Let $X$ and $Y$ be two normal random variables with mean zero, variance 1 and correlation $\rho$. Let $Z=sign(XY)$ with $sign(x)=1$ if $x>0$ and $sign(x)=0$ otherwise.
Now, I calculate $\mathbb{E}[Z]$ as follows \begin{eqnarray}\label{exp} \mathbb{E}\left[Z_{t}\right] &=&\mathbb{P}\left( X\geq 0, Y\geq 0\right)+\mathbb{P}\left( X\leq 0, Y\leq 0\right)\notag\\ &=& 2\mathbb{P}\left( X\geq 0, Y\geq 0\right)\notag\\ &=&2\sqrt{1-\rho^{2}}\frac{1}{2\pi}\int_{0}^{+\infty}\int_{0}^{+\infty} \frac{uv}{|u||v|}e^{-\frac{1}{2}(u^{2}+v^{2}-2uv\rho)}dudv\notag\\ &=&2\sqrt{1-\rho^{2}}\bigg[\frac{1}{\sqrt{2\pi}}\int_{0}^{+\infty}\frac{1}{\sqrt{2\pi}} \int_{0}^{+\infty} e^{-\frac{1}{2}(u-\rho v)^{2}}du e^{-\frac{1}{2}(v^2-\rho^2 v^2)} dv\bigg] \notag\\ &=&2\sqrt{1-\rho^{2}}\bigg[\frac{1}{\sqrt{2\pi}}\int_{0}^{+\infty} \frac{1}{\sqrt{2\pi}}\int_{-v\rho}^{+\infty} e^{-\frac{x^{2}}{2}}dx e^{-\frac{1}{2}(v^2-\rho^2 v^2)} dv\bigg]\notag\\ &=&2\sqrt{1-\rho^{2}}\bigg[\frac{1}{\sqrt{2\pi}}\int_{0}^{+\infty} \Phi(v\rho) e^{-\frac{1}{2}(v^2-\rho^2 v^2)} dv\bigg]\notag\\ &=&2\sqrt{1-\rho^{2}}\bigg[\frac{1}{\sqrt{2\pi}\rho}\int_{0}^{+\infty} \Phi(y) e^{-\frac{y^{2}}{2}(\frac{1-\rho^2}{\rho^2})} dy\bigg]\notag\\ &=&2\sqrt{1-\rho^{2}}\left[\frac{1}{\sqrt{2\pi}\rho} \left[\frac{\rho}{\sqrt{2\pi(1-\rho^{2})}}\arctan\left(\frac{\rho} {\sqrt{1-\rho^{2}}}\right) +\frac{\sqrt{2\pi}\rho}{4\sqrt{1-\rho^{2}}}\right] \right]\notag\\ &=&\frac{1}{\pi}\arctan\left(\frac{\rho} {\sqrt{1-\rho^{2}}}\right)+\frac{1}{2}\,. \end{eqnarray}
However, the solution is $\frac{1}{\pi} \arccos (-\rho)$.
Is any step wrong? Thanks for your kindely helps.