Technology Roundtable: Does asset management need full automation?______

Edward Glyn, Head of Global Markets

Edward Glyn, head of global markets, Calastone
Kate Webber, head of product strategy, Northern Trust
Franck Bonte, global CIO, Société Generale Securities Services

FUNDS EUROPE: The heavy use of fax machines in the funds industry is often maligned, and we regularly hear calls for greater automation. Yet the fax is, to some extent, a tool for automation. Is there a perception problem with how people define automation and automated processes?

FRANCK BONTE, SGSS: The technology around the fax is improving. Before, it was all paper; now, it is, actually, in a sense, digital, connected through the portfolio management system and automatically received by the service provider. But we cannot say this is straight-through-processing (STP).

And nor can we say that automation and digitalisation are the same things. We will see a push towards full digitalisation of processes from the receiving part to the management of entries.

KATE WEBBER, Northern Trust: There has to be a distinction between STP and automation because they’re not the same thing at all. We’ve seen in a post-pandemic world that fax numbers have dropped dramatically. The majority of non-Swift, non-Calastone type activity has moved – but it’s moved to email, bizarrely.

As an industry, we’ve got more secure data files/APIs, and we end up with a different set of issues. So they haven’t necessarily become automated, but they’ve moved on to email, and we’ve partially automated those.

EDWARD GLYN, Calastone: Often, people refer to fax automation as automation when it isn’t automation. Remember, automation has to go through the entire supply chain from front to back, but then all the responses that have to come back from the receiving parties also need to be automated. So, automation is very much bidirectional, and I think there is an issue with both acknowledgement of what happens at one’s own firm and the definition of what automation actually is because if there is a part of a process that isn’t automated, then we can’t call that process fully automated.

FUNDS EUROPE: Does the industry need to be more militant about the difference between automation and STP, as well as pointing out the flaws of partial automation?

WEBBER: You have to look at the process itself and the reason why you would want to automate it. There are elements of processing where it makes no difference if a human being is in the middle of the process, such as order routing or settlement. All those human beings are doing is potentially creating errors. However, there are other processes which are more sensitive, for example, processing an investor’s death. Here we end up with extremely vulnerable people at extremely vulnerable times in their life, and just sending something into a system in a really unempathetic way is not appropriate.

We will also find that when some people choose to invest, they’ll be quite happy to do a lot of research upfront, but at some point, they may wish for that human connection just to go through what they’ve done and make sure it makes sense.

GLYN: If you look at all the different super apps or services that we use online, the majority of the time, we’re very happy for that entire process to be automated. But when something goes wrong, you really want to chat with a human being. That human capital side of things is where a lot of firms can really differentiate themselves from their competitors.

BONTE: This is the challenge when it comes to self-service. If you are using your apps today, you are doing a lot of things that were done by other people on the service side. But we still need to think about where we want to put human connection to make sure that we are bringing added value. Automation will reduce errors but will also reduce the quality of some services for clients, so there is a trade-off to be managed.

FUNDS EUROPE: The Funds Europe Global Automation Survey showed that client service is a big driver for increased automation. Is this a move away from what were previously the traditional drivers of automation, which were cost reduction and risk reduction?

GLYN: You need both. If you look at Amazon, why they did well at the beginning is they were a supply chain, they were a fulfilment shop, so you need to be ruthlessly efficient in every part of the supply chain process in order for your business to be effective, and that still absolutely applies to whatever business you’re in, whether that’s consumer goods, factories, or whether that’s being a custodian.

Only once you’ve got that right can you provide great client service at the front end. In an environment now of competition and basis point compression, everybody’s looking for cost savings wherever possible.

BONTE: Automation should reduce cost and risk, and moving from fax to email is not changing so much, because we still have a risk of error with email. We need to increase not only automation but enable more digitalisation. We need to be able to provide services at any time for our clients so they can access their information and perform transactions. It is about the flexibility that is provided through digital transformation and client interaction.

“We need to discuss this with our clients so they can understand how these transformations are in their interest.”

WEBBER: We must remember that many of our clients’ service activities aren’t necessarily aimed at retail clients but at institutional clients, and so in today’s world, we don’t necessarily need the institutional clients to use an array of portals that we decide to introduce.

The creation of APIs and data solutions such as Snowflake will, in the long run, enable us to create mechanisms of pushing that data into our client systems directly and therefore, automation isn’t just achieved within the context of the asset servicer but also within the context of the pension fund or the asset manager themselves who can access the data they require from the systems that they use on a daily basis. That’s where you’ll start to create higher levels of automation across the industry that will enable investors to access information really easily.

BONTE: It’s not only portals, it’s also open platforms. Typically we can provide services through different channels – fax, email and now APIs. We need to discuss this with our clients so they can understand how these transformations are in their interest. This is because there is a cost involved. Generally, the products that are on the market are not so open, so we need to make the investment that will move us on to the next level. It’s quite important that we can provide different sorts of connections so the client can undergo a cost/benefit analysis and manage their costs and the risks that come with automation and digital transformation. It’s really a platform concept.

FUNDS EUROPE: What role will distributed ledger technology (DLT) play in improving automation rates?

GLYN: The mutual fund product and its structure have been pretty much the same since 1924, when the first mutual fund came out. Apart from the advent of ETFs, there hasn’t been a tremendous amount of innovation for 100 years.

Also, Citibank said that it costs them 8,000 times more to process a fund than equity.

We will see new technology, like DLT, not just enable the digitalisation of the current supply chain but also be used to radically alter the structure of the product itself that uses that supply chain. You will have a digital representation of both the product wrapper as well as a digital representation of the underlying assets within that wrapper.

“The technology could certainly help with processes like KYC and the ability to validate information in real-time.”

WEBBER: What we’ve just described will take a really long time to enact in full across our industry if we use the last 20-30 years as a yardstick, but if we were to start to consult about DLT as an industry, the area that I think it could really help in the short term is around settlement and tokenisation around settlement.

We’ve got to improve our settlement processes. Let’s not just push DLT into things that work, let’s push DLT into things that really don’t work today, and settlement is still one of those areas as an industry where we just haven’t got that licked. DLT could also help with ‘hyper-personalisation’ where instead of just buying a fund, investors pick and choose the stocks they want within certain wrappers.

BONTE: I do not really understand why this technology is not already mainstream in securities services processes. Is the initial investment too high? Should we start with minor processes to test the technology and create some reassurance?

In addition to cost, there are also concerns about the environmental impact of DLT. The technology could certainly help with processes like KYC and the ability to validate information in real-time. I am surprised that we have not seen a strong push in that area.

FUNDS EUROPE: Legacy systems and the culture of manual processes have always been seen as the main barriers to automation. Are these still the main impediments or are we seeing some new barriers emerge?

WEBBER: Legacy systems are indeed still the barrier to change because the costs of change are high and they’re also incredibly functionally rich systems that have built up over an extremely long period of time. It is only when you try to replace them that you realise there are multiple significant levels of functionality behind these systems. This goes back to the concept that a transfer agency is just a shareholder record – well, of course, we know it’s not; it’s got a lot of other rich activities and things it does. The need for data will create things in a fundamentally different way. You are starting to see new modern architectures built around those legacy platforms, which enable you to move the reporting up to create a single version of the data’ truth’, which is incredibly important both in terms of automation and the creation of event-driven architectures and keeping that functionally-rich legacy activity.

Over time I think you will then be able to reduce the legacy platform and potentially replace it with the DLT, but it will take time to do that.

BONTE: The challenge is today not so much the fact that we have to change the legacy system; it is around the data – the capacity to structure it and provide the proper quality so then we can send it back to our client or build on top of those data capacities that will help our clients to move forward. It is also not easy to understand and fully master new technology. So there is a technical challenge to make sure that we are getting the right use of that technology.

