MIGRATION APPROACHES: BY CSD OR BY ISIN

Transcription

1 T2S DIVISION 16 July /2009/ MIGRATION APPROACHES: BY CSD OR BY ISIN 1 Introduction Since the establishment of the Subgroup on Testing and Migration, the migration of different CSDs to T2S was based on a phased approach, excluding a pure big bang because of the unacceptably high risks attached to it. After a short verification, the views of SGTM members centered around a migration by CSD approach, as already stated in the General Principles and High-Level Proposals (June 2007), thereby excluding the by asset class approach. In the AG meeting on June, doubts were raised about this approach and more analysis was requested in order to provide reassurance that the by CSD approach is the most appropriate. In order to do this analysis, it is suggested to look at both options in more detail on the basis of the attached matrix and to list for each type of T2S Actor the pros and cons of both approaches from different perspectives. During the SGTM meeting of July, an attempt would be made to list the major consequences for each of the T2S Actors, with more fine-tuning to be done at the end-of-august meeting, based on further analysis to be conducted in between the two meetings. 2 Possible migration approaches 2.1 Pure big bang versus phased approach For the analysis to be done, the basic assumption is still that a pure big bang approach (meaning a single migration of all CSDs with all transactions in all ISINs) is excluded. Although it is acknowledged that if all goes well the overall costs of such an approach would be lower than with any other approach, the risk of significant business disturbances on the day(s) after go-live is too high and cannot be adequately mitigated by any measures, and can hence not be accepted. A phased approach remains therefore the only reasonable option. T2S will go live with a limited volume, in order to verify that all functionality works as expected (both in T2S and in the CSDs /DCPs internal systems), to fine-tune the technical performance and to streamline the operational procedures. Only after a pre-agreed period, more business is migrated to T2S.

2 2.2 Various types of phasing The phased approach can be implemented in a number of different ways, e.g. by CSD, by ISIN, by type of delivery (DVP, FOP, ), by scope of transaction (intra-csd, cross-csd, with external CSD), by currency etc. However, for various reasons, only two approaches will be further investigated, i.e. by CSD or by ISIN. A migration by CSD would typically imply that a few (low-volume) CSDs would completely migrate on the go-live date, while others would migrate at a later stage, based on a pre-agreed plan. The few migrating CSDs would be selected on the basis of their willingness to migrate and would ideally bring business to T2S with an as-wide-as possible functional scope. A migration by ISIN would typically imply that all CSDs would migrate on the go-live date, but only for a limited number of ISINs. Additional ISINs would be added on subsequent migration dates, based on a pre-agreed plan. The ISINs to be migrated would be selected by the relevant CSDs (issuer and investor CSDs that have committed to join T2S during the migration period), with the aim to limit transaction volumes immediately after the go-live date and to neutralise any competition effects that might arise. Another parameter to consider for both migration approaches is the number of migration dates after the go-live date. From a functional perspective, it would be possible to plan only one additional migration date without creating uncontrollable risks, given that the period after the go-live date is precisely intended to stabilise T2S before the next migration date. However, adding the complete remaining volume in only one additional migration may result in performance issues, because the ratio between transaction volumes and IT processing power is typically not linear, but exponential. This is one of the elements that will also be taken into account in the below comparison between migration approaches.

3 3 Comparison of the by CSD and the by ISIN approach The following tables aim at giving an overview of the pros and cons of each of the two migration approaches, both from and from a short vs. long migration perspective. 3.1 Migration by CSD Project perspective (cost of IT developments (e.g. for links), complexity of tests, ) Business perspective (liquidity reconciliation, competition, ) 2 migration dates (1 pilot group and 1 additional group) 3-4 migration dates (1 pilot group and 2-3 additional groups 2 migration dates (1 pilot group and 1 additional group) 3-4 migra pilot gro additional High-volume CSDs with High-volume CSD with

4 DCPs connected to only few CSDs DCPs connected to majority of CSDs custodians with a dedicated cash account custodians without a dedicated cash account CCPs Exchanges MTFs Euro-area central banks Non-euro area central banks 4CB Payment banks

5 3.2 Migration by ISIN Project perspective (cost of IT developments and maintenance of two applications, complexity of tests, ) 2 migration dates (1 for ISINs with 20% of T2S volume and 2 nd one for full remaining volume of ISINs ) 3 migration dates (1 for ISINs with 20% of T2S volume, 2 nd one for ISINs with 50-60% of T2S volume and 3 rd one for remaining ISINs with % of T2S volume) Business perspective (liquidity reconciliation, competition, ) 2 migration dates (1 for ISINs with 20% of T2S volume and 2 nd one for full remaining volume of ISINs 3 migratio ISINs with volume, 2 n with 50- volume an remaining 30 % of T2 Comment: Is it really relevant for the migration by ISIN? High-volume CSDs with High-volume CSD with DCPs connected to only

6 few CSDs DCPs connected to majority of CSDs custodians with a dedicated cash account custodians without a dedicated cash account CCPs Exchanges MTFs Euro-area central banks Non-euro area central banks 4CB Payment banks