Pensions Pod S8:E2 – Why do pension scheme data projects matter more than ever?

Emma Moore, Associate Partner and Actuary, Aon (00:52)

Yeah, and thank you for having me in this really cool place for the recording. So I am, as you say, I’m an Associate Partner, so I’m a Scheme Actuary. I also lead Aon’s Bristol Retirement team. So, we have 90 consultants now. So that’s across actuaries, governance, defined benefit and defined contribution pension schemes. And then lastly, I also support our Aon’s Administration team. So, the business-as-usual team but also our data solutions team which is our data projects team.

Helen (01:20)

So that’s really what we’re discussing today. So, Emma why do you think data projects are such a hot topic in the industry right now?

Emma (01:26)

You know, I think it’s just data is so integral to everything that goes on with the pension scheme at the moment. And I think probably in the past, you know, administrators were probably seen more as a back-office function. So, you know, processing transactions, doing data projects, but as and when they were needed, you know, on an ad hoc basis, not really thinking about a wider strategy, whereas that’s kind of had to change because there are so many projects going on at the moment in the pensions world. So, you know, we’ve got things like pensions dashboard, GMP equalisation, rectification, we’ve seen huge increases in funding levels and pricing is at a really good level. So a lot of schemes are accelerating that path to buy-in and buy-out transactions. So there’s a lot of projects and all of those involve data. And I think it’s really heightened that importance of if you don’t have high-quality data to do all of those projects, we’ve seen, you know, big costs coming through, extensions to deadlines, it’s having a big impact. So I think that’s really elevated why data projects are so important today.

Helen (02:27)

I think, yeah, you just fall at the first hurdle if you don’t have your data right at the outset. As you mentioned, the pensions dashboard, so our listeners will be aware, but the deadline for that is the 31st of October this year. Well, members will be expecting high qualities of data so that they can rely on what’s in the dashboard, even though they’re not meant to strictly rely on it. I think the regulator’s also mentioning that data quality is such an important thing and actually really needs to be on trustees agendas, and it needs to be a role that they’re thinking about, rather than just outsourcing it to their advisors, and also just have really rising expectations from members about what data they can receive, how they’re involved in their data and what’s going on with the scheme. So having accurate data is really essential to that.

Emma (03:18)

I think it’s really elevated it and made much more awareness for the importance of data.

Helen (03:22)

So what type, you mentioned a few there, but what types of data like projects are you doing at the moment?

Emma (03:30)

Lots I think is probably the answer. So I mentioned sort of pensions dashboard readiness, GMP equalisation and rectification, benefit audits, member tracing exercising, data cleansing, contingent spouse exercises as well. There’s lots of different things there, and if you’re readying for an insurance transaction, there’s lots of work to be done as part of that. So I think there’s lots of projects out there and I would say what we’re seeing from experience is, is that schemes aren’t typically just doing one project. They might have multiple projects on the go at the same time and then you add into that challenge, the data set isn’t static. You know, data is constantly moving as members decide to transfer or retire or update their information. So you’ve got lots of projects on the go at the same time with data that’s constantly changing, so there’s lots of challenges there.

Helen (04:20)

So what do you think, what are the challenges there are for these trustees with their data?

Emma (04:26)

I think, so for me, I think there’s probably two big challenges that I see from my day-to-day work. I think the first one is the big misconception that data cleansing is just changing a couple of numbers.

Helen (04:40)

Yeah.

Emma (04:41)

You know, just updating a spreadsheet for a couple of missing items.

Helen (04:43)

So with data cleansing, do you mean in a buy-in transaction or just in general?

Emma (04:47)

In general. So there’s so many projects that rely on, you know, making sure that data is accurate and that you have all of that information. And a lot of that actually involves a lot of judgement. So you might have multiple data sets. Some of those might have conflicting information, you might have missing information. And actually you need someone there to take a judgement on how do we use the data that we’ve got to make sure that we have the correct output there? So for me, I think it’s a lot more complicated. And if you’ve got a trustee board and a sponsor who really understand the benefit of doing that data cleansing work, I think that makes all of the projects then a lot smoother.

