Celebrating the third anniversary of our Cost Transparency Service
29 May 2020
Q. When did you launch your Cost Transparency product and why did you think it was important to do so?
A. We launched the UK’s first pension fund cost reporting service and dashboard in 2016, when we were still known as KAS BANK. Initially, we worked on a proof of concept with three UK Defined Benefit (DB) schemes. We were looking to leverage our experience in the Dutch pension fund market, which is renowned for its robust governance model, and bring it to the UK market. We identified the rising trend of transparency and saw a unique opportunity to position ourselves as the governance partner of choice for pension funds, fully aware of the data issues within the pensions’ arena and also the lack of transparency on the total cost of ownership, particularly around transaction costs. Our mission was to empower asset owners with data they didn’t have easy access to, and we demonstrated that in our proof of concept. One of our trail clients, SAUL, which provides pensions for higher education employees, signed up full time and is still a client of ours. In fact, we now have four years of data from them and are starting to look at how this cost data trends over time. We then successfully launched the product to the wider market in the summer of 2017.
Q. How does your proprietary cost transparency tool work?
A. By way of background, first we identified a data challenge, particularly around cost transparency and the investment costs associated with running a pension scheme. We learned much from our experience in the Dutch market, where they first identified this data challenge and how unmanaged fees can eat away at savings. The Dutch subsequently created a regulatory framework and rules in 2011 and 2015. In 2011, it was a voluntary code, but by 2015 it was a regulatory requirement for all Dutch DB pension schemes to report on the total cost of ownership to the Dutch regulator.
We identified the same challenge in the UK, and saw that the topic was gaining momentum, so we decided to invest in a solution that would have the same governance standards as the Dutch model, but tailored to meet the specific requirements of UK schemes, including meeting their own internal governance requirements.
Our service and data visualisation dashboard focuses on collecting cost data from the asset managers and various service providers that the pension schemes employ. Over the past three years we have developed a robust end-to-end process that helps asset owners confidently understand the bottom line cost figure of running their schemes, and how these costs can be attributed. This data, in aggregate, captures the total cost of ownership of running the pension scheme. We have a strong emphasis on data cleansing and enrichment, and once our team is confident everything is accounted for, we then build out bespoke on-line reports, board reports and power point presentations for the pension scheme’s internal teams, whether they be investment or governance committees.
Q. Can you tell us about the App you created and how it works?
A. We understood the data challenge that existed with collecting cost data from asset managers. But we also wanted to put ourselves in the shoes of a pension fund trustee. We were collecting a lot of data from a wide range of asset managers and pulling that together into a digestible report. However, we realised that the end user – the pension fund trustee or the investment team, for example – needed to be able to visualise and slice and dice this data to make sense of the costs that were being incurred by the pension scheme. They say a picture paints a thousand words, but in this sense it really does because the App brings the cost data to life and enables trustees to visualise and better understand the cost story of their own portfolios. In the UK, starting with £3 billion assets under management in 2017, we have now conducted cost transparency reports on over £225 billion in UK assets under management.
Q. How is the cost transparency story shifting towards value for money?
A. We started in 2016/2017. At the time, data standards around cost transparency were not ideal. It was a bit like pulling teeth in terms of collecting data on behalf the pension schemes. Some managers were more forthcoming than others, but overall the quality, completeness and willingness of asset managers to report this information was quite low. But within the three-year period I’ve been focused on this, the data standards, quality and willingness has been transformed – behaviour and attitudes to transparency have changed for the better.
This was helped by a series of industry initiatives. In particular, the LGPS code of transparency and the regulatory guidance by the FCA on DC funds. This meant that asset managers working on local government schemes now have to send cost data to their respective pension fund clients if they want to service the local government pension schemes. On the DC side, with the rise of auto-enrolment, more money in DC pots, and the rise in minimum contributions every year, there’s a lot more investment risk on members. Consequently, there’s more of a governance push in terms of making sure that each of these investments in the DC space are representing good value for money to the end user, who is ultimately taking on the risk. The days of a guaranteed DB pension are long gone.
The rise of the Cost Transparency Initiative was the final nail on the head, solidifying cost transparency’s position as a governance practice, both in the pensions and asset management arena. What was key was getting the buy-in of all the key industry bodies, including the PSLA, which represent the pensions industry, the IA, which represents the asset management industry and the LGPS, which represents a pool of local government pension assets, equivalent to around £300bn plus of assets. It was a combination of a clear mandate from the FCA to push cost transparency as an agenda item and the combined forces of these industry bodies. As a result, cost transparency is no longer ‘a nice to have’ but a ‘must have’.
Making this cost data readily available was the great first hurdle. Shifting the focus on value for money and improving members outcomes is the next challenge.
Q. What research have you done into pension fund costs?
A. As I mentioned previously, we have been able to leverage off the experience we gained in the Dutch market. Indeed, our organisation has been working within its regulated cost transparency regime since 2011. When they first started to do cost transparency, we saw a lot of research that suggested there was around a two to three times higher investment cost of running the pension scheme than previously anticipated and reported in various annual reports and accounts.
This was replicated in what we were seeing in the UK as well. To explain the disparity, it’s worth noting the lack of transparency on transaction costs in the past, where pension funds would never have caught sight of these fees before. But that’s not to say these are new fees being incurred – these fees were always there, but there was never really a defined process to report on this data in the past. This is a large driver of where these hidden costs were uncovered, as our research has identified that transaction costs can make up to 25% - 37% of the total investment costs of running a scheme. I also believe that a good governance model and a robust framework for collecting, identifying, analysing and reporting on cost information has ensured that the data quality standards of reporting the total cost of ownership – and all the costs related to running the pension scheme – has revealed things that pension funds have either overlooked or missed when reporting on cost information in the past. It was either previously unavailable, as in the transaction example, or simply the result of a lack of good governance practices around cost management.
