nep-for New Economics Papers
on Forecasting
Issue of 2015‒08‒13
twenty-two papers chosen by
Rob J Hyndman
Monash University

  1. Forecasting, Foresight and Strategic Planning for Black Swans By Kostas Nikolopoulos; F. Petropoulos
  2. Economic theory and forecasting: lessons from the literature By Raffaella Giacomini
  3. On a Bootstrap Test for Forecast Evaluations By Juraj Hucek; Alexander Karsay; Marian Vavra
  4. Modeling and Forecasting Carbon Dioxide Emission Allowance Spot Price Volatility: Multifractal vs. GARCH-type Volatility Models By Segnon, Mawuli; Lux, Thomas; Gupta, Rangan
  5. Stock Return Forecasting: Some New Evidence By Dinh H B Phan; Susan S Sharma; Paresh K Narayan
  6. Dynamic predictive density combinations for large data sets in economics and finance By Roberto Casarin; Stefano Grassi; Francesco Ravazzolo; Herman K. van Dijk
  7. Modelling and Forecasting Branded and Generic Pharmaceutical Life Cycles: Assessment of the Number of Dispensed Units By S. Buxton; Kostas Nikolopoulos; M. Khammash; P. Stern
  8. Forecasting an Aggregate in the Presence of Structural Breaks in the Disaggregates By William Larson
  9. Term Structure Dynamics, Macro-Finance Factors and Model Uncertainty By Joseph P. Byrne; Shuo Cao.; Dimitris Korobilis.
  10. Modeling Dependence Structure and Forecasting Market Risk with Dynamic Asymmetric Copula By Mario Cerrato; John Crosby; Minjoo Kim; Yang Zhao
  11. ISBEM: An econometric model for the Italian State Budget Expenditures By Giuseppe Bianchi; Tatiana Cesaroni; Ottavio Ricchi
  12. The Real-Time Predictive Content of Asset Price Bubbles for Macro Forecasts By Benjamin Beckers
  13. Managerial discretion over initial earnings forecasts By Takuya Iwasaki; Norio Kitagawa; Akinobu Shuto
  14. Prior selection for panel vector autoregressions By Dimitris Korobilis.
  15. A new monthly indicator of global real economic activity By Ravazzolo, Francesco; Vespignani, Joaquin L.
  16. Are Indian Stock Returns Predictable? By Deepa; Paresh K Narayan
  17. Intraday Volatility Interaction between the Crude Oil and Equity Markets By Dinh H B Phan; Susan S Sharma; Paresh K Narayan
  18. Has Oil Price Predicted Stock returns for Over a Century? By Paresh K Narayan; Rangan Gupta
  19. Management Earnings Forecasts as a Performance Target in Executive Compensation Contracts By Shota Otomasa; Atsushi Shiiba; Akinobu Shuto
  20. Inflation forecasting models for Uganda: is mobile money relevant? By Janine Aron; John Muellbauer; Rachel Sebudde
  21. Testing For Stock Return Predictability In A Large Chinese Panel By Joakim Westerlund; Paresh K Narayan; Xinwei Zheng
  22. An Analysis of Sectoral Equity and CDS Spreads By Paresh K Narayan

  1. By: Kostas Nikolopoulos (Bangor University); F. Petropoulos (Cardiff University Business School)
    Abstract: In this research essay we propose a methodological innovation through: a) advocating for the broader use of OR forecasting tools and in specific intermittent demand estimators for forecasting black and grey swans, as a simpler, faster and quite robust alternative to econometric probabilistic methods like EVT; b) demonstrating the use in such a context of a rather popular forecasting paradigm: the Naive method (forecasting short horizons) and the SBA method (foresight long horizons) through the ADIDA non-overlapping temporal aggregation method-improving framework, and c) arguing for a new way for deciding the strategic planning horizon for phenomena prone to the appearance of black and grey swans.
