Introduction

Increasing attention is being paid to the performance of institutional funds; whether public or private, large or small, well-governed or not, these institutions have come to play crucial roles in under-writing the welfare of many citizens of developed and developing countries. In terms of the volume of assets managed by these institutions, it is estimated that, as of 2006, across the world pension funds accounted for $US25,000bn, endowments and foundations $4bn, and the emerging sovereign funds $4bn.¹ The price of poor performance is very high (as is noted by many commentators and academics; see Ambachtsheer (2007a) and Lerner et al. (2007)). Inevitably, institutional performance is conditioned by the inherited practices of various bodies that are responsible for these funds. At the same time, we should not be content with simply relying upon the past for the future.

In this paper, we begin by distinguishing between the inherited structure of investment institutions — normally framed by statute, property rights, and covenants — and the governance of those institutions — often framed by the rules and procedures that sustain their performance. This distinction is owed, in part, to Williamson (1996, pp. 4–5) who noted institutional structure is often difficult to change; by his account, it 'evolves' rather than changes in any substantial sense from one time to the next. Like a number of other theorists of institutional design and performance (eg North, 1990), he suggests that 'governance' is an essential ingredient of any institution's functional performance, being the capacity of an organisation to function in ways consistent with desired goals. Institutional structure is, however, not the only determinant of performance: even 'ideal' institutions fail if poorly governed.

Typically, large institutions are organised by formal arrangements of authority and responsibility. Many organisations can provide the interested researcher with figures and maps demonstrating in theory, at least, the proper relationships between line-officers against hierarchically ordered tasks and functions. But the accumulated evidence suggests that formalism is not sufficient as a description of the life of such organisations whether they be general-purpose corporations, financial institutions or pension funds (witness controversy surrounding the agency problems of funds; see Cocco and Volpin, 2007). For many organisations the governance problem is one of orchestrating collective action in a timely and effective fashion given inherited relationships and systems of control; this is especially important in financial institutions that must be adaptive and responsive to market environments that seem to move at the speed of light.

Social scientists often argue that the functional performance of any institution is dependent upon the clarity of a priori defined tasks and functions — Merton and Bodie (2005) provide a template for institutional design relevant to the financial industry, arguing that well-governed organisations have functional (means-ends) clarity. We share this opinion, but recognise that the global finance industry is replete with all kinds of institutions that share similar if not the same functions, differentiated by history and geography. Institutional investors are not the same the world over because they come from distinctive national political traditions (Roe, 2006) and particular iterations of social organisation (O'Barr and Conley, 1992). The challenge of institutional governance can be thought to be comprised of two related parts: to facilitate adaptation to the functional imperatives driving performance without institutional (re)design in the short term and to build long-term performance through reform and re-design of institutional structure.

Ambachtsheer (2007a, 2007b) is of the opinion that many pension and retirement income institutions are not 'fit-for-purpose' whatever their jurisdiction and inherited institutional form. On the other hand, it is not self-evident what works nor is it self-evident what does not work — for example, do some US endowment funds 'out-perform' because they are endowment funds or because they are better governed or both (see Lerner et al., 2007)? If nation states are to redesign pension and retirement income institutions to cope with 21st century imperatives like demographic ageing, the sustainability of plan sponsors, and the increasing premium on (and visibility of) financial performance, issues of structural design must be considered in relation to institutional governance. In fact, knowledge of governance best-practice may be essential for the institutional design process — an issue we return to in the closing sections of the paper.

The paper proceeds in the following manner. In the next section we consider the status and significance of best-practice noting that our use of exemplars is designed to help to understand the underlying principles of institutional governance rather than the particular details of each and every case. This is followed in the subsequent section with a statement about the challenges facing asset owners, especially in relation to investment practice and the flux and flows of global financial markets. We take seriously the insights of Kahneman and Tversky (1979) regarding the cognitive problems of operating in risk environments; we are more optimistic than some about the role that well-governed institutions can play in promoting best-practice (Engel and Weber, 2006).² In the following section, we spell out 12 principles of best-practice, recognising that no institution is perfect. This is followed in the fifth section with a series of arguments about what works in resource-constrained situations before closing in the Conclusion with comments about the design of investment institutions like sovereign funds.

Scoping best-practice

Management consulting firms, business schools, and sections of the business press are pre-occupied with best-practice; the subject is increasingly important to organisations worldwide with many now including a formal statement in their charters to the effect that they will strive for best-practice. In many situations, best-practice is derived from global experience not simply their national or regional context. Client advice, teaching, and communication rely upon the synthesis of experience, the identification of core principles and practices, and their transfer to relevant situations. It is also apparent that the market-share of organisations striving for global best-practice is large and growing, as global economic and financial integration challenge the robustness and legitimacy of inherited institutions.

Our analysis of best-practice matches a concern in the funds management industry to identify the principles and practices of good governance. Research suggests the impact of good governance may be as much as 100–300 basis points per year (Ambachtsheer, 2007a; Watson Wyatt, 2006). In a number of instances, our exemplar institutions had instituted their own policies on governance designed to foster learning from peers. For some institutions, governance has become part of their sub-committee system being often located with the audit function.

One development has been the adoption of a 'governance budget' framework to promote the management of governance innovation (Urwin and others, 2001). In part, a fund's governance budget is related to size. But, as we show below, even smaller funds can adopt best-practice standards appropriate to their size and capacity. Our conception of governance is based on three principles:

• Governance is a finite and conceptually measurable resource, and the size of this resource — the governance budget — is associated with expected performances.
• A certain size governance budget is best matched with a certain investment style and strategy, consistent with other budgets that recognise limited resources and the need for skill.
• There are ways to adapt the governance budget over time with implications for long-term investment performance and pay-offs.