Helen (05:26)

I would agree, and you said that already, but schemes really do have very crowded agendas. They need to be ready for the dashboard. They need to be, maybe they’re doing GMP equalisation, rectification. Maybe they’ve transacted with an insurer and they’re getting all their data ready to complete those transactions. There’s so much going on, but trustees need to step back and think. The data quality is really essential to all of that, like you said, I think one thing is obviously trustees can lean on their administrators, but also they can lean on their legal advisers as well.

So just this week, we can help lots of trustees with questions that they have around how they can pragmatically approach data, what kind of detailed review they need to do, whether and even- so just yesterday advising trustees on what to do if the data simply doesn’t exist. Yeah. So schemes, some legacy schemes, lots of different sections. And actually you need someone to come in and say this is how it’s reasonable to approach this because people can just not know how to approach it at all.

Emma (06:35)

And I think that’s so important. And I think they need that comfort to understand it’s okay if we take this approach and here’s a way forward when you do have that blocker and you don’t know where to go. And I completely get that trustees, they have so much on their, you know, on their plate at the moment. And it is trying to balance all of those priorities, and I think it’s not just trustees that are really busy administrators, you know, the other biggest challenge for me, I think, is resource.

Helen (06:57)

Yeah.

Emma (06:58)

Administrators at the moment, I think are the busiest they’ve ever been. So you’ve got your business-as-usual work, but then you’ve got all this data project work that’s come out there. So there’s lots of work on the table. And I think the administration market’s probably changed a bit as well I think since Covid we’ve seen much more job mobility. So people moving between different administrators and then we’ve got more competition because we’ve got insurers coming into the market as well, wanting to look at, you know, that in-house.

Helen (07:25)

Yeah.

Emma (07:26)

Exactly. So, I think there’s more competition there. So one of the things that we did actually a few years ago, which I think has made a huge difference for us, is we moved our data projects team out of our administration business and into our wider retirement business. And from resourcing that just meant we had so much more access to a wider, you know, group of consultants. And that team now has actuaries, it has project managers, coders as well as administrators. So for us, opening that up to a wider talent pool has made a huge difference and allowed us to keep pace with just the demand that’s out there at the minute.

Helen (07:57)

Yeah, I think there’s no lack of work out there to be done.

Emma (08:03)

Exactly.

Helen (08:03)

I think just one last challenge that I can think of is really quite a lot of aspects rely on other aspects. So, we can’t do the legal review until certain things have been done, or administrators need to tell us what approach they’ll be adopting before we can say, if that’s reasonable for the trustees to agree to things. There’s all these stacking dominoes, really.

Emma (08:26)

Yeah, exactly.

Helen (08:27)

So we’ve done the doom and gloom. We’ve talked about the challenges so, what do you think trustees can do to make these data projects more successful?

Emma (08:34)

I think, I mean, it links in with part of what you’ve just said there. And I think for me, it’s elevating that administrator role and sort of treating data strategically, I think rather than, you know, sort of the back-office function that we talked about before, I think it’s really understanding how can we look at this holistically rather than as isolated exercises.

So for me, I think trustees can be thinking about what’s our end goal here? What, you know, what are those end objectives, what data issues actually genuinely matter? Which ones do we have to solve? And that’s where, you know, linking with other advisers is really important and legal advisers can be a huge help. And then also, have we got any other future projects on the go and actually, can we combine them and can we look at that all together?

Helen (09:18)

There are people definitely- or at least definitely GMP equalisation and rectification very much done at the same time. And it involves the same people, so that makes sense.

Emma (09:26)

Exactly. And I think we- so we’re seeing much more demand across the industry for what we call sort of strategic data consultants who essentially look at all of the projects that- the data projects that you’ve got on the go and can take more of that strategic view of how do we go about, you know, getting these finished, which runs are best to start with, how do we fill in all the gaps that we’ve got? Are there any bottlenecks that we need to look at? And I think if you can have someone who can do that across the board, it just makes things a lot more efficient.

