14 May 2013

ECB: Optimal CSD reshaping towards T2S

This ECB Working paper uses a game theoretic approach to model the strategic choice of the CSDs. The authors argue that the Nash equilibrium is not robust, and provide a condition under which a given CSD will always reshape, whatever the other CSDs' strategy.

Since reshaping involves a significant lump sum investment, CSDs could be tempted to delay the moment when they reshape. Because higher operating costs in the industry justify higher prices and possibly bigger CSD profit margins, there could be an incentive for CSDs to tacitly collude to avoid reshaping. Hence, besides the problem of knowing what is a CSD’s individual optimal degree of reshaping is the problem of studying the possibility (and plausibility) of tacit collusion not to reshape.

This study makes use of a simple analytical model, cast into a dynamic game theoretic setting, to answer both issues. A closed-form expression for the optimal degree of reshaping is derived from the model, and simulations are provided as an illustration. These simulations involve ranges of plausible parameters derived from public data: For example, the cost of settling in T2S is taken from the T2S price list published by the ECB and the adaptation costs for a large CSD are assumed within the very broad range of €5 million to €50 million. Clearstream and Euroclear, the two largest European CSD groups, have announced that they intend to charge their users a maximum of €30 million and €25 million, respectively, as adaptation costs to T2S. In particular, these simulations allow to experimentally visualise the sense of variation of the optimal degree of reshaping as a function of the various parameters of the model, such as the market size, the costs per transaction of a given CSD, the price elasticities of the demand of settlement services, etc. A simple condition under which tacit collusion not to reshape is an equilibrium is provided. Interestingly, when this condition fails, and when price competition is assumed in each price-setting stage of the model, there can be no such tacit collusion.

The stability of the various equilibria is then discussed. Finally, it is shown analytically that two natural generalisations of the model, i.e. allowing for an arbitrary number of CSDs and introducing a delay in the observability of the reshaping decision of other CSDs, do not affect the results of this paper.

Full paper

© ECB - European Central Bank