Monetary denominational systems, technological changes and cost of cash.

Summary Empirical studies on the social cost of payments all agree to recognize that coins and bills are the most costly payment instruments for society. In recent years, commercial and central banks implemented various strategies to reduce the holdings and the use of cash. Of these, non-pricing strategies such as the improvement of the currency system or the change of the manufacturing technology have been implemented. However, their actual effectiveness is an issue of discussion. This PhD deals with these non-pricing strategies, namely the optimal denominations for coins and banknotes, the cost efficiency of currency systems and payment instruments. The aim is to better understand the implications of the payment behavior, the manufacturing technology and the structure of the monetary denomination system (MDS) on the distribution of cash transactions in the economy and their social cost. First, we analyze the consequences of relaxing some restrictive assumptions of theoretical studies that compare the efficiency of MDS based on a French empirical distribution of cash transactions obtained from a nationally representative household survey. Next, we challenge the widespread belief that an efficient MDS reduces the social cost of cash by differentiating the variable costs of production of currency and fixed costs per denomination. We show that efficient payment solutions in terms of the principle of least effort may be more costly for the central bank, propose an original framework to measure the profitability of the adoption of a new manufacturing technology of currency and finally, parameterize and simulate a search model of money to compare the social cost of multiple MDS.