FINANCE FINANCE, COMPETITIVENESS & INNOVATION EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT Analysis of Pacific National Funds Investment Strategies and Results: Regional Comparative Study © 2021 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved. This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO), http://creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: 2021. Analysis of Pacific National Funds Investment Strategies and Results: Regional Comparative Note. EFI Insight-Finance. Washington, DC: World Bank. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third- party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. Cover design and typesetting: Diego Catto / www.diegocatto.com Cover photo: Marshall Islands, Coastline aerial | Source: https://earthview.withgoogle.com >>> Acknowledgments This note was prepared by Ekaterina M. Gratcheva (lead financial officer, Finance, Competitiveness and Innovation Global Practice) with analytical support from Teal Emery (consultant, Finance, Competitiveness and Innovation Global Practice). The paper has been peer reviewed and incorporates comments from Eric Bouye (lead investment officer, World Bank Treasury Advisory and Partnerships), Aaron Drew (managing director, MyFiduciary Ltd., and formerly with New Zealand Superannuation Fund and New Zealand Institute for Pacific Research), Aaron Levine (senior operations officer, International Finance Corporation). The team is greatly appreciates the feedback received from Gabriel Petre (lead investment strategist, World Bank Treasury Pension Investments), Roberto De Beaufort Camargo (lead financial officer, World Bank Treasury Advisory and Partnerships), Andrius Skarnulis (Young Professional, Finance, Competitiveness, and Innovation Global Practice), Lasse Melgaard (World Bank’s resident representative for the South Pacific), Annette Leith (World Bank’s resident representative for Solomon Islands and Vanuatu), Demet Kaya (senior economist, Macroeconomic and Fiscal Management Global Practice, East Asia and Pacific), Kim Edwards (program lead and senior economist, Macroeconomics, Trade and Investment Global Practice, East Asia and Pacific), Anna Robinson (economist, Macroeconomics, Trade and Investment Global Practice, East Asia and Pacific), Andrew Blackman (senior economist, Macroeconomics, Trade and Investment Global Practice, East Asia and Pacific), Caroline Adams (operations officer, East Asia and Pacific). The report also benefited from inputs and insights from clients and regional partners throughout the team’s engagement in the region and during the technical assistance process. We are thankful to Denton Rarawa (Pacific Islands Forum Secretariat) for his collaboration and for his shepherding this study with FEMM members during the 2021 meeting in the midst of the COVID pandemic. And last but not least, this report would not have been possible without David Gould’s (lead economist, and program leader, Equitable Growth, Finance and Institutions Global Practice for Papua New Guinea and Pacific Islands) tireless support for this project and his ability to navigate complex Pacific landscape to ensure this report is published. The data used for the comparative analysis presenting in the note are a combination of data provided by the funds participating in the study, publicly available information about the funds and their investment strategies or portfolios. The results presented in the note are based on the team’s analysis of provided data, market data from publicly available sources, and specialized market data providers. Acronyms 7 Executive Summary 8 1. Impact of Investment Strategy Decision 11 2. Impact of Active Management Decision and Management Costs 13 3. Key Takeaways 11 1. Note Objective 16 2. Introduction and Pacific Funds Context 19 3. Note on the Use of Individual Fund Data in Comparative Analysis 22 4. Methodology for Comparative Analysis 25 1. Role of Investment Benchmarks 25 2. Defining Reference Portfolios for Funds with Different Risk Tolerance Parameters 27 3. Impact of COVID-19 on Market Volatility and Returns 29 5. Benchmarking of Pacific Funds 31 Contents 1. Pacific Funds’ Peer Comparison 31 >>> 2. Benchmarking Tuvalu’s Investment Approach: OBAA Strategy and Its Effects 33 3. Benchmarking Kiribati’s Investment Approach: 2013–2016 Restructuring 37 and Its Impact. 4. Benchmarking Federated States of Micronesia and the Marshall Islands Investment 40 Process: Gradual Evolution. 5. Nauru: Reestablishing Trust Fund 42 6. Benchmarking of Investment Management Costs 43 7. Conclusions 47 7. References and Other Readings 50 Appendix A. Reference Portfolios: Definition and Construction 53 Appendix B. Morningstar $A Multi-Sector Market Indices 59 Appendix C. Total Costs and Asset Managers Fees for Federated States of 61 Micronesia and Marshall Islands Trust Funds Appendix D. Do Active Managers Add Value Net of Fees Over Time? 65 Appendix E. Comparison with Endowment Model 68 >>> Figures Figure ES.1. Pacific Funds 3- and 5- year Net Returnsa 10 Figure ES.2. Pacific External Funds Management Costs, 2019 10 Figure 4.1. Growth of $100 Over 30-year Horizon 26 Figure 4.2. Return vs. Risk (as Volatility of Returns) of Asset Classes and Reference Portfolios 28 Figure 4.3. Returns vs. Downside Risk of Asset Classes and Reference Portfolios 28 Figure 4.4. Growth of $100 Over 30-year Horizon 29 Figure 5.1. Pacific Funds Peer Comparison: 3- and 5-year Net Returns 31 Figure 5.2. Movement of TTF Asset Category Allocation Since Inception of OBAA Strategy 35 Figure 5.3. Growth of RERF Before 2013–2016 Restructuring 38 Figure 5.4. Pacific External Fund Management Costs, 2019 43 Figure 5.5. Evolution of Investment Fees (Expense Ratios) Since GFC 44 Figure 5.6. Shift in Market Demand for Lower-Fee Products 45 Figure D.1. Percentage of Actively Managed Funds That Outperformed the Market 67 Figure D.2. Migration of Top Quartile Managers Over Investment Horizon 67 Figure E.1. Average Annual Endowment Returns, FY10–FY19 70 Figure E.2. Average 10-year Endowment Returns, FY10–FY19 70 >>> Tables Table ES.1. Key Features of Pacific Trust and Sovereign Funds 9 Table ES.2. TTF’s Investment Returns vs. Alternatives 12 Table ES.3 RERF’s Investment Returns vs. Alternatives 12 Table ES.4. Federated States of Micronesia and Marshall Islands TF Investment Returns vs. Alternatives 12 Table 2.1. Key Features of Pacific Trust and Sovereign Funds 19 Table 3.1. Funds Data Details 23 Table 5.1. Strategic Asset Allocation for Pacific Funds and Comparable Peers 32 Table 5.2. Tuvalu’s Returns vs. Reference Portfolios Table 5.3. Comparison of TTF Returns vs. $A Morningstar Investment Strategies 36 Table 5.4. Kiribati’s Returns vs. Reference Portfolios 37 Table 5.5. RERF Policy (Benchmark) Return vs. Portfolio Net Return Before 2013 SAA Review 38 Table 5.6. Comparison of RERF Returns vs. $A Morningstar Investment Strategies 39 Table 5.7. Federated States of Micronesia and Marshall Islands’ Returns vs. US$ Reference Portfolios 41 Table 5.8. Implementation of Nauru Trust Funds Asset Allocation, 2018–2018. 42 Table A.1. Key Attributes of a Reference Portfolio 54 Table A.2. Definition of Income and Risk Assets 55 Table A.3. Correlation of Monthly Returns for Income Assets and Risk Assets, December 1995–April 2020 56 Table A.4. Asset Composition of Reference Portfolios for Different Risk Levels 57 Risk and Return Characteristics for Income Assets, Risk Assets, and Reference Portfolios, Table A.5. 57 December 1995–April 2020 Table A.6. Treatment of $A Returns for Reference Portfolios 58 Table C.1. Federated States of Micronesia Trust Fund’s Investment and Administrative Costs, FY18 and FY19 62 Table C.2. Marshall Islands Trust Fund’s Investment and Administrative Costs, FY18 and FY19 62 Table C.3. Federated States of Micronesia Trust Fund’s Asset Managers’ Fees, FY18 and FY19 63 Table C.4. Marshall Islands Trust Fund’s Asset Managers’ Fees, FY18 and FY19 64 Table E.1. FY19 Net Investment Returns for Different Size Endowments 70 Table E.2. FY18 Asset Allocation for Different Size Endowments 71 Table E.3. FY18 Asset Class Investment Returns for Different Size Endowments 71 Table E.4. FY18 Variability of Investment Returns for Different Size Endowments 72 >>> Acronyms $A Australian dollar AUM assets under management FEMM Finance and Economic Ministers Meeting (annual) GFC Global Financial Crisis (2008) IMF International Monetary Fund IPS investment policy statement NACUBO National Association of College and University Business Officers NPRT Nauru Phosphate Royalties Trust $NZ New Zealand dollar NZIPR New Zealand Institute for Pacific Research OBAA objective-based asset allocation RAMP Reserve Advisory Management Program (World Bank Treasury) RFP request for proposals RERF Reserve Equalization Revenue Fund (Kiribati) SAA strategic asset allocation TIAA Teachers Insurance and Annuity Association of America TF trust fund TTF Tuvalu Trust Fund US$ US dollar EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 7 >>> Executive Summary At the 2019 Finance and Economic Ministers Meeting (FEMM), policy makers called for improved transparency of the Pacific trust funds’ management practices and investment results as a way to foster its improved management going forward: “Forum recognizes the need to undertake regional analyses on the comparison of the governance, portfolio management and returns of national trust funds.” (FEMM 2019, point 6a). FEMM tasked the World Bank to undertake the study. Whereas significant progress has been made in establishing international best practices for sover- eign or large public funds or both, trust funds in the Pacific are unique in making direct application of those practices challenging. Furthermore, the comparative analysis of the management prac- tices and relevant performance metrics of Pacific trust funds are still lacking. FEMM’s objective for disclosing comparative analysis of Pacific funds investment results is to further stimulate collab- orative discussion on how to continue to strengthen Pacific funds’ management and to inform the design and implementation of relevant reforms. In response to the FEMM request, this note will ad- dress the current vacuum of comparable information about Pacific funds’ investment management Northern Hawaii Mariana Islands Guam Marshall Islands Federated States of Micronesia • Geographically isolated - Palau small population across a large ocean • Small economies dominated Nauru by remittances, tourism, Papa New Kiribati fisheries and agriculture Guinea Tuvalu (subsistence farming) Solomon Islands Tokelau • High import dependency / Christmas Samoa minimal export markets Islands Wallis and Futuna Is. American • Extreme vulnerability to Vanuatu Samoa Cook Islands climate change Fiji Tonga • Populations generally low in New Caledonia Niue capacity Australia Norfolk Island EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 8 practices building on (a) the World Bank engagements with the This comparative analysis covers five national trust funds and funds over the past five years and (b) relevant information pro- a sovereign wealth fund from the following countries: Kiribati, vided by those funds specifically for this analysis. the Marshall Islands, the Federated States of Micronesia, the Republic of Nauru (Nauru), and Tuvalu presented in table This comparative study is meant to inform policy and decision- ES.1. The five funds selected for the study are designed as makers governing the funds on the effect of their investment perpetual funds with a common objective of balancing distri- governance decisions on performance of the funds over the bution of their income for today’s spending while preserving medium to long term. The study builds on World Bank’s own equity across generations. Although seen as heterogenous in extensive practitioner experience with managing US$185 bil- the Pacific community, from the global perspective of publicly lion of its own assets and on behalf of public institutions1 across managed funds, these funds share a common thread in their different investment objectives and mandates. We acknowledge fundamental nature and the similarity of their investment ap- that Pacific countries had the foresight to set up a number of proaches thus providing a sound basis for comparing these funds—provident, pension, and other special purpose funds— funds among one another. We see these funds as distinct that play significant role in the Pacific financial sector and are from provident and pension funds in the Pacific, which face a highly heterogeneous in their setups and investment operations. somewhat different set of challenges that warrant a dedicated This comparative study focuses on the five national funds—four analysis and a separate discussion. trust funds and one sovereign wealth fund—and their investment strategies and performance, and it benchmarks them with rel- Focusing this note solely on these funds, we will refer to the evant practices using their current authorized investment param- five funds throughout this note as “the Pacific funds” and “sov- eters. The analysis does not attempt to assess whether these ereign funds,” encompassing both trust funds and a sovereign parameters are consistent with the set-out mandates for these fund among them. funds nor whether these parameters meet national needs. > > > T A B L E E S . 