Costs are an increasingly contentious concern for clients of the financial services industry, especially given the effects they can have over the long term. Knowing whether costs are reasonable requires an understanding of what types of services add value and what types can destroy value. It also requires understanding of how different pricing structures will affect individuals in different circumstances and the range of ways that the effects of costs can be measured. We propose a new model for both consulting and asset management costs to address the upside down value chain.
Costs matter – particularly when they are compounded over the time frame typically demanded by retirement benefit schemes. But until we grasp the dynamics of the full value chain of delivery to members, we will invariably focus on the wrong debates in our quest to control costs.
What follows is a three-part study that analyses the problem of costs from three different perspectives:
More importantly, we also present a few radical proposals to try to move the debate around costs to a more realistic and effective outcome.
It is time for trustees to rethink what’s really worth paying for.
The issue of costs is multi-dimensional:
To some extent, National Treasury’s paper on costs has already set out many of the issues relating to points 1. and 2.
But a brief summary of the key points are important for our discussion here.
If we were to take a snapshot of the charges allocated in a typical, segregated retirement fund at a given point in time, the distribution of charges for an average 35-year-old fund member will be as shown in the graph below. In the context of this once-off snapshot in time. Most of the fees on the fund go towards asset management and
A crude assessment of how these costs are distributed over the spread of service provided
However if we were to assess how the power of compounding would affect those allocations over the 30-year period these members would have until retirement, the allocation to asset management nearly doubles to 82.4% of total costs by the time the member retires. This translates into a weighted average of 66% for the whole period.
This is primarily a function of the compounding effect of each of the charging types employed across the value chain.
This outcome is a function of the different ways these services are charged, effectively we have three different charging models:
The outcome of the fee impact is a function of the different charging structures and the way they compound over time.
Ironically, we often hear trustees argue that they have no problem paying a premium to fund managers if it means securing the top talent. But here is where we need to be completely clear about how this aspect of the value chain performs over time.
Consider the reality of most long-term investment strategies for defined contribution funds.
Even if we could be fairly certain that we have selected a skilful manager, there is absolutely no assurance that the market will reward that manager’s particular skill set going forward.
The problem is, once multiple managers are employed, the impact of this diversification significantly reduces any short-term performance1.
For example, individual managers may reflect tracking errors of up to 7% to their benchmarks, but the tracking error of the blended managers typically reduces to around 2%. This is not a bad thing – such a diversification strategy provides a significant improvement on the fund’s risk to return ratio. But the total return of the blend will be muted.
Even assuming that manager fees are based on realised performance and not past performance, a fund could end up paying a high performance fee for a particular unspectacular outcome at the aggregate level. Performance fees are hugely problematic with multiple managers.
In reality, fund manager performance tends to go in swings and roundabouts over time. The longer the time frame, the more likely any exceptional manager performance simply evaporates due to the impact of diversification of managers and the flux and flow of manager returns.
Over the long term, performance outcomes are driven more by the long-term strategic asset allocation embedded in the investment strategy than they are by individual manager contributions.
As an example we consider the full value chain of costs in one particular umbrella fund to test the effect of an active against a lower-cost passive investment strategy to the final member outcomes. We use identical member profiles and pricing structures. We also assume that the cost of the active strategy would be 0.85% of assets under management per annum against 0.35% for our passive strategy.
By calculating the reduction in yield, a methodology that allows us to see the value erosion caused by the fee structure over a 40-year period, we can see that a switch to a passive solution would equate to an additional 5.63% points to the member’s replacement ratio. While every additional point in the final replacement ratio is important, keep this number in mind for later when we look at other sources of value erosion.
Note that this analysis says nothing about performance differentials between the two strategies. But assuming that both strategies employ the same long-term strategic asset allocation strategy, for the active strategy to even match the outcome of the passive strategy, it would have to generate a consistent alpha of 50 basis points year after year for 40 years to compensate for the additional costs. Trustees would need to give serious consideration to whether this is realistic.
Should this even be an active against passive debate?
We would like to be provocative here. The government provides a tax incentive for an individual to save for their retirement using pension and provident funds, preservation funds and retirement annuities in compulsory investing. We believe that for the world of compulsory investing to be viable, the issues of performance chasing and performance fees (or high fees) must be taken off the table.
Perhaps it’s time we made a clear-cut separation in asset management servicing between investing to win the top performance sweepstakes and investing to meet a member’s replacement ratio requirement.
