Surveillance of the mean behavior of multivariate time series

In this paper several cumulative sum (CUSUM) charts for the mean of a multivariate time series are introduced. We extend the control schemes for independent multivariate observations of crosier [ Technometrics (1988) Vol. 30, pp. 187–194], pignatiello and runger [ Journal of Quality Technology (1990) Vol. 22, pp. 173–186], and ngai and zhang [ Statistica Sinica (2001) Vol. 11, pp. 747–766] to multivariate time series by taking into account the probability structure of the underlying stochastic process. We consider modified charts and residual schemes as well. It is analyzed under which conditions these charts are directionally invariant. In an extensive Monte Carlo study these charts are compared with the CUSUM scheme of theodossiu [ Journal of the American Statistical Association (1993) Vol. 88, pp. 441–448], the multivariate exponentially weighted moving-average (EWMA) chart of kramer and schmid [ Sequential Analysis (1997) Vol. 16, pp. 131–154], and the control procedures of bodnar and schmid [ Frontiers of Statistical Process Control (2006) Physica, Heidelberg]. As a measure of the performance, the maximum expected delay is used.     

Graphical interaction models for multivariate time series1

In this paper we extend the concept of graphical models for multivariate data to multivariate time series. We define a partial correlation graph for time series and use the partial spectral coherence between two components given the remaining components to identify the edges of the graph. As an example we consider multivariate autoregressive processes. The method is applied to air pollution data. Received: June 1999     

Testing of unit root and other nonstationary hypotheses in macroeconomic time series

Recently proposed tests for unit root and other nonstationarity of Robinson (1994a) are applied to an extended version of the data set used by Nelson and Plosser (1982). Unusually, the tests are efficient (against appropriate parametric alternatives), the null can be any member of the I(d) class, and the null limit distribution is chi-squared. The conclusions vary substantially across the 14 series, and across different models for the disturbances (which, also unusually, include the Bloomfield spectral model). Overall, the consumer price index and money stock seem the most nonstationary, while industrial production and unemployment rate seem the closest to stationarity.     

Boosting nonlinear predictability of macroeconomic time series

We apply the boosting estimation method in order to investigate to what extent and at what horizons macroeconomic time series have nonlinear predictability that comes from their own history. Our results indicate that the U.S. macroeconomic time series have more exploitable nonlinear predictability than previous studies have found. On average, the most favorable out-of-sample performance is obtained via a two-stage procedure, where a conventional linear prediction model is fitted first and the boosting technique is applied to build a nonlinear model for its residuals.     

Bayesian inference in time series models using kernel quasi likelihoods

The paper takes up Bayesian inference in time series models when essentially nothing is known about the distribution of the dependent variable given past realizations or other covariates. It proposes the use of kernel quasi likelihoods upon which formal inference can be based. Gibbs sampling with data augmentation is used to perform the computations related to numerical Bayesian analysis of the model. The method is illustrated with artificial and real data sets.     

Monitoring the cross-covariances of a multivariate time series

In this paper sequential procedures are proposed for jointly monitoring all elements of the covariance matrix at lag 0 of a multivariate time series. All control charts are based on exponential smoothing. As a measure of the distance between the target values and the actual values the Mahalanobis distance is used. It is distinguished between residual control schemes and modified control schemes. Several properties of these charts are proved assuming the target process to be a stationary Gaussian process. Within an extensive Monte Carlo study all procedures are compared with each other. As a measure of the performance of a control chart the average run length is used. An empirical example about Eastern European stock markets illustrates how the autocovariance and the cross-covariance structure of financial assets can be monitored by these methods.     

Forecasting the NN5 time series with hybrid models

We propose a simple way of predicting time series with recurring seasonal periods. Missing values of the time series are estimated and interpolated in a preprocessing step. We combine several forecasting methods by taking the weighted mean of forecasts that were generated with time-domain models which were validated on left-out parts of the time series. The hybrid model is a combination of a neural network ensemble, an ensemble of nearest trajectory models and a model for the 7-day cycle. We apply this approach to the NN5 time series competition data set.     