1 . - Key Features of Pacific Trust and Sovereign Funds Fund name and Source of Total Percentage Base Governing board/ year established revenuea size, mln of GDP currency committee composition Micronesia, Fed. Sts., Micronesia, Fed. Sts. (2 members), Donor contributions US$ 636 167 US$ Compact Trust Fund (2003) United States (3 members) Phosphate, Kiribati RERF (1956) $A 1,171 445 $A Kiribati Cabinet fiscal surpluses Nauru Trust Fund Fiscal surpluses, Australia (1 member), Nauru (1 member), $A 111 75 $A (1968, 2015) donor contributions New Zealand (1 member), Taiwan (1 member) Marshall Islands Compact Marshall Islands (2 members), United States Donor contributions US$ 435 198 US$ Trust Fund (2003) (4 members), Taiwan (1 member) Australia (1 member), New Zealand (1 member), Tuvalu Trust Fund (1987) Donor contributions $A 185 313 $A Tuvalu (1 member) Note: $A = Australian dollar; RERF = Reserve Equalization Revenue Fund (Kiribati); US$ = United States dollar. a. In the case of Pacific trust funds, donors are select national governments that contribute capital to the funds and are represented at the trust fund governing board or committee. Based on the data provided by the funds for this study, figure ES.1 presents 3- and 5-year returns net of fees2 and figure ES.2 their external fund management costs.3 To understand the key factors that contributed to observed differences in Pacific funds investment outcomes over the period, we evaluated the funds’ investment choices along the following parameters. First, we assessed the funds’ specific investment strategy decision against other potential investment strategy choices consistent with their investment authoriza- tions and measured its effect on the funds’ long-term wealth. We further comment on the effect of the funds’ decision on implement- ing their investment policy, notably in relation to the level of active management given to the external fund managers and whether it adds value net of cost to funds over time. We also compare costs of external fund management and discuss financial industry trends. 1 The World Bank, through its Treasury (TRE), manages public assets an asset owner and an asset manager for central banks, sovereign wealth funds, public pension funds, and other public financial institutions. TRE manages a range of portfolios with different investment objectives and mandates across the full spectrum of asset classes, including fixed income, public and private equity, and a wider range of alternative investments. TRE investment operations include both in-house investment management and an external managers program of more than 200 private sector asset managers. TRE brings its own practitioner investment management experience when engaging with public funds in its member countries. 2 It is not clear what costs are excluded by different funds, as asset management costs comprise a number of categories (see next footnote). 3 Total investment management costs include costs of advice, custody, external fund management, internal management and operations, and governance (payments to committee members and travel). As total investment costs are only available from the Marshall Islands and the Federated States of Micronesia funds, we are able to present only the costs of external fund management across the Pacific funds included in this study. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 9 > > > F I G U R E E S . 1 . - Pacific Funds 3- and 5- year Net Returnsa 12 11.3 10 7.7 8.1 7.6 8 Percentage 5.9 6.0 6 4.6 4.9 4 2 0 3 year 5 year Kiribati Micronesia Marshall Islands Tuvalu Source: World Bank calculations based on data provided by funds. a. Tuvalu, Micronesia, and Marshall Islands as of September 30, 2019. Kiribati as of November 30, 2019. Return and other data come from each fund’s respec- tive data or reporting sources, including custodian reports, annual reports, or other investment management reports. The authors are not in the position to ensure the accuracy of the data from these reports. However, our benchmarking against reference portfolio constructed from publicly available market indices over respective investment horizons provides satisfactory validation that the return data are representative of the funds’ investment approaches. > > > F I G U R E E S . 2 . - Pacific External Funds Management Costs, 2019 0.70 0.64 0.60 0.60 0.53 0.50 0.38 Percentage 0.40 0.30 0.20 0.08 0.10 0.00 Tuvalu Nauru Micronesia, Fed. Sts. Marshall Islands Kiribati Source: World Bank calculations based on data provided by funds. As evidenced on figure ES.1, Pacific funds’ investment outcomes have been quite dispersed, and our proposed benchmarking approach helped illuminate the key factors that led to differences in investment outcomes. Direct comparison of the Pacific funds’ investment returns is challenging because of the funds’ different reporting periods and granularity of the available data. However, our approach of using reference portfolios calibrated to specific markets and risk levels allows us to infer the implicit investment benchmarks within each funds’ investment approach and to compare them with the avail- able investment alternatives specific to these funds.4 These implicit investment benchmarks allow us to draw general conclusions about whether the fund’s investment approach is in line with the fund’s investment objective and whether it fully uses its investment authorization within its risk preferences. 4 One reason for the lack of comparative analysis across the Pacific funds has been cited as the funds’ different currencies. We deal with the currencies by selecting ap- propriate reference portfolios tailored to each fund’s circumstance. We also note that the different base currency for these funds is not an issue because the investment portfolios of the considered funds are globally diversified. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 10 Caption: FEMM 2019 | Source: Author 1. Impact of Investment Strategy Decision The most significant factor explaining the differences in invest- and are materially lower than comparable returns of its region- ment outcomes was the investment policies that Pacific funds’ al peers, which have delivered returns in line with balanced or boards pursued to achieve their objectives. The Pacific funds growth investment strategies. In fact, if the TTF board had set followed two different approaches: a traditional strategic asset a simple and easy to implement benchmark, for example, of allocation approach pursued by Kiribati, the Marshall Islands, 50 percent global equities and 50 percent global investment and the Federated States of Micronesia and an objective-based grade fixed income, it would have delivered 10.6 percent per asset allocation (OBAA) strategy in Tuvalu. In the traditional annum returns versus the OBAA strategy’s 6.1 percent per an- strategic asset allocation approach, the boards expressed their num since inception. Thus, our analysis revealed that implicit tolerance for the overall level of risk through an asset allocation benchmark for the OBAA strategy is a highly conservative in- mix that was translated into an investment benchmark. In con- vestment strategy that the TTF board could have pursued by trast, in the OBAA strategy, which the Tuvalu Trust Fund (TTF) implementing a passive portfolio, for example, $A Morningstar implemented in 2012, the decision on asset allocation was del- Conservative Index5 or 80 percent fixed income/20 percent egated from the board to the asset managers with the view that global equity, for a fraction of the cost, as discussed later in this the managers were in a better position to anticipate the market paper. In contrast, the returns of the funds of Kiribati, Marshall movements to deliver superior returns rather than if the board Islands, and Federated States of Micronesia pursued an ex- set the strategic asset allocation mix. plicit investment growth strategy. The implicit benchmarks of these funds were in line with having 50 percent to 80 percent As summarized in tables ES.2, ES.3, and ES.4, OBAA returns growth assets. since inception are comparable with a conservative portfolio 5 Appendix B provides specific details about Morningstar indexes that are constructed for different level of risk tolerance and comprise Australian and global asset classes including cash, fixed income, public equity, and listed and unlisted property. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 11 > > > T A B L E E S . 2 . - TTF’s Investment Returns vs. Alternativesa 3 year 5 year Inception (April 2012) TTF - Tuvalu 4.6% 4.9% 6.1% $A Morningstar — Conservative 4.8% 5.1% 5.8% $A Morningstar — Moderate 6.1% 6.2% 7.0% $A Morningstar — Balanced 8.4% 8.1% 9.1% $A Morningstar — Growth 10.0% 9.2% 10.7% $A Morningstar — Aggressive 12.0% 10.8% 12.4% $A Global Equities 50% / Fixed Income 50% 9.3% 9.0% 10.6% Source: World Bank calculations. Note: TTF = Tuvalu Trust Fund. a. Differences in reporting cycle dates for the Pacific funds and their different base currencies are the reason for differences in returns of reference portfolios across tables ES.2, ES.3, and ES.4. Pacific funds returns are net of fees, whereas reference portfolios returns are based on index returns. Implementation of reference portfolios could range between several basis points for passive index replication strategies to 30 basis points (bps) as assumed by New Zealand Super Fund for its reference portfolio. > > > T A B L E E S . 3 . - RERF’s Investment Returns vs. Alternativesa 3 year 5 year Inception (June 1995) RERF-Kiribati 11.3% 7.6% 6.0% $A Morningstar — Conservative 5.5% 4.8% 6.5% $A Morningstar — Moderate 6.9% 6.0% 6.8% $A Morningstar — Balanced 9.3% 7.9% 7.5% $A Morningstar — Growth 11.0% 9.2% 7.8% $A Morningstar — Aggressive 13.0% 10.8% 8.0% $A Global Equities 50% / Fixed Income 50% 10.0% 8.7% 7.5% Source: World Bank calculations. Note: RERF = Reserve Equalization Revenue Fund. a. Differences in reporting cycle dates for the Pacific funds and their different base currencies are the reason for differences in returns of reference portfolios across tables ES.2, ES.3, and ES.4. Pacific funds returns are net of fees, whereas reference portfolios returns are based on index returns. Implementation of reference portfolios could range between several basis points for passive index replication strategies to 30 basis points (bps) as assumed by New Zealand Super Fund for its reference portfolio. > > > T A B L E E S . 