A nonlinear time series model and estimation of missing observations

This paper formulates a nonlinear time series model which encompasses several standard nonlinear models for time series as special cases. It also offers two methods for estimating missing observations, one using prediction and fixed point smoothing algorithms and the other using optimal estimating equation theory. Recursive estimation of missing observations in an autoregressive conditionally heteroscedastic (ARCH) model and the estimation of missing observations in a linear time series model are shown to be special cases. Construction of optimal estimates of missing observations using estimating equation theory is discussed and applied to some nonlinear models.Authors were supported in part by a grant from the Natural Sciences and Engineering Research Council of Canada.     

Linear relations in time series models. II

In this paper an asymptotic theory is developed for a new time series model which was introduced in a previous paper [5]. An algorithm for computing estimates of the parameters of this time series model is given, and it is shown that these estimators are asymptotically efficient in the sense that they have the same asymptotic distribution as the maximum likelihood estimators.     

Linear relations in time series models. I

A multiple time series is defined as the sum of an autoregressive process on a line and independent Gaussian white noise on a hyperplane that goes through the origin and intersects the line at a single point. This process is a multiple autoregressive time series in which the regression matrices satisfy suitable conditions. It is shown that the maximum likelihood estimates of the line and the autoregression coefficients can be obtained as the values that minimize a given function, and that the remaining maximum likelihood estimates can be computed as simple functions of the first ones. It is also shown that the maximum likelihood estimates are equivariant with respect to the group of bijective linear transformations.     

Innovations algorithm asymptotics for periodically stationary time series with heavy tails

The innovations algorithm can be used to obtain parameter estimates for periodically stationary time series models. In this paper we compute the asymptotic distribution for these estimates in the case where the underlying noise sequence has infinite fourth moment but finite second moment. In this case, the sample covariances on which the innovations algorithm are based are known to be asymptotically stable. The asymptotic results developed here are useful to determine which model parameters are significant. In the process, we also compute the asymptotic distributions of least squares estimates of parameters in an autoregressive model.     

Improved estimates for multivariate complex-normal regression with application to analysis of linear time-invariant relationships

The objective of this paper is the estimation of linear time-invariant relationships for a stationary vector-valued time series using the Finite Fourier Transform as the basic statistic. Since this is asymptotically complex-Normal we are led to consider models of multivariate complex-Normal regression. We propose estimates of regression matrices in the tradition of Stein (shrunken estimates) which improve upon the usual estimates. Some experience with simulated time series is reported.     

基于两步法带宽选择的金融时间序列长记忆参数估计

金融时间序列长记忆参数的半参数估计方法以频域分析为主,带宽选择是其中必不可少的关键环节。不同的带宽可能给出差异明显的长记忆参数估计值,甚至产生矛盾的结论,进而影响时间序列平稳性的判断。本文提出一种两步法,用于金融时间序列长记忆估计的半参数方法的带宽选择,并进一步对长记忆参数进行估计:首先,为了克服半参数方法忽略短期结构的不足,通过信息准则判断ARFIMA(p,d,q)过程的短记忆结构;其次,用短记忆模型拟合差分后的序列,根据拟合效果确定选择带宽及长记忆参数估计值。数值模拟显示以长记忆参数估计值均方根误差最小为标准,两步法优于其他方法。经上证50指数已实现波动率日数据的实证检验,两步法在长记忆模型中的预测误差最小;与短记忆模型相比,两步法在中期提前预测步长上具有优势。     

Conditional least squares and copulae in claims reserving for a single line of business

One of the main goals in non-life insurance is to estimate the claims reserve distribution. A generalized time series model, that allows for modeling the conditional mean and variance of the claim amounts, is proposed for the claims development. On contrary to the classical stochastic reserving techniques, the number of model parameters does not depend on the number of development periods, which leads to a more precise forecasting.Moreover, the time series innovations for the consecutive claims are not considered to be independent anymore. Conditional least squares are used to estimate model parameters and consistency of these estimates is proved. The copula approach is used for modeling the dependence structure, which improves the precision of the reserve distribution estimate as well.Real data examples are provided as an illustration of the potential benefits of the presented approach.     

Forecasting time series with multiple seasonal patterns

A new approach is proposed for forecasting a time series with multiple seasonal patterns. A state space model is developed for the series using the innovations approach which enables us to develop explicit models for both additive and multiplicative seasonality. Parameter estimates may be obtained using methods from exponential smoothing. The proposed model is used to examine hourly and daily patterns in hourly data for both utility loads and traffic flows. Our formulation provides a model for several existing seasonal methods and also provides new options, which result in superior forecasting performance over a range of prediction horizons. In particular, seasonal components can be updated more frequently than once during a seasonal cycle. The approach is likely to be useful in a wide range of applications involving both high and low frequency data, and it handles missing values in a straightforward manner.     