As consultants have become involved in evaluating the effectiveness of fund governance, the definition of governance has evolved. For this paper, the 'governance budget' refers to the capacity to create value from effective actions in the chain of institution-defined tasks and functions (Watson Wyatt, 2004).

There are two rather different approaches to the identification of best-practice. Some analysts rely upon large databases of institutional performance using accepted metrics such as the risk-adjusted rate of return over time to benchmark relative virtue. This approach allows for comparative performance measures across different types of institutions performing similar functions; it also allows for the identification of those types of institutions that are, on average, better performing than others (Lerner et al., 2007). The lessons of this approach are twofold: first, those types of institutions that do better than others ought to be emulated and secondly, those institutions that do better than others can be emulated not withstanding their distinctive attributes and inherited traditions (Gertler, 2001).

At the same time, however, there are acknowledged shortcomings with this approach. Using the risk-adjusted rate of return as the performance measure to discriminate best-practice runs the danger of confusing a common measure of performance with rather different objectives — it is widely appreciated that defined benefit schemes seek to maximise returns subject to their long-term liabilities and government regulations regarding sponsor solvency and mandated funding levels. Defined contribution plans, hybrid schemes, and endowment funds may also seek to maximise returns but do so over very different time horizons and for different purposes. In addition, most performance evaluations have difficulty in using past performance to isolate the relevant determinants of future performance. This issue is exacerbated by the high noise to signal ratio of most measures of investment performance, which weaken the significance of statistical inference studies (Urwin, 1998).

The approach followed in this paper is to rely upon exemplars of best-practice by class of institution, thereby being sensitive to their distinctive attributes while drawing lessons between best-in-class exemplars for industry best-practice. Our selection relied on the authors' extensive knowledge of organisations over a sustained period of time with clear evidence of strong decision-making accompanying success in performance. While performance was not our principle selection criteria, almost all of our best-practice funds had a performance margin of 2 per cent per annum or more over their benchmarks. Our selection of exemplars targeted different types of institutions including corporate pension plans, public pension plans, sovereign funds, and endowment funds. They were also taken from six different countries dispersed across North America, Europe, and Asia-Pacific. As expected, no single country has a dominant position in global best-practice.³

Identifying exemplars of best-practice may be problematic if industry reputation for high-quality governance is the sole criterion for selection. We run the risk of playing favourites with well-known cases rather than challenging the status quo with heretofore unrecognised innovative instances of best-practice. In this project, we went beyond industry reputation, relying upon our shared knowledge of different cases to scope the field for interesting cases. Of course, case study research is challenging for other reasons including the problems sometimes encountered when seeking access to the exemplars deemed most worthy of study. Likewise, care must be taken when reporting results and synthesising experience such that information shared in confidence is not disclosed to the detriment of respondents. Here, we follow social science guidelines regarding respect for confidentiality and anonymity (Clark, 2003). Throughout, no fund or institution is identified by name because we seek to emphasise the principles of best-practice rather than the details of any one institution (see the Appendix for more details).

The challenge of pension fund governance

Notwithstanding the common acceptance of golden rules such as maximising beneficiary (or other stakeholder) interests, pension institutions are subject to many of the same governance problems of the modern corporation (compare Clark (2006) with Jensen (2000)). Such institutions suffer from substantial agency issues, often of a greater order than most corporations. The list of relevant issues includes the following: pension beneficiaries (principals) are unable to monitor the actions of plan administrators and trustees (agents); there may be more than one principal (if we include DB plan sponsors); and there may be an extensive network of agents (such as investment managers) whose motivations and rewards may be difficult to align and difficult to observe (Black, 1992). For most funds, internal investment costs (the direct costs of trustees and their staff) are substantially smaller than external costs (principally of investment managers and other investment agents). The ratio of external to internal costs is generally of the order 10:1 or greater. This external agent expenditure is rarely observed in other corporations (Watson Wyatt, 2006).

Corporate boards of directors do have significant responsibilities, are subject to legal principles such as fiduciary duty (depending upon the jurisdiction), and face formidable rules and regulations as regards their conduct. Even so, in law due deference is paid to the separate operational responsibilities of managers as well as to the myriad of contractual relationships between stakeholders including employees, service providers, and customers. Managers and key associates often receive performance-related pay, especially where their performance is integral to the generation of income distributed to otherwise passive shareholders in the form of stock-price appreciation and dividends (Roberts, 2004). By contrast, the responsibilities of trustee boards are not normally circumscribed by managers' operational responsibilities, and performance-related pay arrangements are very uncommon. Nor are pension beneficiaries normally able to participate in a market for (pension) control. Their reliance on trustee boards for delivery of promised pensions is exceptionally high (Clark and Monk, 2008).

Not surprisingly, trustees are very much aware of their responsibilities. Given that many trustees are only nominally compensated for their roles and responsibilities, an important motive for serving on such boards is the proffered scope of responsibilities in relation to the welfare of others. Well-governed trustee boards segment and prioritise responsibilities, distinguishing (for example) between beneficiaries' claims for special consideration as regards the nature and value of benefits and the investment of plan assets against a target rate of return (see below).

Well-governed trustee boards tend to allocate the routine issues to plan administrators and rely upon reporting systems to oversee the determination and resolution of claims while allocating the available time and resources to issues like investment strategy and management that may affect the long-term integrity of the institution and payment of pension benefits. Well-governed trustee boards also tend to delegate to internal staff and external service providers the execution of tasks and functions governing those relationships by contract and measures of performance (Clark, 2007a). The asset owners in our study showed awareness of these special characteristics, and all made reference to a number of particular challenges of governance that best-practice must surmount.

The challenge of governance is more than the generic issues that afflict all modern organisations — pension funds operate in global financial markets where the management of risk and uncertainty is crucial to the creation of long-term value. Figure 1 demonstrates that governance can create and destroy value shifting the risk-adjusted rate of return above or below the 'gain line' (depending on the risk budget and the governance budget). The implications of this proposition are twofold: first, risk taking against well-defined objectives is an essential ingredient in any well-governed financial institution and secondly, the extent to which risk taking is a deliberate and managed activity depends upon the governance budget allocated to this function within the institution. Poorly governed entities rarely take risk planning seriously and wrongly economise on the governance budget treating it as a cost that limits net financial performance.

Figure 1.

Schematic of governance budget and risk budget

Full figure and legend (49K)

(a) Risk management focus

More generally, in our experience we would contend that pension and retirement institutions must be sensitive to the distinctive attributes of financial markets and behaviour. While theories of financial market structure and performance abound, from our research on pension fund investment it is important that fund decision-makers be able to distinguish between moments of 'normal' risk and moments of uncertainty. In addition, this determination must encompass regime shifts in pricing and risk, and their consequences for non-normal return distributions and investing in extreme conditions. The challenge in risk management could be summarised as employing both quantitative and qualitative disciplines in analysing many fast-moving parts of markets — economic, behavioural, and organisational (Shiller, 2002). The governance challenge here is to function efficiently in the fast changing risk domain, adapting effectively to market signals while recognising that market signals may be subject to unanticipated disruption.

(b) Time horizon focus

Just as importantly, reaping long-term higher than average rates of return requires integrating short-term positions with long-term goals thereby being sensitive to the sequential nature of investment decision-making. Characteristic of poor performing institutional funds, advisors, decision-makers, and stakeholders tend to come to premature conclusions (Wagner, 2002). Long-term optimising must accommodate short-term opportunism while recognising that the short-time horizons of many stakeholders conditions behaviour such that these actions may not be aligned with long-term goals. Therefore, decision-makers have to make a very big adjustment to take a longer-term view, which risks being 'wrong' in the short term. The governance challenge here is to act in the short term with respect to long-term goals, utilising decision-procedures to exploit immediate opportunities but penalise impulsiveness.

(c) Innovative capability

It is widely recognised that financial markets are 'innovation machines' that test investors' fitness to succeed — there are significant rewards for those who are able to identify and exploit unacknowledged market opportunities just as there are enormous rewards for those who create markets and financial products to price and distribute risk (as in alternative investments, infrastructure, and derivates, etc). Recognising the increasing clock speed of markets and strategies places real-time processes at an advantage. Many funds use calendar-time processes, often through a quarterly meeting cycle, making their decisions insufficiently responsive to opportunities and threats. The governance challenge here is to exploit the premium from innovation through the application of judgment and experience to new opportunities, recognising that conventional risk-related procedures may be poorly tuned to the frontiers of finance.

(d) Alignment with a clear mission

Perhaps the greatest governance challenge is to be effective in responding to these governance issues in organisations whose original design, mission, and current size and composition of skills and experience are less than perfect (Clark et al., 2006). For many reasons, pension and retirement income institutions often have a variety of constituents, stakeholders and even competing objectives in the real world of inherited institutions, procedures, and expectations. The governance challenge here is to build alignment behind clear statements of mission-critical goals, particularly in dealing with multiple stakeholders and complex dependencies, recognising that 'reform' is normally an ongoing process of accommodation and only rarely once-and-for-all instances of idealism.

(e) Effective management of external agents

Achievement of goals is unrealistic for the vast majority of funds without external delegation. The characteristic approach of the best-practice fund is to employ external managers in a line-up emphasising diversity to limit risk. In assessing the wide field of choices of investment firms, funds need particular skills and processes for dealing with agency issues, given the informational asymmetries about each manager's value proposition. The processes that have evolved to deal with these decisions have been unsystematic. The governance challenge here is managing the considerable agency issues in using a line-up of managers and other agents that collectively can support the organisation's overall goals.

Governance best-practice

Our project concentrates on three aspects of asset owner best-practice: the ways in which our exemplars organise their governance practices with respect to institutional coherence, their people, and their processes. In summary terms, coherence included consideration of the clarity and focus of investment objectives; people included consideration of those involved in investment decision-making including reference to their skills and expertise; and process included reference to how investment decision-making is organised and implemented. These three aspects of good governance are distinctive but closely related. When screening the available set of case studies to settle on those that represent best-practice, strengths varied; it was found that some institutions are better on institutional coherence than people and process, in other cases, institutions hoped that a strong decision-making process could overcome shortfalls in coherence and people. We contend that the best-governed institutions are those that follow best-practice across all three dimensions.

In this section we concentrate on the lessons learnt about best-practice according to these headings and selectively illustrate those lessons with reference to some of our exemplars. Consideration is also given to the diversity of institutions represented in our study, noting the experience, for example, of public and private pension funds, endowment funds, and sovereign funds. In the Appendix, we list the case study institutions and their attributes and identify them in this manner so as to maintain anonymity. In the following section, we bring together the findings on best-practice by institutional coherence, people, and process to show how these aspects of good governance interact with one another such that best-practice is reinforced and becomes an endogenous element in value-added investment management.

Best-practice — Institutional Coherence

Many investment institutions have a form and structure that is essentially given: in some cases provided by statute and in other cases nonetheless subject to the interests of political and stakeholder constituencies. One way or another, institutional structure is inherited and, if reform is on the agenda, is subject to negotiation and compromise (Roe, 2006). The trick in so many of our institutions, public and private, is to ensure a match between what is inherited with respect to the long-term interests of beneficiaries and other stakeholders. Here, our findings can be summarised as follows.

1. Clarity of the mission and the commitment of stakeholders to the mission statement. In our exemplary cases, abstract golden-rules such as maximising beneficiary welfare were augmented with second-order mission statements such that board members, the senior staff, and stakeholders inside and outside of the institution were able to match the golden rule with an accepted operational goal such as a target yearly real rate of return allowing for liabilities subject to agreed risk parameters. Such funds also developed a set of other supporting goals to support success with their primary goals. The clarity apparent in the funds in our best-practice group is very uncommon in the authors' experience.

2. A highly competent investment function tasked with clearly specified responsibilities and with clear accountability to the institution was characteristic of our exemplars. This arrangement often included an explicit 'map' of institutional authority, distinguishing the responsibilities of trust boards, executives, and service providers. In some cases, there were formal 'charters' providing each element in the governance chain with a mandate for their tasks and functions. In other cases, where charters were not incorporated into standing orders, trust boards sought to provide a clear demarcation of responsibilities, typically distinguishing between strategy and its implementation and execution. The key element for most funds is the executive group, including a Chief Investment Officer with significant delegated responsibility. Our exemplars sought to 'govern' their management by reference to delegated tasks and responsibilities, specified by contract and set in relation to the mission statement and operational goals of the institutions. We have noted elsewhere that many investment institutions blur delegation and deference, often relying upon their senior staff for support in decision-making without clarifying performance objectives, incentives, and sanctions (Clark, 2007a).

3. Most importantly, we observed that institutional coherence was sustained in most cases by resourcing each element in the investment process and governance chain with an appropriate time and resources budget. Unfortunately, resourcing is often seen as a cost to the institution rather than a long-term investment in the coherence of the institution as a functional entity. Our exemplars demonstrated a keen awareness of the value that could be created by internal resources if appropriately targeted.

Best-practice — People

It is, perhaps, a truism that the human capital or talent of any organisation is its most important asset. This is certainly an important theme in contemporary research on industry and firm-related differences in productivity and market performance and is especially important in the financial and service-related industries that overlap with pension and retirement income institutions. Nonetheless, institutions vary a great deal in terms of their ability to select trustees, employ senior staff, and generally govern themselves as human capital-enhancing organisations (Ambachtsheer et al., 2007). Here, our findings were as follows.

4. Leadership has a strong and demonstrable effect on institutional performance, being evident at the board level (particularly in the activities of the chairperson) through to the execution of delegated tasks and functions. Our exemplars sought out highly qualified and respected board chairpersons and charged them with encouraging a culture of accountability and responsibility among board members. This commitment also appears to pay dividends in the selection of senior staff of pension and retirement income institutions, especially when that is matched by a commitment to management by goals and objectives.

5. To the extent trust chairpersons and their boards are able to select their colleagues, three desired qualities guide selection: demonstrable numeric skills, a capacity for logical thinking, and an ability to think about risk in the probability domain. Collegiality is important, but it was often noted that shared competencies combined with peer recognition for experience and ability tended to enhance collective decision-making, whereas disparate and unmatched abilities tend to be a drag on board decision-making (Clark et al., 2006). This issue is under-recognised, with many institutions assuming that commitment, training, and experience can overcome deficiencies. Our exemplars recognised that these competencies are not easy to instil, and so selection of board and staff becomes a critical function. In some cases, our exemplars were able to fashion human resource policies that took advantage of the unique characteristics of their institutions while fashioning tasks and functions that were different than financial institutions.

6. Effective compensation practices are used to build bench-strength and align actions to the mission, with different strategies working according to fund context. Compensation is an important issue. In many cases, our respondents acknowledged that corporate staffing policies and remuneration schemes, public sector scrutiny of salaries and benefits, and the remarkable bonus schemes of bulge-bracket financial firms make head-to-head competition on compensation difficult. This issue has been particularly challenging with respect to key staff members. Different issues arise for board or investment committee members, where in some of our exemplars, payments are set to match the standards set in the mutual fund industry. Whatever the strategy is used, systems of 'reward' are explicitly linked to the mission and performance of the institution and the sense of common responsibility for its performance against objectives. We observed that many funds have acquiesced to a double-standard in compensation — paying limited packages in-house and paying fees that support much more substantial packages externally. Our exemplars have recognised the contradiction implied in this distinction.

Best-practice — Process

By our analysis, institutional coherence and the people involved in decision-making are essential pre-conditions for a high-performance pension and financial institution. Without a clear mission statement and operational goals and the people to frame and implement an appropriate investment strategy, a disciplined investment process will not deliver desired results. On the other hand, with both pre-conditions in place the evidence suggests that the process of investment decision-making was the most important means of reaping the potential value of an institution. It was also noted that this element in the governance chain was that which the institution could most control. Our findings were as follows.

7. Our exemplars rely upon a process centred on strong beliefs and an investment philosophy claiming fund-wide support that aligns with operational goals and informs investment decision-making. Only with a clear and accepted belief structure can an institution sustain its competitive edge in financial markets. In our research, we observed exemplars focus upon four main areas of this issue: (1) asset class and security pricing including the 'fair' prices of investment opportunities, the reasons why mis-pricing can occur, and the degree to which mis-pricing is a systematic fact of life; (2) the fund's ability (or its comparative advantage) in exploiting such identified opportunities; (3) how the fund might develop and integrate these beliefs into its investment strategy; and (4) what these strategies can produce, in value-added and risk terms, across the whole portfolio. Many institutions distinguish between different types of strategic issues and the appropriate location of decision-making relevant to those issues particularly between investment committees and the executive. For example, in many cases the most developed investment beliefs are located at the executive level. But it is still critical for an informed board to build their own beliefs and deal effectively with those of the executive.

