I’ve now attended two major financial conferences in the last three weeks and I’ve got another two or three to go this year. In the last two, pricing in the custody world was a major discussion point and I have no doubt that it will be for the forthcoming events.

Bundled Fee Model

The custody world has many price models but the most prevalent is the ‘bundled fee’, which assesses fees to a custodian based on the assets under management (AUM) using a basis point calculator. Now that model is certainly easy to understand. However, I would have expected that there was some sophistication going on in the background to make sure that the simple end fee has some relationship to the cost of delivering the underlying platform of services and the profit target of the firms concerned.

What I’ve found over the last few years and more pointedly in the last few weeks, is that this is generally not the case. This seems rather strange not just for its obvious flaws, but also for the fact everyone seems to know and acknowledge that the model is broken, yet simple competitive fear has held back the industry from changing it.

Custodian Activites

The issue today is that the activities that go to make up ‘custody’ are very different from each other. Some are commoditised to a large extent e.g. Clearing and settlement, because they are mostly automatable. However, many others are very far from that model. Withholding tax optimisation would be a great example of a custody activity that is at the other end of the spectrum.

So, the industry faces a conundrum. It can either perform proper and detailed cost benefit analyses in order to come up with a basis point model that provides the required margin, or it can unbundle the costs and aggregate all the commoditised elements into a core ‘black box’ service, then offer value add services around that in a menu style package.

There are symptoms of the latter course visible in the account opening arrangements of many custodians where services are selected by customers on a tick box principle, which then affects the bp model slightly. But from the discussions I’ve had, all this is not being done on a scientific and detailed basis, its still almost ‘back of envelope’ stuff.

Approaches

I have performed quite a few cost benefit analyses over the years for banks and brokers. My usual staring point is to ask whether the firm offers tax reclaims and if so, whether it charges its customers. As you’d imagine, some do and some don’t. Of the ones that do offer tax reclaims, most have a very fragmented approach to pricing. Some just include the service in that all encompassing basis point model. Others, the minority, do charge their customers separately, but this too is fragmented. Services are offered reactively not proactively; some clients charged, others not, fee levels negotiated separately by client etc etc.

I’ll take tax reclaims as an example. If you’re a custodian and you do charge your clients, my guess, from experience, is that you don’t charge for relief at source at all, even though you have to do work for it. You charge for long form or standard reclaims at a flat fee in the range €50 – €150. The actual costs of performing this activity on any significant portfolio are usually nowhere near this number and are often variable in workload dependent on the tax status of the client (opaque or transparent). From that fee you have to pay for (i) tax research into a number of markets, (ii) some system – spreadsheets, third party software etc to manage the process, (iii) FTEs at the rate of 1 FTE per 12-25 reclaims, (iv) hardware, (v) software, (vi) training, (vii) space and (viii) recon. Trust me, when you’ve taken all this into account, €150 per reclaim doesn’t even get close. This model also suffers from the problem of being a fixed cost unscalable model, where most custodians will need a variable scalable service if they are to respond to the massive increases in client demand that I’ve seen coming. This relatively simple second level granular analysis relies on assumptions, but is extremely powerful because it clearly shows where the breaks in the model are between the core and the added value.

Paper Trail

The next issue is addressing that almost blinkered view from clients that fees must always go down. The industry is not as automated as it should be. 85% of all corporate actions still involve paper. And from that base, someone has to pay the regulatory costs that are piling up. It’s simply not sustainable for a single segment of the financial chain to carry that burden forever. This is an issue of intelligent education. It starts by explaining the landscape in a much more open way so clients understand what they are buying and what it takes to deliver. That opens the door to the new unbundled approach where the core services can be offered on a competitive basis but where the margins are low – a different negotiation from the now unbundled high value services. These can be offered based on need and perceived value, but importantly the cost base is better known and the service can be offered at a profit.

Future Practices for Custodians

Its about time that the industry took a really hard look at pricing, not just talk about it in conferences, in a much more commercial way. I heard at WFC2015 that the reason its never changed is that everyone is afraid of being the first, there were never enough external pressures to force the issue and clients were able to enforce a continual declining price model.

In other words, clients could argue that prices should always decline because custodians were always focusing on how efficient they were and on the transactional aspects of the custody business. In today’s world, as both the conferences highlighted, custodians are now absorbing regulatory costs at torrid pace with no end in sight. Any simple back of envelope cost benefit analysis now shows starkly that the bundled fee model era must end – and soon.

Image Credit: epSos.de