In an exercise in general relativity, I am trying to show that in the limit as $\hat{r}\rightarrow\infty$,
\begin{equation*} \frac{1-GM/2\hat{r}}{1+GM/2\hat{r}} \approx 1 - \frac {2GM}{\hat{r}},\ \ \text{or}\ \ \frac{1-\phi}{1+\phi} \approx 1-4\phi\ \text{if}\ \phi\ll 1 \end{equation*}
At least, this is the approximation I deduce I need to show. I don't see a path to it, however. I see ways to show through Taylor expansion that to first order it's approximately $1-2\phi$, but I need $1-4\phi$.
What dxiv said in the comment is correct. One more idea for calculating $\frac{1-\phi}{1+\phi}$ is to assume $\tan^2(x)=\phi$ and hence using the formula $\cos(2x)=\frac{1-\tan^2(x)}{1+\tan^2(x)}$ and hence we have $\frac{1-\phi}{1+\phi} = \cos(2 (\tan^{-1}(\sqrt{\phi})))$. This is exact expression and not an approximation.