It could be done – and done in a way that could create a far more robust and diversified asset management industry in general. Imagine an investment world where active managers would provide two types of asset management services:
It’s time we made a clearcut separation in asset management servicing between investing to win the top performance sweepstakes and investing to meet a member’s replacement ratio requirement.
The first mandate would be for the exclusive use of compulsory investment funds. The fee here would be capped at a reasonably low level, possibly halfway between what is currently charged for active and passive strategies for all retirement fund clients. The investment strategy on offer would be whatever the manager felt was feasible at that fee.
But the irony here is that these mandates for compulsory retirement vehicles would be won or lost, not on performance, but on the manager’s ability to follow a targeted, risk-budgeted mandate. This quality of delivery is essential in a blended manager strategy, because it gives the manager of the blend the confidence that their managers will remain diversified. Managers would lose mandates if they couldn’t control their risk budgets.
When compared to purely passive strategies, this blended strategy of risk-controlled mandates, if robustly diversified, should provide incrementally better risk-adjusted returns over the long term than a purely passive strategy – and not necessarily because managers are hugely skilful. It’s actually the combination of the diversified mandates at a lower risk than the index that produces the magic.
The graphs on the next page illustrate this point, employing a real blended manager solution. The first graph shows the variable, rolling one-year performance of the different active managers in the blend over time. In the bottom graph the orange bars represent the aggregate outperformance of the blend of managers.
In 2000, Allan Gray provided exactly this type of portfolio to multi-managers. It was a risk-controlled version of their Classic Equity portfolio, offered at a significantly lower fee. While Classic may have outperformed in certain market phases, this risk-controlled solution offered more robustness over time.
Uncorrelated performance of the underlying managers
Reflecting a rolling 1–year alpha generation of the blend
The other mandate active managers could offer would be exclusively for discretionary clients. This would be the same value proposition they are currently offering. But asset managers would be welcome to charge what they view as fair for their alpha generation potential.
Why preclude compulsory savings funds from using the latter mandate? What retirement investments need is more certainty of outcomes if they are going to meet the liability funding requirements over a 40-year investment period. The reality is that research repeatedly highlights that short-termism, and chasing return, has introduced one of the most important sources of potential value erosion in the active management space. For example, a study by Ron Bird and Jack Gray on the Australian Superannuation Fund industry calculates that these two factors perpetuated as much as 3% of value erosion per year.
Right now the balance of power is on the side of asset management marketing machines that continually hold out a promise of more return. Retirement funds don’t need more; they need consistency, certainty and simplicity – and of course, lower costs.
A proposal like this creates opportunities that are not only fair to all parties, but could develop a far more diversified industry.
While momentum is gathering to promote passive over active strategies in this space, we believe a proposal like this creates opportunities that are not only fair to all parties but would also develop a far more diversified industry. Consider the implications that such a model of separate mandates would have for asset managers – to say nothing of investors themselves:
In truth, we have dealt with only one part of the issue of the cost relative to value discussion – and while the fees that service providers charge have received the most attention to date, we need to understand other factors to address where value is destroyed for members.
In 2012, retirees from our universe of funds retired on average with a replacement ratio of 32% (after an average 21.7 years of pensionable service)2. And this was over a time frame when markets were generally flourishing. For the last 15 years the average active balanced manager has generated around 10% real return – an impressive result.
If we look at only a subset of retirees, for those who have been in the same fund for 30 years or more (and therefore have reasonable savings periods without interruption), the average realised replacement ratio is actually significantly higher, as would be expected. This shows that reasonable outcomes have at least been achievable from the investments – inclusive of current costs – given this favourable environment. So where are things going wrong?
These numbers suggest that we need to examine the issue of cost from a different perspective. Instead of assessing what the service provider provides (and whether they deserve its associated fees), we should turn the analysis on its head by asking: Which decisions have the greatest impact on what ends up in members’ pockets when they retire?
A careful analysis of the sources of value erosion, when measured at member level, suggests that even if the investment strategy was capable of delivering a 75% replacement ratio (after reducing it 5.63 points for cost), the combination of inadequate preservation and a lower pensionable pay percentage proved to be significantly greater value destroyers.
Given that the government has not yet made retirement funds compulsory for all employees, nor has it firmed up its position on preservation or pensionable pay allowances, this leaves a significant burden on employers, trustees and consultants to try to manage these shortfalls.
Clearly to stem this value erosion, we need a more effective consulting framework.
