What is the prudent maximum exposure to any one investment manager for my fund?

Transcription

1 What is the prudent maximum exposure to any one investment manager for my fund? Vanguard commentary May 2010 Executive summary. Trustees of investment funds, who bear a fiduciary duty to protect the interests of investors, often view a multimanager approach as one potential method of mitigating risk. In considering such a strategy, they encounter the question What is the prudent maximum exposure this fund should have to any one investment manager? Vanguard has observed funds that implement a wide variety of responses to this question, assigning maximum exposure limits that range from 10% to 100% of public market assets. Trustees impose such limits primarily because of their concern over risks to the fund from a failure or serious error by the manager. This concern is understandable, but it is harder to know how trustees can quantify those risks so as to set an appropriate manager exposure limit. In this paper, we discuss the main reasons to consider limiting individual manager exposure, assess the difficulties of determining such limits, and provide our recommendations on how to best manage these issues. Connect with Vanguard > global.vanguard.com

2 We conclude that no universal rule can be established for maximum exposure to individual managers, because the types and levels of risk vary so widely among management firms and investment styles. We further find that the use of multiple managers may not mitigate overall risk to the degree that trustees might expect. In all cases, thorough research into managers is essential, as is the use of a truly independent custodian. Diversification and risk Vanguard believes that broad diversification, with exposure to all parts of the stock and bond markets, reduces risk. Constructing a portfolio of active managers will generally reduce investment risk compared to using a single active manager. However, how does the use of such a multimanager strategy affect other risks, such as manager error or failure? This paper examines that question. We consider manager risks in two broad categories: investment risks and enterprise risks. Investment risks Recognising that many factors influence financial markets including economic, political, environmental, and social factors we break investment risk into four components: Market risk (equities, bonds, currency, etc). Portfolio risk relative to market. Key-person risk. Investment process failure risk. Market risk Although market risk is an important component of investment risk, it is not significant to a discussion of maximum exposure to individual managers. Market risks and the corresponding return expectations are generally a core component of the total investment strategy, regardless of whether it is to be implemented by one manager or several. All investment portfolios contain some market risk. The evidence is overwhelming that in the long term, broadly diversified market risks are, on average, rewarded by the corresponding risk premium returns. Portfolio risk relative to market Defined most broadly, risk relative to market is the degree to which a portfolio may be expected to deviate from the performance of its benchmark over time. Investment portfolios vary widely in their levels of this risk. Actively managed portfolios clearly bear a higher level of risk relative to market than do indexed portfolios, as an active manager by definition is attempting to earn higher long-term returns than a benchmark after costs. In pursuit of this goal, active managers will hold assets with attributes that differ from the market to some extent. By contrast, index portfolios seek to track their benchmarks as closely as possible, so their risk relative to market should be very low. Because of this substantial distinction, trustees who are considering the maximum prudent exposure question should distinguish between index and active managers, with limits on indexers being correspondingly higher subject to assessment 2