8. Our exemplars frame the decision-making process by reference to the institution's comparative advantages and disadvantages. Few investment institutions are able to operate effectively in all investment domains (some are better suited to public markets, whereas others may have the capacity to operate most effectively in private markets or exotic products). The best-practice process of decision-making takes into account an institution's own capacities and its acknowledged limits and acts accordingly. This includes deciding on the degree of delegation, choosing to act in a primary investment role, selecting individual investments in some areas, or acting as a manager of managers in other areas where investments would be selected by outside managers. Funds varied in their degree of use of external managers, but it is interesting that none managed all assets in-house.

9. Our exemplars frame the investment process by reference to a risk budget aligned to fund goals incorporating an accurate and integrated view of alpha and beta. Many of our institutions utilise an absolute return ethos, constrained by a risk budget, which is explicit about the desired relative contributions of alpha and beta to overall fund performance. This is a quantitative decision-making framework reinforcing the significance of trustee skills and qualities that sustain consistent method-based decision-making (Clark et al., 2007). While the increased opportunity to separate alpha and beta has attracted many funds' attention, most of the exemplars in our group concentrated on improved alpha and beta transparency in their line-up, but most did not go as far as more formal alpha-beta transport. They all sought clarity over how to manage a judicious mix of alpha and beta consistent with their goals.

10. Recognising the time-dependent nature of investment performance, best-practice institutions utilise decision-making systems that function in real time not calendar time. There are various ways of doing this, including devolving decision-making to expert sub-committees, most of which involve greater delegation of time-dependent decision-making to executives or external firms subject to board over-sight. The authors' contend that calendar-time governance is typical of most funds; the crucial issue here is how that is reconciled with real-time markets.

11. Best-practice masters the effective use of external managers through clearly defined mandates, aligned to goals, and selected with rigorous application of fit-for-purpose criteria. In other words, best-practice institutions distinguish between the nature and types of decision-making by operational entities taking care not to compromise decision-making at one level by poor decision-making at other levels. Characteristically, best-practice asset owners employ external managers in a line-up emphasising diversity so as to limit risk. Mandate specification is one area of importance. Also fit-for-purpose assessment of firms and products is important. Typically, we found three aspects of suitability: (1) investment efficiency, allowing fully for costs, (2) alignment to the fund's needs to achieve sustainability of performance goals, and (3) an appropriate transparency of process, allowing for an assessment of the product according to its manager skills (alpha) and market return (beta) drivers. While acknowledging that the selection of managers is always problematic, our exemplars showed considerable rigour with applying fit-for-purpose assessment of outside firms and their investment products. They also made frequent reference to the importance of the de-selection process as well as the selection process.

12. In terms of investment decision-making, best-practice institutions work within a learning culture that deliberately encourages change and challenges the commonplace assumptions of the industry. In part, this means that past decisions are evaluated against actual outcomes so as to calibrate the decision-making process while allowing appropriately for noise and signal issues (Clark, 2004). In part, this also means that institutions are routinely turned inside-out by challenging trustee boards and senior executive staff to be innovative within the bounds of institutional capacity. Accelerating change is a given of the funds industry; technology is always moving forward and its effect on the growth of knowledge is positive at an accelerating pace. Knowledge of what works and what does not work in investment is at a premium and is time-sensitive; there is added value in changing with new knowledge and opportunities.

The full list of the 12-factor model of best-practice is summarised in Table 1.

Table 1 - Best-practice factors.

Full table

Constrained best-practice governance

We are conscious that these findings, when taken together, are idealistic on two fronts. First, none of our exemplars could be said to be at the leading edge of each and every component of best-practice. Secondly, this list is premised on significant internal resources, which many funds do not have available or are unable to mobilise.

We noted that best-practice institutions were generally aware of their own shortcomings on some or all of these issues. In fact, it could be argued that best-practice funds are those that continuously seek to improve their functional performance whatever their inherited structures and practices (comparing Merton and Bodie (2005) with Roe (2006)). That is, there is a self-critical ethos of institutional learning and best-practice; complacency is the enemy of long-term value creation. One key institutional quality we observe in our exemplars is their use of expertise in investment decision-making and, in particular, whether they utilise in-house investment experts. Based on our experience, we can identify three types of fund structure and organisational design (Figure 2). The simplest type is widely known, being a system of collective deliberation wherein the board makes decisions on a routine basis with the support of a consultant and external service providers. A more sophisticated version utilises an investment sub-committee subject to the final approval at the board relying, again, on collective decision-making according to the regular meeting schedule (Type 2). In Finding 2, we identified in-house investment expertise as a key factor in best-practice our exemplars utilise to drive real-time decision-making (Type 3).

Figure 2.

Governance types

Full figure and legend (28K)

As is widely appreciated, the most prevalent organisational forms are those in which there are no significant internal resources deployed supporting multiple-function boards. Rough calculations suggest that if we consider all institutional funds across the world that have assets that are above $2bn (and there are somewhere over 2,000 of these), we would argue that only around 10 per cent of this group are set up as Type 3 with significant delegated investment authority. In part, this is a product of size — the limited resources available from the plan sponsor and the relatively small volume of assets under management — and the limited time available for decision-making by members of the board. Such funds can hardly embrace the full set of best-practice criteria in the short term or even the long term.

With these considerations in mind, Table 2 separates the best-practice points between the 'exceptional' group associated with the Type 3 structure with an internal executive, and a 'core' group, which are within the range of all funds that seek to strengthen the formal structure of decision-making. In effect, we believe that formal procedures and requirements can compensate to a degree for a lack of institutional capacity and ability. This is the subject of further research, especially in relation to the consequences of such an organisational strategy for investment performance (see below).

Table 2 - Best-practice factors by type of fund.

Full table

Type 1 and Type 2 fund structures face two significant challenges. First, Finding 5 indicates that there is a strong case for selecting members of the board and investment committees based on the task-related competencies needed to be effective asset owners. This finding comes from our exemplars as well as academic research (see Ambachtsheer et al., 2007; Clark et al., 2006, 2007). However, most boards and investment committees are shaped by a variety of agendas including stakeholder representation sometimes leaving funds with competency deficits relative to the specialised skills required to be effective. We do not dispute the value of a representative board (especially in terms of fund sponsorship and motivation). But we do suggest that the criteria for board selection should be balanced against best-practice such that representation reinforces at least the six-point guidelines as summarised in Table 2.

Secondly, Finding 10 sets out the normative role of 'real-time' investing. Type 1 and Type 2 funds cannot employ a real-time approach given their reliance on the periodic meetings of their boards and/or investment committees. This limitation is likely to have its costs, not least involving opportunities that such funds are forced to forego (see below). On the other hand, in these cases, best-practice should recognise the institutional limits of such entities against competing market players eschewing active management for passive management such that the elements listed in Table 2 are focused on beta not alpha activities. Elsewhere, Clark (2004) has argued that size is a real constraint on governance capacity and performance; over the long term, it is arguable that such resource-constrained institutions should seek ways of sharing resources or merging into larger entities.

Best-practice investment strategy

This study concentrates on governance best-practice. In our interviews, it also became apparent that funds' governance budgets are associated with particular institutional capacities and features. At its simplest, an appropriate governance budget is a pre-condition for an effective investment strategy, recognising the limits imposed by fund size and committed resources including time and expertise (noted above). More generally, the governance budget is also a strategic instrument framed according to funds' ambitions in relation to long-term investment objectives. Our exemplars were, more often than not, deliberate about their chosen governance procedures and practices, treating governance as an investment in realising their objectives. Here, a balance is normally struck between short-term cost efficiency and long-term fund performance (even if it is sometimes difficult calibrating the value created by effective governance).

Matching the significance we attribute to formal governance procedures, especially in Type 1 and Type 2 institutions, we suggest that those procedures are matched on the investment side of the equation by certain characteristics. Basically, lower governance budget arrangements are consistent with less complicated or sophisticated arrangements. If this is not the case, we expect some difficulties with such funds' implementation of complex arrangements. The 'value drivers' that funds can use are summarised in Table 3 where, on the left-hand side of the table, the first four drivers of value are deemed appropriate to all types of funds. The second set of four value drivers imply a level of discretion and flexibility with respect to investment policy that Type 1 and Type 2 funds would have difficulty in sustaining given their resources. This is the domain of Type 3 institutions, those characterised by a large volume of assets under management and organisational resources including time, commitment, and real-time investing.

Table 3 - Investment value drivers.

Full table

Of course, operating in this domain is very challenging. For instance, hiring, monitoring, and replacing active managers is a time-consuming process that demands a level of internal expertise among senior staff that is difficult if not impossible to provide in Type 1 and Type 2 organisations. Furthermore, once we move into time-dependent portfolio optimisation using leverage and risk weighting, senior management must be able to make tactical decisions backed by boards that appreciate the nature and scope of the risks assumed. While we are sometimes told that 'trust' between staff and boards is an essential ingredient in investment management in these circumstances, we also note that best-practice is less about trust and more about contract wherein staff are set responsibilities and performance parameters with appropriate levels of compensation.

In these circumstances, our exemplars tend to treat governance as an instrument of management as well as an instrument of control. As a result, some of the most effective Type 3 institutions are those that have made governance design and oversight a standing sub-committee of the board with responsibility for monitoring board performance and its relationships with senior staff and the myriad of consultants and service providers who populate the industry. That is, our exemplars have sought to identify best-practice forms of governance and mechanisms of accountability. These are summarised in Table 4. So, for example, our exemplars are conscious of the costs for decision-making of a large board recognising that many members (normally more than nine) tend to fracture collegiality (Sunstein, 2005) and add a degree of heterogeneity in board member competence that undercuts competent decision-making (Clark et al., 2006). Not surprisingly, our exemplars have become active in the recruitment and nomination of board members even if, in the end, they do not normally control the selection process.

Table 4 - Best-practice board/investment committees.

Full table

Just as our exemplars have developed mechanisms to govern their relationships with senior staff and external service providers, boards have sought to enhance their own systems of accountability. So, for example, some boards have created sub-committees or have used their audit committees to make governance an ongoing issue of scrutiny and oversight. Some of our exemplars have instituted yearly reviews of board member performance in conjunction with longer-term contracts designed to capture the expertise and specialised learning that comes with commitment. Significantly, some funds have introduced trustee compensation schemes, matching the obligations historically associated with the trust institution (Langbein, 1997) with a realistic assessment of the roles and responsibilities of board members.

Conclusions

Governance is on the agenda of many of the world's leading pension and investment institutions. Prompted by the challenges posed by global financial markets and closer scrutiny of performance by sponsors and stakeholders, investment by goals and objectives has demanded innovation in how funds are governed. We note that setting targets and constraints such as rates of return and stable contribution rates has had a salutary effect in many institutions, challenging past practices and encouraging focus upon organisational coherence, the people involved, and decision-making processes. We also note that these initiatives can be found in many different national settings and across a broad array of institutional forms (including sovereign funds, endowment funds, public and private sector funds, etc).

In this paper we have used exemplars to illustrate these developments, drawing inspiration from a select group of institutions that have shared with us their governance strategies and practices. In each and every case study, respondents have emphasised that governance is best treated as an investment in long-term performance rather than a short-term cost to be carved out of sponsor contributions or investment returns. In part, the governance budget is part of the management of risk and return. More importantly, the governance budget is part of any institution's commitment to strategic investment management recognising the frontiers of financial engineering and the challenges that face any institution when operating in the real-time world of financial markets.