4 . - Federated States of Micronesia and Marshall Islands TF Investment Returns vs. Alternativesa 3 year 5 year FSM Inception (Sept 2004) RMI Inception (June 1995) FSM — Micronesia, Fed. Sts. 7.7% 5.9% 5.5% RMI - Marshall Islands 8.1% 6.0% 6.0% US$ Diversified Risk 20% 4.8% 4.9% 5.3% 5.0% US$ Diversified Risk 40% 6.0% 5.5% 6.0% 5.6% US$ Diversified Risk 60% 7.0% 6.2% 6.6% 6.1% US$ Diversified Risk 80% 8.1% 6.8% 7.2% 6.4% US$ Global Equities 50% / Fixed Income 50% 7.1% 5.8% 6.1% 5.7% Source: World Bank calculations. Note: TF = trust fund; US$ = US dollar. a. Differences in reporting cycle dates for the Pacific funds and their different base currencies are the reason for differences in returns of reference portfolios across tables ES.2, ES.3, and ES.4. Pacific funds returns are net of fees, whereas reference portfolios returns are based on index returns. Implementation of reference portfolios could range between several basis points for passive index replication strategies to 30 basis points (bps) as assumed by New Zealand Super Fund for its reference portfolio. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 12 Caption: Kiribati Parliament. | Source: Author Since its inception, the cumulative effect of OBAA strategy 2. Impact of active management on TTF has been significant compared with other investment alternatives. At $A 185 million, TTF is about 313 percent of decision and management costs Tuvalu’s GDP. Thus, every additional 1 percent in investment returns corresponds to about 3 percent of GDP in additional in- come to Tuvalu per annum. Compared with $A 50 percent /50 Another contributing factor to the returns difference was the percent passive strategy that would have costed only several Pacific funds’ different approaches to implementing their in- basis point to implement,6 TTF underperformed by 4.5 percent vestment strategies. All examined investment managers un- per annum, which is equivalent to about 14 percent of GDP derperformed their benchmarks, and in some cases they un- for any year that OBAA has been in place. Since OBAA imple- derperformed significantly (Tuvalu) or suffered outright default mentation in April 2012, $A 50 percent /50 percent strategy (Kiribati). The level of underperformance was directly related would have delivered about 113 percent of GDP in additional to the level of the investment mandate’s complexity or degree wealth for Tuvalu. This amount is significant national wealth of allowable active risk. More complex or active mandates re- that Tuvalu has not accumulated as a result of its choice of the quire significant in-house skills to implement and monitor on OBAA strategy. an ongoing basis. As a result of these experiences, and in line with trends globally, the Pacific funds in our study, with With that in mind, over the past several years the TTF board the exception of Tuvalu, have either moved entirely to passive has been engaging different partners and institutions to as- mandates (Kiribati) or have been reducing the overall level of sess the appropriateness of its current investment approach active risk for the total portfolios (Marshall Islands and Fed- and potential alternatives. As a part of these efforts early in erated States of Micronesia). The level of complexity of the 2020, TTF undertook a competitive selection process for its investment mandates is also the main factor explaining the dif- investment monitor that has a critical role in investment over- ferences in investment costs. For Marshall Islands, Federated sight: the selection continued despite the pandemic, and TTF States of Micronesia, Nauru, and Tuvalu, their higher fees are selected a global firm in June 2020. These steps are important reflective of their investment in alternative asset classes or in improving TTF investment governance, and this compara- higher complexity mandates, whereas Kiribati’s low costs re- tive analysis could help TTF’s design of an appropriate invest- flects fully passive mandates. Furthermore, Kiribati was able ment approach going forward. to further reduce its managers’ fees through competitive pro- cess RFP. This is a standard market practice globally, which would bring value to Pacific funds in general. 6 “One basis point” is 1 percent of 1 percent—that is, 1bp = 0.01%. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 13 Caption: Market in Vanuatu. | Source: IFC K E Y TA K E A W AY S Key takeaways for the Pacific funds from this benchmarking study are consistent with best practices for long-term investors: • To ensure long-term sustainability of Pacific funds, the investment strategy should reflect the fund’s investment purpose, which should have in place a well-defined investment governance framework to ensure the strategy is formalized, implemented, and monitored in line with global best practices. • Investment benchmark, representing replicable strategic asset allocation based on the investment policy, typically accounts for 80 percent to 90 percent of the returns and risk of the portfolio. • More complex investment approaches and mandates are costly and require more sophisticated governance and ongoing efforts to oversee and manage these mandates. Boards should ensure that they have the required time, inclination, and knowledge to oversee such mandates and that these ap- proaches add value net of cost to the fund over time • OUTSOURCING IS NO SUBSTITUTE FOR ACCOUNTABILITY: boards have a fiduciary responsibility to the fund’s beneficiaries and are ultimately responsible for investment decisions, including those that are outsourced. As such, they need to ensure ownership of the risks that are being delegated and en- sure that robust processes are followed to select and monitor service providers in line with the fund’s needs. • HIGH-QUALITY GOVERNANCE of the investment process is necessary for the long-term success of in- vestment funds. Statutory governance—that is, clearly defined rules and investment parameters—should be well articulated to provide clarify and accountability to manage the funds. Operational governance— quality of day-to-day decision making and exercising fiduciary responsibility—will directly affect the financial results of the funds. The frameworks should be reviewed periodically by impartial and independent parties to ensure that investment parameters evolve with the evolving market and policy contexts. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 14 1. Caption: Workday in the Pacific | Source: Climate Investment Funds >>> Note Objective At the 2019 Finance and Economic Ministers Meeting (FEMM), policy makers called for im- proved transparency of the Pacific trust funds’ management practices and investment results as a way to foster their improved management going forward. FEMM adopted the resolution that “Forum recognizes the need to undertake regional analyses on the comparison of the governance, portfolio management and returns of national trust funds” and tasked the World Bank to undertake the study (FEMM 2019). FEMM’s is disclosing the results of the comparative analysis of Pacific funds investment to further stimulate collaborative discussion on how to strengthen the Pacific funds’ management and how to inform the design and implementation of necessary reforms. The first time FEMM considered the management of the Pacific trust and sovereign funds was in 2014 with the World Bank presenting a comparative analysis of the funds with high-level observations of the funds’ governance, investment policy, and portfolio construction against the backdrop of increasing focus on improving the management of national funds. The Asia Pacific Association for Fiduciary Studies (APAFS), established in 2000, has been promoting good gov- ernance, trainings, and sharing of information for its members.7 More recently, a regional coali- tion of provident, superannuation, trust, and sovereign funds—Pacific Islands Investment Forum (PIIF)8—has been set up to improve collaboration and investment practices and to increase cross-border regional coinvestment. Despite this regional progress, publicly available informa- tion about the management of the five Pacific funds was limited as was a comparable analysis of their practices, even though these funds had been managed for several decades and Kiribati’s Revenue Equalization Revenue Fund (RERF), established in 1956, was one of the longest man- aged sovereign funds globally. Whereas significant progress has been made in establishing international best practices for sovereign or large public funds, such as the Santiago Principles,9 the Pacific funds are unique in making direct application of international practices challenging. With input from the World Bank and others, in 2015 the New Zealand government sponsored New Zealand Institute for Pacific Research (NZIPR) to develop best practices tailored specifically to the Pacific funds. The initiative produced a three-part study that (a) examined the role of the sovereign funds in the Pacific Islands; (b) developed an assessment framework and reference portfolios to evaluate the 7 Members of APAFS are found at https://www.apafs.org/default.asp?secID=115. 8 See The Pacific Islands Investment Forum, https://www.pacificinvestmentforum.com/. 9 The Santiago Principles for Sovereign Funds (IWG 2008) outline a set of high-level principles around the legal establishment of a fund and its objectives; its governance structure; investment, and risk management; and a fund’s coordination with broader macroeconomic policies. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 16 management of Pacific funds; and (c) applied the assessment experience with managing US$185 billion of its own assets framework and tools to assess the Tuvalu Trust Fund (TTF), and assets of other public institutions10 across different invest- the Niue International Trust Fund, and the Tokelau Interna- ment objectives and mandates. tional Trust Fund (Drew and Frijns 2017). This comparative study is designed to inform policy and deci- These studies and recommendations have significantly ad- sion-makers governing Pacific funds about how crucial their vanced the understanding of the regional sovereign and trust investment strategy decisions are for the performance of the funds and proposed a framework for their evaluation, serving funds over the medium to long term. Whereas we also provide as a strong foundation for comparative analysis of investment insights from relevant practical experiences and observations practices across the Pacific funds. In response to the FEMM’s that are relevant to the Pacific funds, this study does not at- request, this note addresses the current vacuum of compa- tempt to diagnose investment issues, nor does it provide ad- rable information about Pacific funds’ management practices vice on appropriate reforms to address those issues. Further, building on the NZIPR framework and on the World Bank en- this analysis focuses on investment strategies and implemen- gagements with these funds since 2013. The comparative tation consistent with the respective funds’ current designs, analysis brings together analysis of relevant data provided by authorizing environment, and investment mandates rather those funds specifically for this analysis, publicly available in- than attempting to assess whether these design parameters formation, and the World Bank’s own extensive practitioner are meeting the national needs. 10 The World Bank, through its Treasury (TRE), manages public assets both an asset owner and an asset manager for central banks, sovereign wealth funds, public pen- sion funds, and other public financial institutions. TRE manages a range of portfolios with different investment objectives and mandates across the full spectrum of asset classes, including fixed income, public and private equity, and a wider range of alternative investments. TRE investment operations include both in-house investment management and an external managers program of more than 200 private sector asset managers. TRE brings its own practitioner investment management experience when engaging with public funds in its member countries. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 17 2. Caption: Tuvalu resident. | Source: IFC >>> Introduction and Pacific Funds Context Our comparative analysis covers a select set of national trust funds and a sovereign fund pre- sented in table 2.1 from the following countries: Kiribati, the Marshall Islands, the Federated States of Micronesia, Nauru, and Tuvalu. Throughout the note we will refer to these funds col- lectively as “the Pacific funds” and “sovereign funds,” encompassing both trust funds and a sovereign fund among them. The Pacific countries had the foresight to set up a number of funds—provident, pension, and other special purpose funds—that play a significant role in the Pacific financial sector and that are highly heterogeneous in their setups and investment operations. We are limiting our comparative study to the five national funds—trust funds and a sovereign wealth fund— for the following reasons. First, over $A 3 billion is currently invested in these funds, which is equivalent to 230 percent of these countries’ combined GDP, making them collectively a significant source of public capital for these countries. This is similar to the relative size of Norway’s sovereign funds (US$1,063 billion), which is about 255 percent of its GDP. In contrast, sovereign funds of Australia (US$161 billion) and New Zealand (US$23 billion) are just over 10 percent of these countries’ economies. Except for the Nauru trust fund, which was recently reestablished following its depletion in the 1980s, the four other funds range between 75 percent to nearly five times of these countries respective GDPs. Kiribati’s RERF is the largest in absolute size with the current balance of about $A 1.2 billion. Most of the other funds are about $A 100 million to nearly $A 1 billion. > > > T A B L E 2 . 1 . - Key Features of Pacific Trust and Sovereign Funds Fund name and Source Total Percentage Base Board year established of revenue size, mln. of GDP currency composition Micronesia, Fed. Sts., Compact Donor Micronesia, Fed. Sts. (2 members), US$636 167 US$ Trust Fund (2003) contributions United States (3 members) Phosphate, Kiribati RERF (1956) $A 1,171 445 $A Kiribati Cabinet fiscal surpluses Australia (1 member), Nauru Nauru Trust Fund Fiscal surpluses, donor $A 111 75 $A (1 member), New Zealand, (1 member), (1968, 2015) contributions Taiwan (1 member) Marshall Islands Compact Marshall Islands (2 members), United Donor contributions US$ 435 198 US$ Trust Fund (2003) States (4 members), Taiwan (1 member) Sources: Fund size taken from annual reports or other funds’ reporting provided to the World Bank team. GDP figures from IMF, 2019. Note: RERF = Reserve Equalization Revenue Fund. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 19 Second, although seen as heterogenous in the Pacific com- vestment results over the medium to long term, while accept- munity, from the perspective of publicly managed funds glob- ing that unique features of individual funds would contribute ally, these Pacific funds are quite homogeneous despite some to some variations in investment results over the short term. differences in their institutional set-up. From the perspective Thus, we compare investment outcomes of Pacific funds over of international sovereign wealth funds, their governance is the recent period. similar because their boards are composed of government of- ficials with high reliance on external advisers. This arrange- To understand the key factors that contribute to observed dif- ment is strikingly different from international best practices ferences in Pacific funds investment outcomes over the period, where board members appointed typically have relevant ex- we evaluate the funds’ investment choices along the following perience and are at arm’s length from the government. In dimensions. First, we propose a methodology to assess the contrast, in these Pacific funds most people in governance funds’ investment strategy decisions against clearly defined roles have little if any industry experience, and many have and implementable alternatives. We further comment on the had no formal training in investment governance. In particu- effect of the funds’ decisions in implementing their investment lar, the board of TTF consist of Tuvalu’s finance minister and policy, notably in relation to the level of active management representatives from Australian and New Zealand aid agen- given to the external fund managers and whether that man- cies, supported by the TTF secretariat housed in Tuvalu’s fi- agement adds value net of cost to funds over time. We also nance ministry. Similarly, the boards of the Marshall Islands compare the funds’ costs of external fund management and and Federated States of Micronesia are composed of do- discuss financial industry trends in that regard. nor countries and national representatives, supported by an executive administrator in Washington, D.C., who also runs The rest of the paper is structured as follows. “Note on the Use day-to-day operations for both funds. In contrast, the govern- of Individual Fund Data in Comparative Analysis” presents the ment of Kiribati, as both the trustee and a beneficiary, has data that were used for the comparative study and the team’s complete control over the RERF. The Kiribati Cabinet, chaired treatment of the data to ensure comparability across the funds. by the president, is the highest-level governing body respon- The section also addresses the limitations of the existing data- sible for key decisions and is supported by a small team in the set and its effect on the results of the analysis. “Methodology Finance and Economic Development Ministry that also has for Comparative Analysis” (a) presents the case for focusing other responsibilities. the comparative analysis on the funds’ selection of their invest- ment benchmarks, (b) outlines the overall methodology used Despite these differences in governance setups, these Pacific for the study, and (c) addresses the effect of the recent market funds share commonality in their fundamental nature11 and volatility owing to COVID-19 on the interpretation of the study. similarity in investment approaches that provide a sound ba- “Benchmarking of Pacific Funds” presents the peer compari- sis for comparing them with one another. The five funds se- son of the five funds’ investment outcomes and benchmarking lected for the study are designed as perpetual funds with a of each of the fund’s investment outcomes against relevant common objective of balancing distribution of their income for investment alternatives. The section also shares insights today’s spending while preserving equity across generations. from the evolution of the funds’ investment practices based The geographical remoteness of Pacific islands from financial on provided documents and on the World Bank engagement centers has further affected their national funds’ access to fi- with these funds since 2013. Finally, the section provides an nancial markets and quality of provided services, whereas low overview of the funds’ investment management costs and dis- technical capacity makes these funds dependent on external cusses trends in the investment management industry. The expertise and specialized services providers. final section concludes with key takeaways from this study for the Pacific funds consistent with best practices for long-term Finally, given their common objectives, it’s not surprising that investors globally. these funds follow similar investment approaches. In particu- lar, to grow their capital base and to fund their national devel- The following four sections are tailored to policy and decision- opment needs, 100 percent of the funds’ capital is invested makers at the highest level of governance of the Pacific funds abroad in international markets across the full risk spectrum responsible for determining the funds’ overall investment of financial assets. strategies and level of risk. Appendixes A through E provide extensive technical details and the underlying analyses that Consequently, this study focuses on comparative analysis of informed the paper’s findings that would interest investment the main investment decisions that drive the Pacific funds in- practitioners and technical advisers. 11 There are some variations in the design of trust funds in the Pacific. For example, Nauru’s Compact Trust Fund is designed as a sinking fund. Nauru’s fund predates the Compact Trust Funds of the Federated States of Micronesia and the Marshall Islands that are designed as perpetual funds. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 20 3. Caption: Marshall Islands, Coastline aerial | Source: https://earthview.withgoogle.com >>> Note on the Use of Individual Fund Data in Comparative Analysis The World Bank team requested relevant data from each of the funds to be used in this com- parative analysis. Monthly return data and granular breakdown of asset management costs were provided by the funds’ administrators of the Federated States of Micronesia and the Marshall Is- lands. Tuvalu and Kiribati shared investment reports from their custodian or service providers that, among other data, provide aggregated 1-year, 3-year, 5-year rolling returns over their respective fiscal years (which are different), but not monthly returns. The team also got the costs of external asset management for these funds, which is a small component of total asset management costs that were provided by the Marshall Islands and the Federated States of Micronesia. Nauru referred the team to the fund’s website with its 2019 annual report and publicly available information, which is significantly more limited than information that was made available by other funds. We are not in the position to ensure the accuracy of the data from these reports. However, our benchmarking against reference portfolios constructed from publicly available market indices over respective investment horizons provides satisfactory validation that the return data are representative of the funds’ investment approaches. Table 3.1 provides details about each fund’s data availability, inception, frequency, and sources. FREQUENCY OF AVAILABLE DATA. As the frequency of the provided data and the level of aggregation are different across the funds in this study, a granular analysis with a high degree of accuracy is not possible at this stage. To provide a comparative analysis with the current data limitations and without the benefit of monthly granular data, the team—to the extent possible— aligned the return data for different funds across compared periods and validated results of the analyses through publicly available market indices as appropriate. DIFFERENT REPORTING CYCLES. Another data challenge with the current data given its an- nual frequency is different reporting cycles for the funds. To address this issue for the aggregate 1-year, 3-year, 5-year returns and returns since inception that are linked to the funds’ respective reporting dates, we aligned returns of reference portfolios with each of the respective funds’ reporting periods. The difference in reporting cycle dates is one of factors contributing to differ- ences in returns of reference portfolios in the respective funds Investment Returns vs. Alterna- tive tables presented in the Executive Summary and in Section V. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 22 > > > T A B L E 3 . 1 . - Funds Data Details Financial Performance Financial Performance Data Fund Name Fees Data Source Fees Data Description Data Sources Description Detailed breakdown of Annual Returns from FY06 to FY19. administrative expenses, 1-year, 3-year, 5-year, and since Federated States of broken down by the inception returns for the fund, asset Micronesia Compact Annual Report FY19 Annual Report FY19 subcomponents of Investment classes. Fund reporting inception Trust Fund Expenses and Administrative September 30, 2004, data as of Expenses on a fiscal year September 30, 2019 basis, since inception (FY04) 1-year, 3-year, 5-year, 10-year, and Manager fees from RFP, Performance Report from since inception returns for the fund, Investment expense data weighted by assets. No data Kiribati RERF fund custodian, as of asset classes. Fund reporting inception from Request for Proposal available on administrative November 30, 2019 June 1, 1995, data as of November (RFP). costs. 