社会治理支出与经济增长：抑制还是促进？

本文基于1990—2012时间序列数据,对社会治理支出与经济增长进行 Johansen 协整和 Granger 因果检验,并进行了脉冲响应分析和方差分解。为了弥补时间序列数据只包含时间和指标两维信息的缺陷,进一步基于2000—2012年中国31省域面板数据,对二者进行了面板数据单位根检验、协整检验和固定效应变系数模型估计分析。研究结果表明,社会治理支出会促进经济增长,而经济增长对社会治理支出促进作用十分有限,各省域社会治理支出对经济增长促进作用不尽相同。     

Forecasting correlated time series with exponential smoothing models

This paper presents the Bayesian analysis of a general multivariate exponential smoothing model that allows us to forecast time series jointly, subject to correlated random disturbances. The general multivariate model, which can be formulated as a seemingly unrelated regression model, includes the previously studied homogeneous multivariate Holt-Winters’ model as a special case when all of the univariate series share a common structure. MCMC simulation techniques are required in order to approach the non-analytically tractable posterior distribution of the model parameters. The predictive distribution is then estimated using Monte Carlo integration. A Bayesian model selection criterion is introduced into the forecasting scheme for selecting the most adequate multivariate model for describing the behaviour of the time series under study. The forecasting performance of this procedure is tested using some real examples.     

Prediction intervals in conditionally heteroscedastic time series with stochastic components

Differencing is a very popular stationary transformation for series with stochastic trends. Moreover, when the differenced series is heteroscedastic, authors commonly model it using an ARMA-GARCH model. The corresponding ARIMA-GARCH model is then used to forecast future values of the original series. However, the heteroscedasticity observed in the stationary transformation should be generated by the transitory and/or the long-run component of the original data. In the former case, the shocks to the variance are transitory and the prediction intervals should converge to homoscedastic intervals with the prediction horizon. We show that, in this case, the prediction intervals constructed from the ARIMA-GARCH models could be inadequate because they never converge to homoscedastic intervals. All of the results are illustrated using simulated and real time series with stochastic levels.     

A two–stage approach to additive time series models

For nonlinear additive time series models, an appealing approach used in the literature to estimate the nonparametric additive components is the projection method. In this paper, it is demonstrated that the projection method might not be efficient in an asymptotic sense. To estimate additive components efficiently, a two–stage approach is proposed together with a local linear fitting and a new bandwidth selector based on the nonparametric version of the Akaike information criterion. It is shown that the two–stage method not only achieves efficiency but also makes bandwidth selection relatively easier. Also, the asymptotic normality of the resulting estimator is established. A small simulation study is carried out to illustrate the proposed methodology and the two–stage approach is applied to a real example from econometrics.     

Automatic neural network modeling for univariate time series

Artificial neural networks (ANNs) are an information processing paradigm inspired by the way the brain processes information. Using neural networks requires the investigator to make decisions concerning the architecture or structure used. ANNs are known to be universal function approximators and are capable of exploiting nonlinear relationships between variables. This method, called Automated ANNs, is an attempt to develop an automatic procedure for selecting the architecture of an artificial neural network for forecasting purposes. It was entered into the M-3 Time Series Competition. Results show that ANNs compete well with the other methods investigated, but may produce poor results if used under certain conditions.     

Numerical issues in threshold autoregressive modeling of time series

This paper analyses the contribution of various numerical approaches to making the estimation of threshold autoregressive time series more efficient. It relies on the computational advantages of QR factorizations and proposes Givens transformations to update these factors for sequential LS problems. By showing that the residual sum of squares is a continuous rational function over threshold intervals it develops a new fitting method based on rational interpolation and the standard necessary optimality condition. Taking as benchmark a simple grid search, the paper illustrates via Monte Carlo simulations the efficiency gains of the proposed tools.     

