The EU is looking to bring in a new directive on CSDs. As often is the case, the document produced is thorough and thoughtful, but that does not necessarily mean that it is entirely directionally correct. Ever since a stream of research showed, hardly surprisingly, that it was cheaper to operate a CSD in the US than through the multiple infrastructures in Europe, there has been an almost total denial of the reasons for Europe’s fragmentation and the US consolidation.
The US is one country, with a single currency, centralised regulation, uniform securities laws, a single (if somewhat challenging for cross border investors) fiscal code and a brokerage and custodial industry heavily entrenched in New York. Europe is a collection of nation states, with a single currency for some, different regulators working on common principles rather than uniform regulation, localised securities laws more dependent on local property law than European ones, widely variant fiscal codes and a distributed brokerage and custodial industry.
The EU notes that there are more than 30 CSDs and two ICSDs in Europe against DTC and Fed Wire in the US. They could have also noted the proliferation of securities trading platforms and CCPs in Europe but they are, on this occasion, silent on that issue. The EU does not appear fully convinced of the value of trading or clearing platform consolidation despite the fact that the cost effects of that fragmentation are far higher than in the CSD world.
The EU view of Europe is much driven by T2S despite the fact that it appears to be a cost rather than a benefit. And the EU is assuming greater interconnectivity across Europe when most commentators do not believe CSDs will be effective cross market consolidators. Indeed the consensus is that T2S has changed the value of all the smaller CSDs from a positive net present value to a negative or at best marginal one, due to the cost of upgrading platforms to the T2S environment. Thus, paradoxically, T2S has reduced the likelihood of future CSD consolidation.
The CSDs are systemically important and the EU notes their EUR 920 trillion of annual transactions (and at even one basis point that would create a tax of EUR 92 billion per annum, well above any figure cited by the supporters of the apparently aborted idea of a pan European FTT!). And they account for some EUR 32 trillion of assets under custody.
It is not surprising that those figures cause concern, for they are the source of genuine systemic risk, and so we have the good parts of the EU proposal.
First they advocate the need for consistent settlement discipline procedures to avoid the prospect of regulatory arbitrage. That is a valid consideration but my own experience, especially in the CCP world, is that the major players look for efficient platforms with sound rules to eliminate the risk of contagion from delinquent market players. And CSD arbitrage, even in a truly connected environment, is far harder than many suppose. Settlement and custody is rather driven by critical mass and client requirements than back office failures or preferences.
The EU also commends the adoption of T+2 as a standard settlement period. The fail experience in most markets is sufficient to allow this and foreign exchange challenges of such an environment are feasible within this timeframe in the major cross currencies.
The desire to allow issuer choice of CSD is a valid proposal and will improve competition both between CSDs and between CSDs and their ICSD colleagues. Such a proposal will reduce the country-centric monopolies, although adoption of that structure may be fraught with difficulties from a legal and political perspective. One can imagine the frissons in government if a national champion in one country sought to make its primary listing in another market!
The EU avoids talking about interoperability although it does mention the much discussed, but much less effective, “Code of Conduct” signed by most infrastructures. At CCP level, interoperability is challenged by risk management considerations; at CSD level, it appears to be a political barrier as it is operationally and technically far from challenging. And it would bring user value albeit at the cost of revenue flow in many small and mid-size CSDs. However, the major benefactors of that user value are the major cross border players rather than domestic entities and thus we have another example of localism versus globalism. The
EU has steered away from some risk issues. The liability of CSDs in settlement processes is not clarified. Their need for insurance or capital is not addressed (other than for the banking issue I mention below) and, most importantly, the structure of their regulator is only noted and these proposals create a real need to define the role of ESMA versus local regulators in CSD policy management.
And there is one major challenge for the industry. The EU believes in segregation of banking services. That means credit provision through a separate banking entity with adequate capital. The level of capital needed for any such banking function is unclear as the major providers of this service (the ICSDs) tend to limit the scope of their product to secured intraday exposures which are not explicitly covered by any capital adequacy regime. The cash and securities link, especially at ICSD level, stretches beyond credit into the risk profile for settlement liquidity, collateral management or asset financing. An unintended consequence of any fundamental change in structure (as distinct from a requirement to allocate dedicated capital to different risks) could be yet another bifurcation of process in the securities life cycle and that would lead to more risk and cost rather than the reverse. And, worryingly, we cannot rule out, despite the silence on the issue, that the segregated approach for the CSD’s will, in time, also be applied to systemically important custodian banks.
John Gubert
Chairman Global Securities Services Executive Committee UniCredit