An important question emerges: If these factors are at the heart of the value erosion to members (and not poor performance or potentially costs of fund managers), then who should be accountable for getting it right? From a legal perspective, that answer, of course, is the board of trustees or the management committee. But then who, in turn, is providing this incredibly broad aggregating function to the fiduciaries?
Compare a board of directors of a publically listed company to a board of trustees of a retirement fund. The critical difference is that corporates typically have an executive management team that reports to the directors who are accountable for the day-to-day activities of the company. They provide the reality check on what is achievable, what isn’t, and what needs to be done to fulfil the board’s wishes. They bring technical depth and insight about the business to the table and can even be dismissed for not delivering the goods.
A board of trustees is presented with a different challenge. With the exception of a handful of parastatal and large corporate funds, most retirement fund boards have no accountable executive that possesses a depth of technical insight and perspective on all the complex moving parts of delivery to members. While the new industry organisation, Batseta, is promoting the professionalisation of the principal officer role as a way of filling that gap, few boards have the tools and resources that can provide them with a holistic picture on where they are winning or losing in this delivery.
This, in turn, means that boards of trustees are hugely dependent on their service providers – and not always in the healthiest way. The problem is, the way our industry is structured or ‘fragmented’, the most powerful entity in the value chain, from a compensation point of view, is the asset manager, but providing that critical linkage is clearly not their role.
If the true source of value erosion is not with fund managers, then who should be accountable for getting it right?
The truth is, if we are going to get this right for members, we need to define a completely new role that integrates all these aspects of delivery.
1. This suggests that to get these decisions right, what trustees and management committee members really need is a sort-of über-consultant who can:
2. Create the critical link between the employer (and their HR departments) and the fund’s fiduciaries to eliminate any policy gaps.. Provide the relevant analysis of the membership demographics and behaviours to establish:
3. Define and manage an appropriate risk budget for the fund’s investment strategy.
4. Consider appropriate platform and cost structures to address member profiles.
5. Monitor member progress against established goals.
6. Measure and monitor the potential impact of trustee decisions on specific member outcomes.
7. Ensure appropriate communications of all these points to all relevant stakeholders.
The problem is that providing this type of service requires significant resources: member monitoring tools, asset-liability modelling tools, risk budgeting tools, aggregated reporting and attributions analysis for funds, tools and models for member projections and for building asset class return assumptions. None of these tools are typically found in an asset manager’s range of capabilities, which begs the question once again: Who pays for this capability?
Essentially, we have a vicious circle here. Trustees may be right to question the value that consultants add if consultants don’t have the necessary technology or the opportunity to provide the holistic picture that trustees so desperately need. But not paying for a service that has the greatest potential impact on outcomes perpetuates the problem because consultants see no way to cover the cost of being properly resourced.
The reality is that globally, boards of trustees appear to be either turning their backs on asset and employee benefits consultants or pushing down fees to unsustainable levels – probably with cause. Ultimately, the way these services are segregated in many consulting operations means there is simply no way the holistic picture, which seems so eminently sensible, can realistically be delivered.
Two factors exacerbate this outcome:
Clearly it’s time for a new model of accountability and governance!
We would like to believe that by creating a solution that provides a more efficient and focused way of delivering end value to the member, we can create a much more equitable value chain.
We believe that the retirement industry as a whole desperately needs to rethink its value proposition to members. The reality is that a member’s fund credit can only withstand so much reduction in yield as a result of costs over the 40 years a member needs to be invested or drawing on risk benefits.
We would like to believe that by creating a solution that provides a more efficient and focused way of delivering end value to the member, we can create a much more equitable value chain.
We recognise too that everyone who enters this debate is conflicted at some level, including the authors, and readers may therefore treat this debate with some circumspect. That consideration is encouraged, but it should not preclude the debate.
Trustees need to rethink how to fairly allocate that maximum cost to ensure that both critical points – the liability management and the asset management – are properly serviced and adequately resourced to get the job done comprehensively. We also need to recognise that only by insisting that these two skill sets become more effectively integrated will funds be able to get greater certainty that they will get the outcomes they require.
Combined, the über-consultant or über-asset manager concept provides a powerfully appealing way of addressing our problems of delivering targeted outcomes at controlled prices. In total, we should be able to reduce costs. While asset manager fees would most definitely come down, we should also be able to compensate investors with better risk-adjusted returns that are meaningfully sustainable in a long-term investment strategy.
