1: \begin{definition}\label{defYM} 

2:  Let $M$  be the cadlag version  of the martingale

3:  $t \mapsto  \mathbbm{E}\left[\xi + \int_0^T \hat{f}_r d\hat V_r\middle|\mathcal{F}_t\right]-\mathbbm{E}\left[\xi + \int_0^T \hat{f}_rd\hat V_r\middle|\mathcal{F}_0\right]$.

4: 

5: $M$ is square integrable by Proposition \ref{L1}. 

6: It admits a cadlag version taking into account

7:  Theorem 4 in Chapter IV of \cite{dellmeyerB}, since the stochastic basis fulfills

8:  the usual conditions.

9: We denote by $Y$ the cadlag process defined by \\

10: $Y_t = \xi+\int_t^T\hat{f}_r d\hat V_r - (M_T - M_t)$.

11: This will be called the {\bf cadlag reference process} and we omit its 

12: dependence to $(\dot U,N)$.

13: \end{definition}

14: