'No company without technology': Jupiter's Paul Kay on data science

'No company without technology': Jupiter's Paul Kay on data science 1Paul Kay, Jupiter’s head of technology, reveals some lessons from his previous roles in the banking sector, lays out his ambitions for improving the group’s capabilities in data science and admits his bias in believing technology to be critical for asset managers to remain competitive.

How much of ‘change’ across asset management is driven by the need to stay at the forefront of technology and how much is the technology there to enable the sector’s requisite changes?

There’s a spectrum of examples; things like cybersecurity, where you absolutely need to stay right at the forefront and if you don’t there are serious ramifications. And there are things like our sales people having up-to-date mobile devices so they can do things as simple as making a phone call but also connect back to the office, have data at their fingertips, for instance, that is also really important.

With areas like order management systems or portfolio management systems, if you’re not keeping up with your peers you’re going to have some serious issues; you may struggle with efficiency, operations on a day-to-day basis or dealing with the next regulations.

And I think at the more exciting end of the spectrum are things like robotic process automation (RPA), developments related to data science such as AI and machine learning. If you’re not thinking about these emerging technologies, you’re probably going to find in time that you’re falling behind.

What I’m trying to push for at Jupiter is a bit more focus on innovation, where there is a lot of scope. There’s also the reality, where things come up that require a technology solution, such as new regulations like MiFID II and GDPR. You could probably deal with both of them manually, but you will have a more difficult time doing so.

Can you tell me about one of the larger change initiatives over the past year, describing the challenges and benefits to the business at Jupiter?

The biggest thing we did last year was putting in Aladdin – BlackRock’s proprietary investment management platform, or what they call an IBOR, an investment book of record.

By putting it in last year we made the decision it would form part of our solution for addressing MiFID compliance, which required us to make sure it was implemented before January 1.
The short timeframe was the first challenge, and the second was managing the change itself. Not just the implementation and all the integration challenges, but just managing the internal changes and yet we landed it successfully on November 27.

If you talk to the users you’re going to get mixed reviews, some are still getting used to it as will always be the case with a big operational overhaul. But in terms of the benefit to the business, we took a bold step because we didn’t need this at this point in time, we could have run our business on the current platform. The reason it’s a bold move is it’s putting it in ahead of time, ahead of growth, ahead of real complexity. We’re already seeing some immediate benefit in terms of operational efficiencies of things such as corporate actions, but the real benefit is yet to come.

How do you see technology developing into the other business areas, such as sales, marketing or execution, for example?

I could be a biased because I’m a technology person, but I fundamentally believe that no company can be successful without technology. I see that every single area of our business needs to not just adopt technology but embrace it or they’ll fall behind in time.

Sales could benefit massively from understanding sales data better, be it their pipeline, current customer base or trends. Who’s buying our funds, selling our funds, why are they buying and why are they selling? Looking deeply at that data and correlating that to market events and seeing if there’s any interesting trends or patterns could prompt sales people to make optimally timed calls to clients, which can help with asset stickiness.

In marketing, it is probably quite straightforward. I think it’s about understanding your clients and serving up really relevant personal information versus one broad bit of marketing material that goes out to everybody irrespective of whether they’re an individual, a corporate, an institution or ignoring their level of permitted investment.

Do you think the types of collateral currently being used will become increasingly digitised versions, or rather there’ll be completely left-field communications coming in to market funds, perhaps taking lessons from more consumer-facing sectors?

If you went to the Jupiter website once a day for the next 10 days, you would have a very different experience to if you were to visit Amazon 10 days in a row. Repeat visits to Amazon see you hit with all sorts of information nudges. I expect we will see more personalisation; perhaps the site knows it’s you, tracks what you’re interested in and starts to serve up very specific marketing material that is aligned to your interests, behaviours and fund-buying patterns. I see that first and foremost as the change that’ll occur.

Are there any areas across asset management more generally, that you see as quite ‘broken’ in terms of business processing and how technology might serve to fix some of those issues?

One is the basic movement of data. If I go back to our Aladdin implementation and its challenges, we put this big system in, but we actually have to take data from Aladdin – out of the BlackRock environment – and bring it back into Jupiter. We store it on something we call EDM – an enterprise data management system and from there it goes on to client reports, regulatory reports and is used internally for various things.