    Keywords: Forecasting, Black Swans, Intermittent Demand, Temporal Aggregation, Forecasting Horizon
    Date: 2015–03
  2. By: Raffaella Giacomini (Institute for Fiscal Studies and cemmap and UCL)
    Abstract: Does economic theory help in forecasting key macroeconomic variables? This article aims to provide some insight into the question by drawing lessons from the literature. The definition of "economic theory" includes a broad range of examples, such as accounting identities, disaggregation and spatial restrictions when forecasting aggregate variables, cointegration and forecasting with Dynamic Stochastic General Equilibrium (DSGE) models. We group the lessons into three themes. The first discusses the importance of using the correct econometric tools when answering the question. The second presents examples of theory-based forecasting that have not proven useful, such as theory-driven variable selection and some popular DSGE models. The third set of lessons discusses types of theoretical restrictions that have shown some usefulness in forecasting, such as accounting identities, disaggregation and spatial restrictions, and cointegrating relationships. We conclude by suggesting that economic theory might help in overcoming the widespread instability that affects the forecasting performance of econometric models by guiding the search for stable relationships that could be usefully exploited for forecasting.
    Date: 2014–09
  3. By: Juraj Hucek (National Bank of Slovakia, Economic and Monetary Analyses Department); Alexander Karsay (National Bank of Slovakia, Economic and Monetary Analyses Department); Marian Vavra (National Bank of Slovakia, Research Department)
    Abstract: This occasional paper considers the problem of forecasting, nowcasting, and backcasting the Slovak real GDP growth rate using approximate factor models. Three different versions of approximate factor models are proposed. Forecast comparison with other models such as bridge equation models and ARMA models is also provided. Our results reveal that factor models clearly outperform an ARMA model and can compete with bridge models currently used at the Bank. Therefore, we tend to incorporate factor models into the regular forecasting process at the Bank.Finally, we hold the view that future research should be devoted to further improvements of bridge models since these models are simple to construct, easy to understand, and widely used in central banks.
    Keywords: factor models, principal components, bridge equations, short-term forecasting, GDP
    JEL: C22 C38 C52 C53 E27
    Date: 2015–07
  4. By: Segnon, Mawuli; Lux, Thomas; Gupta, Rangan
    Abstract: This paper applies Markov-switching multifractal (MSM) processes to model and forecast carbon dioxide (CO2) emission price volatility, and compares their forecasting performance to the standard GARCH, fractionally integrated GARCH (FIGARCH) and the two-state Markov-switching GARCH (MS-GARCH) models via three loss functions (the mean squared error, the mean absolute error and the value-at-risk). We evaluate the performance of these models via the superior predictive ability test. We find that the forecasts based on the MSM model cannot be outperformed by its competitors under the vast majority of criteria and forecast horizons, while MS-GARCH mostly comes out as the least successful model. Applying various VaR backtesting procedures, we do, however, not find significant differences in the performance of the candidate models under this particular criterion. We also find that we cannot reject the null hypothesis of MSM forecasts encompassing those of GARCH-type models. In line with this result, optimally combined forecasts do indeed hardly improve upon the best single models in our sample.
    Keywords: carbon dioxide emission allowance prices,GARCH,Markov-switching GARCH,FIGARCH,multifractal Processes,SPA test,encompassing test,backtesting
    JEL: Q47
    Date: 2015
  5. By: Dinh H B Phan (Deakin University); Susan S Sharma (Deakin University); Paresh K Narayan (Deakin University)
    Abstract: This paper makes three contributions to the literature on forecasting stock returns. First, unlike the extant literature on oil price and stock returns, we focus on out-of-sample forecasting of returns. We show that the ability of the oil price to forecast stock returns depends not only on the data frequency used but also on the estimator. Second, out-of-sample forecasting of returns is sector-dependent, suggesting that oil price is relatively more important for some sectors than others. Third, we examine the determinants of out-of-sample predictability for each sector using industry characteristics and find strong evidence that return predictability has links to certain industry characteristics, such as book-to-market ratio, dividend yield, size, price earnings ratio, and trading volume.
    Keywords: Stock returns; Oil price; Predictability; Forecasting; Out-of-sample.
