More to do on settlement
Alongside safekeeping, settlement is one of the foundation stones of the classic ‘hold and settle’ custody model, and in theory one of the most straightforward functions performed by a custodian bank. Yet figures produced by UK-based Amaces show that missed settlement deadlines remain a concern.
The firm currently provides data via its CMS analytical and benchmarking service to more than 50 clients who over the last year have executed a total of 750,000 trades worth £600bn (€872bn). Over that period, the average value of late settlements ranged from 3-4% each month – or approximately £25bn.
The CMS service allows Amaces to compare individual funds and/or portfolios within a specific asset class. Amaces’ data highlights that material variations in the timeliness (or otherwise) of settlement occur at the portfolio level each month even within what are generally considered to be efficient markets.
In the case of US equities, for example, some funds had late settlement rates of up to 20% over the course of the year, and in two particular months recorded figures of over 40%. Similarly, while most Japanese equities transactions fell within the 0-3% range, a sizeable minority were between 10-20%, and in one month hit 60%.
Of course, the blame for late settlements should not automatically be laid at the custodian’s door. Numerous factors that are beyond the custodian’s control come into play, not least brokers and their clearing agents. That said, the CMS analysis - by highlighting any trend towards deterioration, especially in the context of comparable firms and their respective portfolios - allows the root causes to be identified and approaches to tackling any problems devised.
One particularly opaque area is how some custodians may take advantage of shortcomings in the settlement process. Most custodians offer contractual settlement date accounting, in which the cash element of a trade will occur irrespective of whether the trade has settled on contractual date or has failed. Effectively, a failing sale will involve the custodian advancing cash to a client, while in the case of a failing purchase the custodian will debit the client’s cash account.
The proverbial visitor from Mars might expect that a relationship and a correlation exist between failing purchases/sales and the total number of purchases/sales. While variations over time are inevitable, over the longer term such a correlation is nonetheless expected. Yet, in reality there are huge variations in the marketplace.
With a number of custodians there is a close correlation which holds up well over time and across many different types of client. There are,
however, a number of custodians where this is absolutely not the case. The ratio of sales to purchases is in the region of 1 to 1.1 – yet that of failed sales to failed purchases is 1to 8, and in some in instances as high as 1 to 15.
What this means is that, while those purchases for which the custodian has debited the client are left languishing, the failing sales are cleared up very speedily. The benefit that the custodian derives from this practice in terms of interest payments is called ‘fail float’. As one can imagine, when custodians are settling billions of dollars worth of transactions, the earnings accrued from just one day’s interest can quickly mount up.
Not only is a client stung financially, but they are then exposed to credit risk as well as any subsequent operational problems that might arise. In theory a market claims process should go some way to tackling this, but that is not something which has been adopted by all markets, or even by many custodians.
Some will argue that there are always good reasons why purchases, especially large value ones, would lead to a fail. But it is interesting to note that, when looking across all custodians, there are unquestionably variations in behaviour. Custodians can quite fairly point out that brokers can cause purchases to fail, and there is little they can do about it. This should, however, affect all custodians equally; yet over the last two years Amaces has analysed over 1.5m trades, and variations between different custodians are marked. Clearly, there remains scope for improvement in certain instances, and firms will need to work with all their counterparties to ensure they are not losing out.
Amaces was established in 2002 and provides analytical and benchmarking services to pension funds and fund managers. The first module of CMS covers Custody and Treasury services and the analysis quoted in the article relates to 52 clients subscribing to this module of CMS.