Many years ago, I took part in a panel where the concept of unbundled pricing was discussed as the next new development. The subject is now back on the agenda with a twentyyear gap since my first experience of the then described “inevitable phenomena”. So is it a concept that has at last become possible; or is this yet another example of an industry looking for a way to beat the apparently inevitable downward curve of pricing and upward surge of risk?
There appear three ways to unbundle pricing, either to sub-divide services into components and charge for each component, or to adopt risk-based pricing, or to adopt a combination of the two. It is clear that buy and sell side proponents of this concept have diametrically opposed and irreconcilable objectives. The buy side, especially on the broker dealer front, see it as a vehicle for cost reduction, whilst the sell side expect more revenue sources to lead to more revenue!
Unbundling is, as I noted, no novelty. It has already happened to some extent. The ICSDs offer a form of unbundled price and I challenge any network manager who can easily and accurately compare the cost of intermediating through one or another of these institutions. Many custodians have unbundled settlement to the extent that they differentiate between STP and non STP transactions. But, although there are some end-to-end markets where a pure STP and confirm back is viable, defining and agreeing STP numbers is still often a challenge.
Operational unbundling has a problem to the extent that the current bundled proposition means certain functions are “free” or, in reality cross-subsidised. The cross-subsidy is usually from the ad valorem fee rather than the settlement one. Account opening, reporting, corporate actions, income collection or messaging are all often embraced by the ad valorem. Tax reclamation, proxy voting and more recent additions may be more commonly separate charges. So the first challenge for the unbundled proposition is what is to be charged and what exactly does the charge cover?
Risk-based unbundling also creates a problem. Generic product risk is covered by the ad valorem charge. It is true that the major risk, at least by historic loss experience, is in the corporate action field. But it would be wrong to assume that this is the only risk covered. There are a raft of other risks, from KYC process risk through messaging volume/cost risk or infrastructure outage risk that are also covered through this charge (by default as there is normally no separate charge for such risks). So it may be simplistic to assume that there should be a nominal account set-up charge and then risk based pricing just on corporate actions.
And risk-based pricing is also difficult to assess. As a former chairman of more than one risk committees in both infrastructure and financial institutions, I would look at risk quite simply. I would want to know the quantum, the term, the scope and the method of repayment. How can this be applied to a corporate action? A voluntary corporate action is more risky than the mandatory ones. A multi-choice corporate action is more difficult than a single-choice one. The timetable for the corporate action can change the risk profile. Regulation in the host country will affect the process. And the size of the holding impacts the quantum of the risk, as does the duration of the action. If the bank prefunds the action, then the source of repayment is dependent on client credit worthiness, for sale of entitlement is a fall-back in the event of non-payment only. Quite simply, there are a lot of moving parts in every corporate action. Should each event be treated as a separate one for pricing purposes or should one bundle events into categories? And would each category of event be subject to an agreed fixed charge or a variable one depending on the sell side risk assessment from time to time.
And settlement unbundling could be even more traumatic. At one level it is beautifully simple. The transaction will be charged at a fixed cost for STP or non-STP settlement. But should non STP settlement be further sub-divided? One charge in markets with telephone matching, another for those with file exchanges and a third for those where there is a need to physically meet the counterpart. And how should DKs be charged? One price for a message alert of a DK with error code (as often received from a CSD), but what charge for a repair at the agent without calling the client? And how much for telephone intervention? I, once, was given an analysis that showed for every dollar of cost of an STP trade, we had $15 of cost for one we could repair in house (usually missing ISIN or counterpart code) and $50 when we had to consult directly with the client. Do clients prefer such costs to be averaged or to be charged on a case-by-case basis? If the latter, then the cost of operation will be dwarfed by the cost of billing challenges!
But that is not the worst part of settlement. If I charge separately for risk, primarily counterparty (as the underlying transaction acts as collateral rather than the source of repayment), how does the bank do so? They need to measure risk. Therefore they need limits. And limits will change from time to time as the banks assess their appetite for the relevant counterpart. As importantly, so will the price of risk to that counterpart. Logically that price should be geared to the CDS price (if available). In reality a fluctuating price for settlement is a challenge for the buy side to manage, fails to align to their business model and would most likely disrupt settlement flow in markets. The concept is only feasible if it ensures adequate funding, low pricing volatility and dependability.
There is much to say about the debate on unbundling. Recognising efficiency by lower prices to match lower costs would be admirable. Pricing for client risk is also commendable, and is already reflected in many of today’s term bundled pricing models. But before being bewitched by the mirage of lower pricing, aficionados of the unbundled world should carefully consider market stability and operational efficiency rather than heading blindly for the unknown!
John Gubert
Chairman
Global Securities Services Executive Committee