This response is from Pension PlayPen. We provide a digital guidance service that allows employers to choose the right workplace pension based on a number of metrics.
We welcome this call for evidence as it allows us to input our thoughts that result from our work with small and medium sized employers.
Henry Tapper, the Founder of Pension PlayPen and the responder to these questions is also a Director of First Actuarial. First Actuarial advise large employers on the selection and governance of workplace pensions.
In this capacity, Henry has worked with Novarca on transaction costs for two years. As part of this work, First Actuarial has introduced Novarca to tPR and the DWP resulting in the introduction to the FCA.
Both Pension PlayPen and First Actuarial can show considerable involvement in improving governance in workplace pensions and beyond.
Question 1: Should the requirements for standardised, comparable disclosure
of transaction costs apply only to those schemes that will be subject to the
new governance and charges measures from April 2015?
Initially yes, we need proof of concept before we deal with legacy but – once we have proved this can work on the relatively simple workplace schemes of today, we should tackle the legacy. We hope that the ABI’s legacy review will embrace this work.
As regards other types of schemes – typically large DB schemes , our work suggests that precisely the same issues surrounding value to money pertain. It is simply a different entity bearing the costs.
If not, are there differences that should be taken into account when considering transparency
in other schemes?
At the provider level yes, different costs apply to running different structures (DB,DC and other collectives)
Question 2: What are the advantages and disadvantages of capturing and
reporting bid-ask spreads?
There is no point in capturing this information unless there is a benchmark to compare the spread to. It will take a long time to develop the benchmarks but recording the spreads is a relatively easy process. Ultimately the benchmark should be to set high, in the short term a practical benchmark might be to be in the top quartile of spreads for the type of transaction.
Do you have any views on the ease of identifying
bid-ask spreads, or modelling them?
We are told that spreads are easy to measure but the information is hard to evaluate due to lack of benchmarks
What practical challenges are there in
calculating bid-ask spreads?
We don’t have experience in this and cannot comment,
Do you have any views on estimation models of
Question 3: What are the advantages and disadvantages of capturing and
reporting market impact?
We see those transacting with skill and knowledge neutralising market impact or actually supplying liquidity to the market. This is a component of a manager’s performance or alpha. We think the management of market impact should be part of an assessment of a manager’s competence and information on achieved minimisation should be available to IGCs and trustees as part of their governance duties.
Do you have any views on the ease of identifying
market identifying market impact costs?
We don’t ourselves, but we work with Novarca who have shown us their methodology. We understand it and explain it to our fiduciaries.
What practical challenges are there in
calculating market impact costs? Do you have any views on the possible
estimation models of market impact? Do you have any views on the availability
of these models, their consistency, and the costs providers charge to access
We are not competent to properly answer these questions.
Question 4: Do you believe that missed trade “opportunity costs” and “delay
costs” are transaction costs?
No , these impact on the performance of a fund but relate to the skill of the person executing the trades. They are not measurable in the same way as a spread. Over time managers whose transactions are timely will benefit from better performance and IGCs and trustees will need , in their attribution analysis , to pay attention to the precision of execution as part of performance.
Do you believe that there is merit in reporting
them as part of the disclosure regime and in governance bodies reviewing
Yes , but as part of the performance measurement not as part of a cost analysis.
Do you believe that the practical issues, for example around the
subjective nature of some of the inputs needed to calculate them could be
Over time yes, consistently getting the timing right suggests that a transaction team have skill, knowledge and experience. This will feed through to better performance and can be an attributable factor in that performance, but judging each trade on its variance from the best possible trade creates a problem for those conducting governance (not least because of the subjectivity mentioned in the question).
Question 5: Do you have any further thoughts on the analysis of transaction
costs outlined in this chapter? Are there any alternative approaches to
identifying transaction costs, or other considerations to take into account?
This is an area for experts. Our work with Novarca makes us aware that the heavily intermediated world of transactions needs highly skilled forensic consultants capable of keen analysis. The cost of this analysis is and should be high. Such analysis should only be carried out on large funds where the quantum of efficiencies can justify the cost of the analysis.
For this reason, we suggest that this work can only be carried out by Trustees and IGCs with substantial funds under governance.
