Why account segregation might be a bad idea

The debate between the protagonists for and against segregation of client accounts has been driven more by technological reach, especially at CSDs, rather than risk assessment. There has been relative silence on the latter issue and that is strange for segregation may not be the panacea its supporters pretend.

The debate between the protagonists for and against segregation of client accounts has been driven more by technological reach, especially at CSDs, rather than risk assessment. There has been relative silence on the latter issue and that is strange for segregation may not be the panacea its supporters pretend.

It is true that end to end, client by client segregation would create an enormous additional amount of data flow, both from a settlement, reconciliation and reporting perspective. It is true that many CSDs would not be able to absorb the additional volumes, neither the added static data nor the related transaction data, without a major reengineering of their processes. It is correct to assume that total segregation would need new allocation products in areas where these do not exist or where the mechanism for allocation is either manual or definitely not at the requisite industrial strength. And it is accurate to assume that prices for those impacted would need to rise substantially to accommodate the resultant proliferation of named accounts. It is also possible that end to end segregation could lead to calls for changes to the relationship between parties in the value chain.

It is therefore appropriate to assess what could be done and whether it is practical without destroying, especially direct retail involvement, in the securities world. And the answer is, most likely, that the latter segment would need to change its approach and, perhaps, withdraw from direct stock exchange investment.

The end to end segregated product has an interesting history. It used to be the mantra of the retail and transfer agency sectors in the UK but was abandoned in favour of nominee account structures for cost reasons. The Nordic countries effectively operate their definitive share registers at CSD level with millions of registered investor account, but they are, on a global scale, small markets. Operating designation at every stage of the value chain in every market, and especially the larger ones, is problematical. On the negative side, over and above transactional and data volume considerations, allocation of block trades would need to be advised at each stage and this could create increased short term reconciliation breaks. There could also be questions around the KYC process and laws would need to change to ensure that any duplication of end investor records does not require multiple duplicative client appraisals. And, notification of end investor in some jurisdictions may change the liability relationship between parties, whilst in others it raises interesting questions around client confidentiality. On the positive side, fraud may be more difficult as holdings, especially at CSD level, are disaggregated. Inadvertent intra account stock loans would become less possible, although so would formal stock lending or similar operations. It is true, on the positive side, that ownership could be clearly established with accounts legally segregated in the event of an intermediary insolvency. However, mandating segregation is unhelpful.

Clients select intermediaries for several reasons, and a risk assessment is part of their due diligence process. Segregation is already possible. Surely, one can argue, with some strong logic on one’s side, that the choice is a commercial one for the individual and not an area for regulatory incursion?
The cost of any mandated segregation is difficult to assess. At intermediary level, one has to decide two things. The simple one is to structure a segregated account to ensure ownership of the assets lodged there are secure in the event of bankruptcy. The difficult one is to protect the account holder from intermediary error or fraud. At CSD level, much can be learned from the Nordic experience but it is no global template. The scale of those markets, in terms of investor numbers, investment counters and transactional activity means they cannot be equated to the ICSD or other major market infrastructures in terms of the platform dimensions needed to accommodate the data and transactional volumes that would result. Elsewhere, there could also be risks. Would segregation lead to a major reduction in trading lot sizes? And how would it impact Exchanges? How does this flow through to trade reporting? And can the regulators accommodate such a change at a time when they appear to have problems handling existing and disparate reports from the different trade repositories.

So is it feasible? It would need a major reengineering of the CSD IT platforms and, most likely, a rewriting of their core applications. That is no simple task and scale demands could even lead to a reversal of the trend to consolidation of infrastructures. It will lead to a major review of the viability of retail, and some smaller wholesale, clients as the costs of segregation may be the death knoll for those sectors. At one level, it may impact the attraction of specific markets, assuming, as is likely to be the case, that any mandatory segregation will not be universal. But the main policy issue, though, is that it could price direct investment out of the market for the smaller investor. It is possible it could be a boon for collective investment schemes as long as they, and the intermediaries involved in their flows, too could cope with segregated accounts!

It may be that regulators are being driven towards segregation by a desire to reduce their risks from compensation funds, and perhaps the public purse, as they see the process as removing insolvency risk. But they need to be aware that it retains operational risks, as intermediaries are not going to be eliminated from the transaction flows. And, in reality, operational risks are the biggest challenge to the process for title can be assured in the current system. 

 

«