Forecasting papers 2009-01-26

In this issue we have: Business surveys and inflation forecasting in China ; Ripple Effects and Forecasting Home Prices ; Predicting Stock Volatility Using After-Hours Information ; Stochastic Dominance Approach to Evaluate Optimism Bias in Truck Toll Forecasts ; and more.

    Date: 2009-01
    By: Sonali Das (CSIR, Pretoria)
    Rangan Gupta (Department of Economics, University of Pretoria)
    Alain Kabundi (Department of Economics and Econometrics, University of Johannesburg)
    This paper analyzes whether a wealth of information contained in 126 monthly series used by large-scale Bayesian Vector Autoregressive (LBVAR) models, as well as Factor Augmented Vector Autoregressive (FAVAR) models, either Bayesian or classical, can prove to be more useful in forecasting real house price growth rate of the nine census divisions of the US, compared to the small-scale VAR models, that merely use the house prices. Using the period of 1991:02 to 2000:12 as the in-sample period and 2001:01 to 2005:06 as the out-of-sample horizon, we compare the forecast performance of the alternative models for one- to twelve-months ahead forecasts. Based on the average Root Mean Squared Error (RMSEs) for one- to twelve-months ahead forecasts, we find that the alternative FAVAR models outperform the other models in eight of the nine census divisions.
    Keywords: Dynamic Factor Model, BVAR, Forecast Accuracy
    JEL: C11 C13 C33 C53
  • Business surveys and inflation forecasting in China
    Date: 2009-01-13
    By: Kaaresvirta, Juuso (BOFIT)
    Mehrotra, Aaron (BOFIT)
    We use business survey data collected by the People's Bank of China for inflation forecasting. Some survey indicators lead to enhanced forecasting performance relative to the univariate benchmark model, especially for a period of moderate inflation. However, the estimated models do not do a good job of tracking the recent pickup in Chinese inflation, due to increases in food prices.
    Keywords: inflation forecasting; business surveys; China
    JEL: C53 E31
  • "Ripple Effects" and Forecasting Home Prices In Los Angeles, Las Vegas, and Phoenix
    Date: 2009-01
    By: Rangan Gupta (Department of Economic, University of Pretoria)
    Stephen M. Miller (College of Business, University of Las Vegas, Nevada)
    We examine the time-series relationship between housing prices in Los Angeles, Las Vegas, and Phoenix. First, temporal Granger causality tests reveal that Los Angeles housing prices cause housing prices in Las Vegas (directly) and Phoenix (indirectly). In addition, Las Vegas housing prices cause housing prices in Phoenix. Los Angeles housing prices prove exogenous in a temporal sense and Phoenix housing prices do not cause prices in the other two markets. Second, we calculate out-of-sample forecasts in each market, using various vector autoregessive (VAR) and vector error-correction (VEC) models, as well as Bayesian, spatial, and causality versions of these models with various priors. Different specifications provide superior forecasts in the different cities. Finally, we consider the ability of theses time-series models to provide accurate out-of-sample predictions of turning points in housing prices that occurre! d in 2006:Q4. Recursive forecasts, where the sample is updated each quarter, provide reasonably good forecasts of turning points.
    Keywords: Ripple effect, Housing prices, Forecasting
    JEL: C32 R31
  • Predicting Stock Volatility Using After-Hours Information
    Date: 2009-01
    By: Chun-Hung Chen (KPMG)
    Wei-Choun Yu (Winona State University)
    Eric Zivot (University of Washington)
    We use realized volatilities based on after hours high frequency returns to predict next day volatility. We extend GARCH and long-memory forecasting models to include additional information: the whole night, the preopen, the postclose realized variance, and the overnight squared return. For four NASDAQ stocks (MSFT, AMGN, CSCO, and YHOO) we find that the inclusion of the preopen variance can improve the out-of-sample forecastability of the next day conditional day volatility. Additionally, we find that the postclose variance and the overnight squared return do not provide any predictive power for the next day conditional volatility. Our findings support the results of prior studies that traders trade for non-information reasons in the postclose period and trade for information reasons in the preopen period.