Question 6: Do you have any comments about the different frameworks within
which information might be reported and their respective strengths and
We have seen the Novarca framework for reporting, it makes sense to us.
Question 7: How should transaction costs incurred at product level be
captured and reported?
These transaction costs we take to be the buying and selling of units rather than individual assets. The most efficient way of doing this is within a fund- for instance a target date fund, where a transition over time from one asset allocation to another can be achieved with minimal disruption to the underlying asset structures. Where units have physically to be bought and sold (as in most insured lifestyle programs) the physical sale of units can be kept to a minimum by the careful management of the book, ensuring that the vast majority of units are effectively being re-registered, rather than triggering the sale of physical assets.
Would there be merit in splitting out costs incurred for
The disclosure of costs to members should be as simple as possible. We favour a gross and net basis , where people can see how their funds would have done without any of the costs and again with all the costs added in.
For fiduciaries, a detailed breakdown of how costs were incurred and who benefited is important. Where costs were incurred and no value arose, trustees and IGCs can act, but where costs can be seen to be adding value (value for money) then trustees and IGCs need a means to formulate what value and what money was created and incurred.
How could this be achieved in practice? Are there any other
costs incurred at a product level that are not administration charges, and that
could potentially be considered transaction costs?
In workplace pensions, there are member borne costs and employer borne costs, some of these costs relate to matters which have nothing to do with investment, such as the cost of inter-acting with payroll and HR systems. The general rule is that a cost can only be member borne (eg impacting on the fund) when the beneficiary of the work charged for is the member. We see this distinction as right and proper. However, when assessing value for money at an employer level (eg are employers getting a fair deal, a trustee or IGC must also bear in mind the costs of such external factors such as middleware. These costs should be considered separately from the member borne charges (including transaction costs).
Following this distinction (employer v member borne) leads to a clear test for considering what is impacting an employer (and their capacity to fund the scheme) and what is impacting the member (and impacting member outcomes).
Question 8: Do you have any views on whether pension schemes should be
required to look through to the transaction costs of all listed, exchange-traded
We think that all securities and funds in which a member’s fund is invested should come under scrutiny including ETFs.
Do you have any particular comments on how the
transaction costs incurred by property, (and other real asset investments),
private equity and hedge funds should be identified and disclosed? Is separate
guidance needed on how to disclose transaction costs in these areas, or can
the principles used in securities markets be applied?
We know these to be very hard to measure, but as with the recent work by Railmen, it is not impossible to get to the underlying costs of these asset classes. When the true costs of these funds are known, a value for money assessment can be made. If costs cannot be known then we don’t think that trustees or IGCs should be investing in these asset classes.
Question 9: Do you have any comments on the treatment of derivatives?
Should the costs of derivatives be disclosed separately somewhere within the
disclosure reports? Do you have any comment about the transaction costs
associated with structured products?
We don’t fully understand derivatives but we are that they can be costed and that these costs could and should be recorded as any other assets and liabilities.
Question 10: Do have any views on the different approaches to calculating
Philosophically we prefer an approach which differentiates between cost that are incurred explicitly to pay intermediaries (of all kinds) and those that occur as a result of incompetence (market impact etc).
Do you agree that a principles-based approach is
appropriate to set how transaction costs should be reported for each type of
We think that people need to know the total cost of ownership of their assets. As Financial Educationalists, we want people to understand the cost of owning a fund relative to owning a property or holding cash. The principle we want to follow is total transparency of information presented in such a way that people understand what they are paying for. We believe in the principle of informed choice
Do you have any comments on the reporting of negative transaction
costs? Is there such a thing as a positive transaction cost – other than to intermediaries?
Question 11: Should portfolio turnover rates be reported alongside transaction
costs? Yes, this is essential and an absolute no-brainer. Portfolio turnover rates are indicative of a style of management. People should be able to choose managers using this information.
If so, do you have any comments on the best methodology to use to
ensure comparability of portfolio turnover and transaction costs? No
Question 12: Do governance bodies need risk and return information to be
reported alongside transaction costs, or is it sufficiently readily available to
them from other sources, considering the balance of costs and benefits that
such new requirements may impose?