30, 2019 Returns data available for 1 year, and Calculated as fees paid as a Nauru Trust Fund Annual Report FY19 since inception (April 2016). Data as of Annual Report FY 2019 percentage of fund size. June 30, 2019 Detailed breakdown of Annual Returns from FY04 to FY19. administrative expenses, 1-year, 3-year, 5-year, and since broken down by the Republic of the Marshall inception returns for the fund, asset Annual Report FY19 Annual Report FY 2019 subcomponents of Investment Islands Trust Fund classes. Fund reporting inception Expenses and Administrative September 31, 2005, data as of Expenses on a fiscal year September 30, 2019 basis, since inception (FY05) 1-year, 3-year, 5-year, and since “Tuvalu Trust Fund - inception returns for the fund, and Investment Review” by Calculated as fees paid as a Tuvalu Trust Fund Quarterly Report 3Q19 its two managers. Fund reporting Vinstar, November 2011, percentage of fund size. inception April 1, 2012, data as of and information on asset September 30, 2019 managers websites DIFFERENT BASE CURRENCIES. The Pacific funds includ- recommendations for how to improve the management of indi- ed in the analysis include both US$-based funds (Marshall vidual funds. As monthly return data for each of the participat- Islands and Federated States of Micronesia) and $A-based ing funds becomes available, the team stands ready to extend funds (Kiribati, Nauru, Tuvalu). When benchmarking funds this comparative analysis on a risk- and cost-adjusted basis. against publicly available market indices, we compared US$- based funds with respective indices in US$ currency and $A- TREATMENT OF RETURNS DATA NET OF COSTS. The Pa- based funds with $A indices or returns hedged into $A. The cific funds returns data provided to the team are net of fees, currency difference is another contributing factor to the refer- although it’s not clear what costs are excluded by different ence portfolios’ returns being different from the Pacific funds in funds as the total asset management costs include investment the tables in the Executive Summary and Section V. advice, custody, external fund management, internal manage- ment and operation, and governance (payments to commit- The use of the lower frequency aggregated data over slightly tee members, travel, other relevant expenses). In constructing different time periods introduces some level of distortion when reference portfolios, we use the index return data and assume comparing funds’ investment returns. However, the team is these returns to be before fees. Thus, readers need to assume confident that proposed framework is robust enough to as- some costs for implementing the reference portfolios. By way sess fundamental differences driving the funds’ long-term val- of guidance, the New Zealand Superannuation Fund (NZ Su- ue because of the high-level investment benchmark decisions per Fund) estimates implementation of its reference portfolio of the Pacific funds over the medium- to long-term horizons benchmark is around 29 basis points (NZ Super Fund 2020). (three years and more). Using the monthly data would have For the passive $A 50 percent /50 percent portfolios that could allowed the team to provide risk-adjusted analyses and further be implemented through passive index funds, implementation insights into investment management practices of each fund. cost including asset management and custody would amount Still, using monthly data for comparative analysis would not to several basis points. fundamentally change the study’s conclusions, nor specific EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 23 4. Caption: Road in Kiribati. | Source: Author >>> Methodology for Comparative Analysis 1. Role of Investment Benchmarks It has been widely documented that selection of an appropriate benchmark is the most crucial investment decision for a fund or a portfolio that will determine its future performance (Scott et al. 2017). In selecting and following a benchmark, a fund basically defines its strategic asset al- location, the level of its expected future return and the degree of uncertainty around investment outcomes—that is, the level of risk. There are variations in internal processes across investment funds globally to come up with an appropriate investment benchmark, as well as in the use of terminology depending on the purpose of a fund, its size and investment philosophy. Some funds define their investment benchmarks through the process of strategic asset allocation, others through the construction of a reference portfolios, while others define specific invest- ment benchmark within their investment policy objective statements. While there are nuances to each of these approaches, their underlying objective is to come up with an implementable portfolio to guide the funds’ investment management operations. In this note we will use the term “benchmark” to represent a benchmark portfolio against which performance of a fund could be assessed, regardless of the process through which it was established. While there is a growing supply of complex investment products promising to outperform the market, the choice of benchmarks continues to explain 80 percent to 94 percent of their return variations across global markets and all types of portfolios, as presented in the table 4.2. As a result, the investment benchmark is the most crucial decision particularly for long-term funds because of the cumulative impact of every additional percentage point of investment returns over its investment horizon. As illustrated on figure 4.1, cumulative wealth of a hypothetical US$100 invested with different rate of returns have remarkably different outcome at the end of the 30-year investment period. For example, investment return of 5 percent leads to nearly twice the size of the portfolio (US$432) as opposed to 3 percent return (US$243), whereas 7 percent returns to US$761. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 25 > > > F I G U R E 4 . 1 . - Growth of $100 Over 30-year Horizon $800 $761 $700 $600 $500 $432 Growth of $100 $400 $300 $242 $200 $135 $100 $- 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Year 1% Growth 3% Growth 5% Growth 7% Growth Source: World Bank calculations. > > > F I G U R E 4 . 1 . - Growth of $100 Over 30-year Horizon Brinson, Hood, and Beebower 1986 United States Canada United Kingdom Australia Japan Number of balanced funds 91 709 303 743 580 406 in each market sampe Median percentage of actual-return 91.1% 86.0% 80.5% 89.1% 87.9% 93.6% variation explained by policy return Sources: Vanguard calculations, using data from Morningstar, Inc.; Scott et al., 2017. Note: For each fund in our sample, a calculated adjusted R2 represented the percentage of actual-return variation explained by policy-return variation explained by policy-return variation. Percentages shown in the figure—91.1% for the United States, 86.0% for Canada, 80.5% for the United Kingdom, 89.1% for Australia, and 87.9% for Japan—represent the median observation from the distribution of percentage of return variatin explained by asste allocation for balanced funds. For the period January 1990—September 2015, the sample included the following: for the United States, 709 balanced funds; for Canada, 303; for the United Kingdom, 743; for Australia, 580; and for Japan, 406. Calculations were based on monthly net returns, and policy allocations were derived from a fund’s actual performance compared with a benchmark using returns-based style analysis (as developed by William F. Sharpe) on a 36-month rolling basis. Funds were selected from Morningstar’s Multi-Sector Balanced category. Only funds with at least 48 months of return history were considered in the analysis. The policy portfolio was assumed to have a U.S. expense ratio of 1.5 basis points (bps) per month (18 bps annually, or 0.18%) and non-U.S. expense ratio of 2.0 bps per month (24 bps annually, 0r 0.24%). Thus, the focus of our comparative analysis is on the Pacific funds’ decisions with respect to their investment strategies, expressed through their investment benchmarks. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 26 Caption: School kids in Tuvalu | Source: IFC 2. Defining Reference Portfolios for Funds with Different Risk Tolerance Parameters To compare investment outcomes across the Pacific Funds would require more sophisticated governance and technical we build on the reference portfolio approach recommended expertise to implement and manage on an ongoing basis. by Drew and Frijns (2017). We construct several reference These reference portfolios also allow us to scale the level of portfolios with different risk profiles representing alternative in- market risk from low exposure to risk assets (say, 20 percent vestment choices for the funds against which we benchmark of the total portfolio) to high risk (100 percent of the portfolio), their decision. The key advantage of the reference portfolio as such representing a broader range of investment alterna- approach is that it incorporates best practice features and can tives for governing boards to express their differing fund pur- be implemented easily and cost effectively (Drew and Frijns poses and necessary risk levels. 2017). Appendix A provides details of how we construct ref- erence portfolios for benchmarking both US$-based and $A- Figures 4.2 and 4.3 present historical returns versus two based Pacific funds, their asset composition, historic risk and risk measures for these reference portfolios over the period return characteristics and other relevant parameters. 1995–2020. Not surprising, portfolios with a higher level of risk generated a higher level of returns over the period. Fur- We used two different types of reference portfolios that are ther, reference portfolios with a broader range of asset classes relevant for the Pacific funds: (a) a simplified and cost effec- benefited from diversification when compared with the passive tive to implement option and (b) a more diversified range of 50 percent/50 percent portfolio with higher returns for a given options for different risk tolerance levels that would require level of risk across both risk measures. Presented risk and more technical capacity to design and implement. returns data incorporate the effects of COVID 19 on the global markets up to April 30, 2020. The first and most simple reference portfolio is constructed as a passive 50 percent global equity/50 percent global fixed The key takeaway from these figures is that while diversified income exposure, which allows for a relatively high level of portfolios provide somewhat better risk adjusted returns, in global diversification at very low cost. This simple structure downside protection and in volatility of returns over an invest- also does not require a significant level of technical expertise ment horizon, the main decision that an investor makes that to maintain and monitor. drives its investment results is the allocation to growth (risky) versus income (or defensive) assets. The second set of reference portfolios are constructed using a broader mix of asset classes to increase diversification but EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 27 > > > F I G U R E 4 . 2 . - Return vs. Risk (as Volatility of Returns) of Asset Classes and Reference Portfolios FI - EM Sovereigns Average annualized return, percentage 8% Diversified US Small-Cap Equities Diversified Diversified Risk 100% Diversified Risk 80% Risk 60% Public Real Estate Diversified Risk 40% Risk 20% 6% Global Equities Passive 50% FI/50% Equity FI - Global Agg 4% 0% 5% 10% 15% 20% Average annualized risk, percentage 50% FI/50% Equity Income Assets Reference Portfolios Risk Assets Source: World Bank calculations. > > > F I G U R E 4 . 3 . - Returns vs. Downside Risk of Asset Classes and Reference Portfolios FI - EM Sovereigns Average annualized return, percentage 8% US Small-Cap Equities Public Real Estate Diversified Risk 60% Diversified Risk 100% Diversified Risk 40% Diversified Risk 80% Diversified Risk 20% 6% Global Equities Passive 50% FI/50% Equity FI - Global Agg 4% 0% 20% 40% 60% Largest 12-month loss, percentage 50% FI/50% Equity Income Assets Reference Portfolios Risk Assets Source: World Bank calculations. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 28 3. Impact of COVID-19 on Market Volatility and Returns As this note is being prepared based on the analysis conduct- As figures 4.4 and 4.5 demonstrate, the immediate market ed during April 2020 through June 2020, global markets have drop owing to the pandemic has been dramatic and presents been volatile owing to the ongoing global pandemic. At the on- significant market shock and volatility challenges over the im- set of the pandemic in March 2020, the markets experienced mediate term; yet as of June 2020 measured over the medium a significant drop with equity markets losing nearly 30 percent to longer term, the pandemic’s effect is comparable with other over the short period. market events that long-term investors have dealt with in the past. As of June 15, 2020, the US equity market has been fluc- The Pacific funds’ risk and return numbers provided to the tuating around its prepandemic levels, and equity markets in team for the comparative analysis as of late 2019 preceded other regions have recovered from their lows as presented in the market drop during the pandemic. The team would be figure 4.4. Figure 4.5 presents the three-year rolling returns in- ready to update this comparative analysis when investment corporating current market volatility up to April 30, 2020, when performance numbers for individual funds incorporating the the market was near its lows, and shows that even for port- COVID-19 period become available. In the meanwhile, we folios with 80 percent risk allocation the returns are positive. place our comparative analysis in the current market context Over 10-year horizons, the portfolios with the highest level of and address the potential effects of the current market drop on risk still demonstrate the highest level of returns. Pacific funds’ investment returns. > > > F I G U R E 4 . 4 . - Movement of Equity Market (January 2020 to June 2020) Equity Market Performance, 100 Indexed to 100 90 80 70 Jan Feb Mar Apr May Jun S&P 500 Nikkei Global Equities DAX Hang Seng FTSE 100 Source: World Bank calculations. > > > F I G U R E 4 . 5 . - Effects of COVID-19 on Rolling 3-year Return for Reference Portfolios (April 2018 to April 2020) Equity Market Performance, 10% Indexed to 100 7.5% 5.0% 2.5% 2018—07 2019—01 2019—07 2020—01 Passive 50%/50% Diversified Risk 20% Diversified Risk 40% Diversified Risk 60% Diversified Risk 80% Source: World Bank calculations. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 29 5. Caption: Kiribati Copra Mill | Source: Author >>> Benchmarking of Pacific Funds 1. Pacific Funds’ Peer Comparison In this section we provide comparative statistics for the funds’ investment returns over medium to longer time periods and explain the variation across returns. As presented in figure 5.1, Pacific funds demonstrated quite different returns with two $A-based funds, Kiribati and Tuvalu, exhibiting the most contrasting outcomes over the considered period of five years, but particularly over the more recent periods. The two US$-based funds of Marshall Islands and Federated States of Micronesia displayed quite similar returns over each of those horizons, illustrating similarity of the overall investment strategy, whereas small deviations illus- trate variations in the funds’ individual fund portfolio implementation. > > > F I G U R E 5 . 1 . - Pacific Funds Peer Comparison: 3- and 5-year Net Returns 12 11.3 10 7.7 8.1 7.6 8 Percentage 5.9 6.0 6 4.6 4.9 4 2 0 3 year 5 year Kiribati Micronesia Marshall Islands Tuvalu Source: World Bank calculations based on data provided by funds. Note: The Tuvalu, Federated States of Micronesia, and Marshall Islands funds are as of September 30, 2019. The Kiri- bati fund is as of November 30, 2019. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 31 When comparing the Pacific funds’ allocation across different university endowments, the Marshall Islands, and the Feder- asset classes (table 5.1), it’s not surprising that different funds’ ated States of Micronesia; or an alternative OBAA strategy returns differ given their different exposures to risky assets.12 pursued by Tuvalu. These numbers are illustrative to provide a Funds based on US$—the Marshall Islands and the Feder- static insight into these funds’ current asset allocation, as they ated States of Micronesia—have similar risk levels to the large are not sufficient to explain significant difference in returns dis- regional sovereign wealth funds, such as the Australian Gov- played on figure 5.1. ernment Future Fund ($A) and New Zealand’s Super Fund ($NZ), and similar to that of US university endowments, espe- As some of these funds have been considering whether to cially those of comparable size to the Marshall Islands and the adopt an endowment model, in appendix E we provide rel- Federated States of Micronesia funds. The difference in out- evant information about US universities and colleagues en- comes for Kiribati versus Tuvalu is quite striking considering dowments’ investment approaches and detailed results. As a that these funds have similar exposure to risky assets (about point of comparison, US university and college endowments a half of the portfolio) and, as $A-based funds, manage across of similar size to these Pacific funds (US$250 million to US$1 similar investable universe. Table 5.1 represents the current or billion), which are quite similar in their purpose and investment recent asset composition, whether as a strategic asset alloca- approach, have generated an average return of 8.9 percent tion (SAA), such as in Australia and Kiribati; reference portfolio over a three-year period and around 5.2 percent over the five- as in New Zealand; actual portfolio compositions, such as in year period.13 > > > T A B L E 5 . 1 . - Strategic Asset Allocation for Pacific Funds and Comparable Peers University Small Micronesia, Marshall Kiribati Nauru Tuvalu Australia New Zealand endowments, university Fed. Sts. Islands >US$1 billion endowments Fixed income 18% 50% 18% 25% 20% 20% 10% 22% Equities 58% 50% 60% 66% 35% 80% 32% 56% Private markets 24% 22% 9% 45% 58% 22% OBAA 100% Income assets 18% 50% 15% 25% 20% 20% 10% 22% Risk assets 82% 50% 85% 75% 80% 80% 90% 78% Sources: Trust Fund for the People of the Republic of the Marshall Islands, 2020; Trust Fund for the People of the Federated States of Micronesia, 2020; Kiribati from November 30, 2019 Custodian Report; Tuvalu 3Q19 Investment Report, with “Defensive” allocated to fixed income, “Growth” allocated to equities, and “Di- versified” allocated to Private Markets; Australian Government Future Fund, 2019—cash added to fixed income and figures rounded; New Zealand from the NZ Super Fund, 2014; Nauru from the 2019 Annual Report, with the Mercer Growth fund allocated among the neutral weights as listed on the Mercer Growth fund’s product disclosure statement; university endowment data from the NACUBO-TIAA Study of Endowments 2019. Note: OBAA = objective-based asset allocation. Comparing returns of these funds against reference portfolios within their respective currencies ($A for Kiribati and Tuvalu and US$ for Marshall Islands and Federated States of Micronesia) over different investment horizons reveals different levels of total risk across the Pacific funds, which explains variations in their returns. In the next section, we examine the Pacific funds’ individual investment choices and the effects of those decisions on investment outcomes over the period. 12 Definition of risky assets for the purposed of this study are defined in appendix A. There could be differences in qualification of respective asset classes into risk versus other assets leading to potentially different percentages across income and risk assets. 13 We provide results of US universities and college endowments owing to their high level of transparency and availability of the periodic analysis of 802 institutions manag- ing endowments ranging from US$25 million assets under management to more than US$1 billion, making it the largest and most comprehensive study of its kind. As there are some regional biases in investment approaches across asset owners, we suggest that the Universe of Australasian midsize foundations could be a better comparator for the Pacific funds. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 32 Caption: Tuvalu Delegation to Tokyo Olympics 2021 | Source: Tuvalu.TV 2. Benchmarking Tuvalu’s Investment Approach: OBAA Strategy and Its Effects The Tuvalu Trust Fund was established in 1987 to provide As set out in part I, article 2 of the trust fund agreement, TTF a source of recurrent revenue to the government of Tuvalu, has the objectives to (a) assist the government to achieve which had extremely limited alternative sources of revenue at greater financial autonomy in the management of its recurrent its disposal when it achieved independence in 1978. Tuvalu budget; (b) enable the government to maintain, and if pos- started developing a case for a trust fund with its donor part- sible, improve existing levels of social infrastructure and ser- ners in 1982. Following negotiations, the International Trust vices; (c) enhance the capacity of the government to receive Fund Agreement was signed June 16, 1987, by Australia, New and effectively use external capital development and techni- Zealand, Tuvalu, and the United Kingdom as the original par- cal assistance; (d) enable the government to meet long-term ties. The fund itself was invested on August 21, 1987, with an maintenance and operating costs of social and economic in- initial value of $A 27.1 million of which $A 1.6 million was con- frastructure and services; and (e) assist the government to tributed by Tuvalu, $A 8 million by Australia, NZ$ 8.3 million by develop the economy of Tuvalu (TTF 2019). New Zealand, $A 8.5 million by United Kingdom, $A 0.7 mil- lion by Japan, and $A 31,000 by the Republic of Korea. Since As presented in table 5.2, since 2012 TTF returns are compa- inception, the fund has been receiving additional contributions rable with the most conservative of reference portfolios. The from the parties to the fund (TTF 2019). defining decision that affected TTF returns over this period was the implementation of the objective-based asset alloca- tion (OBAA) strategy in 2012. > > > T A B L E 5 . 2 . - Tuvalu’s Returns vs. Reference Portfolios 3 year 5 year Inception (April 2012) TTF 4.6% 4.9% 6.1% US$ Diversified Risk 20% 5.7% 6.6% 7.5% US$ Diversified Risk 40% 7.7% 8.3% 9.5% US$ Diversified Risk 60% 9.8% 9.9% 11.5% US$ Diversified Risk 80% 11.8% 11.5% 13.4% Global Equities 50% 9.3% 9.0% 10.6% Source: World Bank calculations. Note: TTF = Tuvalu Trust Fund. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 33 The foundation of the OBAA strategy is that instead of the was an incumbent managing an Australian equity mandate, TTF board deciding on strategic asset allocation mix consis- and the second was given a mandate without a competitive tent with their view of the TTF risk tolerance and investment process. Since the OBAA strategy was put in place, several objective, the board delegates the asset allocation decision reviews of TTF have occurred, including a desk review by to investment managers. The asset managers are given sig- the World Bank’s Reserve Advisory Management Program nificant leeway in changing allocation across different market (RAMP) team in 2016, that found that OBAA strategy may not sectors. In embracing the OBAA strategy, the TTF board ad- be an appropriate strategy for TTF because it was not aligned opted the view that asset managers are in a better position with the funds objective to deliver real returns over a long-term to deliver required returns by actively managing the portfolio, investment horizon. The review by the New Zealand Institute rather than by setting the strategic asset allocation mix for the for Pacific Research (NZIPR) found that “there has been an fund itself. Implementation of the OBAA strategy resulted in unusually high level of delegation down of investment gover- TTF restructuring its governance structure, investment man- nance responsibilities (e.g. asset allocation, benchmarks, per- agement, and portfolio monitoring. The terms of reference mitted asset classes, etc.) by the Board to Fund Managers, of relevant governance and technical bodies, including the the Monitoring Consultant and Investment Committee repre- board, the advisory committee, investment committee, invest- sentatives” (Drew 2017). ment monitor, and the secretariat were revised to reflect the shift of decision making from the governing structures to the The two managers’ publicly available information on this strat- implementing structures, that is, from the public sector fidu- egy describes their investment objectives as to “increase ex- ciary to the private sector agents. posure to rising assets and reduce risk by avoiding or selling unattractive assets”14 with the expectation that this strategy Before the 2012 restructuring, TTF followed a more traditional will capture the upside of the market (that is, will deliver high asset allocation approach in which the TTF board was respon- returns when markets rally) and will protect the portfolio from sible for setting the risk tolerance of the fund and expressing it investment losses at the times of the market sell-offs. The two through the fixed allocation between “growth” and “defensive” managers have an investment target, or a benchmark, such assets (or “income” assets in our terminology), which was 60 as the consumer price index (CPI) + 5 percent. The 2012 TTF percent in growth assets and 40 percent in fixed assets at the review that advocated (successfully) for transitioning to the time of the OBAA strategy implementation. Before that, the OBAA strategy also promised that, in addition to reducing the target asset allocation varied over time depending on board downside risk, the OBAA strategy would reduce the overall decisions from relatively conservative 30 percent /70 percent volatility of TTF portfolio returns. to 70 percent /30 percent —the highest level of risk taken by TTF. This asset allocation was implemented with passive In fact, despite having significant discretion to reallocate mandates (for example, for Australian equities) with asset across different asset classes, on the aggregate the two man- managers replicating respective market indices for their sec- agers have maintained about 50 percent allocation to defen- tors. Performance of asset managers was reviewed regularly, sive assets, about 20 percent to diversified assets and about and they were periodically replaced for underperformance. 30 percent to growth assets as figure 5.2 illustrates. Small temporary deviations from those ranges have occurred while To implement the OBAA strategy, the TTF reduced its num- the market had significant movements during the period.15 ber of external managers from seven to two. One manager 14 See AMP Capital Multi-Asset Class fund, https://www.ampcapital.com/au/en/investments/funds/multi-asset-and-diversified/amp-capital-multi-asset-fund. 15 This abstracts from any derivative positions the managers may have had in place over time to manage risk levels. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 34 > > > F I G U R E 5 . 2 . - Movement of TTF Asset Category Allocation Since Inception of OBAA Strategy 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Oc 016 Oc 017 Oc 018 9 Oc 013 Oc 014 Oc 015 Ap 015 Ap 016 Ap 17 Ap 018 Ap 019 Oc 012 Ap 013 Ap 014 Ju 14 Ju 15 Ju 16 Ju 17 Ju 18 Ju 19 Ju 13 Ja 014 Ja 015 Ja 16 Ja 017 Ja 018 Ju 12 Ja 012 Ja 013 01 20 0 0 0 0 0 0 0 0 0 l2 r2 t2 l2 l2 2 r2 r2 t2 t2 l2 l2 2 2 r2 r2 t2 t2 l2 l2 2 l2 2 r2 2 r2 t2 r2 t2 n n n n n n n Ap Defensive Diversified Growth Source: TTF Quarterly Report, September 2019. Note: OBAA = objective-based asset allocation; TTF = Tuvalu Trust Fund. Thus, the assumption that if given leeway, investment manag- a scenario, we also provide comparison of TTF returns versus ers would have the incentive to make changes to their asset $A Morningstar investment portfolios with comparable levels allocation and the foresight to anticipate the market move- of risk, including conservative, moderate, balanced, growth ments—both of which would be required to deliver on the and aggressive portfolios. We also included these indices be- OBAA strategy—have not materialized. Both OBAA manag- cause they are used by the TTF investment monitor to com- ers are paid a fixed fee (relatively high, as discussed in the pare TTF returns against in their quarterly investment reports. section “Benchmarking of Investment Management Costs”) Appendix B provides specific details on the composition of whether or not they deliver on their performance objectives. these portfolios, their characteristics and historical risk and To introduce incentives for these types of mandates, the mar- return statistics. ket practice is to add a performance component to the asset management fee, making the payment depend on them deliv- Morningstar indices and our reference portfolios share the ering their return objectives. The level of market foresight and common approach to building portfolios with increasing level the skill required to deliver value net of fees over the broad of exposure to risky assets, while only differing on the geo- market—both required to deliver on such a mandate—are dis- graphic origin of these assets. Returns of Morningstar indi- cussed in great depth in appendix D. ces and our reference portfolios of similar risk levels corre- late between 0.8–0.85 confirming once again that it’s the risk Our reference portfolios are constructed to reflect the global level that drives returns over investment horizons rather than nature of financial markets and were constructed using global geographic exposure in this case. Table 5.3 presents com- risk and income assets reflective of the full spectrum of invest- parisons of TTF returns versus the $A Morningstar investment ment options accessible to the Pacific funds. Still, we acknowl- strategies and, consistently with the results using our refer- edge that $A-based Pacific funds might want to have greater ence portfolios, TTF returns correspond to the most conserva- exposure to regional $A investments. To accommodate such tive of the Morningstar portfolios. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 35 Caption: School Kids in Tonga | Source: Climate Investment Funds > > > T A B L E 5 . 3 . - Comparison of TTF Returns vs. $A Morningstar Investment Strategies 3 year 5 year Inception (April 2012) TTF - Tuvalu 4.6% 4.9% 6.1% $A Morningstar — Conservative 4.8% 5.1% 5.8% $A Morningstar — Moderate 6.1% 6.2% 7.0% $A Morningstar — Balanced 8.4% 8.1% 9.1% $A Morningstar — Growth 10.0% 9.2% 10.7% $A Morningstar — Aggressive 12.0% 10.8% 12.4% $A Global Equities 50% / Fixed Income 50% 9.3% 9.0% 10.6% Source: World Bank calculations. Note: $A = Austalian dollar; TTF = Tuvalu Trust Fund. Furthermore, the OBAA strategy significantly underperformed Tuvalu. This is significant national wealth that Tuvalu has not a simple easy to implement strategy of 50 percent global eq- accumulated as a result of its choice of the OBAA strategy. uities and 50 percent fixed income. Since its inception, the cumulative effect of OBAA strategy on TTF has been signifi- With that in mind, over the past several years, the TTF board cant compared with other investment alternatives. At $A 185 has engaged different partners and relevant institutions to as- million, TTF is about 313 percent of Tuvalu’s GDP. Thus, ev- sess the appropriateness of its current investment approach ery additional 1 percent in investment returns corresponds to and potential alternatives. As a part of these efforts in early about 3 percent of GDP in additional income to Tuvalu per 2020, TTF undertook a competitive selection process for its annum. Compared with $A 50 percent /50 percent passive investment monitor that plays a critical role in investment over- strategy, which would have cost only several basis points to sight: the selection continued despite the pandemic and TTF implement, TTF underperformed by 4.5 percent per annum, selected a global firm in June 2020. These important steps which is equivalent to about 14 percent of GDP for any year will improve TTF investment governance, and this compara- that OBAA has been in place. Since OBAA’s implementation tive analysis could inform TTF efforts in designing appropriate in April 2012, $A50 percent/50 percent strategy would have investment approach going forward. delivered about 113 percent of GDPP in additional wealth for EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 36 3. Benchmarking Kiribati’s Investment Approach: 2013–2016 Restructuring and Its Impact. The sovereign wealth fund of Kiribati—the Reserve Equaliza- bati’s Public Finance [Revenue Equalization Reserve Fund] tion Revenue Fund (RERF)—is one of the longest functioning Rules, 1983). The legislation did not address the purpose of sovereign wealth funds in the world. It was set up in 1956 by the the fund nor its investment universe, which were subsequently then colonial government of the Gilbert (now Kiribati) and Ellice periodically reviewed and defined by the cabinet chaired by (now Tuvalu) Islands to hold revenue from phosphate mining the president as described later in this section. with an initial allocation of US$555,580.16 Additional payments into the fund between 1973 and 1979 were from tax revenue. Comparing RERF returns over different investment horizons Since the end of phosphate mining operations in 1979, returns presented in table 5.4 reveals distinct investment patterns from RERF, particularly the investment returns, have been used over the medium horizon versus over the past 25 years: over to finance recurrent operations of the government. the past three to five years, RERF returns were consistent with medium- to high-risk reference portfolios, while before Legislation passed in 1983 formalized RERF’s governance that its returns had been lower than the conservative strategy. structure and the roles of relevant bodies and institutions (Kiri- > > > T A B L E 5 . 4 . - Kiribati’s Returns vs. Reference Portfolios 3 year 5 year Inception (June 1995) RERF 11.3% 7.6% 6.0% US$ Diversified Risk 20% 6.4% 6.1% 7.6% US$ Diversified Risk 40% 8.4% 7.8% 7.8% US$ Diversified Risk 60% 10.5% 9.4% 8.0% US$ Diversified Risk 80% 12.5% 11.1% 8.0% Global Equities 50% 10.0% 8.4% 7.5% Source: World Bank calculations. Note: RERF = Revenue Equalization Reserve Fund. The explanation for such different outcomes over these years is that, similarly to TTF, RERF had evolved its investment approach over time and, in particular, undertaken significant restructuring during 2013–2016 following a review by the International Monetary Fund (IMF). In 2011, the IMF reviewed the management of the RERF and found significant issues with how the fund was managed. The IMF concluded that unless these issues were addressed, the RERF could be depleted in the foreseeable future. According to the IMF, at the time of their review the RERF strategic asset allocation reflected the advice received from an external consultant in 1995, albeit with some modifications, and that the advice had been too conservative. The IMF further detected large risks ow- ing to active risk taking by asset managers that went undetected. Those risks were eventually materialized as losses in large part because of the Global Financial Crisis. The IMF also identified unsustainable drawdowns to finance the government’s recurrent expenditures. The level of drawdowns increased significantly during the first decade of the 2000s and was compounded by invest- ment losses during financial crises in 2001 and 2007–2008 as figure 5.3 illustrates. 16 Amount of initial allocation is approximate owing to missing information on what currency was used. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 37 > > > F I G U R E 5 . 