In a study involving a parameter $\theta$, you compare the Wald statistic $W = \frac{(\hat{\theta} - \theta_0)^2}{Var(\hat{\theta})}$ and the Likelihood Ratio Test statistic $\lambda = -2 \ln \frac{L(\theta_0)}{L(\hat{\theta})}$. If the log-likelihood surface is highly asymmetric and skewed in the neighborhood of $\theta_0$, which of the following is true regarding these tests?