Smooth transition exponential smoothing

Adaptive exponential smoothing methods allow a smoothing parameter to change over time, in order to adapt to changes in the characteristics of the time series. However, these methods have tended to produce unstable forecasts and have performed poorly in empirical studies. This paper presents a new adaptive method, which enables a smoothing parameter to be modelled as a logistic function of a user‐specified variable. The approach is analogous to that used to model the time‐varying parameter in smooth transition models. Using simulated data, we show that the new approach has the potential to outperform existing adaptive methods and constant parameter methods when the estimation and evaluation samples both contain a level shift or both contain an outlier. An empirical study, using the monthly time series from the M3‐Competition, gave encouraging results for the new approach. Copyright © 2004 John Wiley & Sons, Ltd.     

Robust forecasting with exponential and Holt–Winters smoothing

Robust versions of the exponential and Holt–Winters smoothing method for forecasting are presented. They are suitable for forecasting univariate time series in the presence of outliers. The robust exponential and Holt–Winters smoothing methods are presented as recursive updating schemes that apply the standard technique to pre‐cleaned data. Both the update equation and the selection of the smoothing parameters are robustified. A simulation study compares the robust and classical forecasts. The presented method is found to have good forecast performance for time series with and without outliers, as well as for fat‐tailed time series and under model misspecification. The method is illustrated using real data incorporating trend and seasonal effects. Copyright © 2009 John Wiley & Sons, Ltd.     

Robust exponential smoothing

The paper is devoted to robust modifications of exponential smoothing for time series with outliers or long-tailed distributions. Classical exponential smoothing applied to such time series is sensitive to the presence of outliers or long-tailed distributions and may give inadequate smoothing and forecasting results. First, simple and double exponential smoothing in the L₁ norm (i.e. based on the least absolute deviations) are discussed in detail. Then, general exponential smoothing is made robust, replacing the least squares approach by M-estimation in such a way that the recursive character of the final formulas is preserved. The paper gives simple algorithmic procedures which preserve advantageous features of classical exponential smoothing and, in addition, which are less sensitive to outliers. Robust versions are compared numerically with classical ones.     

Exponential smoothing: The state of the art

This paper is a critical review of exponential smoothing since the original work by Brown and Holt in the 1950s. Exponential smoothing is based on a pragmatic approach to forecasting which is shared in this review. The aim is to develop state-of-the-art guidelines for application of the exponential smoothing methodology. The first part of the paper discusses the class of relatively simple models which rely on the Holt-Winters procedure for seasonal adjustment of the data. Next, we review general exponential smoothing (GES), which uses Fourier functions of time to model seasonality. The research is reviewed according to the following questions. What are the useful properties of these models? What parameters should be used? How should the models be initialized? After the review of model-building, we turn to problems in the maintenance of forecasting systems based on exponential smoothing. Topics in the maintenance area include the use of quality control models to detect bias in the forecast errors, adaptive parameters to improve the response to structural changes in the time series, and two-stage forecasting, whereby we use a model of the errors or some other model of the data to improve our initial forecasts. Some of the major conclusions: the parameter ranges and starting values typically used in practice are arbitrary and may detract from accuracy. The empirical evidence favours Holt's model for trends over that of Brown. A linear trend should be damped at long horizons. The empirical evidence favours the Holt-Winters approach to seasonal data over GES. It is difficult to justify GES in standard form–the equivalent ARIMA model is simpler and more efficient. The cumulative sum of the errors appears to be the most practical forecast monitoring device. There is no evidence that adaptive parameters improve forecast accuracy. In fact, the reverse may be true.     

Exponential smoothing: Behavior of the ex-post sum of squares near 0 and 1

When using simple exponential smoothing on a given time series the nature of the relationship between the optimal smoothing constant and the autocorrelation structure of the series remains an unresolved question. Although numerical search routines can easily be used to find optimal values of the smoothing constant, they offer little insight into the nature of the relationship between the estimated smoothing constant and the structure of the underlying time series. We suggest that renewed investigations of the ex-post sum of squares function may prove helpful in this pursuit. Results are presented that illustrate how the optimal smoothing constant depends upon the value used to initialize the smoothing and upon the sample autocorrelation coefficients of the observed series. These results are based on a new formula for the derivative of the ex-post sum of squares function. In particular, the derivative is examined near 0 and 1, where great simplifications occur in its form, thereby facilitating investigations near these points. A necessary and sufficient condition is stated for when the ex-post sum of squares must have a positive derivative at 0 and the autocorrelation coefficients of the differenced series are shown to affect the sign of the derivative near 1. Based on these results, a general algorithm is presented as an alternative to grid search routines for minimizing the ex-post sum of squares.     