With more data becoming available, we’re now starting to identify the nature of these costs, how they trend over time and how asset manager selection or asset allocation can impact the overall fees and resulting returns of pension schemes. A lot of this research is now feeding into the educational work we are doing on cost transparency, and we hope that this will better inform asset owners, enabling them to make the most of this newly available data and empowering them in their value for money assessments.
Q. Can you provide a breakdown on the total investment costs?
A. Yes. Our research has found that on average 42% are management costs, 37% are transaction costs, 10% are performance fees, 6.5% are ongoing charges, and 4.5% are for custody, fund administration and investment advice. Private equity, real estate and hedge fund mandates are the highest costing asset classes, but they are typically (but not always) associated with higher returns.
The majority of transaction costs are driven by implicit transaction costs. These capture the spread of the transaction and the market impact costs. These implicit costs are not direct costs to the scheme. Instead, they represent how efficient each fund and its managers are in their trading activities. There is also a growing focus on best execution and this data can help facilitate further analysis.
Q. What role are asset managers playing? I understand they are now obliged to provide information on costs to pension funds.
A. An asset manager’s primary function is to manage an investment mandate on behalf of the pension scheme. That’s its primary role and function and in the past I don’t think they thought of themselves as providers or reporters of data. Historically, that role was reserved for the fund administrators, and potentially the custodians as well, providing high level data on what’s been going on within the fund. However, I now think that asset managers are recognising that the needs of their client base have evolved. There’s more of a focus on governance and data related to the assets the managers are managing themselves. It’s not just focusing on the mandate at hand, but softer information as well, which UK clients now require. Cost transparency has been a great example of this, and we see that the availability of ESG related data is now also hot on the agenda.
As I have previously mentioned, behaviour has changed and we’re now at a place where assets managers recognise that being transparent on costs and what’s going on in the fund is the new normal and they see it as potentially a competitive advantage that they can advertise relative to other asset managers providing a similar service. It also helps that there’s regulatory momentum around this and the general theme of transparency is growing across all industries, not just the asset management industry.
Q. How successful have you been in obtaining cost data from fund managers?
A. We find that fund managers are now forthcoming with most of the data we require for the cost transparency reports. However, it’s not a finished product yet. Although we have the CTI template and the format is now standardised across the industry, the data quality is still not 100% there yet. We have found that even with the CTI, only 20% of the data comes in right first time - so we have taken a flexible approach in terms of what we collect from a cost data perspective. Not everyone is up to speed with the CTI, with around 40% of submissions still referring to legacy formats.
As well as that there’s a real emphasis within our operation in ensuring that the quality of the data is as high as possible, and that all costs are accounted for. We achieve this through a combination of technology, which can efficiently identify and map all the templates to the new CTI, and a team of on the ground analysts with ‘hands on’ experience in working on cost data on a daily basis. These two factors have helped us to achieve a 99.85% success rate in collecting cost data. To build on an often-repeated adage, we strongly believe that you can’t ‘effectively’ manage what you can’t ‘accurately’ measure.
Q. Do you use Machine Learning to analyse the cost data you collect?
A. Yes. The use of machine learning has helped us to efficiently solve the various data challenges. We wanted to shine a light on the costs spent on running a pension scheme. We were able to carry that out manually working with one or two clients back in 2017, but as the product and the requirements of pension schemes grew, we have had to evolve to meet that demand.
We’ve spent a lot of time thinking about and implementing a technology solution that will help us automate the collection and analysis of this cost data from a range of asset managers. What we’re doing is essentially using a machine learning platform, which enables us to ‘drag and drop’ cost files in a variety of formats to produce a CTI template. This is now the industry standard and within that process we have a database of billions worth of cost information across our UK and Dutch universes.
What we can do with this technology is essentially cross-reference and sense check the data we’re receiving for each mandate and the various asset classes it’s in relative to our large database to ensure that any anomalies are picked up. If any anomalies are picked up, a flag is raised to our analyst team to follow up with the asset manager to explain why that has occurred. Usually, there is a reasonable explanation as to why costs change and this speaks to how costs can go up and down each year, depending on the market cycle and where the economy is at that time.
Overall, this process enhancement ensures that we are accurately collecting and reporting on all that we can, giving our clients confidence when making decisions on their cost data.
Q. Are Defined Benefit schemes required to carry out full cost transparency yet?
A. Not yet. But there is a House of Commons Work and Pensions Committee paper that was written by Frank Field. He led and chaired this paper. It made a recommendation that DB pension schemes should follow the same level of cost analysis, reporting and scrutiny as DC pension schemes. The reason for that is that it leads to better governance practice throughout the entire pension scheme arena. Many pension schemes are struggling with their deficit levels and funding ratios, and we believe better governance around costs are helping to plug the gap in ensuring a stable and well-functioning scheme, and a better pension system as a whole. We hope that access to more data and more transparency will ultimately help pension schemes achieve their goal of providing the promised retirement income for their members.
In conclusion, it’s important to add that the paradigm shift from DB to DC has led to this focus on governance and transparency, which should ultimately benefit the end consumer. The introduction of auto-enrollment, the increased investment risk on members and the pressure to deliver a healthy retirement income has highlighted the need for all schemes – DB and DC alike - to ensure good value.