I examine the implementation of the Friedman rule under the assumption that age dependent lump sum transfers are possible and private intermediation is costly. This is done both in an infinitely lived agents model and in an overlapping generations model. I argue that in addition to a zero nominal-interest-rate policy (the so called Friedman rule) a transfer to young agents, or a government loan program is required for satiating agents with real balances. The paper also contributes to the understanding of Friedman's original article and discusses related questions about the size of the financial sector. It is shown that the adoption of the (modified) Friedman rule will crowd out private lending and borrowing. I also look at the social value of a market for contingent claims and argue that resources spent on operating a market for accidental nominal bequests are a waste from the social point of view in spite of the fact that individuals have an incentive to trade in such markets.