ARARMA models for time series analysis and forecasting

Methods of time series forecasting are proposed which can be applied automatically. However, they are not rote formulae, since they are based on a flexible philosophy which can provide several models for consideration. In addition it provides diverse diagnostics for qualitatively and quantitatively estimating how well one can forecast a series. The models considered are called ARARMA models (or ARAR models) because the model fitted to a long memory time series (t) is based on sophisticated time series analysis of AR (or ARMA) schemes (short memory models) fitted to residuals Y(t) obtained by parsimonious‘best lag’non-stationary autoregression. Both long range and short range forecasts are provided by an ARARMA model Section 1 explains the philosophy of our approach to time series model identification. Sections 2 and 3 attempt to relate our approach to some standard approaches to forecasting; exponential smoothing methods are developed from the point of view of prediction theory (section 2) and extended (section 3). ARARMA models are introduced (section 4). Methods of ARARMA model fitting are outlined (sections 5,6). Since‘the proof of the pudding is in the eating’, the methods proposed are illustrated (section 7) using the classic example of international airline passengers.     

Exponential smoothing of seasonal data: a comparison

A parsimonious method of exponential smoothing is introduced for time series generated from a combination of local trends and local seasonal effects. It is compared with the additive version of the Holt–Winters method of forecasting on a standard collection of real time series. Copyright © 2001 John Wiley & Sons, Ltd.     

Automatic monitoring of forecast errors

This paper evaluates a variety of automatic monitoring schemes to detect biased forecast errors. Backward cumulative sum (cusum) tracking signals have been recommended in previous research to monitor exponential smoothing models. This research shows that identical performance can be had with much simpler tracking signals. The smoothed-error signal is recommended for α = 0.1, although its performance deteriorates badly as α is increased. For higher α values, the simple cusum signal is recommended. A tracking signal based on the autocorrelation in errors is recommended for forecasting models other than exponential smoothing, with one exception. If the time series has a constant variance, the backward cusum should give better results.     

A new Bayesian formulation for Holt's exponential smoothing

In this paper we propose a Bayesian forecasting approach for Holt's additive exponential smoothing method. Starting from the state space formulation, a formula for the forecast is derived and reduced to a two‐dimensional integration that can be computed numerically in a straightforward way. In contrast to much of the work for exponential smoothing, this method produces the forecast density and, in addition, it considers the initial level and initial trend as part of the parameters to be evaluated. Another contribution of this paper is that we have derived a way to reduce the computation of the maximum likelihood parameter estimation procedure to that of evaluating a two‐dimensional grid, rather than applying a five‐variable optimization procedure. Simulation experiments confirm that both proposed methods give favorable performance compared to other approaches. Copyright © 2008 John Wiley & Sons, Ltd.     

Prediction intervals for exponential smoothing using two new classes of state space models

Three general classes of state space models are presented, using the single source of error formulation. The first class is the standard linear model with homoscedastic errors, the second retains the linear structure but incorporates a dynamic form of heteroscedasticity, and the third allows for non‐linear structure in the observation equation as well as heteroscedasticity. These three classes provide stochastic models for a wide variety of exponential smoothing methods. We use these classes to provide exact analytic (matrix) expressions for forecast error variances that can be used to construct prediction intervals one or multiple steps ahead. These formulas are reduced to non‐matrix expressions for 15 state space models that underlie the most common exponential smoothing methods. We discuss relationships between our expressions and previous suggestions for finding forecast error variances and prediction intervals for exponential smoothing methods. Simpler approximations are developed for the more complex schemes and their validity examined. The paper concludes with a numerical example using a non‐linear model. Copyright © 2005 John Wiley & Sons, Ltd.     

Exponential smoothing: Estimation by maximum likelihood

In this paper several forecasting methods based on exponential smoothing with an underlying seasonal autoregressive-moving average (SARIMA) model are considered. The relations between the smoothing constants and the coefficients of the autoregressive and moving average polynomials are used. On that basis, a maximum likelihood procedure for parameter estimation is described. The approach rules out the need for initial smoothed values. Prediction intervals are also obtained as a by-product of the approach and a fast algorithm for implementing the method is outlined.     

Disaggregation of annual flow data with multiplicative trends

Various methods based on smoothing or statistical criteria have been used for constructing disaggregated values compatible with observed annual totals. The present method is based on a time‐series model in a state space form and allows for a prescribed multiplicative trend. It is applied to US GNP data which have been used for comparing methods suggested for this purpose. The model can be extended to include quarterly series, related to the unknown disaggregated values. But as the estimation criteria are based on prediction errors of the aggregated values, the estimated form may not be optimal for reproducing high‐frequency variations of the disaggregated values. Copyright © 1999 John Wiley & Sons, Ltd.     

A comparison of approximate Bayesian forecasting methods for non‐Gaussian time series

We present the results on the comparison of efficiency of approximate Bayesian methods for the analysis and forecasting of non‐Gaussian dynamic processes. A numerical algorithm based on MCMC methods has been developed to carry out the Bayesian analysis of non‐linear time series. Although the MCMC‐based approach is not fast, it allows us to study the efficiency, in predicting future observations, of approximate propagation procedures that, being algebraic, have the practical advantage of being very quick. Copyright © 2000 John Wiley & Sons, Ltd.     