The Stochastically Subordinated Poisson Normal Process for Modelling Financial Assets

The method of stochastic subordination, or random time indexing, has been recently applied to Wiener process price processes to model financial returns. Previous emphasis in stochastic subordination models has involved explicitly identifying the subordinating process with an observable quantity such as number of trades. In contrast, the approach taken here does not depend on the specific identification of the subordinated time variable, but rather assumes a class of time models and estimates parameters from data. In addition, a simple Markov process is proposed for the characteristic parameter of the subordinating distribution to explain the significant autocorrelation of the squared returns. It is shown, in particular, that the proposed model, while containing only a few more parameters than the commonly used Wiener process models, fits selected financial time series particularly well, characterising the autocorrelation structure and heavy tails, as well as preserving the desirable self-similarity structure, and the existence of risk-neutral measures necessary for objective derivative valuation. Also, it will be shown that the model proposed fits financial times series data better than the popular generalised autoregressive conditional heteroscedasticity (GARCH) models. Additionally, this paper will develop a skew model by replacing the normal variates with Lévy stable variates.     

Anisotropic Brown-Resnick space-time processes: estimation and model assessment

Spatially isotropic max-stable processes have been used to model extreme spatial or space-time observations. One prominent model is the Brown-Resnick process, which has been successfully fitted to time series, spatial data and space-time data. This paper extends the process to possibly anisotropic spatial structures. For regular grid observations we prove strong consistency and asymptotic normality of pairwise maximum likelihood estimates for fixed and increasing spatial domain, when the number of observations in time tends to infinity. We also present a statistical test for isotropy versus anisotropy. We apply our test to precipitation data in Florida, and present some diagnostic tools for model assessment. Finally, we present a method to predict conditional probability fields and apply it to the data.     

Adaptive Filtering: a Theoretical and Simulation Study of the Convergence of the Parameter Estimates

The existence of bias in the final parameter estimates using adaptive filtering is demonstrated theoretically. For observations generated by an autoregressive model of order one, an approximate theoretical expression for the bias is derived which is valid for long series of observations. The validity of the expression is investigated by simulation and comparing the theoretical bias with the simulated bias at the end of one major iteration. By carrying out a number of major iterations, it is shown that the bias reaches a stable value which is a function of the learning constant. The magnitude of the stable bias may be made as small as desired by taking smaller values for the learning constant. However, as the learning constant decreases, the number of major iterations required to achieve stability increases. By means of simulation experiments, the existence of bias in the final parameter estimates is demonstrated for shorter series of observations generated by an AR(1) model, and for long series of observations generated by an AR(2) model. Again the bias appears to increase with the magnitude of the learning constant. It is argued that the presence of this bias need not be a drawback in the practical application of the method since, by systematic reduction of the training constant between major iterations, the bias may also be reduced, while reasonably rapid convergence can still be maintained.     

A Levin-type algorithm for accelerating the convergence of Fourier series

A nonlinear sequence transformation is presented which is able to accelerate the convergence of Fourier series. It is tailored to be exact for a certain model sequence. As in the case of the Levin transformation and other transformations of Levin-type, in this model sequence the partial sum of the series is written as the sum of the limit (or antilimit) and a certain remainder, i.e., it is of Levin-type. The remainder is assumed to be the product of a remainder estimate and the sum of the first terms oftwo Poincaré-type expansions which are premultiplied by two different phase factors. This occurrence of two phase factors is the essential difference to the Levin transformation. The model sequence for the new transformation may also be regarded as a special case of a model sequence based on several remainder estimates leading to the generalized Richardson extrapolation process introduced by Sidi. An algorithm for the recursive computation of the new transformation is presented. This algorithm can be implemented using only two one-dimensional arrays. It is proved that the sequence transformation is exact for Fourier series of geometric type which have coefficients proportional to the powers of a numberq, |q|<1. It is shown that under certain conditions the algorithm indeed accelerates convergence, and the order of the convergence is estimated. Finally, numerical test data are presented which show that in many cases the new sequence transformation is more powerful than Wynn's epsilon algorithm if the remainder estimates are properly chosen. However, it should be noted that in the vicinity of singularities of the Fourier series the new sequence transformation shows a larger tendency to numerical instability than the epsilon algorithm.     

Generalized discrete autoregressive moving‐average models

This article proposes the generalized discrete autoregressive moving‐average (GDARMA) model as a parsimonious and universally applicable approach for stationary univariate or multivariate time series. The GDARMA model can be applied to any type of quantitative time series. It allows to compute moment properties in a unique way, and it exhibits the autocorrelation structure of the traditional ARMA model. This great flexibility is obtained by using data‐specific variation operators, which is illustrated for the most common types of time series data, such as counts, integers, reals, and compositional data. The practical potential of the GDARMA approach is demonstrated by considering a time series of integers regarding votes for a change of the interest rate, and a time series of compositional data regarding television market shares.     