Helen (09:54)

Yeah. And it’s beneficial for the trustees, but also for the sponsors, particularly if schemes are looking to buy-in or buy-out. They want to achieve the goal, they don’t want to get stuck at certain hurdles.

Emma (10:08)

Exactly. And where we’ve seen, you know, where we’ve done this in practice, it can save hundreds and thousands of pounds. You know, we’re not talking about small amounts here. And it can, I think, like you say for sponsors that it can bring forward that time to your end goal. Whereas actually sometimes we’ve seen that’s had to be pushed back if the data quality is not there. But I think all of that involves elevating administration, bringing it to the table, getting the administrator in the room when you’re talking about what’s the timeline look like, what resources do we need? Because they can provide input into what’s actually realistic, and then they can also link in with the legal advisers and the actuaries as well, to make sure that everyone’s working together, everyone knows what that common goal is and you can sort of pull out those problems at an early stage.

Helen (10:48)

No, definitely. I think well, the theme here really is that investing in data is always going to lead to more successful schemes. The regulator really emphasises, well, I guess today’s episode is really saying this isn’t a one-off thing and the regulator has advocated that improving data is a continuous exercise, it’s not a one-off exercise.

Emma (11:11)

Exactly.

Helen (11:12)

And trustees really need to take an active role in this kind of data management, checking that everything is going in the right direction. So, the regulators initiative recently have said many trustees defer these kind of data queries and data outcomes to their administrators, but really they need to have an active role in this as well. So, the regulator suggested perhaps having it as a standard item on trustee meeting agendas, or even where there are bigger schemes that do have more of a resource, creating a subcommittee to, kind of, control data quality and just have it as a continuously evolving thing rather than just this one-off thing that we think about when you’re reaching the end of the scheme, because, as you say, data changes all the time.

Emma (11:59)

Exactly. And we’ve seen that, you know, where schemes and trustees have done that, it’s just made everything so much smoother and much more efficient.

Helen (12:06)

Well, on smoothness and efficiency, I guess I can come to the next question, which is how can AI and automation really get involved in this conversation?

Emma (12:17)

So, I think AI and automation, it can have significant efficiencies and advantages, and there’s no doubt about that. We’re definitely looking at using it, for example with file reviews. So it can be done and can have benefits.

I would say I don’t think we yet have it as a magic solution.

Helen (12:37)

No.

Emma (12:37)

I think there’s probably a difference between what AI can do in theory and what it can currently do in practice.  I think with administration and if you’re using AI, you need the underlying data to be correct for AI to produce something that has the right output. And with administration, you know, there might be multiple, as I mentioned, multiple records with different conflicting information that needs to be thought through. There might be things in people’s heads that haven’t been written down.

Helen (13:07)

Yeah, there are- we do have schemes that have a challenge of someone who’s been in place for a long time. Maybe it’s in-house administration and actually, what would happen if that person retires or has to stop working? You need these things written down and recorded.

Emma (13:23)

And the same with, you know, trustee discretions as well and admin practice. Those things aren’t always, you know, put into rules. So, it’s things like that there’s lots of information that’s kind of floating out there which hasn’t yet been recorded. So I think it’s very much, use it, but I think use it and work out on a scheme specific basis how does that work for you? And have some plans and controls around how you use it.

Saying that, it’s evolved very quickly in the recent years and it will do again, I’m sure, in, you know, in the next couple of years. So I think watch this space. We will see it being used more. And, you know, as we do, we’ll get more examples of how we can use it efficiently and the right way, but as long as there are those, safety controls and measures around it.

Helen (14:06)

I think we have seen a big increase in demand of trustees asking for AI training on their meeting agendas.

Emma (14:13)

Which is great.

Helen (14:14)

Which is great, and from all players, not just from the legal advisers. We want input from everyone and actually trustees looking to understand how they themselves can use AI, but also how their advisors are using AI. So administrators being encouraged to ask, well, on behalf of the trustees to ask all the advisors, what is their AI policy? How will my data be used? And what, kind of, what are the right uses of it? Can you put in confidential information? Do you have a policy that’s going to control how trustees themselves are going to use it?