The movement of data between organisations is painful to do and every firm that implements systems like Aladdin has the same challenge.

I think as an industry we need to build some utilities to make this easier, or take advantage of emerging data technologies, as we are all basically mapping and modelling and translating the same data but on different systems.

In very simple terms the challenge is similar to translating languages, and of course, for clients we actually do need to translate into different languages. More consistency and more data standards in terms of how we move data around would really help.

From a client perspective, I think we can take a page out of the banks’ book. If you logged into your online banking account today, you’d have your balance that same day.  If you got a note that said you couldn’t get it until the end of the month, or five or 10 days after the end of the month, it wouldn’t be acceptable.

Yet in the asset management world we still have processes where clients have to wait to the fifth or the 10th day of the business month to actually get information that was relevant to the previous month.

The best example would be factsheets of our funds.  Every month we produce a fund factsheet, which shows the top 10 holdings, some performance data and commentary from the fund manager. And we just don’t do that in a timely manner – they are out of date by the time they are released. I think that’s a massive challenge for the industry and it’s been an issue for over 20 years.

Is there scope for a more personalised or interactive factsheet then or are people wary of things that veer too much from what they are used to?

I hope in the future, rather than just a commentary from the fund manager, we might also see things like embedded video, that would be an improvement. Perhaps a link where you can FaceTime, at a pre-set time into a conference with the fund manager and ask some questions. It would certainly be nice to see some evolution from static pages to a more interactive, engaging format. Also, we typically display the top 10 holdings so again maybe something more interactive that allowed you to drill down and understand those holdings a little bit better. Perhaps the tools that wrap around that particular fund, could allow you to look at peer products? Maybe you’re interested in US equities, what else is there in the sector and how do they compare? These are all possible future developments.

What about insourcing or outsourcing? Are there certain areas that lend themselves more to one model or the other?

I come from the big banks where you even have outsourced arrangements within firms; big technology department in India and China with contracts with providers, and consultancies within that. I think what Jupiter has in place, where other than the sales folks everybody is in the same building is a powerful model.

If people are on the help desk or service desk, or running projects internally, having them close to the business and being able to interact, go to the desk and ask in person what’s going on is absolutely critical for me. I know technology allows us to be distant but I always try to stress what a powerful model it is having everybody together.

But there are things we should consider outsourcing. One example is pushing some of our infrastructure out to one of the major cloud providers, such as Microsoft Azure, Amazon AWS, Google Cloud Platform or IBM rather than doing that ourselves.

Maybe have a framework to the side. As you take on new technologies to solve business problems, such as Aladdin, consider what you get from the provider when looking at what you bring in-house. Aladdin is basically like a platform-as-a-service – it’s their support desk not mine that deals with the bulk of issues. I prefer that to a full outsourcing arrangement where you’re lifting and shifting your teams to somebody else for some salary arbitrage or tightened cost efficiency, which I don’t see as a good service model.

With regard to talent, how do you see the skills you are looking for when you come to hire evolving over the next five years?

For us, because of our size and because of some choices we’re making, we’ll have fewer people with pure development and coding skills and more people who are integrators and configuring platforms. There are so many great products out there, we will probably limit our development to the things that really will give Jupiter an edge. There are the specifics as well, such as our old investment management platform was Charles River, so that required people with Charles River skills. And now we have Aladdin so we need people with Aladdin skills.

I see over a five-year period a massive upskilling in a couple of things, one being cloud technology.  We’ve literally just appointed someone as a cloud architect. It’s the first person with that title in the history of Jupiter and we’re starting to address the skills and the roles and the people we need for roles like that.

The other area in which we’ll need a massive upskilling is data. One of my priorities for Jupiter is building a data science capability. I fundamentally believe that understanding our data, how to analyse data, take advantage of new technologies, and apply data science itself but also have end users that can use tools like that Tableau or Power BI. It’s critically important, so I see that changing as well.

As an active manager, data science could help fund managers with investment decisions by bringing them insights from current data they have, maybe behavioural patterns within how they work and trade. It can also help provide insights from alternative data sources, such as footfall traffic to Starbucks or Legoland through geolocation data.