3 of other manager risks. This principle could be extended to distinguish between active investment styles and managers levels of active risk. Key-person risk Trustees also must assess the potential for harm should a key member of the investment team leave the manager. Such a departure potentially could affect the subsequent returns of the managed portfolio. Key-person risk is much more relevant to active management than to indexing, as the latter relies primarily on an investment process rather than on the decisions of any individual. The degree to which active managers are subject to key-person risk needs to be assessed for each firm, and the assessment should be updated through time. Key-person risk is another factor that should lead trustees to distinguish between specialist index managers and active managers, with the maximum exposure to indexers being set higher. Investment process failure risk Here we focus primarily on index managers because, as noted above, indexing relies largely on the implementation of a process. For a specialist index manager, investment process failure would cause the portfolio to become materially out of line with the index benchmark in the worst case, systemically. The risk of a systemic failure of the indexing process across asset classes is considered to be extremely small, because most index managers employ independent decision-making processes for different asset classes. This independence is necessitated by the different ways in which target benchmarks are constructed. The management of an index portfolio is largely driven by the difference between portfolio data and the characteristics of the relevant index. Data for the standard global and Australian equity and bond indexes are provided by a diverse range of organisations: MSCI Barra, S&P/ASX, Citigroup, Barclays Capital, and UBS. This diversity makes it very unlikely that portfolios tracking indexes in different asset classes would experience the same types of process failure with similar degrees of tracking error across the organisation. Leaving aside the risk of systemic failure, a number of other relevant factors should be examined. These include: The robustness of the manager s processes, including the effectiveness of its cross-checks and independent monitoring. The adequacy of the separation of investment management from operations and compliance. The existence of adequate backup for key personnel within the team. The existence and effectiveness of a welldesigned and functional pre-trade compliance system. Given this range of factors and the material differences in practices of investment managers, the risk of an investment process failure needs to be individually assessed for each investment manager both indexed and active. Investment risk: The implications for manager limits How should these investment risk considerations affect limits on the exposure to either individual specialist index managers or active managers? In our view, any attempt to determine a universal maximum exposure is of little value because it cannot properly account for the often substantial differences between managers. Clearly, a manager deemed at risk of process failure would either be removed from the portfolio or given very little responsibility. Otherwise, from the perspective of investment risk, the maximum allowable exposure to any one manager largely depends on comparing average levels of active investment risk in its portfolios, with key-person risk as an attendant consideration. As a result, the 3

4 maximum exposure to individual index managers should be correspondingly higher than that of individual active managers. This preliminary result, however, does not account for enterprise risks, which we consider next. Enterprise risks From our experience in dealing with large investment funds, we believe that trustees concern about enterprise risks is a major reason for the imposition of limits on exposure to any one manager. The key categories of enterprise risk are: (1) operational process risk; (2) fraud; (3) loss of business continuity, which could entail loss of infrastructure (systems, building), loss of organisation (people), or both; and (4) business/competitive risk. We will address each of these categories individually and then comment on other factors relevant to enterprise risk. When assessing enterprise risks, it is very important to begin by considering the role of the independent custodian for the assets to be managed. The custodian is responsible for the safekeeping of all financial assets, as well as for trade clearance and settlement and other administrative services. The custodian should be an organisation completely independent of the fund s investment management service providers. The separation of these two functions investment management from the safekeeping and administration of financial assets is a critical safeguard for investors. For trustees, ensuring that a fully independent custodian is in place is important to the mitigation of fiduciary risk. In the context of this paper, it is also noteworthy that the custodian establishes a precedent for one entity having responsibility for 100% of the assets of an investment fund. Operational process risk Operational process failures can take many forms, including a failure to adhere to an existing process, the failure of a process itself, negligence, and human error. (Fraud is categorised separately within this paper.) What are the safeguards against operational process failures within a fund manager? Trustees of course cannot study a manager s processes in detail, but they can check that the manager follows known best practices across the organisation. For example, an effective separation between the investment management, operations, and compliance teams is desirable not only to deter fraud, but because it keeps responsibilities clear and provides valuable checks and balances on the investment and administration processes. Before employing a manager, trustees should evaluate the processes and experience of the compliance team, which has a critical role in effectively monitoring client portfolios relative to mandate constraints. (In many cases the custodian also monitors client portfolios.) The effectiveness, robustness, and efficiency of the operations function should be assessed for managers individually, both active and indexed. The oversight of the operational risks within an investment manager can also be strengthened through regular and thorough internal and external audit reviews of the investment operations, investment processes, and compliance controls of the manager. Trustees should also recognise that using multiple managers will not reduce the probability of manager errors and failures. In fact, by adding complexity, it increases the probability of experiencing a process failure, as each manager represents an additional opportunity for an error or failure. The overall risk is thus greater than would exist with a single highestquality manager strategy. 4

