Forecasting inflation through a bottom-up approach: How bottom is bottom?

The aim of this paper is to assess inflation forecast accuracy over the short-term horizon, using Consumer Price Index (CPI) disaggregated data, through a bottom-up approach. That is, aggregating forecasts is compared with aggregate forecasting. A new dimension to the question of to bottom-up or not is introduced by considering different levels of data disaggregation, namely a higher disaggregation level than the one considered up to now. This raises modelling issues that one has to cope with. In particular, it is suggested the use of a new strand of models, the Factor-Augmented SARIMA models. Considering as case-study the Portuguese one, we find an inverse relationship between the forecast horizon and the amount of information underlying the forecast, when minimizing the RMSFE.     

基于ARIMA与GARCH模型的国际油价预测比较分析

在分析影响油价波动因素的基础上,利用1986年1月至2010年12月的WTI国际原油价格月度数据,分别建立ARIMA和GARCH模型对油价进行预测。并通过对2011年1月至2012年4月WTI原油价格进行外推预测,检验模型的预测效果。比较分析发现,在短期预测中,ARIMA和GARCH模型对油价的预测均比较准确,但当油价由于受到重大事件的影响而有较大波动时,模型的预测精度下降;在长期预测中,GARCH模型的预测效果优于ARIMA模型;整体来看,GARCH模型预测的精度高于ARIMA模型。因此,在国际油价预测中,用GARCH模型是比较合适的。     

Do disaggregated CPI data improve the accuracy of inflation forecasts?

In this paper, we evaluate the role of using consumer price index (CPI) disaggregated data to improve the accuracy of inflation forecasts. Our forecasting approach is based on extracting the factors from the subcomponents of the CPI at the highest degree of disaggregation. The data set contains 54 macroeconomic series and 243 CPI subcomponents from 1992 to 2009 for Mexico. We find that the factor models that include disaggregated data outperform the benchmark autoregressive model and the factor models containing alternative groups of macroeconomic variables. We provide evidence that using disaggregated price data improves forecasting performance. The forecasts of the factor models that extract the information from the CPI disaggregated data are as accurate as the forecasts from the survey of experts.     

Combination forecasts of tourism demand with machine learning models

The main objective of this study is to analyse whether the combination of regional predictions generated with machine learning (ML) models leads to improved forecast accuracy. With this aim, we construct one set of forecasts by estimating models on the aggregate series, another set by using the same models to forecast the individual series prior to aggregation, and then we compare the accuracy of both approaches. We use three ML techniques: support vector regression, Gaussian process regression and neural network models. We use an autoregressive moving average model as a benchmark. We find that ML methods improve their forecasting performance with respect to the benchmark as forecast horizons increase, suggesting the suitability of these techniques for mid- and long-term forecasting. In spite of the fact that the disaggregated approach yields more accurate predictions, the improvement over the benchmark occurs for shorter forecast horizons with the direct approach.     

Does the vector error correction model perform better than others in forecasting stock price? An application of residual income valuation theory

This paper employs a multi-equation model approach to consider three statistic problems (heteroskedasticity, endogeneity and persistency), which are sources of bias and inefficiency in the predictive regression models. This paper applied the residual income valuation model (RIM) proposed by Ohlson (1995) to forecast stock prices for Taiwan three sectors. We compare relative forecasting accuracy of vector error correction model (VECM) with the vector autoregressive model (VAR) as well as OLS and RW models used in the prior studies. We conduct out-of-sample forecasting and employ two instruments to assess forecasting performance. Our empirical results suggest that the VECM statistically outperforms other three models in forecasting stock prices. When forecasting horizons extend longer, VECM produces smaller forecasting errors and performs substantially better than VAR, suggesting that the ability of VECM to improve VAR forecast accuracy is stronger with longer horizons. These findings imply that an error correction term (ECT) of the VECM contributes to improving forecast accuracy of stock prices. Our economic significance analyses and robustness tests for different data frequency are in favor of the superiority of VECM estimator.     

