In a simple linear regression analysis with an intercept, let $\hat{y}_i$ be the fitted (predicted) values and $e_i$ be the residuals. What is the value of the sample covariance between $\hat{y}_i$ and $e_i$?