1: \begin{abstract}

2: We consider the estimation of  a structural function which models a

3: non-parametric relationship between a response and an endogenous

4: regressor given an instrument in presence of dependence in

5: the data generating process. Assuming an independent and identically

6: distributed (\iid) sample it has been shown in

7: \cite{JohannesSchwarz2010} that a least squares estimator based on

8: dimension reduction and thresholding can attain minimax-optimal rates

9: of convergence up to a constant. As this estimation procedure requires

10: an optimal choice of a dimension parameter with regard amongst others to certain

11: characteristics of the unknown structural function we investigate its

12: fully data-driven choice based on a combination of model selection and

13: Lepski’s method inspired by \cite{GoldenshlugerLepski2011}. For the

14: resulting fully data-driven thresholded least squares estimator a

15: non-asymptotic oracle risk bound is derived by considering  either an

16: \iid sample or by dismissing the independence assumption. In both

17: cases the derived risk bounds coincide up to a constant  assuming

18: sufficiently weak dependence characterised by a fast decay of the

19: mixing coefficients. Employing the risk bounds the minimax optimality

20: up to constant of the estimator is established over a variety of

21: classes of structural functions. 

22: \end{abstract}