GLYN: The technology is not the issue. There are still a lot of unanswered questions on how the future landscape is going to operate from a regulatory perspective, and the operating landscape of the future hasn’t yet been fully figured out in terms of what role each of the players is going to play. The same goes for each of the participants in the supply chain, and that really needs to be worked through properly, as well as the exact economic business case of this future model.

How is everybody going to get paid? Is it genuinely worth it for firms to do it from an upside perspective instead of just cost-cutting? Because no firm ever shrunk its way to greatness, certainly not in our space, and asset management is very much a scale game.

FUNDS EUROPE: Will we see at some point in the near future a more uniform level of automation on a global scale rather than the regional disparities we see at the moment?

BONTE: We have to rethink the client interaction model and the range of new capacities we can bring to it. In many developing countries, they were probably not at the right level of automation at some point, but they decided to move several steps forward.

I am a relatively new arrival in the asset management industry, but it is a very small step at a time when it comes to innovation. We have not pushed operational transformation to the level that could be reached, and perhaps it’s time now to see if we can push it one step or multiple steps further to really make a difference. We have experts on the processes and people who know very well how everything operates. And it’s clear that we know where systems are inefficient, and it’s clear where we would like to have more connections. But there needs to be the will to move to the next stage.

WEBBER: In future, we are all going to be technology companies – whether we sell financial services or cars. Instead of seeing IT as a completely separate entity, we’re all starting to see IT as very much part of who we are, regardless of what region we are in.

There will still be regulatory nuances creating a little bit of friction, even where you have global operating models for processes like fund accounting. But in regions where you don’t have legacy technology, there is the ability to take a leap forward. For example, Amazon didn’t have all the legacy platforms and systems, so it was able to start with a blank piece of paper.

Is it regions that will drive innovation forward, or is it the emergence of new players, globally or regionally, which will drive forward? I don’t know, but I think the industry will change regardless because it has to.

FUNDS EUROPE: Will we ever get to 100% automation, and would this be a good or bad thing? Are there any potential downsides to automation that we’re perhaps not aware enough of?
GLYN: Some people are scared of automation because it equals immediacy, and if something has gone wrong, it’s much harder to fix it. If there’s a problem, you can’t unwind it, as opposed to in the olden days when things were done on paper, and they were going from person to person to person. If there were an issue, you could probably catch it before the money went out the door. Although intellectually, that’s a very fair point; it will not put the brakes on the various automation projects.

We will still see regional differences because of how clients in each country wish to be serviced. Some people like going into a bank for advice, and that is going to be less automated, but the next generation of investors do everything online and are much more likely to get advice from social media than from a financial advisor. As that comes to the fore and we see different investor trends and different distribution models in each country, we will still see a regional disparity in automation levels, even though the back-end processes themselves will be more automated.

BONTE: We will not be able to manage the cost of 100% automation, so that should not be a dream. Keep on pushing standardisation and being able to manage exceptions. You want to be close to your client when there is an issue, and if it is just a matter of saying, ‘I cannot manage your transaction because it’s not standard any more,’ you can expect him to move somewhere else. It is about finding a balance between standardisation and added value.

“We will still see regional differences because of how clients in each country wish to be serviced.”

WEBBER: I still see financial services as a relationship business. What we’ve seen over the decades is that people choose their financial services partners because they trust them; they don’t tend to choose them just based on automation.

If we think about the way that wealth is being transferred, not just in the immediate future to women but in the long-term future to the next generation, it will remain a relationship business – but in order to do that effectively and create the hyper-personalisation that we were talking about earlier, it’s essential to get rid of manual processes where they don’t add value. I don’t ever believe we’re going to get to 100% automation, but I also believe that things like settlement, order routing, dividend payments, all of the things that don’t need human intervention to make them work well, that is where we should be focused.

Edward Glyn, Head of Global Markets

Featured articles