1: \begin{abstract}\noindent

2: We consider a recurrent Markov process which is an Itô semi-martingale. The Lévy kernel describes the law of its jumps. Based on observations $X_0,X_\Delta,\dotsc,X_{n\Delta}$, we construct an estimator for the Lévy kernel's density. We prove its consistency (as $n\Delta\to\infty$ and $\Delta\to0$) and a central limit theorem. In the positive recurrent case, our estimator is asymptotically normal; in the null recurrent case, it is asymptotically mixed normal. 

3: Our estimator's rate of convergence equals the non-pa\-ra\-me\-tric minimax rate of smooth density estimation. The asymptotic bias and variance are analogous to those of the classical Nadaraya--Watson estimator for conditional densities. Asymptotic confidence intervals are provided.

4: %{We consider a recurrent Markov process which is an Itô semi-martingale. The law of its jumps is described by the L\'evy kernel. Based on the discrete-time observations $X_0,X_\Delta,\dotsc,X_{n\Delta}$, we construct an estimator for the density of the L\'evy kernel. We prove its consistency (as $n\Delta\to\infty$ and $\Delta\to0$) as well as a central limit theorem. At the core of our research, we also study the fundamental case where we observe a sample path of the process and, in particular, all its jumps.} In the positive recurrent case, our estimator is asymptotically normal; in the null recurrent case, our estimator is asymptotically mixed normal. Our estimator's rate of convergence equals the non-pa\-ra\-me\-tric minimax rate of smooth density estimation. The asymptotic bias and variance are analogous to those of the classical Nadaraya--Watson estimator for conditional densities. Asymptotic confidence intervals are provided.

5: \end{abstract}

6: