1: \begin{proof}

2:   We only have to turn the formal discussion of Section~\ref{sec:fixpoint}

3:   into rigorous mathematics. The small factor $T^{\delta}$ is obtained from a

4:   scaling argument and while this does not require any new insights it is

5:   somewhat lengthy and we refer to~{\cite{Gubinelli2012,Gubinelli2014}} for

6:   details.

7:   

8:   Let us just indicate how to iterate the construction to obtain the existence

9:   of solutions up to the explosion time $\tau$. Let us assume that we

10:   constructed $u$ on $[0, T_0]$ for some $T_0 > 0$. Now we no longer have $X

11:   (T_0) = 0$, and also the initial condition $u (T_0)$ is no longer in $\CC^{2

12:   \gamma}$. But we only used $X (0) = 0$ to derive the initial conditions for

13:   $u'$ and $u^{\sharp}$, and we only used $u_0 \in \CC^{\gamma}$ to obtain

14:   $\CC^{2 \gamma}$ initial conditions for $u^{\sharp}$. Since now we already

15:   know $(u' (T_0), u^{\sharp} (T_0)) \in \CC^{\gamma} \times \CC^{2 \gamma}$,

16:   we do not need this anymore.

17:   

18:   As for the continuity in $(u_0, \mathbb{X})$, let $(\tilde{u}_0,

19:   \tilde{\mathbb{X}})$ be another set of data also bounded by $M$. Then the

20:   solutions $u$ and $\tilde{u}$ both are bounded in $\CD^{\gamma}_T$ by some

21:   constant $C = C (F, M) > 0$. So by the continuity properties of the

22:   paracontrolled product (and the other operations involved), we can estimate

23:   \[ d_{\CD^{\gamma}_T} (u, \tilde{u}) \leqslant P (C) \left( \| u_0 -

24:      \tilde{u}_0 \|_{2 \gamma} + d_{\mathcal{X}_{\tmop{pam}} (T)}

25:      (\mathbb{X}, \tilde{\mathbb{X}}) + T^{\delta} d_{\CD^{\gamma}_T} (u,

26:      \tilde{u}) \right) \]

27:   for a polynomial $P$. The local Lipschitz continuity on $[0, T]$ immediately

28:   follows if we choose $T > 0$ small enough. This can be iterated to obtain

29:   the local Lipschitz continuity on ``macroscopic'' intervals. 

30: \end{proof}

31: