Should pension scheme sponsors implement their pension promise through dedicated, internal ‘subsidiaries’? Or should they outsource the implementation function to external service providers? There is no simple ‘yes’ or ‘no’ answer here. However, we can build a strategic framework to explore the pros and cons of outsourcing all or part of the implementation function. For example, pension scheme design, potential scale economies, and effective governance should all be key factors in any serious strategic assessment of the outsourcing question. This article shows how these factors can impact outsourcing strategies, describes how outsourcing decisions should be framed, and sets out the conditions under which outsourcing decisions can be part of ‘win-win’ strategies for all pension scheme stakeholders.
Deciding the ‘Pension Deal’
The quest for insights into the outsourcing question should begin with the ‘pension deal’ itself. Specifically, what is the nature of the pension promise, where are the underlying risks, and who is to bear them? Game theory urges us to build pension deals that start out as ‘win-win’ propositions, and stay that way. However, game theory also tells us that plain vanilla defined benefit (DB) and defined contribution (DC) pension plans both suffer from considerable potential to turn into adversarial ‘win-lose’ games.
In DB contexts, ‘win-lose’ conflicts usually revolve around the ‘ownership’ of balance sheet surpluses and deficits. In DC contexts, conflict arises when the accumulated pension capital is not enough to buy a ‘decent’ pension, or when participants outlive their retirement capital. Fortunately, the game theory framework not only identifies the design flaws in traditional DB and DC schemes, but also shows how these problems can be mitigated through transparent risk sharing ‘deals’. The point is that while pension scheme sponsors can get outside professional advice on fundamental pension scheme design issues, they cannot outsource the ultimate design decision itself. Some things just can’t be outsourced.
The ‘scale’ factor
With a clearly spelled-out ‘pension deal’ in mind, the scale factor looms large in determining the optimal implementation strategy. The pension investment and administration functions are both subject to major scale economies. Generally speaking, ‘big’ means low unit costs, ‘small’ means high unit costs. Our benchmarking firm Cost Effectiveness Measurement (CEM) has quantified these scale impacts on unit costs. Three key findings are relevant to our discussion here:
o In a universe of 260 US, Canadian, European, and Australian DB pension schemes, the average ‘scale’ factor for their investment programmes over the course of the last five years was -18. Specifically, over the dollar size range of the funds in the CEM database (from $100m (E103m) to $150bn), a 10-fold increase in assets resulted in an average 18 basis point decrease in unit cost, all other things (eg, asset mix) equal.
o In a universe of 50 US, Canadian, European, and Australian DB pension schemes, the average ‘scale’ factor for their pension administration programmes was -51. Specifically, over the membership size range of the pension schemes in the CEM database (from 50k members to 2m members), a 10-fold increase in plan membership resulted in an average $51 decrease in annual per member cost, all other things (eg, service levels) equal.
o In a universe of 86 US401(k) DC pension schemes, the average ‘scale’ factor for the total programme (ie, including both investments and administration) over the last five years was -16. Specifically, over the dollar size range of the schemes in the CEM database (from $47M to $21B), a 10-fold increase in assets resulted in an average 16 basis point decrease in unit cost, all other things (eg, asset mix) equal.
Clearly, scale has a material impact on all pension scheme unit costs. An important corollary of this conclusion is that the larger the pension scheme (whether measured in terms of assets or plan membership), the more likely it becomes that the ‘internal subsidiary’ option will win out over the ‘total outsourcing’ option on a cost effectiveness basis. Another corollary is that if an employer does choose to totally outsource the pension scheme implementation function, it should look for suppliers who have large scale economies, and who are willing to pass on that advantage in the form of lower fees.
The ‘governance’ factor
Aware or not, every organisation must engage in three types of activities to achieve its purpose: governance, management, and operations. Pension schemes are no exception. So with a decision to establish an ‘internal pension scheme implementation subsidiary’ come three further strategic decisions:
o How will the pension scheme’s governance function be designed and implemented?
o Who will the scheme’s governors hire to develop and implement the scheme’s strategic plans in the investment and in the plan administration areas? In other words, how will the executive function be organised?
o As part of the scheme’s strategic plans, who on the scheme’s executive team will be accountable for which implementation decisions on the investment side, and on the administration side?