  6. By: Roberto Casarin (University Ca’ Foscari of Venice); Stefano Grassi (University of Kent); Francesco Ravazzolo (Norges Bank (Central Bank of Norway) and Centre for Applied Macro and Petroleum economics at BI Norwegian Business School); Herman K. van Dijk (Econometric Institute Erasmus University Rotterdam, Econometrics Department VU University Amsterdam and Tinbergen Institute)
    Abstract: A Bayesian nonparametric predictive model is introduced to construct time-varying weighted combinations of a large set of predictive densities. A clustering mechanism allocates these densities into a smaller number of mutually exclusive subsets. Using properties of the Aitchinson’s geometry of the simplex, combination weights are defined with a probabilistic interpretation. The classpreserving property of the logistic-normal distribution is used to define a compositional dynamic factor model for the weight dynamics with latent factors defined on a reduced dimension simplex. Groups of predictive models with combination weights are updated with parallel clustering and sequential Monte Carlo filters. The procedure is applied to predict Standard & Poor’s 500 index using more than 7000 predictive densities based on US individual stocks and finds substantial forecast and economic gains. Similar forecast gains are obtained in point and density forecasting of US real GDP, Inflation, Treasury Bill yield and employment using a large data set.
    Keywords: Density Combination, Large Set of Predictive Densities, Compositional Factor Models, Nonlinear State Space, Bayesian Inference, GPU Computing
    JEL: C11 C15 C53 E37
    Date: 2015–07–31
  7. By: S. Buxton (Swansee University); Kostas Nikolopoulos (Bangor University); M. Khammash (University of Sussex); P. Stern (University of Exeter)
    Abstract: In this paper, modelling and forecasting pharmaceutical life cycles are investigated, specifically at the time of patent expiry when the generic form of the product could be introduced to the market, while the branded form is still available for prescription. Assessing the number of dispensed units of branded and generic forms of pharmaceuticals is increasingly important due to the irrelatively large market value in the United Kingdom and the limited number of new ÔblockbusterÕ branded drugs. As a result, pharmaceutical companies make every effort to extend the commercial life of their branded products and forecast their sales in the future, while public health institutes seek insights for effective governance as the use of a branded drug, when a generic form is available. In this paper, eleven methods are used to model and forecast drugs life cycles: Bass Diffusion, Repeat Purchase Diffusion Model, Exponential Smoothing, Holt Winters Exponential Smoothing, Moving Averages, ARIMA, Regression over t, Regression over t-1, Robust Regression, Na•ve and Na•ve with drift. The results suggest a difference depending on the forecasting horizon with the ARIMA model and Holt Winters Exponential Smoothing both producing accurate short term forecasts. However for 3-5 year forecasts the results suggest that a very simple forecasting model blended with an addition of a small trend provides the most accurate forecasts for both branded and generic pharmaceuticals
    Date: 2015–04
  8. By: William Larson (Federal Housing Finance Agency)
    Abstract: There is a debate in the literature on the best method to forecast an aggregate: (1) forecast the aggregate directly, (2) forecast the disaggregates and then aggregate, or (3) forecast the aggregate using disaggregate information. This paper contributes to this debate by suggesting that in the presence of moderate-sized structural breaks in the disaggregates, approach (2) is preferred because of the low power to detect mean shifts in the disaggregates using models of aggregates. In support of this approach are two exercises. First, a simple Monte Carlo study demonstrates theoretical forecasting improvements. Second, empirical evidence is given using pseudo-ex ante forecasts of aggregate proven oil reserves in the United States.
    Keywords: Model selection; Intercept correction; Forecast robustification
    JEL: C52 C53 Q3
    Date: 2015–07
  9. By: Joseph P. Byrne; Shuo Cao.; Dimitris Korobilis.
    Abstract: This paper extends the Nelson-Siegel linear factor model by developing a flexible macro-finance framework for modeling and forecasting the term structure of US interest rates. Our approach is robust to parameter uncertainty and structural change, as we consider instabilities in parameters and volatilities, and our model averaging method allows for investors’ model uncertainty over time. Our time-varying parameter Nelson-Siegel Dynamic Model Averaging (NS-DMA) predicts yields better than standard benchmarks and successfully captures plausible time-varying term premia in real time. The proposed model has significant in-sample and out-of-sample predictability for excess bond returns, and the predictability is of economic value.
