This article develops ways to endogenize the borrowing constraints use
d in a class of computable incomplete markets models. We allow the con
straints to depend on an investor's characteristics such as time prefe
rence, risk aversion, and income streams. The proposed constraint can
be interpreted as a borrowing Limit;within which an investor has no in
centive to default. Using a numerical algorithm, we find that for an a
rray of structural parameters, the endogenous borrowing constraints ca
n be much less stringent than the ad hoc borrowing constraints adopted
by the existing studies.