These questions, and their answers, provide the framework for continuing our quest to place outsourcing decisions in context, and to see when they become part of an overall ‘win-win’ strategy. Why? Because now we have a clear organisational context in which such decisions can be made. The importance of such a context will become clear below.

The executive function: internal or external?
For example, the pension scheme’s governors could choose to hire a CEO to manage the pension ‘subsidiary’, or not. If they choose to, it should be done with the understanding that the minimum cost of an effective internal executive function will be in the $1m per annum range. Further, if the governors choose to go this route, it should be clearly done with the understanding that it is the CEO who will be accountable for the construction of, and the implementation (with their blessing of course) of the strategic investment and scheme administration plans. The degree to which specific functions will be performed internally, or outsourced, and how such decisions will be made, will be key elements of the strategic plan.
If the governors choose not to go the internal CEO route, the worst possible thing they could do is to retain the executive function at the board level. Trying to simultaneously govern and manage a pension scheme is equivalent to trying to suck and blow at the same time. It can’t be done. In a pension scheme context, the typical result is accountability confusion, poor operational performance, and high unit costs. In other words, a ‘lose-lose’ outcome all the way around. Without an internal CEO, the only sensible alternative is to outsource the CEO function. We are pleased to note that most major pension consulting organisations are also coming around to this point of view. Many have begun to offer ‘CEO for hire’ services. At the same time, we challenge them to stop ‘consulting’ to the boards of trustees of pension schemes that have usurped the executive function unto themselves. Such practice is value-destroying, and hence pure folly.

Pension scheme operations: internal or external?
Which specific activities required to operate a pension scheme should be conducted directly by the ‘pension subsidiary’, and which should be outsourced? By now, our answer to this question should be clear. ‘Inside versus outside’ decisions about specific activities should depend only on the judgment of the person accountable for the cost effective implementation of that specific aspect of the strategic business plan for the pension scheme. Thus, to state the obvious, ‘lowest cost’ can be an extremely poor decision rule for some internal versus outsourcing decisions.
How successful have these ‘internal versus outsourcing’ decisions been, on average? The CEM database offers a clue. The statistical coefficient between net risk-adjusted investment performance and the proportion of fund assets with outsourced investment mandates was -0.3. Specifically, in a universe of 155 funds with five years of continuous return data there was a negative relationship between investment performance and proportion of asset management outsourced (ie, for every 0.1 increase in proportion outsourced, net risk-adjusted performance fell by 0.03% or 3 bps, on average). However, the ‘t’ statistic was only -1.0, suggesting that we cannot reject the hypothesis that the true underlying relationship was in fact non-existent. This lack of statistical significance contrasts with the results of previous research where we correlated the net risk-adjusted investment performance of 80 pension funds with their quality of governance and organization design. Here, a statistically significant positive relationship was found.
Thus in the end, pension scheme sponsors are left with the challenge to build a tangible ‘cost effectiveness’ culture inside their ‘pension subsidiary’. Such a culture must start at the top. The scheme’s governors must understand what their job is, and what it is not. The executive function must be clearly delegated, either internally or explicitly outsourced. The strategic plan should be clear in its goals and in accountabilities for results. Performance measurement systems must clearly align what is measured with what must be managed. Compensation must reflect the achievement (or not) of key elements of the strategic plan. Within this framework, we can rest assured that outsourcing decisions will be part of ‘win-win’ strategies that will benefit all pension scheme stakeholders.
Keith Ambachtsheer is president of KPA Advisory Services, in Toronto, which provides strategic advice to major pension schemes around the world. He is also the co-founder of Cost Effectiveness Measurement Inc. (CEM), providing strategic benchmarking services.