How to standardise value for money assessments

Additional evidence to support AgeWage’s submission to the DWP’s consultation on Improving Outcomes for Members of DC schemes.


Summary of submission

AgeWage has submitted answers to the three questions asked about Chapter 1 – encouraging consolidation. In summary

1.We agree with the idea behind net returns but suggest internal rates of return generated from scheme data is a better idea. The timeframe for reporting should align with the time members have been contributing (assuming a full contribution history is available)

2. We agree that the amending regulations support the DWP’s policy intentions

3. We think the guidance to trustees will create over complex, over expensive and over long assessments that will not facilitate comparisons between schemes. We suggest an alternative way of reporting.

An alternative approach

Over the past year, AgeWage has been encouraging employers, trustees, IGCs and pension platforms to request data from their record keepers and request from us an analysis. We require anonymised contribution histories and pot values (net asset values). This data analysis allows us to produce internal rates of return for each pot analysed . We have produced over 1m IRRs and analysed 132 data sets.

Once we have completed the IRR calculation we create a synthetic IRR using a benchmark index we have co-created with Morningstar. Their UK pension index provides a daily price going back to 1980. The index aims to track the price variations of the average managed fund used by UK pension savers

By investing the contributions into this virtual fund , we create a new net asset value – being what the pot would have been worth had it been invested in the “average fund”. This in turn allows us to calculate the benchmark IRR.

We have created an algorithm that allows us to compare the actual with the benchmark IRR. If value has been created for the money invested, the algorithm will create a score of between 51 and 100. If value has been lost the score will be 1-49, a 50 score suggests the investor has the average return.

The algorithm reflects the amount of value created and so tends to reward consistent outperformance over time. Short term out-performance is less likely to result in a high score and the converse is also true.

Case studies

  1. A large DC scheme with a mature profile

A large Occupational scheme with a slightly below par score.  Assets in the default were highly diversified resulting in some opportunity cost against the benchmark. The Trustees concluded that they would have delivered better outcomes with a higher equity allocation. A further analysis suggested that for the risk   taken, the default fund had created value, however – relative to the market , the scheme was not delivering value

2. A contract based  pension which had consistently underperformed the market 

This data set of around 11,000 records showed underperformance over 20 years. There was an acceptance that little attention had been paid to the default over that time and that until recently members had paid above average charges.

3. A contact based pension which consistently outperformed the benchmark over time

Here the outperformance is over a similar period to the severely under performing data set. The default turned out to have been regularly reviews and charges had reduced over the 16 years of the scheme.

4. A multi -employer occupational scheme (master trust)

Although the level of out performance was small, it had been achieved consistently over time.

5. A small and immature multi-employer scheme that has underperformed

Although the scale of the under -performance is high, this is a relatively young scheme with only just over three year’s contribution history. It falls into the category of small schemes that might consider consolidation based on net performance (if IRR is used as the basis for measure).

Observations for the DWP.

You can see from these examples that we have seen occupational schemes deliver high value and low value in the £100m + space. We have also seen extremes of high and low value from contract -based workplace plans. The chart below shows that volatility in scoring is more marked in schemes with less than £100m in assets (to the left of the red line

numbers in the grid indicate size of membership within the bubble

Schemes that use the AgeWage scoring system are immediately able to see if absolute value has been achieved against the average saver using our benchmarking service.

The results are simple to understand and it is easy to compare one scheme’s net performance with another without the use of complicated tables.

As all schemes are analysed against a single benchmark, this system creates a common definition of value for money.

Provenance of the benchmark

The research conducted by AgeWage and validated by Morningstar shows that the typical pension saver would, over time , be invested in a mix that averages 80% equities and 20% bonds. Within these broad categories , sub-classes of assets such as index-linked gilts and corporate bonds  international and UK equities, have varied much more.

Morningstar maintain the index and review developments. Since inception in 2019, we have already seen a shift towards ESG factors in many DC defaults, if continued, we expect to this be picked up in the index.

Quality of Data

An analysis of contribution histories can also provide insights into the quality of data. Each data set gives up “outliers”, data items that for one reason or another, do not give rise to a credible IRR. Some outliers are benign, very recent histories need to be excluded as annualized IRRs cannot be relied on when less than a year’s data is available. Often records are shown with no value on the pot – these matters can be easily dealt with. More worrying are histories where the IRR is out of all proportion to the benchmark, here there may be evidence of poor data recording.

For Trustees conducting a value assessment , the level of these concerning outliers is a mark of data quality.

Here is an example of an outlier report

We consider these reports an important indicator to trustees , not just for their value for members assessment but also to assess data readiness for the pension dashboard.

Communicating with members

We are currently trialing the communication of individual scores to members. This is happening in the FCA sandbox and the final results may not be available within the timeframe of the consultation. We have received confirmation  from the FCA that they do not consider the scores – even if influencing individual decision making – are in themselves “advice”.  We know of several employers and trustee boards who are sharing their scores with members and the general public.

We believe this will become more common over time , especially where employers have good scores to communicate. Poor scores present different challenges. We will not disclose the scores of any data set without the express permission of the organization commissioning our analysis.

Benchmarking and comparisons

As can be seen from the case studies, it is extremely easy for comparison to be made using the simple set of metrics on these dashboards.

We would like more trustees to publicize their scores. We expect the commercial master trusts to advertise their scores as an inducement to transfer to them.

We also expect some employers wishing to publicize their “employer scheme” scores. Whether participating in contract based schemes or in multi-employer master trusts, employers will be keen , if their scores are good – to publicist their schemes to their staff.

We note the FCA’s intention for value for money to become the basis for employers to decide whether to keep a provider or switch to another. We see a single definition of value for money as critical to this.


About henry tapper

Founder of the Pension PlayPen,, partner of Stella, father of Olly . I am the Pension Plowman
This entry was posted in pensions. Bookmark the permalink.

Leave a Reply