We cannot think of any better way of considering the performance of a manager other than through a holistic view of the risk and return achieved. If a manager is not prepared to share information ratios and other risk related measures with fiduciaries, it is hard to see why a fiduciary should justify using that manager
If you think risk information should be
reported, do you have any feedback on the best risk measures to use when
considering transaction costs?
We’d like to see a restatement of the performance figures that would have been reported without the charges to Net Asset Value. This restatement would not include the cost to the fundholders of market impact and at a later stage, reporting against a benchmark cost for market impact would be helpful (e.g. did the manager’s performance suffer from poor execution or benefit from good execution)
Question 13: Do you have any views on the value and/or costs of
benchmarking? Are there any other issues to be taken into account when
We are for benchmarking. But we don’t think the benchmark should be set simply by averaging good and bad practice in the universe of funds in which they compete. Data exists in Europe and from other continents that can allow a benchmark to be created around global best practice. We’d favour such a benchmark rather than a parochial benchmark that rewards local bad behaviour.
Question 14: Do you have any feedback on the reporting of the costs of
securities lending, foreign exchange and related activities, and on how these
should be reported? The difficulty with reported costs is that it will focus on the knowns. The need is for an independent expert to continue to work with Government to disturb the unknowns and the new unknowns created by an innovative financial services industry that will find knew ways of taking money our of funds that gets around fiduciary scrutiny. We’d suggest that particular focus is given to the areas of stock-lending, the more obscure areas of custodial activity (particularly in the area of FX and the bundling of research into dealing costs (soft commissions).
Are there any other areas or practices that you would
highlight where providers are imposing additional costs or generating
The costs outlined above relate to intermediation at the fund level. There are new ways to generate costs at the product level, in particular these are open to those managing vertically integrated workplace pensions where the managers are also intermediaries. The provision of advice to the fund which is charged to Net Asset Value is an area we’d like to see explored.
We believe this gives advisers the opportunity to subvert the 0.75% charge cap and the restrictions of the RDR and consultancy charging in one go.
Question 15: Do you have any comments on the practical issues with
presenting costs and charges information?
The presentation of this information to IGCs and Trustees must be consistent, intelligible but not dumbed down. We need to educate up our fiduciaries to understand it. This means fewer fiduciaries of better quality.
The presentation of information to members of workplace schemes needs to be easy to absorb for those with limited time, energy or knowledge, a simple traffic light system is all that is needed to tell a member whether a manager is good ,bad or indifferent at managing transaction costs. Technology now means that with a click of a link, more curious members can read the detail if they want to. We think that IGCs and Chairs of Trustees should report on the management of transaction costs within the funds and declare the cost of ownership as we suggest above, within their reports to members. We’d like to see the traffic lights and links to the detailed information at the front of the reports and not buried in an appendix.
Do you have any comments on the
degree of standardisation that will both enable governance bodies to take
decisions on their scheme and achieve comparability across the market?
We have seen the template suggested by Novarca and like it. We think that the information in this template should form the basis of a standard report that would be consistent whether it came from the Chair of an IGC or a board of Trustees.
Are there any other factors in the presentation of transaction costs in a report that
would enable governance bodies to make better decisions?
We think that governance bodies need to be absolutely sure of their data and this can only be achieved by making the reporting statutory and consistent. If the reporting is wrong and governing bodies take decisions that subsequently turn out to be based on erroneous information, they should not be liable for the error. The liability for any loss from the error should rest with those producing the information. Accountability for accuracy of reporting must rest with those managing the funds,
Question 16: Do you agree with the use of portfolio turnover rates and unit
transaction costs to enable better prediction of likely transaction costs?
We think that the actual turnover rates and transaction costs achieved are predictive of what are likely to be achieved going forward. Managers trade as a matter of style, if they do not know why they had high unit costs or high portfolio trading they do not know what they are doing. If they do know why their costs and trading are as they are, they should be able to justify those numbers and predict they are maintained.
The argument that past costs cannot be used to predict future costs is totally specious.