Forecasting house prices for the four census regions and the aggregate US economy in a data-rich environment

This article considers the ability of large-scale (involving 145 fundamental variables) time-series models, estimated by dynamic factor analysis and Bayesian shrinkage, to forecast real house price growth rates of the four US census regions and the aggregate US economy. Besides the standard Minnesota prior, we also use additional priors that constrain the sum of coefficients of the VAR models. We compare 1- to 24-months-ahead forecasts of the large-scale models over an out-of-sample horizon of 1995:01–2009:03, based on an in-sample of 1968:02–1994:12, relative to a random walk model, a small-scale VAR model comprising just the five real house price growth rates and a medium-scale VAR model containing 36 of the 145 fundamental variables besides the five real house price growth rates. In addition to the forecast comparison exercise across small-, medium- and large-scale models, we also look at the ability of the ‘optimal’ model (i.e. the model that produces the minimum average mean squared forecast error) for a specific region in predicting ex ante real house prices (in levels) over the period of 2009:04 till 2012:02. Factor-based models (classical or Bayesian) perform the best for the North East, Mid-West, West census regions and the aggregate US economy and equally well to a small-scale VAR for the South region. The ‘optimal’ factor models also tend to predict the downward trend in the data when we conduct an ex ante forecasting exercise. Our results highlight the importance of information content in large number of fundamentals in predicting house prices accurately.     

Labour market forecasting in Germany: is disaggregation useful?

This paper examines whether labour market forecasts can be improved by using disaggregated information. We construct vector-autoregressive models for employment by sector in order to produce out-of-sample forecasts of aggregate employment. Forecast accuracy is compared to univariate models by using Clark/West tests. In an application to German data, it is evident that disaggregation significantly improves the employment forecast. Moreover, using fluctuation-window tests we find that disaggregation yields superior results especially in phases with strong and sustained employment changes.     

Forecasting the US real house price index: Structural and non-structural models with and without fundamentals

We employ a 10-variable dynamic structural general equilibrium model to forecast the US real house price index as well as its downturn in 2006:Q2. We also examine various Bayesian and classical time-series models in our forecasting exercise to compare to the dynamic stochastic general equilibrium model, estimated using Bayesian methods. In addition to standard vector-autoregressive and Bayesian vector autoregressive models, we also include the information content of either 10 or 120 quarterly series in some models to capture the influence of fundamentals. We consider two approaches for including information from large data sets — extracting common factors (principle components) in factor-augmented vector autoregressive or Bayesian factor-augmented vector autoregressive models as well as Bayesian shrinkage in a large-scale Bayesian vector autoregressive model. We compare the out-of-sample forecast performance of the alternative models, using the average root mean squared error for the forecasts. We find that the small-scale Bayesian-shrinkage model (10 variables) outperforms the other models, including the large-scale Bayesian-shrinkage model (120 variables). In addition, when we use simple average forecast combinations, the combination forecast using the 10 best atheoretical models produces the minimum RMSEs compared to each of the individual models, followed closely by the combination forecast using the 10 atheoretical models and the DSGE model. Finally, we use each model to forecast the downturn point in 2006:Q2, using the estimated model through 2005:Q2. Only the dynamic stochastic general equilibrium model actually forecasts a downturn with any accuracy, suggesting that forward-looking microfounded dynamic stochastic general equilibrium models of the housing market may prove crucial in forecasting turning points.     

Bottom-up or direct? Forecasting German GDP in a data-rich environment

In this paper, we investigate whether there are benefits in disaggregating GDP into its components when nowcasting GDP. To answer this question, we conduct a realistic out-of-sample experiment that deals with the most prominent problems in short-term forecasting: mixed frequencies, ragged-edge data, asynchronous data releases and a large set of potential information. We compare a direct leading indicator-based GDP forecast with two bottom-up procedures—that is, forecasting GDP components from the production side or from the demand side. Generally, we find that the direct forecast performs relatively well. Among the disaggregated procedures, the production side seems to be better suited than the demand side to form a disaggregated GDP nowcast.     