5 Some would argue that the employment of multiple managers reduces the potential consequences of any single operational process failure. This is simply because in a multimanager structure, each individual firm controls only a portion of the fund s assets. However, that argument assumes that the consequences of a process failure would be proportional to the size of the manager s mandate, which is not necessarily always the case. Manager culture The culture of a management firm is very relevant to an assessment of process risks, in part because it bears on the performance of employees at every level. We strongly believe that trustees should consider this factor when seeking to estimate the likelihood and consequence of a manager error or failure, and hence the maximum permissible exposure to that manager. In particular, the transparency of operations, the degree of emphasis on ethical standards, and the known integrity of the manager are paramount to understanding how the firm operates and how it responds to challenging circumstances. Fraud Unfortunately, one cannot dismiss the possibility that fraud may occur within an investment manager. The steps that trustees can take to reduce the possibility and potential consequences of fraud are similar to those relevant to other forms of manager failure: 1. Employ a quality custodian that is an organisation independent of all investment management service providers. 2. Confirm that an effective separation exists between a manager s investment, operations, and compliance teams. 3. Conduct an exhaustive due-diligence assessment of the manager s internal controls and processes, and update this review regularly. 4. Ensure that the manager regularly undergoes thorough audits. 5. Be wary of extraordinary investment performance and of investment schemes that are opaque or difficult to understand. 6. Assess the integrity, ethics, and behavioural characteristics of the manager and its executives. 7. Check that the manager has adequate insurance coverage. Compared with the use of a single highest-quality manager, would a multimanager structure reduce the likelihood of fraud? Probably not: As is true for the risk of an operational process failure, the use of multiple managers is likely to increase the chance of a fraud simply because each individual manager brings its own risk of experiencing fraud. Would the consequence of fraud be higher in the case of one manager? Potentially yes, but quite likely not. The scale of a fraud is likely to be only loosely related to the size of the portfolio under management. This is because most perpetrators try to hide their activity by limiting its scope. Business continuity risk The major risks to business continuity for an investment manager are: Loss of infrastructure. Regardless of what might cause such a loss, it is essential for the manager to have in place a high-quality, documented, and fully tested disaster recovery plan. Each manager s plan should be carefully assessed by technically qualified people to verify that it can be effectively implemented in an emergency situation. Loss of organisation. This would entail a catastrophic loss of people who comprise the organisation. In such an event, absent effective replacements to take on the key roles and assuming all client assets are held by an independent custodian trustees would need to appoint another manager. 5

6 Business/competitive risk This is the risk of the financial failure of a manager. In such an event, the failed manager may be taken over or merged with a financially stronger firm, or it may simply cease operations. The impact on clients can vary widely, depending on the circumstances. If the client had a separate account managed by the failed manager, then a new investment manager could be appointed to manage the portfolio. The new manager probably would want to divest some holdings and acquire others, and thus a transition cost is likely. Where the client is invested in a fund of the failed manager, redeeming can be problematic, in terms of the timing of the redemption, the value of assets redeemed, and the form of assets redeemed (cash and/or fund assets in kind). Trustees of course cannot foresee such eventualities. They can, however, assess as far as possible a manager s financial solidity and the strength of its position and business strategy in a very competitive industry. Other factors relevant to enterprise risk Insurance coverage Adequate insurance coverage is an important mechanism to provide protection against the costs of manager failure or error. Enterprise risk: The implications for manager limits Of course, managers are not identical with respect to the risk of an operational process failure, or to any other aspect of enterprise risk. Some managers have demonstrably superior compliance, governance, operational, and audit systems, together with higher standards of business integrity. If, from this perspective, the trustees are able to identify the best manager among a group of candidates, then using that manager alone would likely entail a lower overall enterprise risk than would the multimanager equivalent. That is, imposing a maximum exposure limit to any investment manager can have the effect of forcing funds to hire managers with poorer operational and governance structures. In addition to the forms of enterprise risk we have discussed, there are many other aspects of a manager s operation that trustees would need to consider with respect to potential selection. A thorough due diligence assessment of the candidate manager s operations is very important. The necessarily individual nature of such assessments provides another argument against attempting to impose a single maximum exposure limit that would cover all managers, in both index and active management and across asset classes. Management fees Critical consideration of fees is a given for trustees. Not everyone recognizes, however, how portfolio costs may be affected by the imposition of maximum exposure limits for managers. Investment management fee structures are often scaled; that is, fees decrease in percentage terms as the size of the mandate increases. The existence of exposure limits is likely to deprive a portfolio of the benefit of such scaled decreases. This cost is one factor to consider when contemplating the imposition of such a limit. 6