As suggested, our exemplars provide insights about best-practice governance. These are summarised in Figures 1 and 2 and Tables 1, 2, 3 and 4. Nonetheless, we recognise that these insights represent a high hurdle for any institution and its board members and senior staff; it is arguable that our exemplars have certain advantages such as size and structure that have allowed them to develop their distinctive approaches to governance. In many cases, a large pool of assets or the prestige of the fund concerned has provided an effective platform for developing best-practice governance. In other cases, leadership has been an essential ingredient in institutional innovation. Nonetheless, we could identify, as readers could identify, similarly sized institutions that seem to lack a commitment to best-practice (as implied by Lerner et al., 2007).

We would also suggest that these findings could be used to inform debate over the design of the new sovereign funds that have come to occupy an important place in national savings programmes. In some countries, these institutions have become a means of realising the apparent advantages of scale and scope in the context of declining coverage rates by occupational and industry pension plans. In other countries, sovereign funds have been a mechanism for mobilising social security assets for placement in global financial markets in the hope of reaping higher rates of return. In yet other countries, sovereign funds are strategic investment vehicles for foreign reserves and earnings. Whatever their origins, sovereign funds are likely to be as important for 21st century markets as Anglo-American pension funds were for the second half of the 20th century (Clowes, 2000).

Sovereign funds face significant challenges in implementing best-practice governance. In some cases, the objectives of such institutions are unclear and subject to unresolved debate. In other cases, governing boards are quite large and heterogeneous in terms of the skills and aptitudes of those appointed. When combined with poorly specified responsibilities, the process of investment decision-making can become a competition for power and influence rather than a process responsive to the five challenges of governance that underpin this paper. In these cases, inadequate governance may translate into unrealised promises of national wealth. Just because an institution controls a large volume of assets does not mean that it is a well-governed entity.

Our evaluative framework and 12 findings have significant implications for the design of these types of institutions. For example, we have suggested that return targets and contribution constraints have played a vital role in focusing boards on the nature and scope of their governance processes (Finding 1). Likewise, we have suggested that investment institutions deserve to be governed in a deliberate manner with formal charters or mandates used to set roles, responsibilities, and accountabilities (Finding 2). And we have emphasised that governance is an investment not just a cost (Finding 3). Most importantly, we focused on people and process emphasising that those involved ought to have certain types of skills and aptitudes (Findings 4–6) as well as carry out well-defined responsibilities in a disciplined investment process (Findings 7–12).

Notes

¹ Here, we refer to data collected by Watson Wyatt through the Global Pension Asset Study of 2007; further details are available at www.watsonwyatt.com.

² See also the recent commitment shown by the CFA Institute in promoting a code of conduct for pension scheme governing bodies wherein the code will require members to 'take actions that are consistent with the established mission of the scheme' and 'regularly review the efficiency and effectiveness of the scheme's success in meeting its goals'. See www.cfainstitute.org.

³ Note that we focus upon governance principles and policies in this paper and ignore, for the moment, the distinctive regulatory and legislative environments within which our chosen funds operate. This is not because we think this is irrelevant, quite the contrary. Rather, our emphasis on principles and policies is such that we believe that over the long term the regulatory environment ought to enable best-practice rather than constrain best-practice. See Clark (2007a, 2007b).

