Differentiability of transition semigroup of generalized Ornstein-Uhlenbeck process: a probabilistic approach

Abstract

Let $P_s\phi(x)=\mathbb{E}\, \phi(X^x(s))$, be the transition semigroup on the space $B_b(E)$ of bounded measurable functions on a Banach space $E$, of the Markov family defined by the linear equation with additive noise $$ d X(s)= \left(AX(s) + a\right)ds + BdW(s), \qquad X(0)=x\in E. $$ We give a simple probabilistic proof of the fact that null-controlla\-bility of the corresponding deterministic system $$ d Y(s)= \left(AY(s)+ B\mathcal{U}(t)x)(s)\right)ds, \qquad Y(0)=x, $$ implies that for any $\phi\in B_b(E)$, $P_t\phi$ is infinitely many times Fr\'echet differentiable and that $$ D^nP_t\phi(x)[y_1,\ldots ,y_n]= \mathbb{E}\, \phi(X^x(t))(-1)^nI^n_t(y_1,\ldots, y_n), $$ where $I^n_t(y_1,\ldots,y_n)$ is the symmetric n-fold It\^o integral of the controls $\mathcal{U}(t)y_1,\ldots \mathcal{U}(t)y_n$.