    Keywords: Term Structure of Interest Rates; Nelson-Siegel; Dynamic Model Averaging; Bayesian Methods; Term Premia.
    JEL: C32 C52 E43 E47 G17
    Date: 2015–02
  10. By: Mario Cerrato; John Crosby; Minjoo Kim; Yang Zhao
    Abstract: We investigate the dynamic and asymmetric dependence structure between equity portfolios from the US and UK. We demonstrate the statistical significance of dynamic asymmetric copula models in modelling and forecasting market risk. First, we construct “high-minus-low" equity portfolios sorted on beta, coskewness, and cokurtosis. We find substantial evidence of dynamic and asymmetric de- pendence between characteristic-sorted portfolios. Second, we consider a dynamic asymmetric copula model by combining the generalized hyperbolic skewed t copula with the generalized autoregressive score (GAS) model to capture both the multivariate non-normality and the dynamic and asymmetric dependence between equity portfolios. We demonstrate its usefulness by evaluating the forecasting performance of Value-at-Risk and Expected Shortfall for the high-minus-low portfolios. From back- testing, we find consistent and robust evidence that our dynamic asymmetric copula model provides the most accurate forecasts, indicating the importance of incorporating the dynamic and asymmetric dependence structure in risk management.
    Keywords: asymmetry, tail dependence, dependence dynamics, dynamic skewed t copulas, VaR and ES forecasting
    JEL: C32 C53 G17 G32
    Date: 2015–02
  11. By: Giuseppe Bianchi; Tatiana Cesaroni; Ottavio Ricchi
    Abstract: In this paper we describe a pilot econometric model for the Italian State Budget Expenditures (ISBEM). In search for leading indicators, we consider a newly available data set of the Italian State Budget financial microdata at monthly frequency that we use to estimate and forecast annual budget data. Early work on the issue is encompassed with the provision of a dynamic multiple equations model for the budget cycle linking data coming various budget phases (i.e. appropriations, expenditures commitments and payments) and disaggregated by budget macro aggregates. The model, that consists of several “pseudo” behavioral equations and identities, can be used for simulation exercises as well as forecasting purposes.
    Keywords: Fiscal forecasting, budget state expenditures, intra annual cash data, econometric models.
    JEL: C53 E62 H50
    Date: 2015
  12. By: Benjamin Beckers
    Abstract: This paper contributes to the debate of whether central banks can \lean against the wind" of emerging stock or house price bubbles. Against this background, the paper evaluates if new advances in real-time bubble detection, as brought forward by Phillips et al. (2011), can timely detect bubble emergences and collapses. Building on simulations, the paper shows that the detection capabilities of all indicators are sensitive to their exact specifications and to the characteristics of the bubbles in the sample. Therefore, the paper suggests a combination approach of different bubble indicators which helps to account for the uncertainty around start and end dates of asset price bubbles. Additionally, the paper then investigates if the individual and combination indicators carry predictive content for inflation and output growth when the real-time availability of all variables is taken into account. It finds that a combination indicator is best suited to uncover the most common stock and house price bubbles in the U.S. and shows that this indicator improves output forecasts.
    Keywords: Asset price bubbles, financial stability, leaning-against-the-wind, monetary policy, real-time forecasting, unit root monitoring test
    JEL: C22 C53 E44 E47 G12
    Date: 2015
  13. By: Takuya Iwasaki (Kansai University); Norio Kitagawa (Kobe University); Akinobu Shuto (The University of Tokyo)
    Abstract: The main purpose of this study is to investigate managerial discretion over managers’initial management forecasts issued concurrently with earnings announcements. The unique reporting system for management forecasts in Japan, systematic bundled management forecasts, creates an earnings benchmark (i.e., forecast innovations) to which earnings management research has not paid much attention. This study investigates whether and why firm managers manage their initial forecasts to avoid negative forecast innovations. First, we find that managers engage in forecast management through discretionary forecasts to avoid negative forecast innovations. Second, we reveal that 1) firms that avoid negative forecast innovations enjoy a higher return even when they use discretionary forecasts to do so and that 2) the relation between forecast innovations and return is S-shaped. These results suggest that the market rewards firms that achieve a forecast innovation benchmark, providing a sound rationale for managers’ use of forecast management. Finally, our additional analyses suggest that managers are not likely to conduct forecast management to convey their private information on future performance to investors, suggesting opportunistic managerial behaviors concerning their earnings forecasts.