Nowcasting quarterly GDP growth in a monthly coincident indicator model

This paper presents an extension of the Stock and Watson coincident indicator model that allows one to include variables available at different frequencies while taking care of missing observations at any time period. The proposed procedure provides estimates of the unobserved common coincident component, of the unobserved monthly series underlying any included quarterly indicator, and of any missing values in the series. An application to a coincident indicator model for the Portuguese economy is presented. We use monthly indicators from business surveys whose results are published with a very short delay. By using the available data for the monthly indicators and for quarterly real GDP, it becomes possible to produce simultaneously a monthly composite index of coincident indicators and an estimate of the latest quarter real GDP growth well ahead of the release of the first official figures. Copyright © 2005 John Wiley & Son, Ltd.     

Modelling a traffic network with missing data

Whitlock and Queen (1998) developed a dynamic graphical model for forecasting traffic flows at a number of sites at a busy traffic junction in Kent, UK. Some of the data collection sites at this junction have been faulty over the data collection period and so there are missing series in the multivariate problem. Here we adapt the model developed in Whitlock and Queen ( 1998 ) to accommodate these missing data. Markov chain Monte Carlo methods are used to provide forecasts of the missing series, which in turn are used to produce forecasts for some of the other series. The methods are used on part of the network and shown to be very promising. Copyright © 2000 John Wiley & Sons, Ltd.     

Causal forces: Structuring knowledge for time-series extrapolation

This paper examines a strategy for structuring one type of domain knowledge for use in extrapolation. It does so by representing information about causality and using this domain knowledge to select and combine forecasts. We use five categories to express causal impacts upon trends: growth, decay, supporting, opposing, and regressing. An identification of causal forces aided in the determination of weights for combining extrapolation forecasts. These weights improved average ex ante forecast accuracy when tested on 104 annual economic and demographic time series. Gains in accuracy were greatest when (1) the causal forces were clearly specified and (2) stronger causal effects were expected, as in longer-range forecasts. One rule suggested by this analysis was: ‘Do not extrapolate trends if they are contrary to causal forces.’ We tested this rule by comparing forecasts from a method that implicitly assumes supporting trends (Holt's exponential smoothing) with forecasts from the random walk. Use of the rule improved accuracy for 20 series where the trends were contrary; the MdAPE (Median Absolute Percentage Error) was 18% less for the random walk on 20 one-year ahead forecasts and 40% less for 20 six-year-ahead forecasts. We then applied the rule to four other data sets. Here, the MdAPE for the random walk forecasts was 17% less than Holt's error for 943 short-range forecasts and 43% less for 723 long-range forecasts. Our study suggests that the causal assumptions implicit in traditional extrapolation methods are inappropriate for many applications.     

Modeling Compositional Time Series with Vector Autoregressive Models

Multivariate time series describing relative contributions to a total (like proportional data) are called compositional time series. They need to be transformed first to the usual Euclidean geometry before a time series model is fitted. It is shown how an appropriate transformation can be chosen, resulting in coordinates with respect to the Aitchison geometry of compositional data. Using vector autoregressive models, the standard approach based on raw data is compared with the compositional approach based on transformed data. The results from the compositional approach are consistent with the relative nature of the observations, while the analysis of the raw data leads to several inconsistencies and artifacts. The compositional approach is extended to the case when also the total of the compositional parts is of interest. Moreover, a concise methodology for an interpretation of the coordinates in the transformed space together with the corresponding statistical inference (like hypotheses testing) is provided. Copyright © 2015 John Wiley & Sons, Ltd.     

The wavelet scaling approach to forecasting: Verification on a large set of Noisy data

In the paper, we undertake a detailed empirical verification of wavelet scaling as a forecasting method through its application to a large set of noisy data. The method consists of two steps. In the first, the data are smoothed with the help of wavelet estimators of stochastic signals based on the idea of scaling, and, in the second, an AR(I)MA model is built on the estimated signal. This procedure is compared with some alternative approaches encompassing exponential smoothing, moving average, AR(I)MA and regularized AR models. Special attention is given to the ways of treating boundary regions in the wavelet signal estimation and to the use of biased, weakly biased and unbiased estimators of the wavelet variance. According to a collection of popular forecast accuracy measures, when applied to noisy time series with a high level of noise, wavelet scaling is able to outperform the other forecasting procedures, although this conclusion applies mainly to longer time series and not uniformly across all the examined accuracy measures.     

Estimation and forecasting of long-memory processes with missing values

This paper addresses the issues of maximum likelihood estimation and forecasting of a long-memory time series with missing values. A state-space representation of the underlying long-memory process is proposed. By incorporating this representation with the Kalman filter, the proposed method allows not only for an efficient estimation of an ARFIMA model but also for the estimation of future values under the presence of missing data. This procedure is illustrated through an analysis of a foreign exchange data set. An investment scheme is developed which demonstrates the usefulness of the proposed approach. © 1997 John Wiley & Sons, Ltd.     

Linking series generated at different frequencies

This is a report on our studies of the systematical use of mixed‐frequency datasets. We suggest that the use of high‐frequency data in forecasting economic aggregates can increase the accuracy of forecasts. The best way of using this information is to build a single model that relates the data of all frequencies, for example, an ARMA model with missing observations. As an application of linking series generated at different frequencies, we show that the use of a monthly industrial production index improves the predictability of the quarterly GNP. Copyright © 2008 John Wiley & Sons, Ltd.