On the consulting or actuarial side we’ve streamlined and focused the service to deliver measurable outcomes. A service this comprehensive and with this much greater potential for delivering necessary outcomes, given the current environment of reform, now warrants a more serious compensation consideration.
Lowest cost is not always the best deal for the member, when we consider value. Likewise, lowest average cost is not always the best deal for the individual member. In our last solution, we suggested that the long-term reduction in yields outcomes for individual members could vary significantly depending on that member profile, their income levels and their fund credit. To properly round out our assessments about cost structures, we need to understand:
The important point is that services are often priced with the end user in mind. As such, we need to consider which pricing structure works best for which market segment.
Let’s use umbrella fund pricing structures as an excellent case in point. Comparing one umbrella fund’s cost structure with another can be fairly complicated, but some insight into how providers determine their pricing structure helps in determining which fee structure truly offers value for money for members, given their particular demographic make-up.
We need to consider which pricing structures work best for which member profiles.
There are four broad categories of fees which a member in an umbrella fund will face: administration fees, governance fees, consulting fees or broker commission, and investment fees. These fees may be charged in a few different ways (initial or ongoing; fixed rand amount, percentage of salary, percentage of contribution, or percentage of assets). This makes it hard to compare apples with apples. For example, if Fund A charges a lower administration fee (as a percentage of salary) than Fund B, but a higher investment fee (as a percentage of assets), which fund is cheaper? And over what period?
Increasing the complexity is the fact that many umbrella funds charge multiple layers of fees. For example, instead of charging one administration and governance fee per member, some umbrella funds charge an additional scheme expense or contingency reserve account levy to cover governance-related expenses (such as audit fees, trustee expenses, FSB levies, and fidelity insurance premiums). These funds may also charge additional administration-related fees such as participating employer fees, investment administration fees or asset-based fees, resulting in two or three different administration related fees being charged.
When comparing umbrella funds, consider the total cost per member across all the different fees (administration, governance, consulting and investments) and not only the administration fee.
National Treasury3 has commented that “employers may be easily persuaded by low up-front charges (that is, those deductible from contributions) to select a commercial umbrella fund that in fact, because of higher recurring charges [in other words, high investment fees], represents lower ‘value-for-money’ than other commercial umbrella funds”.
This point is best illustrated through an example.
Joe’s Manufacturing Ltd is a small manufacturing firm with 100 employees. These employees have a combined monthly pensionable salary of R1 million and total retirement fund savings of R30 million. The company is looking to join an umbrella fund and is comparing various options.
There is a significant difference in the administration fees quoted by two of the options: Fund A will charge a monthly fee of 0.5% of the member’s monthly pensionable salary, whereas Fund B will charge only 0.25%. The investment fees of these two funds are fairly similar: 1.0% per annum of assets for Fund A against 1.1% per annum for Fund B. So which option should Joe Manufacturing Ltd select?
Our initial reaction is that Fund B offers significantly better value thanks to the lower administration fee. We pay little attention to the investment fees as the difference between them is relatively small. However, if we calculate the combined rand administration and investment fee, we find that this fee is the same for both funds (R30 000 per month).
Companies whose employees have managed to accumulate significant retirement fund savings need to be particularly aware of high investment fees. If the member has a choice of investment portfolios with differing investment fees, they should not choose the investment option with the lowest fees, as that option would typically be a money market fund offering a lower expected investment return. Rather the member should assess the potential benefit of such a portfolio against its cost.
Note that there is not necessarily an optimal cost structure that suits all participants. The employer should balance the various trade-offs and select the cost structure most equitable to the majority of their employees. We can illustrate the point through two examples with Joe Manufacturing Ltd.
Administration fees are typically charged as a percentage of salary, whereas investment fees are charged as a percentage of assets. These different charging structures affect members with small and large retirement fund savings differently.
Two employees at Joe Manufacturing Ltd each earn R420 000 per annum. The first employee has worked for 30 years and saved R3 million in his retirement fund. The second employee has worked for 10 years and saved R420 000 in his retirement fund. Although the company is indifferent to the two options at the outset from the previous example (Fund A or Fund B), the two employees should have strong preferences on which fund is selected.
The first employee, who has already saved a significant amount towards retirement, would prefer Fund A. This fund has higher up-front charges (charges that are calculated as a percentage of his monthly salary or contribution) and lower recurring charges (charges that are calculated as a percentage of his total assets). Fund A would result in this employee paying approximately R160 per month less in administration and investment fees combined (at the outset).