References

1. Ambachtsheer, K. (2007a) Pension Revolution: A Solution to the Pensions Crisis, John Wiley, New York.
2. Ambachtsheer, K. (2007b) 'The Three Grades of Pension Fund Governance Quality: Bad, Better, and Best', mimeo, Rotman International Center for Pension Management, University of Toronto, Toronto.
3. Ambachtsheer, K., Capelle, R. and Lum, H. (2007) 'The State of Global Pension Fund Governance Today: Board Competency still a Problem', mimeo, Rotman International Center for Pension Management, University of Toronto, Toronto.
4. Black, B. (1992) 'Agents Watching Agents: The Promise of Institutional Voice', UCLA Law Review, 39, 811–893.
5. Clark, G. L. (2003) European Pensions & Global Finance, Oxford University Press, Oxford.
6. Clark, G. L. (2004) 'Pension Fund Governance: Expertise and Organisational Form', Journal of Pension Economics and Finance, 3, 233–253. | Article |
7. Clark, G. L. (2006) 'Regulation of Pension Fund Governance', in G. L. Clark, A. Munnell and M. Orszag (eds.) The Oxford Handbook of Pensions and Retirement Income, Oxford University Press, Oxford, 483–500.
8. Clark, G. L. (2007a) 'Expertise and Representation in Financial Institutions: UK Legislation on Pension Fund Governance and US Regulation of the Mutual Fund Industry', 21st Century Society: Journal of the Academy of Social Sciences, 2, 1–23.
9. Clark, G. L. (2007b) 'Governing Finance: Reconciling Functional Performance with Stakeholder Representation', Working Paper 07-08, Oxford University Centre for the Environment, Oxford.
10. Clark, G. L., Caerlewy-Smith, E. and Marshall, J. (2006) 'Pension Fund Trustee Competence: Decision Making in Problems Relevant to Investment Practice', Journal of Pension Economics and Finance, 5, 91–110. | Article |
11. Clark, G. L., Caerlewy-Smith, E. and Marshall, J. (2007) 'The Consistency of UK Pension Fund Trustee Decision-Making', Journal of Pension Economics and Finance, 6, 67–86. | Article |
12. Clark, G. L. and Monk, A. H. B. (2008) 'Conceptualizing the Defined Benefit Pension Promise: Implications from a Survey of Expert Opinion', Benefits Quarterly, 24, 7–18.
13. Clowes, M. J. (2000) The Money Flood: How Pension Funds Revolutionized Investing, Wiley, New York.
14. Cocco, J. F. and Volpin, P. F. (2007) 'Corporate Governance of Pension Plans: The UK Evidence', Financial Analysts Journal, 63(1), 70–83. | Article |
15. Engel, C. and Weber, E. U. (2006) 'The Impact of Institutions on the Decision How to Decide', MPI Collective Goods Preprint No. 2006/19. Available at SSRN: http://ssrn.com/abstract=93502.
16. Gertler, M. S. (2001) 'Best Practice? Geography, Learning and the Institutional Limits to Strong Convergence', Journal of Economic Geography, 1, 5–26. | Article |
17. Jensen, M. (2000) A Theory of the Firm: Governance, Residual Claims, and Organizational Forms, Harvard University Press, Cambridge, MA.
18. Kahneman, D. and Tversky, A. (1979) 'Prospect Theory: An Analysis of Decisions Under Risk', Econometrica, 47, 263–291. | Article | ISI |
19. Langbein, J. (1997) 'The Secret Life of a Trust: The Trust as an Instrument of Commerce', Yale Law Journal, 107, 165–189. | Article |
20. Lerner, J., Schoar, A. and Wongsunwai, W. (2007) 'Smart Institutions, Foolish Choices: The Limited Partner Performance Puzzle', Journal of Finance, 62, 731–764. | Article |
21. Merton, R. and Bodie, Z. (2005) 'The Design of Financial Systems: Towards a Synthesis of Function and Structure', Journal of Investment Management, 3, 1–23.
22. North, D. C. (1990) Institutions, Institutional Change and Economic Performance, Cambridge University Press, Cambridge.
23. O'Barr, W. and Conley, J. (1992) Fortune and Folly, Wiley, New York.
24. Roberts, J. (2004) The Modern Firm: Organizational Design for Performance and Growth, Oxford University Press, Oxford.
25. Roe, M. J. (2006) 'Legal Origins, Politics, and Stock Markets', Harvard Law Review, 120, 460–527.
26. Shiller, R. J. (2002) 'Bubbles, Human Judgement, and Expert Opinion', Financial Analysts Journal, 58(3), 18–26. | Article |
27. Sunstein, C. (2005) 'Group Judgements: Statistical Means, Deliberation, and Information Markets', New York University Law Review, 80, 962–1049.
28. Urwin, R. (1998) Avoiding Disappointment in Investment Manager Selection, International Association of Consulting Actuaries, London.
29. Urwin, R., Breban, S., Hodgson, T. and Hunt, A. (2001) 'Risk Budgeting in Pension Investment', Annals of the Faculty of Actuaries, 7, 319–364.
30. Wagner, R. K. (2002) 'Smart People Doing Dumb Things: The Case of Managerial Incompetence', in R. J. Sternberg (ed.), Why Smart People Can Be So Stupid, Yale University Press, New Haven, 42–63.
31. Watson Wyatt (2004) Remapping, Reigate.
32. Watson Wyatt (2006) Changing Lanes, Reigate.
33. Williamson, O. (1996) The Mechanisms of Governance, Oxford University Press, Oxford.

Appendices

Appendix

Case study exemplars

Shown below are brief sketches of the institutions that were the basis of our case studies. Inevitably, these sketches are shallow and indicative rather than definitive, and properly so given our undertakings regarding confidentiality. Whereever possible, interviews were conducted with the CEO or CIO of the institution, or nominee; certainly, someone with insight regarding its investment performance and knowledge of the nature and scope of the governance issues encountered therein. The procedures governing the interviews including the stages of the process are explained in more detail in Clark (2003) and conform to standard social science practices.

Funds ranged in size from $5bn to $100bn. Five funds were located in North America, three in Europe and two in Asia-Pacific.

Fund A: a large, multi-employer, state-sponsored fund investing on behalf of the participating public sector defined benefit pension plans.

Fund B: a large, multinational company cross-listed between three stock exchanges, with substantial consolidated pension liabilities principally DB in nature.

Fund C: a very large, multi-employer industry fund offering a range of retirement plans including hybrid versions of defined benefit and defined contribution plans.

Fund D: an industry fund operating in a competitive national market for investment management and related services in the defined contribution environment.

Fund E: a corporate defined benefit pension plan with an in-house investment division to manage its pension assets.

Fund F: a global company with significant worldwide pension assets in particular with large US DB plans.

Fund G: a major endowment fund with a long-term commitment to the growth and stability of its university sponsor.

Fund H: a leading global endowment fund with a mandate in perpetuity in the interests of research.

Fund I: a national pension and retirement savings institution operating on behalf of national and the second pillar of government investment provision.

Fund J: a national pension fund operating in a unitary state, with responsibility for the investment of mandatory individual contributions for supplementing the basic pension.

Acknowledgements

The authors wish to thank the funds and institutions that participated in the study recognising that most requested anonymity and, as such, cannot be formally identified. The authors also wish to thank their colleagues at both institutions including Emiko Caerlewy-Smith, Tessa Hebb, Lisa Hagerman, John Marshall, and Dorothee Franzen (at Oxford) and Michael Orszag, Carole Judd and Lloyd Raynor (at Watson Wyatt). Keith Ambachtsheer, Charles Ellis, Adam Dixon, and Ashby Monk made useful comments on a previous draft. Research on pension fund governance at Oxford has been supported by the National Association of Pension Funds (United Kingdom), the Lupina Foundation (Toronto), and the Rockefeller and Ford foundations. None of the above should be held responsible for any comments or opinions expressed herein, especially any misapprehension of the lessons to be drawn from particular instances of best-practice value-creating investing.