and Eastern European currencies: Czech Koruna, Hungarian Forint, Polish Zloty, Romanian Leu and Slovakian Koruna. First, different term-structure models are fitted on the survey forecasts. Then, the forecasting performances of the fitted forecasts are compared. The fitted forecasts for the 5 months horizon and beyond are proved to be significantly better than the random walk on the pooled data of the five currencies. The best performing term-structure model is the one that assumes an exponential relationship between the forecast and the forecast horizon, and has time-varying parameters.
    Keywords: evaluating forecasts, exchange rate, survey forecast, time-varying parameter, term-structure of forecasts
    JEL: F31 F36 G13
    Date: 2011
  3. By: Caporin, M.; McAleer, M.J.
    Abstract: In the last 15 years, several Multivariate GARCH (MGARCH) models have appeared in the literature. Recent research has begun to examine MGARCH specifications in terms of their out-of-sample forecasting performance. In this paper, we provide an empirical comparison of a set of models, namely BEKK, DCC, Corrected DCC (cDCC) of Aeilli (2008), CCC, Exponentially Weighted Moving Average, and covariance shrinking, using historical data of 89 US equities. Our methods follow part of the approach described in Patton and Sheppard (2009), and the paper contributes to the literature in several directions. First, we consider a wide range of models, including the recent cDCC model and covariance shrinking. Second, we use a range of tests and approaches for direct and indirect model comparison, including the Weighted Likelihood Ratio test of Amisano and Giacomini (2007). Third, we examine how the model rankings are influenced by the cross-sectional dimension of the problem.
    Keywords: covariance forecasting;model confidence set;model ranking;MGARCH;model comparison
    Date: 2011–05–31
  4. By: Anna Naszódi (Magyar Nemzeti Bank (central bank of Hungary))
    Abstract: This paper proposes a new test for the asset pricing model of the exchange rate. It examines whether the way market analysts generate their forecasts is closer to the one implied by the asset pricing model, or to any of those implied by some alternative models. The asset pricing model is supported by the test since it has significantly better out-of-sample fit on survey data than simpler models including the random walk. The traditional test based on forecasting ability is applied as well. The asset pricing model proves to have better forecast accuracy in case of some exchange rates and forecast horizons than the random walk.
    Keywords: asset pricing exchange rate model, present value model of exchange rate, survey data
    JEL: F31 F36 G13
    Date: 2011
  5. By: J. A. Giesecke; N. H. Tran; G.A. Meagher; F. Pang
    Abstract: We forecast detailed trends for employment by industry, occupation and qualification in Vietnam for the period 2010 - 2020. The forecast is conducted using VNET - a large-scaled computable general equilibrium (CGE) model of the Vietnamese economy. Inputs into the forecast include independent projections for changes in macroeconomic variables; trend movements in variables describing the details of industry input requirements and household preferences; assumptions relating to Vietnam's foreign trading environment; and projections for government policies. A decomposition analysis is used to identify the contribution of each of the exogenous shocks to the forecast outcomes. This analysis facilitates transparency in forecasting by clearly distinguishing and ranking the factors responsible for generating a particular forecast outcome. It also helps researchers to focus research effort towards improving estimates for those inputs to the simulation that have the most bearing on the outcomes forecast for the labour market.
    Keywords: labour market forecasting, long-run forecasting, decomposition analysis, Vietnamese economy
    JEL: C68 D58 E47 J21
    Date: 2011–04
  6. By: Chen, C.W.S.; Gerlach, R.; Hwang, B.B.K.; McAleer, M.J.
    Abstract: Value-at-Risk (VaR) is commonly used for financial risk measurement. It has recently become even more important, especially during the 2008-09 global financial crisis. We propose some novel nonlinear threshold conditional autoregressive VaR (CAViar) models that incorporate intra-day price ranges. Model estimation and inference are performed using the Bayesian approach via the link with the Skewed-Laplace distribution. We examine how a range of risk models perform during the 2008-09 financial crisis, and evaluate how the crisis affects the performance of risk models via forecasting VaR. Empirical analysis is conducted on five Asia-Pacific Economic Cooperation stock market indices and two exchange rates????. We examine violation rates, back-testing criteria, market risk charges and quantile loss function to measure the forecasting performance of a variety of risk models. The proposed threshold CAViaR model, incorporating range information, is shown to forecast VaR more efficiently than other models, which should be useful for financial practitioners.