Should providers be required to provide reasons if turnover rates are likely to
be different in the forthcoming period? Yes they should. But this should be part of normal reporting. If a fund traded higher or lower than was expected in the previous year, we would expect fiduciaries to be told of it and why it happened. If the turnover rates are likely to increase in future, fiduciaries should be told and to take action accordingly
Is there any other information that
would enable the governance body or scheme members to understand
potential future transaction costs?
If a manager has a strategy around costs and is set on adopting benchmarks for performance, he or she should share these. If there is no strategy to better manage costs, fiduciaries should know why not. The answer may be that there is no scope to improve – that is rarely accepted as a good answer!
In the absence of there being no strategy, fiduciaries should raise concern about the potential for costs to escalate.
Question 17: Do you have any comments on whether a transaction cost
disclosure regime will have any other consequences for the way that pension
schemes and their agents transact?
At a time when the governance of DC pension schemes is coming under greater scrutiny, the thoroughness with which this issue is being dealt with, is bound to have knock on consequences into other areas of pension management- administration, advice and the development of new decumulation products. We can only see these as being positive and warmly congratulate the FCA and the DWP on the work they are doing in this area.
Question 18: Should regulations and rules on transaction cost disclosure only
directly apply to pension providers and trustees? If not, on whom would
additional disclosure requirements be necessary to ensure that transaction
costs are reported accurately to relevant people?
The disclosure at fund level are of course equally important for those holding money outside of the pension wrapper and in the management of DB pension schemes where the risk of high costs only falls on the member indirectly. Whatever the structure, we would expect the regulations that come out of this work to be applicable to other parts of the funds industry and to protect a wider range of beneficiaries than simply those in workplace pensions.
We would like to see disclosure regulations extend in time to the managers of all collective arrangements and for the fundamentals to be reported by those dealing with individually managed portfolios as well.
Question 19: What information on transaction costs would be useful to
employers and members?
We have great difficulty getting small and micro employers to engage in the governance of the workplace pensions they establish. Simply getting them to think about the macro-measures of choice is hard enough!
Employees (members) are likely to be more engaged, as it is their outcomes that are impacted by these costs. We think that as DC member balances increase in value, pressure from members to get proper information of this kind will become more intense.
What employers need is sufficient information to satisfy themselves and their insurers that they have exercised their obligations. The answer is therefore dependent on what duties Government imposes on them. We would suggest that no further obligations are placed on employers.
How and when should this be reported to them?
We suggest that a figure of total cost of ownership be reported to members and employers on the fund. This should be expressed as a percentage and as a £sd figure. The £sd figure should be illustrated per thousand pound of fund. E.g.
“you are paying £28 in costs each year for every £1000 that xyz manages on your behalf. By comparison the average amount payable across all workplace pensions is £30, you are getting low transaction costs”.
This should be reported once a year in the annual statement sent to all members.
Question 20: What information on costs and charges should be made publicly
available? When and how should this be information be provided?
The information should form part of the Trustee Chair and IGC Chair’s report. This information will be put in the public domain. The information should be freely available on provider websites and find its way into the advisory chain. www.pensionplaypen.com has already got the capability of taking this information and showing the impact to employers and members in terms of member outcomes.
Question 21: Are there any areas that you would highlight where firms, trustees
or asset managers may not comply with the disclosure regime in the way
intended? We feel our response has answered this question.
If you are concerned that this may be the case, are there steps that
could be taken to reduce the incentive to get around reporting transaction
costs? Yes, see answer below
Would third-party oversight of reports enhance their value and
We think that this reporting regime should be incorporated into the master trust assurance framework and that adoption of this Framework should be compulsory for all master trusts being used as workplace pensions under auto-enrolment.
The extra level of scrutiny provided by MAF is important as many mastertrusts are vertically integrated and vulnerable to the bad practices resulting from allowing an adviser to manage, advise and advise on a product.
Question 22: Do you have any comment on the likely costs involved in
implementing transaction cost disclosure along the lines described in this call
We do not have the resources to answer this question but suggest that the costs must be borne by the managers of the funds and not passed on to members of workplace pensions as member borne charges. In particular we would like to see the cost of this reporting specifically reported in the long scheme and accounts as a cost that fell to the manager and not to the fund’s net asset value.