I study how boundedly rational agents can learn the solution to an infinite horizon optimal consumption problem under uncertainty and liquidity constraints. I present conditions for the existence of an optimal linear consumption rule and characterize it. Additionally, I use an empirically...