    Date: 2015–07
  14. By: Dimitris Korobilis.
    Abstract: There is a vast literature that speciÖes Bayesian shrinkage priors for vector autoregressions (VARs) of possibly large dimensions. In this paper I argue that many of these priors are not appropriate for multi-country settings, which motivates me to develop priors for panel VARs (PVARs). The parametric and semi-parametric priors I suggest not only perform valuable shrinkage in large dimensions, but also allow for soft clustering of variables or countries which are homogeneous. I discuss the implications of these new priors for modelling interdependencies and heterogeneities among di§erent countries in a panel VAR setting. Monte Carlo evidence and an empirical forecasting exercise show clear and important gains of the new priors compared to existing popular priors for VARs and PVARs.
    Keywords: Bayesian model selection; shrinkage; spike and slab priors; forecasting; large vector autoregression
    JEL: C11 C33 C52
    Date: 2015–04
  15. By: Ravazzolo, Francesco (Norges Bank); Vespignani, Joaquin L. (Unversity of Tasmania)
    Abstract: In modelling macroeconomic time series, often a monthly indicator of global real economic activity is used. We propose a new indicator, named World steel production, and compare it to other existing indicators, precisely the Kilian’s index of global real economic activity and the index of OECD World industrial production. We develop an econometric approach based on desirable econometric properties in relation to the quarterly measure of World or global gross domestic product to evaluate and to choose across different alternatives. The method is designed to evaluate short-term, long-term and predictability properties of the indicators. World steel production is proven to be the best monthly indicator of global economic activity in terms of our econometric properties. Kilian’s index of global real economic activity also accurately predicts World GDP growth rates. When extending the analysis to an out-of-sample exercise, both Kilian’s index of global real economic activity and the World steel production produce accurate forecasts for World GDP, confirming evidence provided by the econometric properties. Specifically, a forecast combination of the three indices produces statistically significant gains up to 40% at nowcast and more than 10% at longer horizons relative to an autoregressive benchmark.
    JEL: C1 C5 C8 E1 E3
    Date: 2015–06–01
  16. By: Deepa (Deakin University); Paresh K Narayan (Deakin University)
    Abstract: In this paper we show that Indian stock returns, based on industry portfolios, portfolios sorted on book-to-market, and on size, are predictable. While we discover that this predictability holds both in in-sample and out-of-sample tests, predictability is not homogenous. Some predictors are important than others and some industries and portfolios of stocks are more predictable and, therefore, more profitable than others. We also discover that a mean combination forecast approach delivers significant out-of-sample performance. Our results survive a battery of robustness tests.
    Keywords: Stock Returns; Predictability; Profits; Sectors; Rational asset pricing; India.
  17. By: Dinh H B Phan (Deakin University); Susan S Sharma (Deakin University); Paresh K Narayan (Deakin University)
    Abstract: This paper investigates the price volatility interaction between the crude oil and equity markets in the US using five-minute data over the period 2009 to 2012. Our main findings can be summarised as follows. First, we find strong evidence to demonstrate that the integration of the bid-ask spread and trading volume factors leads to a better performance in predicting price volatility. Second, trading information, such as bid-ask spread, trading volume, and the price volatility from cross-markets, improves the price volatility predictability for both in-sample and out-of-sample analyses. Third, the trading strategy based on the predictive regression model that includes trading information from both markets provides significant utility gains to mean-variance investors.
    Keywords: volatility; trading volume; bid-ask spread; cross-market; predictability; forecasting.