The second employee would prefer higher investment fees as his retirement savings account is relatively small. Fund B would result in this employee paying approximately R50 per month less in administration and investment fees.
Employee 1 has already saved a lot towards his retirement and should prefer Fund A. This fund has higher up-front charges and lower recurring charges. Employee 2 would prefer higher investment fees and lower up-front fees as his retirement savings account is relatively small.
Employee 1 (R3 million in savings)
Employee 2 (R420 000 in savings)
Different approaches to charging for the administration fee affect high-income and lowincome members differently.
Fund A charges each member a monthly administration fee of 0.5% of their monthly pensionable salary. Fund C charges each member a monthly administration fee of R50. The total administration fee is the same for both umbrella funds (R5 000 per month) but the decision on which umbrella fund to select has significant implications to the individual members.
Fund C charges all members the same administration fee (R50 per month), irrespective of the size of their salaries or their total contribution. Administration fees may be charged for collecting a member’s contribution and allocating it to their account, for example. Many of the expenses incurred in providing administration services have a fixed cost per member, irrespective of the size of that member’s contribution.
Therefore, charging a fixed rand amount is viewed as the most equitable approach from a financial or costing perspective. The issue with this approach is that the administration fee may become prohibitive for low-income members, significantly reducing the amount allocated towards their retirement savings and their expected replacement ratio outcomes.
If a factory worker at Joe Manufactoring Ltd earns R60 000 per annum, of which he contributes R500 per month to his retirement fund, he ends up paying 10% of his monthly contribution towards administration fees.
Fund A’s charging structure is more equitable from a social perspective, as the administration fee would reduce all members’ contributions by the same percentage (assuming the contribution rate is the same).
The factory worker at Joe Manufactoring Ltd would pay an administration fee of R25 per month (0.5% of his monthly salary), reducing his retirement fund contributions by 5%. The CEO earning R1.2 million per year and contributing R10 000 per month to his retirement fund would pay a far larger administration fee (R500 per month or 5% of his monthly contributions).
Although this approach is more equitable, it introduces cross-subsidies between high-income and low-income members within that employer group. Cross-subsidies need to be understood and managed over time.
All the examples make important points about how differential charging structures can affect members of a fund who may have different fund credits, annual salaries, or years to retirement. While these types of comparisons are easy to calculate and understand, they only represent a snapshot in time.
When it comes to the long-term impact on the outcome to member, we need to be able to see how these charges compound over time. The starting point for most decision-makers is the graph below. This graph provides a composite picture of the total charges for each type of umbrella fund pricing model. In this regard Fund X appears to be the cheapest. Often the decisionmaker will stop at this point and simply select the cheapest option.
But the real test comes by moving beyond the limits of this snapshot. To add the dimension of time and how compounding impacts the outcomes for various types of members, the chart on the next page provides the really critical information.
This measures the reduction in yield in the member’s annual return after the required period of saving. So, for example, although the ‘average’ employee would appear to get the lowest reduction in fees from Fund X. In truth, though, for those members of the fund who either have a high fund credit or a high annual salary, Fund Y would have turned out to have the lowest impact on member out comes.
The key message here is that employers need to use more comprehensive measures in making their assessments of pricing differentials between umbrella fund fees. A tool that can take one through the debates we have described above can fill that need.
Although costs are important, value needs to be the determining factor. Good value in an umbrella fund is all about ensuring that the fund parameters are appropriate, investment options are suitable and appropriately priced, the fund improves member engagement and financial education, and there is firm and fair governance structure to mitigate against catastrophic failures or conflicts of interest.
Costs do matter. But paying the right amount for the value on offer also matters. It’s a delicate balance. But with the right tools and the right understanding of the retirement fund value chain we should all be able to get to the right decisions about how to appropriately redress the imbalances that still exist in the market today.
The important point, though, is that it is the aggregate of all these costs that the member or investor actually bears. That means that this aggregate cost has to be held up against the aggregate potential value that can be added – and this is where is gets complicated. Until consumers are in a powerful enough position to shake down the value chain and demand fees that are commensurate with the value that the service contributes, the current cost hierarchy will remain.
There is only so much total cost a retirement fund or investment solution can shoulder and still provide investors with reasonable outcome. The travesty is that in a world that believes emphatically in the value of active management, trustees are more likely to try to force down the costs in the other building blocks than take on the price-making top managers. But they do so at the risk of cutting back on the very services that are most important in that final delivery.
1 Bird & Grey (2009)
2 Benefits Barometer 2013
3 National Treasury (2014)