The law of one price and the modelling of disaggregated trade flows

This paper argues that the application of a differentiated good model to disaggregated commodity trade, and in particular to primary commodity trade, is feasible. Price data for as narrowly defined a commodity as wheat are shown to violate the law of one price. An Armington-type model which allows prices of a commodity to vary by supplier is then applied to world wheat trade. Two issues concerning the Armington model are addressed. First, it is shown that the theoretical implications of the model are plausible in the case of a disaggregated commodity. Second, as an example, 1973–1974 wheat trade flows and prices forecast by an Armington-type model are shown to be consistent with actual trade patterns and prices.     

Forecasting US consumer price index: does nonlinearity matter?

The objective of this article is to predict, both in sample and out of sample, the consumer price index (CPI) of the US economy based on monthly data covering the period of 1980:1–2013:12, using a variety of linear (random walk (RW), autoregressive (AR) and seasonal autoregressive integrated moving average (SARIMA)) and nonlinear (artificial neural network (ANN) and genetic programming (GP)) univariate models. Our results show that, while the SARIMA model is superior relative to other linear and nonlinear models, as it tends to produce smaller forecast errors; statistically, these forecasting gains are not significant relative to higher-order AR and nonlinear models, though simple benchmarks like the RW and AR(1) models are statistically outperformed. Overall, we show that in terms of forecasting the US CPI, accounting for nonlinearity does not necessarily provide us with any statistical gains.     

Macroeconomic expectations and time varying heterogeneity:evidence from individual survey data

ABSTRACT

The goal of this paper is to investigate forecast heterogeneity and time variability in the formation of expectations using disaggregated monthly survey data on macroeconomic indicators provided by Bloomberg from June 1998 to August 2017. We show that our panel of forecasters are not rational and are moderately heterogeneous and thus confirm that previously well-established results on asset prices hold for macroeconomic indicators. We propose a flexible hybrid forecast model defined at any time as a combination of the extrapolative, regressive, adaptive and interactive heuristics. Controlling for endogenous structural breaks, we find that experts adjust their forecast behaviour at any time with some inertia in extrapolative and adaptive profiles. Changes in the formation of expectations are triggered mostly by financial shocks, and uncertainty is dealt with by using complex processes in which the fundamentalist component overweighs chartist activity. Forecasters whose models combine different relevant rules and display high temporal flexibility provide the most accurate forecasts. Authorities can then stabilize the domestic markets by encouraging fundamentalists’ forecasts through increased transparency policy.     

Constructing a new leading indicator for unemployment from a survey among German employment agencies

The article investigates the predictive power of a new survey implemented by the Federal Employment Agency (FEA) for forecasting German unemployment in the short run. Every month, the CEOs of the FEA’s regional agencies are asked about their expectations of future labour market developments. We generate an aggregate unemployment leading indicator that exploits serial correlation in response behaviour through identifying and adjusting temporarily unreliable predictions. We use out-of-sample tests suitable in nested model environments to compare forecasting performance of models including the new indicator to that of purely autoregressive benchmarks. For all investigated forecast horizons (1, 2, 3 and 6 months), test results show that models enhanced by the new leading indicator significantly outperform their benchmark counterparts. To compare our indicator to potential competitors, we employ the model confidence set. Results reveal that models including the new indicator perform very well at the 10% level.     

A boosting approach to forecasting gold and silver returns: economic and statistical forecast evaluation

We use a boosting algorithm to forecast the returns of gold and silver prices. We then study the implications of using different information criteria to terminate the boosting algorithm in terms of the statistical and economic performance of a forecasting model. Our findings demonstrate that information criteria that select parsimonious forecasting models perform better in statistical terms than information criteria that select relatively complex forecasting models, but this good performance does not necessarily survive an economic performance evaluation.     

