In a simple linear regression $Y = \beta_0 + \beta_1 X + \epsilon$, if the variance of the error term is heteroscedastic, specifically $Var(\epsilon_i|X_i) = \sigma^2 X_i^2$, which statement is true regarding the Ordinary Least Squares (OLS) estimator $\hat{\beta}_1$?