  18. By: Paresh K Narayan (Deakin University); Rangan Gupta (University of Pretoria)
    Abstract: This paper contributes to the debate on the role of oil prices in predicting stock returns. The novelty of the paper is that it considers monthly time-series historical data that span over 150 years (1859:10-2013:12) and applies a predictive regression model that accommodates three salient features of the data, namely, a persistent and endogenous oil price, and model heteroskedasticity. Three key findings are unraveled: First, oil price predicts US stock returns. Second, in-sample evidence is corroborated by out-sample evidence of predictability. Third, both positive and negative oil price changes are important predictors of US stock returns, with negative changes relatively more important. Our results are robust to the use of different estimators and choice of in-sample periods.
    Keywords: Stock returns; Predictability; Oil price.
  19. By: Shota Otomasa (Kansai University); Atsushi Shiiba (Osaka University); Akinobu Shuto (The University of Tokyo)
    Abstract: This paper investigates whether and how Japanese firms use management earnings forecasts as a performance target for determining executive cash compensation. Consistent with the implications of the principal?agent theory, we find that the sensitivity of executive cash compensation varies with the extent to which realized earnings exceed initial management forecasts. In particular, we find that, for a sample of Japanese firms comprising 14,899 firm-year observations from 2005 to 2012, the executive cash compensation is positively related to management forecast error (MFE). Moreover, we show that the relationship between executive cash compensation and MFE strengthens when realizing positive MFEs despite aggressive initial forecasts. Overall, we find that initial management forecasts can be used as a performance target in executive compensation contracts. These findings also suggest that management earnings forecasts are important for improving contract efficiency as well as for providing useful information to investors in the capital market.
    Date: 2015–07
  20. By: Janine Aron; John Muellbauer; Rachel Sebudde
    Abstract: Forecasting inflation is challenging in emerging markets, where trade and monetary regimes have shifted, and the exchange rate, energy and food prices are highly volatile. Mobile money is a recent financial innovation giving financial transaction services via a mobile phone, including to the unbanked. Stable models for the 1-month and 3-month-ahead rates of inflation in Uganda, measured by the consumer price index for food and non-food, and for the domestic fuel price, are estimated over 1994-2013. Key features are the use of multivariate models with equilibrium-correction terms in relative prices; introducing non-linearities to proxy state dependence in the inflation process; and applying a ‘parsimonious longer lags’ (PLL) parameterisation to feature lags up to 12 months. International influences through foreign prices and the exchange rate (including food prices in Kenya after regional integration) have an important influence on the dependent variables, as does the growth of domestic credit. Rainfall deviation from the long-run mean is an important driver for all, most dramatically for food. The domestic money stock is irrelevant for food and fuel inflation, but has a small effect on non-food inflation. Other drivers include the trade and current account balances, fiscal balance, terms of trade and trade openness, and the international interest rate differential. Parameter stability tests suggest the models could be useful for short-term forecasting of inflation. There is no serious evidence of a link between mobile money and inflation.
    Keywords: Error Correction Models; Model Selection; Multivariate Time Series
    JEL: E31 E37 E52 C22 C51 C52 C53
    Date: 2015
  21. By: Joakim Westerlund (Deakin University); Paresh K Narayan (Deakin University); Xinwei Zheng (Deakin University)
    Abstract: This paper proposes a simple panel data test for stock return predictability that is flexible enough to accommodate three key salient features of the data, namely, predictor persistency and endogeneity, and cross-sectional dependence. Using a large panel of Chinese stock market data comprising more than one million observations, we show that most financial and macroeconomic predictors are in fact able to predict returns. We also show how the extent of the predictability varies across industries and firm sizes.
    Keywords: Panel data; Bias; Cross-section dependence; Predictive regression; Stock return predictability; China.
    JEL: C22 C23 G1 G12
  22. By: Paresh K Narayan (Deakin University)
    Abstract: In this paper, we find that CDS return shocks are important in explaining the forecast error variance of sectoral equity returns for the USA. The CDS return shocks have different effects on equity returns and return volatility in the pre-crisis and crisis periods. It is the post-Lehman crisis period in which the effects of CDS return shocks are the most dominant. Finally, we construct a spillover index and find that it is time-varying and explains a larger share of total forecast error variance of sectoral equity and CDS returns for some sectors than for others.
    Keywords: Equity Returns; CDS Spread; Forecast Error Variance; Spillover.

This nep-for issue is ©2015 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.