Forecasting demand for a newly introduced product using reservation price data and Bayesian updating

Forecasting demand during the early stages of a product's life cycle is a difficult but essential task for the purposes of marketing and policymaking. This paper introduces a procedure to derive accurate forecasts for newly introduced products for which limited data are available. We begin with the assumption that the consumer reservation price is related to the timing with which the consumer adopts the product. The model is estimated using reservation price data derived through a consumer survey, and the forecast is updated with sales data as they become available using Bayes's rule. The proposed model's forecasting performance is compared with that of benchmark models (i.e., Bass model, logistic growth model, and a Bayesian model based on analogy) using 23 quarters' worth of data on South Korea's broadband Internet services market. The proposed model outperforms all benchmark models in both prelaunch and postlaunch forecasting tests, supporting the thesis that consumer reservation price can be used to forecast demand for a new product before or shortly after product launch.     

Forecasting house prices using dynamic model averaging approach: Evidence from China

Forecasting house price has been of great interests for macroeconomists, policy makers and investors in recent years. To improve the forecasting accuracy, this paper introduces a dynamic model averaging (DMA) method to forecast the growth rate of house prices in 30 major Chinese cities. The advantage of DMA is that this method allows both the sets of predictors (forecasting models) as well as their coefficients to change over time. Both recursive and rolling forecasting modes are applied to compare the performance of DMA with other traditional forecasting models. Furthermore, a model confidence set (MCS) test is used to statistically evaluate the forecasting efficiency of different models. The empirical results reveal that DMA generally outperforms other models, such as Bayesian model averaging (BMA), information-theoretic model averaging (ITMA) and equal-weighted averaging (EW), in both recursive and rolling forecasting modes. In addition, in recent years it is found that the Google search index, instead of fundamental macroeconomic or monetary indicators, has developed greater predictive power for house price in China.     

On winning forecasting competitions in economics

To explain which methods might win forecasting competitions on economic time series, we consider forecasting in an evolving economy subject to structural breaks, using mis-specified, data-based models. ‘Causal’ models need not win when facing deterministic shifts, a primary factor underlying systematic forecast failure. We derive conditional forecast biases and unconditional (asymptotic) variances to show that when the forecast evaluation sample includes sub-periods following breaks, non-causal models will outperform at short horizons. This suggests using techniques which avoid systematic forecasting errors, including improved intercept corrections. An application to a small monetary model of the UK illustrates the theory.     

Think national,forecast local: a case study of 71 German urban housing markets

In this article, we examine whether the local indicators are able to predict the city-level housing prices and rents better than national indicators. For this purpose, we assess the forecasting ability of 126 indicators and 21 types of forecast combinations using a sample of 71 large German cities. There are several predictors that are especially useful, namely price-to-rent ratios, national-level business confidence, and consumer surveys. We also find that combinations of individual forecasts are among the top forecasting models. On average, the forecast improvements attain about 20%, measured by a reduction in root mean square error, compared to the naive models.     

Learning,forecasting and optimizing: An experimental study

Rational Expectations (RE) models have two crucial dimensions: (i) agents on average correctly forecast future prices given all available information, and (ii) given expectations, agents solve optimization problems and these solutions in turn determine actual price realizations. Experimental tests of such models typically focus on only one of these two dimensions. In this paper we consider both forecasting and optimization decisions in an experimental cobweb economy. We report results from four main experimental treatments: (1) subjects form forecasts only, (2) subjects determine quantity only (solve an optimization problem), (3) they do both and (4) they are paired in teams and one member is assigned the forecasting role while the other is assigned the optimization task. All treatments converge to Rational Expectation Equilibrium (REE), but at different speeds. We observe that performance is the best in treatment 1 and the worst in Treatment 3. We further find that most subjects use adaptive rules to forecast prices. Given a price forecast, subjects are less likely to make conditionally optimal production decisions in Treatment 3 where the forecast is made by themselves, than in Treatment 4 where the forecast is made by the other member of their team, which suggests that “two heads are better than one” in term of the speed of finding the REE.