In a survey, if the auxiliary variable $X$ is perfectly correlated with $Y$ ($\rho = 1$), what is the variance of the ratio estimator $\hat{Y}_R$ compared to the regression estimator $\hat{Y}_{lr}$?