In this correspondence we deene a notion of principal angles between two linear autoregressive (AR) models by considering the principal angles between the ranges of their innnite observability matrices. We show how a recently deened metric for these models, which is based on their cepstra, is related to the subspace angles between the two models.

In the autoregressive process of first order AR(1), a homogeneous correlated time series ut is recursively constructed as ut = q ut−1 + σ t, using random Gaussian deviates t and fixed values for the correlation coefficient q and for the noise amplitude σ. To model temporally heterogeneous time series, the coefficients qt and σt can be regarded as time-dependent variables by themselves, leading ...

We use recent developments on threshold autoregressive models that allow deriving endogenously threshold effects to analyse the evolution of the US stock price–dividend relation over the period 1871 to 2004. More specifically, a mean-reverting dynamic behaviour of the stock price–dividend ratio should be expected once such threshold is reached. Our empirical results showed that significant adju...

We develop and estimate a model of the dynamic behavior of aggregate corporate dividends as a function of the change in permanent earnings of firms. Although structured along the lines of the Lintner-Brittain-FamaBabiak models of individual-firm dividend behavior, the model uses changes in stock prices instead of accounting earnings to measure permanent earnings changes. The performance of the ...

1.1. Point Processses and Intensities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1.1. Stochastic Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2. The Autoregressive Conditional Duration Model . . . . . . . . . . . . . . . . . . . . . . . 2 1.3. The Autoregressive Conditional Intensity Model . . . . . . . . . . . . . . . ...