Emma (14:53)

Yeah.

Helen (14:54)

But I really do think it’s kind of become well, yeah, it’ll be involved in all areas of these processes and the regulator kind of anticipates that. So, the regulator’s come up with its AI plan. And it says that it expects pension schemes to have good governance and frameworks around AI. But it’s clearly anticipating that schemes use AI. And in autumn this year, the regulators are going to produce some further guidance on the use of AI, so I think that will be really game changing for the industry, for trustees to know how, well what’s expected of them really.

Emma (15:31)

Yeah, exactly. And I think there’s the flip side so there’s, you know, what are advisers and administrators and trustees, how are they using AI? Also how are members using AI? And I know that’s a completely separate conversation but that again, there’s some real advantages for members there. But there are also some disadvantages and areas to be on the lookout for with, for trustees on what a member is doing with that information that they get from AI.

Helen (15:53)

And whether, even just in a practical sense, maybe there’ll be a lot more member complaints or things, because people can use AI to speed up the process.

Emma (16:03)

Yeah.

Helen (16:04)

Just lastly on this, obviously Burges Salmon and Aon are working very closely in the AI sphere. So Chris, my co-host spoke on Aon’s webinar recently about governance and then your colleague Dominic Makemson had a slot at our recent AI conference with Quietroom so, really looking forward to continuing to stay involved in this as it evolves.

Emma (16:26)

Definitely.

Okay, Helen, I’ll hand one over to you I think this time then, I guess as a sort of close. What practical advice would you give trustees that are embarking on a data project?

Helen (16:38)

I think, three approaches I would say, all the things I think we’ve said before. So, data audit trails and documentation is so significant for trustees, I think. Particularly if they’re seeking to take a pragmatic approach to certain questions around data, it really needs to be recorded so that in the future you can know why you reach that decision. I think I say that on every podcast but it’s just so important.

I’d also say that, as we’ve mentioned, data projects aren’t just a one-off exercise, only to be used in certain instances with the lifetime of your scheme. It’s a constantly evolving thing. It really needs to be a standing agenda item, you need to be thinking about it regularly and working with your advisors.

And then lastly, I would say seek to lean on your advisors if you want to as a trustee. So we’re here, we can, if you feel like you’ve got a knowledge gap, we can give you training not just on things such as AI, but also just on how to approach trustees decision making. These questions that come up and they are difficult, and sometimes These questions that come up and they are difficult, and sometimes lawyers are only involved in the later stages, whereas actually you can involve us at any point and get some advice really. And that’s not just from the legal advisers, also your administrators, also you’re actuary, just, there are people out there to help make this more understandable.

Emma (18:08)

Yeah. I think, I completely agree that, I think the bits I’d add is, please, please get your administrator, you know, into that room having that discussion from an early point. They can provide so much information on what’s realistic from their side, you know, does it work for time scales because you don’t want to start off already delaying your project because you haven’t looped that in.

And then just the last one is, think about, you know, think holistically about all of those data projects that you’ve got. Have you got someone who can who can consider this from a from a top-down view? For example, if you’re going to be accessing member files for your data project, you only want to be going through individual files once. So if you’ve got other projects that are coming up, think about what data they might need and do it all in one go.

Helen (18:50)

I guess the same with tracing exercises and things that will be so- there are tracing agents out there to help with these things just because data is missing doesn’t mean that it isn’t out there. It just takes the time to find it.

Emma (19:01)

Exactly.

Helen (19:03)

Well, thank you so much Emma for coming on the podcast.

Emma (19:05)

Thank you for having me.

Helen (19:07)

Thank you for listening to the Burges Salmon Pensions Pod. So that was our second episode of our eighth season. In our next episode, we’ll be discussing pensions in divorce, so that should be a really good one to look out for and I hope that you enjoyed this episode and you can find any of our previous episodes on Spotify or wherever you listen to your podcasts. Thanks for listening.