I suspect some millions of words have been written over the years about the challenge of shorter settlement cycles. As a Londoner (by adoption), I have lived through the trauma of the UK domestic market moving from a structure of account settlement (a two week long trading session where all transactions eventually settled several days after the end of that period) to the current, and much more logical, T+3 process; and now onto T+2. I have worked in markets with T+0 and T+1; some with high settlement fail rates and others which overcame this with settlement pre-funding or draconian rules forcing settlement as advised by the relevant broker. I have railed against the illogicality of retaining the two day spot cycle for foreign exchange although there is now a liquid market in next day settlement in most major currencies.
The EU plans to move almost all markets to a T+2 cycle later this year, although certain ones will defer the operation by a few months because it clashes with other developments. As in any such change in the settlement period, two days trading will be concentrated into one day’s settlement. There is a risk, if markets see peak volumes on the relevant days, of severe settlement overload. Although we do not need a lengthy launch period for a move to T+2, splitting markets into Groups to reduce potential peak flow risk would appear a logical and sensible policy.
The rationale for T+2 is simple. Every day that is taken out of the trade to settlement lag reduces counterparty risk. As volumes have ballooned, even with the palliative impact of the growth of CCPs, the amount of counterparty risk has grown exponentially. However, the desire to eliminate credit risk must not be at the expense of an increase in operational risk. It seems unlikely that this will happen with the latest T+2 moves, but there are changes that are needed and regulatory attention should be focused on encouraging such change.
The market faces two fundamental problems, namely the inherent inefficiency of the client to global custodian link and the fundamental flaws in market side trade matching in many jurisdictions. Both are solvable from a technical perspective, but the market needs encouragement to invest in the solutions. The lack of focus on this issue in many markets is very short sighted, for the risks and costs of the current structures are substantial; perhaps more than half of all settlement cost in the affected markets.
The first problem relates to the quality and speed of transmission of data from the investor through to the global custodian. There are facilities to overcome this, with OMGEO especially having the product to enable rapid broker to investor matching and onward transmission to the administrator or global custodian. Although I may be in a minority, I cannot understand why electronic trade matching in one guise or another is not made mandatory for all regulated entities. And I would also suggest, as commission disputes can be the cause of many delays, for markets to structure matching to allow trades to move ahead to settlement at the lower of any disputed commission rates with any balance being negotiable off line.
Furthermore, although I appreciate the risk issues involved, I am amazed that matching engines are often not permitted to feed matched trades through to the designated securities services provider although they have the technical capacity to do so. The reality is that we could be agnostic about routing, whether from client directly or through the intermediation of the client approved matching engine, as long as, alongside tighter settlement cycles, regulators mandate timely delivery of instructions to the service provider who stands next in the chain. In reality in T+2 environments, these instructions are needed on T+0 or early on T+1 at the latest.
The global custodian to local agent routing is highly efficient and, with the help of SWIFT messaging and standards, is rarely the cause for delay. And local custodians, at least in volume markets, have the ability to pass the trades efficiently on into their market domain. It is here that the second problem arises, namely the inefficiency of settlement matching. There are multiple reasons for this. A key one is the quality, or rather poor quality, of static data. The advent of data repositories has alleviated the problem to some extent with a material increase in the number of LEI’s in issue. But matching to fund name, especially, remains a nightmare in many markets. The process often involves manual exception processing and “dead” periods where market side systems churn through their data. And then late in the day, with the results known, the concentration of settlement data to be sent to clients may cause communication stresses in either sender or receiver gateways.
The reality is that many markets have been keen to develop their capacity without focusing on the post trade world. The problem arises to some extent with the advent of high frequency trading, remote brokerage, and similar facilities. Obviously, where there are efficient netting engines, these flows are managed quite efficiently but not all markets accepting such activity have the appropriate CCP. Again the regulators need to monitor volumes in markets and insist on, firstly, automation of the matching process, and, secondly, performance standards among regulated firms.
In cash markets, as they moved to RTGS processes, there was a tendency by some delinquent operators, to wait for their receipts before sending out their payments, thus avoiding a need for liquidity. Central Banks stamped down hard on this practise and it is a minor problem these days in most markets. The same Central Banks recognised the interbank risk of spot settlement in the days when each transaction was settled as dealt. But they encouraged the market to develop the excellent CLS structure to ensure a sound technical, operational, and legal infrastructure to accommodate massively increasing trading volumes at the time.
Now is the moment in global securities markets to act similarly, albeit with ESMA and its counterparts in the driving seat. We need to have electronic trade matching with global custodian advice for 99% plus of transactions on trade date. We need to raise the bar in so far as it concerns the management of static data in securities markets. And we need to ensure that market side matching of settled as dealt transactions is as efficient as its peers in netting structures. Now that would really reduce risk in settlement!