Functional-coefficient partially linear regression model

In this paper, the functional-coefficient partially linear regression (FCPLR) model is proposed by combining nonparametric and functional-coefficient regression (FCR) model. It includes the FCR model and the nonparametric regression (NPR) model as its special cases. It is also a generalization of the partially linear regression (PLR) model obtained by replacing the parameters in the PLR model with some functions of the covariates. The local linear technique and the integrated method are employed to give initial estimators of all functions in the FCPLR model. These initial estimators are asymptotically normal. The initial estimator of the constant part function shares the same bias as the local linear estimator of this function in the univariate nonparametric model, but the variance of the former is bigger than that of the latter. Similarly, initial estimators of every coefficient function share the same bias as the local linear estimates in the univariate FCR model, but the variance of the former is bigger than that of the latter. To decrease the variance of the initial estimates, a one-step back-fitting technique is used to obtain the improved estimators of all functions. The improved estimator of the constant part function has the same asymptotic normality property as the local linear nonparametric regression for univariate data. The improved estimators of the coefficient functions have the same asymptotic normality properties as the local linear estimates in FCR model. The bandwidths and the smoothing variables are selected by a data-driven method. Both simulated and real data examples related to nonlinear time series modeling are used to illustrate the applications of the FCPLR model.     

Nonparametric time series prediction: A semi-functional partial linear modeling

There is a recent interest in developing new statistical methods to predict time series by taking into account a continuous set of past values as predictors. In this functional time series prediction approach, we propose a functional version of the partial linear model that allows both to consider additional covariates and to use a continuous path in the past to predict future values of the process. The aim of this paper is to present this model, to construct some estimates and to look at their properties both from a theoretical point of view by means of asymptotic results and from a practical perspective by treating some real data sets. Although the literature on the use of parametric or nonparametric functional modeling is growing, as far as we know, this is the first paper on semiparametric functional modeling for the prediction of time series.     

An assessment of interfuel substitution by electric utilities

This paper applies a translog model to a pooled sample in order to measure the extent of regional interfuel substitution effects in the electric power industry. The results obtained indicate that relative changes in fuel prices both regionally and nationally have significant effects on fossil fuel consumption. This, in turn, has important implications for public policy. In particular, the market system appears better able to deal with exogenous shifts in energy supplies than has frequently been assumed in the formulation of public policies toward the energy crisis.Further, compared to aggregate United States time series estimates, a more elastic fuel price response is found thus questioning whether full long-run adjustment is being measured in the pure time series estimates. Given the importance of the latter in energy tax policy analysis for example, the question is indeed more than academic.     

On the estimation of nonlinear time series models

In the present paper, a framework for parametric estimation in nonlinear time series is developed. Strong consistency and asymptotic normality of minimum Hellinger distance estimates for a determined class of nonlinear models are investigated. The main Interest for these estimates is motivated by their robustness under perturbations as it has been emphazized in Beran [2]. The first part of the paper is devoted to the study of some probabilistic properties which ensure the existence and the optimal properties of the estimates     

Hellinger distance estimation of SSAR models

The present paper deals with the statistical inference of the simultaneous switching autoregressive (SSAR) model. This model has been introduced by Kunitomo and Sato (Jpn. Econ. Rev. 50 (2) (1996) 161) in order to take into account the asymmetry in financial and economical time series modelling. Under some conditions which ensure some probabilistic properties of the model, we establish, under other mild assumptions, the asymptotic properties of the minimum Hellinger distance estimates of the parameters. An application to a true data is also given.     

The estimation of M4 processes with geometric moving patterns

There are many parameters in multivariate maxima of moving maxima processes—or M4 processes. However, the more parameters there are, the more difficult it is to estimate them. It is not just an issue of numerical stability, of course. The statistical precision of the estimates will be poor if the number of parameters is too large. We consider asymmetric geometric structures which correspond to special moving patterns of extreme observations in observed time series. We study the model identifiability and propose parameter estimators. All proposed estimators are shown to be consistent and asymptotically joint normal. Simulation study and real data modeling of North Sea wave height data are illustrated.     

Resampling DEA estimates of investment fund performance

Data envelopment analysis (DEA) is attractive for comparing investment funds because it handles different characteristics of fund distribution and gives a way to rank funds. There is substantial literature applying DEA to funds, based on the time series of funds’ returns. This article looks at the issue of uncertainty in the resulting DEA efficiency estimates, investigating consistency and bias. It uses the bootstrap to develop stochastic DEA models for funds, derive confidence intervals and develop techniques to compare and rank funds and represent the ranking. It investigates how to deal with autocorrelation in the time series and considers models that deal with correlation in the funds’ returns.