    Keywords: Value-at-Risk;CAViaR model;Skewed-Laplace distribution;intra-day range;backtesting;Markov chain Monte Carlo
    Date: 2011–06–30
  7. By: Peter Mavromatis; Marnie Griffith
    Abstract: This paper examines methods aimed at improving baseline economic forecasts using a dynamic CGE model. Forecasting can be used to test the validity of such models, as well as to highlight possible improvements, by investigating the discrepancies between the forecast and actual outcomes. The model employed here is USAGE - a recursive dynamic, 500-industry CGE model of the U.S. USAGE generates baseline forecasts by incorporating expert forecasts for certain macro variables and extrapolating historical trends in technology, consumer preferences, positions of foreign demand curves for U.S. products, and numerous other naturally exogenous variables. In instances where important trends either dissipate or reverse, large forecast errors can arise. This paper seeks to provide explanations and guidance as to whether these various trends from the period 1992 to 1998 would continue for the 1998 to 2005 USAGE forecast. The twenty largest errors on a relative and/or absolute basis are examined. It is found that for some commodities, had all publicly available information by 1998 been appropriately utilised, certain important trends should not have been expected to continue. Hence, a better forecast could have been generated had the projection of certain trends been nullified. More generally, the findings suggest that there is a case to be argued against projecting forward large values relating to import-domestic preference twist factors in particular. It is also shown that for commodities in the trade-exposed textile, clothing and footwear industries moderately better results could have been produced by implementing import price forecasts in a way that is more in line with historical trade policy. This was achieved by projecting forward real basic import prices. However, the key drivers behind these errors were usually the significant underestimation of the impact of import-domestic preference twist factors, as well as the overestimation of factor input cost savings. In relation to forecasts for commodities in the oil and mining sectors as well as industries that service these cyclical industries, it is concluded that these typically could not have been improved in the absence of strong convictions (by 1998) about an impending mining "super-cycle" or extended boom. For the construction-related commodities demand was fuelled by virtually unprecedented low borrowing costs. In these instances, it is difficult to conclusively argue that the modeller could have produced a better forecast. Moreover, while large improvements in forecast accuracy can be obtained for some industries and sectors, the overall economy-wide forecast error does not fall greatly due to the sheer volume of commodities. While it is disappointing that the error is not very reducible, it is also reassuring because it implies that the default implementation of the model is quite powerful. In all about 4% of all commodities were specifically examined to assess the potential for error reduction. After due consideration about 7.5% of commodities were in some way directly re-projected. To generate a large reduction in the forecast error would require an extensive amount of work and probably call for the input of numerous industry specialists.
    Keywords: CGE, forecasting, validation
    JEL: C68 D58
    Date: 2011–04
  8. By: Alfonso Mendoza-Velázquez; Pilar Gómez-Gil
    Abstract: Credit risk ratings have become an important input in the process of improving transparency of public finances in local governments and also in the evaluation of credit quality of state and municipal governments in Mexico. Although rating agencies have recently been subjected to heavy criticism, credit ratings are indicators still widely used as a benchmark by analysts, regulators and banks monitoring financial performance of local governments in stable and volatile periods. In this work we compare and evaluate the performance of three forecasting methods frequently used in the literature estimating credit ratings: Artificial Neural Networks (ANN), Ordered Probit models (OP) and Multiple Discriminant Analysis (MDA). We have also compared the performance of the three methods with two models, the first one being an extended model of 34 financial predictors and a second model restricted to only six factors, accounting for more than 80% of the data variability. Although ANN provides better performance within the training sample, OP and MDA are better choices for classifications in the testing sample respectively.
    Keywords: Credit Risk Ratings, Ordered Probit Models, Artificial Neural Networks, Discriminant Analysis, Principal Components, Local Governments, Public Finance, Emerging Markets
    JEL: C25 C63 H79
    Date: 2011–06–28
  9. By: Carlos Serrano-Cinca; Begoña Gutiérrez-Nieto
    Abstract: This paper uses Partial Least Square Discriminant Analysis (PLS-DA) for the prediction of the 2008 USA banking crisis. PLS regression transforms a set of correlated explanatory variables into a new set of uncorrelated variables, which is appropriate in the presence of multicollinearity. PLS-DA performs a PLS regression with a dichotomous dependent variable. The performance of this technique is compared to the performance of 8 algorithms widely used in bankruptcy prediction. In terms of accuracy, precision, F-score, Type I error and Type II error, results are similar; no algorithm outperforms the others. Behind performance, each algorithm assigns a score to each bank and classifies it as solvent or failed. These results have been analyzed by means of contingency tables, correlations, cluster analysis and reduction dimensionality techniques. PLS-DA results are very close to those obtained by Linear Discriminant Analysis and Support Vector Machine.
    Keywords: bankruptcy; financial ratios; banking crisis; solvency; data mining; PLS-DA
    Date: 2011–06
  10. By: Frijters, Paul (University of Queensland); Johnston, David W. (Monash University); Shields, Michael A. (University of Melbourne)
    Abstract: In this paper we address the question of how much of adult life satisfaction is predicted by childhood traits, parental characteristics and family socioeconomic status. Given the current focus of many national governments on measuring population well-being, and renewed focus on effective policy interventions to aid disadvantaged children, we study a cohort of children born in a particular week in 1958 in Britain who have been repeatedly surveyed for 50 years. Importantly, at four points in their adult lives this cohort has been asked about their life satisfaction (at ages 33, 42, 46, and 50). A substantive finding is that characteristics of the child and family at birth predict no more than 1.2% of the variance in average adult life satisfaction. A comprehensive set of child and family characteristics at ages 7, 11 and 16 increases the predictive power to only 2.8%, 4.3% and 6.8%, respectively. We find that the conventional measures of family socioeconomic status, in the form of parental education, occupational class and family income, are not strong predictors of adult life satisfaction. However, we find robust evidence that non-cognitive skills as measured by childhood behavioural-emotional problems, and social maladjustment, are powerful predictors of whether a child grows up to be a satisfied adult. We also find that some aspects of personality are important predictors. Adding contemporaneous adulthood variables for health and socio-economic status increases the predictability of average life satisfaction to 15.6%, while adding long-lags of life satisfaction increases the predictive power to a maximum of 35.5%. Repeating our analyses using data from the 1970 British Cohort Study confirms our main findings. Overall, the results presented in the paper point to average adult life satisfaction not being strongly predictable from a wide-range of childhood and family characteristics by age 16, which implies that there is high equality of opportunity to live a satisfied life, at least for individuals born in Britain in 1958 and 1970.
    Keywords: childhood, socioeconomic status, life satisfaction, non-cognitive, cognitive
    JEL: I1 J1
    Date: 2011–06

This nep-for issue is ©2011 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.