  • Stochastic Dominance Approach to Evaluate Optimism Bias in Truck Toll Forecasts
    Date: 2008
    By: Sen Gupta , Rajorshi
    Vadali , Sharada R
    Optimism bias is a consistent feature associated with truck toll forecasts, à la Standard & Poor's and the NCHRP synthesis reports. Given the persistent problem, two major sources of this bias are explored. In particular, the ignorance of operating cost as a demand-side factor and lack of attention to user heterogeneity are found to contribute to this bias. To address it, stochastic dominance analysis is used to assess the risk associated with toll revenue forecasts. For a hypothetical corridor, it is shown that ignorance of operating cost savings can lead to upward bias in the threshold value of time distribution. Furthermore, dominance analysis demonstrates that there is greater risk associated with the revenue forecast when demand heterogeneity is factored in. The approach presented can be generally applied to all toll forecasts and is not restricted to trucks.
    Keywords: Forecast Bias; Operating costs; Risk assessment; Savings; Stochastic Dominance; Tolls;Trucks
    JEL: D81 C15 R41
    Date: 2008-01
    By: Wolfgang Polasek (IHS, Austria and The Rimini Centre of Economic Analisys, Italy)
    Richard Sellner (IHS, Austria)
    Completing data that are collected in disaggregated and heterogeneous spatial units is a quite frequent problem in spatial analyses of regional data. Chow and Lin (1971) (CL) were the rst to develop a uni ed framework for the three problems (interpolation, extrapolation and distribution) of predicting disaggregated times series by so-called indicator series. This paper develops a spatial CL procedure for disaggregating cross-sectional spatial data and compares the Maximum Likelihood and Bayesian spatial CL forecasts with the naive pro rata error distribution. We outline the error covariance structure in a spatial context, derive the BLUE for the ML estimator and the Bayesian estimation procedure by MCMC. Finally we apply the procedure to European regional GDP data and discuss the disaggregation assumptions. For the evaluation of the spatial Chow-Lin procedure we assume that only NUTS 1 GDP is known and predict it at NUTS 2 by using employment and spatial information available at NUTS 2. The spatial neighborhood is de ned by the inverse travel time by car in minutes. Finally, we present the forecast accuracy criteria comparing the predicted values with the actual observations.
  • Infrastructure for Sustainable Growth: A Demand Projection Exercise for India
    Date: 2008
    By: Majumder, Rajarshi
    Critical precondition for attaining growth and sustainable development is availability of a host of infrastructural facilities in adequate quantity and of reliable quality. The association between the latter and growth is well documented and a large number of theoretical propositions conclude that the association is quite strong and runs from the former to the latter. India, on attaining independence, accorded highest importance to the development of infrastructural facilities and lion's share of the plan outlays were on this sector. This resulted in a remarkable growth in such facilities. But the recent spurt in actual and target growth rates has been associated with substantial shortages in the physical availability of infrastructural facilities. To achieve and sustain the growth targets such shortages must be removed. This should start with determining the likely demand for these facilities both at current le! vels of economic intensity and at levels corresponding to desired growth rates. In this paper we seek to forecast the demand for selected infrastructural facilities for India over the next decade and a half so that we have an idea regarding the magnitude of the task facing the economy. In addition to projecting physical quantum of demand for those facilities, we also attempt at indicating the financial implications of realising those levels. The projected demand is substantially larger than the present availability and the task becomes harder as not only population will rise in future but the per capita demand would also increase. The Capacity Addition required would call in for huge investment amounting to a Capital outlay of 6-6.2 per cent of GDP for the five selected sectors only. One possible way to dent into this awesome job is to use a dual strategy. Along with heavy investment in creation of new physical stock of infrastructural facilities, one must also aim at impro! ving the utilization rate and operational efficiency of existing stock .
    Keywords: Infrastructure; Demand Projection; Sustainable Growth;
    JEL: H54 C53 C33 C21 O21
  • Bayesian posterior prediction and meta-analysis: an application to the value of travel time savings.
    Date: 2008-12-31
    By: Moral-Benito, Enrique
    In the evaluation of transportation infrastructure projects, some non-tradable goods such as time are usually key determinants of the result. However, obtaining monetary values for these goods is not always easy. This paper introduces a novel approach based on the combination of bayesian posterior prediction and meta-analysis. This methodology will allow to obtain predictive distributions of the monetary values for this type of goods. Therefore, uncertainty is formally considered in the analysis. Moreover, the proposed method is easy to apply and inexpensive both in terms of time and money. Finally, an application to the value of travel time savings is also presented.
    Keywords: Bayesian Prediction; Meta-Analysis; Uncertainty; Value of Travel Time Savings.
    JEL: L91 D61
  • Taken from the NEP-FOR mailing list edited by Rob Hyndman.