What are you doing to prepare Northern Trust for the future?

Huge changes are coming in our industry – how we position ourselves for the future is very, very important. Today, financial services is not a utility – when a manager buys something such as a brand new OTC derivative they can’t just snap their fingers and everything just operates. It will not become commoditised, not for a long time, but we have to prepare for where the world is going. Fees continue to be pushed down and we will continue to be asked to do more for less. I have been in the business 30 years – there is a constant evolution and as long as innovation occurs there will be new things to focus on. All the literature says to invest in indexes but if the market is producing a return below what you need, you need alpha. Right now the whole world is moving towards alternatives– so that is where the innovation is happening.

What is being done to facilitate investment in these areas?

We picked one client problem that fits with a blockchain application and said ‘let’s run with it and actually get it live’, rather than having a technology lab work on 15 proof-of-concept projects. A blockchain is not hard to implement from a technology perspective. If everything matches you have a transaction and an immutable record, then you are done. The hard part is everything around it. Private equity has been a very manual asset class. In February we launched our first blockchain application in Guernsey. For an offshore fund to qualify certain transactions need to happen on the island. We worked with the government to get to the point where as long as the blockchain keys are on a server in Guernsey, the transaction is considered to be happening there. It’s very simple but it took a while. We then put together the technology and then demoed it and carried out security audits. People will not just accept it – you need to go through a process and that takes time.

How does blockchain change the role of the incumbent service providers?

It will automate about 30% of the fund administrator’s role. You have a perfect document storage facility, one document that is an immutable record, and audits become much easier. But the biggest benefit is that it shrinks the time from when commitments are made to actually spending money. People with dry powder want to spend it and we can shorten that time. The investment manager, audit firm, legal firm, administrator and regulator all have access. In Guernsey the regulator wanted to be able to ‘break the glass’ to audit it when needed. The next stage is to build in electronic capital calls and then finish off the whole life cycle of the asset class, before converting more clients. The key is to create a minimum-viable product and go live – doing things in a proof-of-concept world is just not the same. Suddenly, issues such as legal compliance come up. It has really driven our innovation.

How much of the technology does it make sense to develop in-house?

To get the talent you have got to have a certain work environment. We have an association with Pivotal Labs, which separates our teams out from the bank infrastructure. They can dress down, have an open workspace and developers and users interact to build on an iterative basis. It can be within a bank – you just need a new environment. Call it a lab or whatever you want – it is an important aspect of doing things differently. There is a reason people sit in a garage and develop things that are successful – they are unencumbered by bureaucracy. We can emulate that, but then put in controls. You need to create your minimum-viable product, and then make sure it works and is secure.

How do you attract the necessary skills?

If you ask ‘where would you rather work, Google or a custodian bank?’ the reality is the best and brightest kids are more attracted to technology firms and start-ups. We need to hire quality people and tap into those going elsewhere. Executives from Northern Trust meet with Silicon Valley representatives twice a year. Half of each meeting is to discuss technology investment and the other is about how we could use the technology. We get the time because we have the cash and the brand. We get access to the best and the brightest, to figure out what innovations we can use and bring them back, sometimes as an investor and sometimes as users.

How does this process work in practice?

As an example, FX relationship dealing has come to a crashing end over the last 10-15 years – spreads are lower and the need for proof of best execution is getting more heated. We are investing in a company called BEx and created a digital trading platform that allows clients to create alpha through their proprietary algorithms. We ensure there are perfect records and full transparency of every transaction and they can prove they have best execution. Building that ourselves would have been really hard so we took an ownership stake. Having the developers handle the innovation aspect, while we have a stake and can benefit from that, has worked very well for us.

How much further does the push for transparency have to run?

It’s just started. People want transparency. We have got to be able to create data hubs that allow access to all of the data that clients are interested in. There is a rush to become as transparent as possible – and that is huge. The nature of transparency is that fees get squeezed – you have to be prepared. 

Are investment management fees still too high?

It’s a high margin business. However, who determines that the current level is the proper margin? At some point, someone is always asking ‘is that too much?’ On the indexing side fees are getting squeezed. For active management in the middle, that conversation is happening, though it hasn’t hit the private equity side. There are more people chasing alpha than catching it. Managers that figure out how to get alpha for their clients will be paid well, because it is hard to create alpha.

Is investor services feeling the effect of fee pressure on clients?

It all rolls downhill. We are constantly being pressured on the fee side and we have to offset that with technology and new value-add products. We have a global model so we are able to balance labour costs but the reality is that technology is allowing this business to be done more easily. On the flipside it is an opportunity. We are in the outsourcing business and I am convinced that over time everything post-execution, perhaps even execution, will be outsourced. It may take 10 years, it may take 50 years. The capital needed to invest in a technology will be great and the scale to drive efficiencies will be such that we will be able to provide a much better product at a much lower price. People do not outsource when they are making 50% margins – they outsource when fees are shrinking and a capital investment is needed. It is going to become tougher for mid-sized providers to compete. They don’t have the capital to invest in technology or the global footprint. They can service niche markets but it’s going to get harder to truly be a player.

Are you concerned that technology could disintermediate your business?

Today, people like to bundle fund accounting and custody as there are synergies. The doomsday scenario is if clients no longer need a custodian bank to get your data, if all you need is something like a Paypal account and to deal directly with a blockchain. If a technology provider could see what is held, in all markets, it could put all the data together and do the middle office work but wouldn’t need to be a registered bank, as it is not actually holding any cash. Suddenly a huge barrier to entry goes away. It could be harder for a large organisation, with huge platforms, to change than for a small firm to build all of this. It won’t happen in the short term, but will we get to that position someday? Maybe.

How can custodian banks ensure they remain relevant?

You have to create a model that insulates you. Today, we compete on custody. If the custody layer is going to become much more automated and data-centric then we have to compete on information. We will be competing on data, how we make it available, how we manage it and how we present it. It doesn’t matter if we get that from doing the custody, holding it physically, or getting it from a blockchain. Hedge funds have administrators not custodians – all their assets are held by the prime broker – and those administrators do all their work without any custody. A pension plan may have OTC derivatives, private equity and investment funds – there is actually no custody there. Global custodians pull the data from all the investments together and produce valuable information. It’s a misnomer even now that you need to be a global custodian.

What problems are emerging as a result of the increasing centrality of technology?

It’s getting harder and harder to find people that understand the whole process, end-to-end. As a fund accountant, I could pick up a pencil and tell you the net asset value (NAV). Today, things are so automated that nobody knows how to do that. How do you train people to solve problems if they don’t know the end-to-end? It’s really hard to find the root cause and fix it. If we do not invest enough in training we will get stuck as an industry – people will not know how to make things better. Anyone at a senior level needs to understand technology, or they will have a huge issue.

How much more information do asset managers and regulators actually need?

We could sell you all the data in the world but it could be expensive to put together – so you need to focus on the action you might actually take. If you are going to make an investment decision based on daily data of your underlying holdings and achieve alpha, great. If not, it is probably too much. You need to take some time, figure out what decisions you would actually make, and then what data you need to drive those decisions. To understand the risks in the market, regulators are going to need more data in a more consumable structure. We already do a lot of reporting – but what is being done with that data? The regulators have the right idea and the industry is on the right path, but we are at step one. If there is opaqueness and you have a run, as with Lehman, you are done.