We analyze collusion in an infinitely repeated version of a standard auction with a continuum of types. Because of the lack of efficiency results in this setting the literature has focused on determining and comparing benchmarks on how well bidders can collude. Aoyagi (2003) has shown that the...