I think then your claim is that the true risk and the perception of risk are very different, not that default is necessarily likely (i.e. > 50%). If you were really confident that default was likely within 20 years, you could make an awful lot of money shorting T-bonds (or something).