Consistent estimation of a general nonparametric regression function in time series

We propose an estimator of the conditional distribution of X_t|X_t−1,X_t−2,…, and the corresponding regression function , where the conditioning set is of infinite order. We establish consistency of our estimator under stationarity and ergodicity conditions plus a mild smoothness condition.     

Sensitivity to autocorrelation in judgmental time series forecasting

How well can people use autocorrelation information when making judgmental forecasts? In Experiment 1, participants forecast from 12 series in which the autocorrelation varied within subjects. The participants showed a sensitivity to the degree of autocorrelation. However, their forecasts indicated that they implicitly assumed positive autocorrelation in uncorrelated time series. Experiments 2 and 2a used a one-shot single-trial between-subjects design and obtained similar results. Experiment 3 investigated the way in which the between-trials context influenced forecasting. The results showed that forecasts are affected by the characteristics of previous series, as well as those of the series from which forecasts are to be made. Our findings can be accommodated within an adaptive approach. Forecasters base their initial expectations of series characteristics on their past experience and modify these expectations in a pseudo-Bayesian manner on the basis of their analysis of those characteristics in the series to be forecast.     

Investigating the accuracy of cross-learning time series forecasting methods

The M4 competition identified innovative forecasting methods, advancing the theory and practice of forecasting. One of the most promising innovations of M4 was the utilization of cross-learning approaches that allow models to learn from multiple series how to accurately predict individual ones. In this paper, we investigate the potential of cross-learning by developing various neural network models that adopt such an approach, and we compare their accuracy to that of traditional models that are trained in a series-by-series fashion. Our empirical evaluation, which is based on the M4 monthly data, confirms that cross-learning is a promising alternative to traditional forecasting, at least when appropriate strategies for extracting information from large, diverse time series data sets are considered. Ways of combining traditional with cross-learning methods are also examined in order to initiate further research in the field.     

The asymmetry of judgemental confidence intervals in time series forecasting

In forecasting a time series, one may be asked to communicate the likely distribution of the future actual value, often expressed as a confidence interval. Whilst the accuracy (calibration) of these intervals has dominated most studies to date, this paper is concerned with other possible characteristics of the intervals. It reports a study in which the prevalence and determinants of the symmetry of judgemental confidence intervals in time series forecasting was examined. Most prior work has assumed that this interval is symmetrically placed around the forecast. However, this study shows that people generally estimate asymmetric confidence intervals where the forecast is not the midpoint of the estimated interval. Many of these intervals are grossly asymmetric. Results indicate that the placement of the forecast in relation to the last actual value of a time series is a major determinant of the direction and size of the asymmetry.     

Measuring and forecasting retail trade in real time using card transactional data

We build big data retail trade indicators for Spain using high-dimensional card transaction data from one of the country’s biggest banks. The resulting indicators replicate the dynamics of the Spanish retail trade indices (RTI), regional RTIs (Spain’s autonomous regions), and RTI by retailer type (distribution classes) released by the Spanish National Statistics Institute. The new indicators not only have a higher frequency (daily data) and higher geographical and sectorial breakdown but are also shown to improve nowcasting and forecasting power for the official RTI, making them key variables to monitor consumption.     

On forecasting cointegrated seasonal time series

We analyze periodic and seasonal cointegration models for bivariate quarterly observed time series in an empirical forecasting study. We include both single equation and multiple equation methods for those two classes of models. A VAR model in first differences, with and without cointegration restrictions, and a VAR model in annual differences are also included in the analysis, where they serve as benchmark models. Our empirical results indicate that the VAR model in first differences without cointegration is best if one-step ahead forecasts are considered. For longer forecast horizons however, the VAR model in annual differences is better. When comparing periodic versus seasonal cointegration models, we find that the seasonal cointegration models tend to yield better forecasts. Finally, there is no clear indication that multiple equations methods improve on single equation methods.