7 Conclusion It is clearly prudent for trustees of a fund to carefully assess the various risks associated with their exposure to its service providers, including investment managers. The trustees of many funds have decided that employing more than one investment manager, and limiting the portfolio s exposure to any individual manager, is one way to control risk. For most large investment funds, there is already a precedent for a single firm having responsibility for 100% of portfolio holdings the fund s custodian. This assignment of responsibility is particularly noteworthy given that the custodian has safekeeping of the fund s actual financial assets. In our view, the ideal of a uniform maximum exposure to any manager is too simplistic, as no universal rule can properly account for each manager s level of potential investment and enterprise risk. Material differences exist between managers investment styles and asset mixes, their operational and investment processes, and their organisational structures, governance, and internal controls. All of these issues are relevant to a manager s potential investment and enterprise risk and hence the maximum exposure a fund may be prepared to allocate to that manager. Also important, in our view, is the manager s business culture, which has implications regarding the likelihood and consequence of problems. In particular, a manager s reputation for ethical standards and integrity can shed light on how the firm operates and how it will respond to challenging circumstances. Compared with employing a single high-quality manager, spreading assets among multiple managers actually increases the likelihood of encountering a manager failure or error. This is because each additional manager presents an additional opportunity for error or failure. But would a multiple-manager structure reduce the potential consequence of a failure? Such a benefit cannot be assumed, as the consequences of an error need not be proportionate to the size of a manager s mandate; the fund and its investors may be affected in unforeseeable ways. Our recommendation is for a thorough assessment by technically qualified professionals to be conducted on the candidate managers, to assess the robustness of each manager s investment and operational processes, organisational structure and governance, culture and internal controls. To assist investors, we have provided an example checklist in the Appendix. This will permit a more comprehensive assessment of the risks associated with individual investment managers. In this way, highly informed decisions can be made on the prudent maximum exposure to each manager. 7

8 Appendix Checklist of matters to investigate when assessing manager risks Investment risks 1. What is the manager s expected risk relative to benchmark (measured by tracking error) for its products? Is the manager an active or index manager? 2. To what degree is the manager exposed to key-person risk? 3. What are the internal checks and oversight within the investment team? Does the manager have a functional pre-trade compliance system? 4. Are the investment, operations, and compliance teams effectively separated? 5. Does adequate backup exist for members of the investment team? Does the investment team genuinely operate as a team? Enterprise risks 1. How efficient and robust is the operations function of the manager? 3. Does the manager have a strong culture of transparency, very high ethical standards, and integrity in the management of the firm s business? 4. Is the custodian an organisation completely independent from all investment management services? 5. Are the manager s operations and investment products transparent and readily understood? 6. Does the manager have adequate insurance coverage? 7. Does the manager have a high-quality, documented, and fully tested disaster recovery plan? 8. Does the manager have effective short-term replacements for key staff, in the event of a significant loss of people? 9. What is the probability of the manager s business experiencing a downturn that could challenge its ongoing viability? 2. Does the manager ensure that regular, thorough internal and external audit reviews are conducted of the investment processes, operations, and compliance systems and controls? 8

9

10

11

12 P.O. Box 2600 Valley Forge, PA Connect with Vanguard > global.vanguard.com (non-u.s. investors) Vanguard research > Vanguard Center for Retirement Research Vanguard Investment Counseling & Research Vanguard Investment Strategy Group > This presentation contains general information and is intended to assist you. This presentation was prepared in good faith and we accept no liability for the contents of this presentation or for any errors or omissions. You should consider your circumstances and those of any other relevant person and, if relevant, the particular PDS, before you make any investment decision. You can access Vanguard s suite of PDSs at Vanguard.com.au. Past performance is not an indication of future performance. We are the trustee of: Vanguard Personal Superannuation Plan ABN / Vanguard LifeStrategy Index PST Conservative ABN / Vanguard LifeStrategy Index PST Balanced ABN / Vanguard LifeStrategy Index PST Growth ABN / Vanguard LifeStrategy Index PST High Growth ABN Vanguard Investments Australia Ltd. All rights reserved. ICRPME