This paper analyses how borrower liquidity constraints and home equity relate to the realized loss given default (LGD) using the quarterly U.S. residential mortgage loan-level data observed from Q2 2005 to Q1 2015. We define defaulted loans with zero-LGD as cure loans and those with non-zero LGD as non-cure loans. We find robust evidence that the borrower liquidity constraints and positive equity are explaining cure, while negative equity explains non-zero loss. However, a relationship between borrower liquidity constraints and the non-zero LGD is not economically meaningful. Our findings support to separate cure and non-cure loans in mortgage loss risk models.