Solution (source code)

= Solution

\b[The printed pathwise conclusion is false without specifying the initial value.] We first prove exactly what the hypotheses do imply, by <stationary drift saturation>.

Stationarity gives $\mathbb E Z_s^2=1$. The <Itô formula> yields
$$
d(Z_s^2)=2\sqrt{2\lambda}Z_s\,dW_s+(2Z_sg_s+2\lambda)ds.
$$
The <stochastic integral> is a true <square-integrable> <martingale> on every finite horizon, since $\mathbb E\int_0^tZ_s^2ds=t$. Also $\mathbb E|Z_sg_s|\le\lambda$ by the <Cauchy-Schwarz inequality>. Taking expectations therefore gives
$$
0=2\int_0^t\bigl(\mathbb E[Z_sg_s]+\lambda\bigr)ds.
$$
Every integrand is nonnegative, so $\mathbb E[Z_sg_s]=-\lambda$ for almost every $s$. For those times,
$$
\mathbb E(g_s+\lambda Z_s)^2
=\mathbb Eg_s^2+2\lambda\mathbb E(Z_sg_s)+\lambda^2\mathbb EZ_s^2\le0.
$$
It follows that $g_s=-\lambda Z_s$ for almost every time and outcome. Equality of these integrable drift terms gives the exact solution
$$
Z_t=e^{-\lambda t}Z_0+\sqrt{2\lambda}\int_0^te^{-\lambda(t-s)}dW_s,
\qquad
\boxed{Z_t-Y_t=e^{-\lambda t}(Z_0-B_1).}
$$
Thus $Z$ solves the same equation, but stationarity has not identified its initial random variable with $B_1$.

For an explicit counterexample, take
$$
Z_0=-B_1,\qquad
Z_t=-e^{-\lambda t}B_1+\sqrt{2\lambda}\int_0^te^{-\lambda(t-s)}dW_s,
\qquad g_t=-\lambda Z_t.
$$
This is adapted to $\mathcal G_t$. Its independent normal initial value and Brownian noise make it a centered stationary <Gaussian process> with <covariance> $e^{-\lambda|t-s|}$, exactly as for $Y$. Hence $\mathbb EZ_0^2=1$ and $\mathbb Eg_t^2=\lambda^2$, satisfying every printed hypothesis. Nevertheless $Z_t-Y_t=-2e^{-\lambda t}B_1$, which is nonzero almost surely for every finite $t$.

The corrected result is \b[indistinguishability if one adds $Z_0=Y_0$ almost surely]. Without that addition, there is equality in law of the stationary Ornstein-Uhlenbeck processes, not necessarily equality of their coupled paths. To see that the initial law is forced, independence of <Brownian increments> from $\mathcal G_0$ and stationarity give, for its <characteristic function> $\varphi$,
$$
\varphi(u)=\varphi(e^{-\lambda t}u)\exp\!\left[-\tfrac12(1-e^{-2\lambda t})u^2\right].
$$
Letting $t\to\infty$ proves $\varphi(u)=e^{-u^2/2}$. This is the distinction captured by <same-noise stationary Ornstein-Uhlenbeck processes>.