Why custodians need to go beyond data management

Custodians may be deluding themselves about their ability to overcome the challenge of data management and the long-term sustainability of data-linked revenue flows.

Data is regularly cited as a new source of income by custodians. They are right about their assumption. But they may be deluding themselves about two key issues; namely their ability to overcome the challenge of data management and the long-term sustainability of data-linked revenue flows.

Data is needed to enable more effective fund distribution, accurate regulatory reporting, portfolio analytics and operational efficiency. I am sure there are other uses but these appear the dominant ones. The wide scope of the data needed creates a challenge for custodians. The data has to be held in coherent and compatible data bases, and rationalising data within a firm is onerous and one of the poorest performing areas for both the securities services providers and their counterparts. The data held privately also has to be linked into public data sets, especially for distribution analytics. The geography of data is important. The sectoral provenance of data is critical. And the understanding of data dependencies can change its value and meaning.

In fund distribution, with ever greater liability placed on originators and distributors for ensuring the absolute correlation of fund with client profile, we have a classical need to reconcile data from public sources, sectoral analytics and client specific information banks. The old philosophy that a postcode was the best data for fund targeting is no longer relevant in an environment where specific analytics are needed rather than general assumptions. In fund distribution, this will lead to an ever-growing demand for personal data, assuming, as is likely, that the regulatory trend continues towards “caveat venditor” or let the seller beware. This in turn implies that the retail market must move to on line apps to gather the specified information and, more importantly, assess it for the fund being proposed.  

That leads to a challenge for the industry. Although it can be argued that data at the point of sale is appropriate for one-off transactions, long term investment plans will need data updates on a regular basis. This is far removed from the reactive data storage function of current day transfer agents or fund administrators. In reality it is a new industry; a logical stop for a blockchain-style application as long as the data inputs can be trusted and, perhaps, distributors believe they can outsource securely the management of much client data to third parties or more open platforms.

Regulatory reporting, including transaction reporting, is perhaps a more logical venue for the fund administrator. As a price calculator for fund valuations they need to have transactional data within tight time frames. But it is questionable if their timeframes are sufficiently tight to meet regulatory prerequisites or that they have the data, especially on counterparties of their fund managers, in the correct format. The trade reporting industry has started with bite sized chunks, intermediating data on a market by market basis.

Universal banks, with a greater internalisation of transactions, will have an advantage over more liberal marketplaces in sourcing data. Long term, all clients will be looking for holistic solutions. The multi-custodian will want a single pipeline for all their funds and to all their markets. It is debateable who can provide such a service. Logically, it could be one of three candidates. Would it be a major global custodian, perhaps through alliances rather than on their own? Is this a space for a new utility, perhaps in alliance with a technology provider? Or could it be space for a FinTech company? The liability issue will be as fundamental as content management and data delivery. But my intuitive thought would be that an infrastructure consortium is the most likely long-term candidate and the one with which regulators would feel most comfortable dealing.

But how does one handle data that relates to operational efficiency and portfolio analytics? Operational efficiency benefits any party to a transaction and so it is illogical that the holder of the data charges the originator or other intermediates for data that enhances efficiency unless the cost of data collection and collation is greater than the benefit they extract. Global custodians can create universes and compare client efficiency, even rate it and, in theory, adopt differential pricing with discounts for efficiency. But they would need to be careful to ensure that the populations in the different universes are truly comparable. 

The efficiency of a broker-dealer will be vastly different from that of a traditional fund manager. But the old-fashioned bugbears of the business, such as continued automation phobia in parts of the fund management industry, the replacement broker-dealer trade without cancellation of the original, partial information, late allocations, abuse or misuse of LEIs and other indicators as well as insistence on paper based information for new issuance could continue to make true STP and efficient data management an impossible goal for most. Who could provide the universes? Are they the successor to the custodian surveys that are still closely followed? Or, once again, will there be space for independent providers to commingle multiple source data that individual suppliers can use as their benchmarks?

Portfolio analytics remain more client-specific especially with tracker funds becoming more popular and benchmarks thus much clearer. However, analytics must include assessment as well as data production. Fund prospectus compliance, especially in tracker or surrogate tracker funds, must be more carefully scrutinised for liberal interpretations of a fund’s remit can destroy the credibility of the chosen benchmark. In the end, I personally believe that the critical benchmark will always be the real return of a fund over time in excess of the cash alternative. Comparing that, on a five year-plus basis across funds is valid although, once again, interpretation has to be predicated by the risk appetite of the relevant fund and its investors.

In conclusion, I suspect that, longer term, we will see the emergence of trusted data banks far beyond the current fairly static offering for KYC, SSI or securities and corporate indicators. There are evident openings for the major infrastructures, most likely through collegiate undertakings, especially in the regulatory reporting field.  And I see openings for new entrants with smart FinTech competencies, especially in the fund distribution space, although any solutions will require close collaboration between regulators and suppliers. Space remains in the operational data and in the traditional performance analytics fields for the custodian, but I suspect that the golden chalice of a major new income stream from data is far from a certainty and perhaps, without a radical change in structure, far removed from the core competencies of most in our industry.