We consider a kernel based approach to nonlinear canonical correlation analysis and its implementation for time series. We deduce various diagnostics for reversible processes and gaussian processes.

The focus of the paper is the nonparametric estimation of an instrumental regression function * defined by conditional moment restrictions stemming from a structural econometric model: E[Y- *(Z) | W]=0, and involving endogeneous variables Y and Z and instruments W.

We consider a kernel based approach to nonlinear canonical correlation analysis and its implementation for time series. We deduce various diagnostics for reversible processes and gaussian processes. The method is first applied to a stimulated series satisfying a diffusion equation allowing us to...

We consider a kernel based approach to nonlinear canonical correlation analysis and its implementation for time series. We deduce a test procedure of the reversibility hypothesis. The method is applied to the analysis of stochastic differential equation from high frequency data on stock returns.

We decompose a stationary Markov process (X^t) as: X^t = a^o + [Sommation from j=1 to infinity) a^j Z^(j,t), where the Z^j 's processes admit ARMA specifications. These decompositions are deduced from a nonlinear canonical decomposition of the joint distribution of (X^t, X^(t-1)).