3 . - Growth of RERF Before 2013–2016 Restructuring RERF Net Values, Drawdowns, and Deposits 60,000,000 800,000,000 700,000,000 50,000,000 600,000,000 40,000,000 500,000,000 30,000,000 400,000,000 300,000,000 20,000,000 200,000,000 10,000,000 100,000,000 - - 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Deposits (US$) Drawdowns (US$) Value (US$) Source: RERF data. Note: RERF = Revenue Equalization Reserve Fund. At the time of the IMF review, RERF assets were managed by Over time, both asset managers had significantly underper- two external asset managers each of whom managed rough- formed their benchmarks before fees. One of the managers ly one-half of the portfolio. The RERF’s assets were actively suffered a large irrecoverable loss when the three largest managed with the expectation that active management would banks in Iceland, in which the manager had a sizable position, add value. One of the managers had managed the RERF’s as- defaulted on their debt in the fall of 2008, and the Icelandic sets since the fund’s inception and the second one was hired króna became nonconvertible. Management fees—higher for in 1995. Initially, the asset managers had similar investment active management than for passive management—had fur- agreements with identical benchmarks, but after a number ther eroded the RERF’s balances. Table 5.5 presents the ex- of changes over the years their respective portfolios started tent of portfolio underperformance versus investment bench- to diverge. These differences had grown over time, initiated marks soon after the IMF review. partially by the asset managers proposing changes to their benchmarks or allowable investment universe. > > > T A B L E 5 . 5 . - RERF Policy (Benchmark) Return vs. Portfolio Net Return Before 2013 SAA Review 1-year return 3-year return 5-year return Kiribati (policy return) 15.40% 12.90% 8.00% Kiribati (portfolio net return) 7.70% 7.85% 7.26% Source: FEMM 2014. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 38 The IMF recommended that RERF’s conservative asset allo- percent bonds and 50 percent global public equities. Follow- cation be revised to be better aligned with the fund’s long-term ing further evaluation of potential options and with ongoing purpose. It also recommended switching from active invest- support from RAMP, over the course of 2014–2016 RERF un- ment strategies to passive strategies so the risk of underper- dertook full portfolio restructuring implementing the cabinet’s formance would be significantly reduced, asset management decisions, which resulted in termination of prevailing active costs would be cut, and ongoing monitoring and oversight of managers and selection of new passive managers through a the managers would be simplified. public tender process. With capacity building from the World Bank’s Reserve Advi- Thus, the three-year returns in table 5.6 and 5.7 correspond to sory Management Program (RAMP), in June 2013 RERF In- the revised investment approach centered on explicitly defined vestment Committee recommended to cabinet and the cabi- asset allocation target and passive investment mandates. The net approved to (a) confirm the objective of the fund as budget five-year return still captures the transition from conservative support and sustaining the value of the fund for future genera- asset allocation and underperforming asset managers to the tions; (b) limit eligible asset classes to government bonds and new approach. Inception returns reflect predominately invest- public equities; and (c) set the strategic asset allocation to 50 ment before the 2013–2016 restructuring. > > > T A B L E 5 . 6 . - Comparison of RERF Returns vs. $A Morningstar Investment Strategies 3 year 5 year Inception (June 2015) RERF — Kiribati 11.3% 7.6% 6.0% $A Morningstar — Conservative 5.5% 4.8% 6.5% $A Morningstar — Moderate 6.9% 6.0% 6.8% $A Morningstar — Balanced 9.3% 7.9% 7.5% $A Morningstar — Growth 11.0% 9.2% 7.8% $A Morningstar — Aggressive 13.0% 10.8% 8.0% $A Global Equities 50% / Fixed Income 50% 10.0% 8.7% 7.5% Source: World Bank calculations. Note: $A = Australian dollar; RERF = Revenue Equalization Reserve Fund. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 39 4. Benchmarking Federated States of Micronesia and the Marshall Islands Investment Process: Gradual Evolution. The Compact of Free Association17 provides for the establish- . . . maximize investment returns and assets in the Trust ment of the Marshall Islands and the Federated States of Mi- Fund through the period ending fiscal year-end Septem- cronesia compact trust funds in accordance with the trust fund ber 30, 2023, subject to the total portfolio risk param- agreement between the original parties: (a) the United States eters defined in the section Risk Definition. Over shorter government and the Federated States of Micronesia govern- periods, outperformance will be sought relative to the ment and (b) the United States government and the Marshall notional return on a benchmark portfolio designed to re- Islands government. Sections 215 through 218 of the amend- flect the risk profile according to which the assets are ed compact set forth the funding to be contributed to the funds invested at the time. Post-fiscal year 2023, the broad by the three governments through 2023.18 investment objective will be to produce a level of return sufficient to maintain levels of spending from the Trust The Marshall Islands and the Federated States of Micronesia Fund consistent with the desires of the members and funds were incorporated as nonprofit corporations under the consistent with risk parameters and spending policy to laws of the District of Columbia in August 2004. The two trust be defined at that time.20 fund agreements are supported by a set of bylaws initially ap- proved by the respective committees on March 24, 2006, and IPSs are prepared by the investment adviser in consultations August 19, 2005, respectively. Resolutions are considered with the TF committees. and approved periodically to improve the overall management and operations of both funds, as determined by the respective Both funds have been investing across a diverse set of global committees. The investment policy statements (IPSs) provide asset classes and, in contrast to RERF and TTF, currently in- both funds with main investment guidance. The Federated vest in alternative asset class through private equity (Marshall States of Micronesia IPS was amended in April 2019 and the Islands and Federated States of Micronesia) and hedge funds Marshall Islands IPS was amended June 15, 2017, and imple- (Federated States of Micronesia). The list of asset managers mented in July 2017 (Trust Fund for the People of the Feder- for both funds, their investment products and costs are includ- ated States of Micronesia 2020; Trust Fund for the People of ed in appendix C. Whereas the Marshall Islands and the Fed- the Republic of the Marshall Islands). erated States of Micronesia compact trust funds’ investment approaches have evolved, their changes have been gradual The trust fund agreements define the purpose of both trust with marginal impact on the variability of returns during the funds as to “contribute to the economic advancement and tenure of the current administrator since 2011. Before that, long-term budgetary self-reliance” of the two states “by pro- the Marshall Islands was investing in public equity and fixed viding an annual source of revenue, after Fiscal Year 2023.”19 income and has since added two private equity funds to that Both funds contract an executive administrator who serves in allocation. The Marshall Islands had a hedge fund allocation support of the governance, administration, and operations of that was liquidated with proceeds allocated to passive low- the fund. The current executive administrator is based in the cost products in 2017. The Federated States of Micronesia Washington, D.C., area and has been providing services since had an earlier experience with a complex portfolio of private April 2011. equity, hedge funds, real estate, and some debt funds that was later restructured for underperformance and optimization IPSs for both funds codify their investment objectives and of costs and management. specify their strategic asset allocation, which are reviewed an- nually. IPS investment objectives are identical for both funds Most recent investment changes in the compact trust funds to do as follows: have continued to reduce active risk taking and move to pas- 17 The Compact of Free Association, amended as the “Amended Compact,” is codified in the Compact of Free Association Amendments Act of 2003 (U.S. Public Law 108- 188, December 17, 2003; “the Amended Compact Act”) under Title Two: Economic Relations, Sections 215 and 2016, 18 Federated States of Micronesia TF FY19 Annual Report, March 2020; the Marshall Islands TF FY19 Annual Report, March 2020. 19 FSM FY19 annual report, 4. 20 RMI FY19 annual report, 4–5. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 40 sive mandates. In particular, in 2019 the Federated States of As presented in table 5.7, over the past three to five years, Micronesia compact trust fund asset allocation was adjusted returns for both compact trust funds are in line with the funds’ with “the objective of improving performance and reducing in- growth approach corresponding to returns of reference port- vestment fees” (Trust Fund for the People of the Federated folios with 60 percent to 80 percent risk levels. Before that, States of Micronesia 2020, 11). The changes were to (a) con- the Federated States of Micronesia returns were lower and vert the US small/mid cap and emerging market asset strate- investments in more complex and costly investment products gies from active to passive and (b) reduce the opportunistic that may have contributed to that outcome. Implementing a fixed income asset strategy from 16 percent to 5.0 percent more robust investment process, through annual reviews and and a corresponding increase in the core fixed income fund optimization of complexity and costs have led to returns being (Trust Fund for the People of the Federated States of Micro- in line with the growth objectives for both compact trust funds. nesia 2020). > > > T A B L E 5 . 7 . - Federated States of Micronesia and Marshall Islands’ Returns vs. US$ Reference Portfolios 3 year 5 year FSM Inception (Sept 2004) RMI Inception (June 1995) FSM — Micronesia, Fed. Sts. 7.7% 5.9% 5.5% RMI - Marshall Islands 8.1% 6.0% 6.0% US$ Diversified Risk 20% 4.8% 4.9% 5.3% 5.0% US$ Diversified Risk 40% 6.0% 5.5% 6.0% 5.6% US$ Diversified Risk 60% 7.0% 6.2% 6.6% 6.1% US$ Diversified Risk 80% 8.1% 6.8% 7.2% 6.4% US$ Global Equities 50% / Fixed Income 50% 7.1% 5.8% 6.1% 5.7% Source: World Bank calculations. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 41 5. Nauru: Reestablishing Trust Fund The Intergenerational Trust Fund for the People of the Re- rather than financial assets such as bonds and equities. In- public of Nauru, or Nauru Trust Fund, in its current incarnation vestments in property saw significant losses in the 1990s and is the newest addition to the Pacific trust and sovereign fund Nauru’s property portfolio was seized by receivers in 2006 as family, reestablished on November 6, 2015, with “the objective the country’s debts rose. The Nauruan government also bor- to generate investment earnings that can be used to provide rowed at high interest rates against the NPRT to finance pub- a source of revenue to the Republic of Nauru post 2033 (or lic expenditures. In 2009 Nauru’s public debt was at $A 869 at a time sooner as determined by the Committee) for invest- million, around 30 times GDP, and NPRT was wound down ment in education, health, environment and infrastructure” (In- (Gould 2014). tergenerational Trust Fund for the People of the Republic of Nauru 2019, 3). Since its reestablishment in 2015, Nauru trust fund has been developing its investment operations and building its invest- The original Nauru trust fund—the Nauru Phosphate Royal- ment portfolio. Nauru’s finance ministry referred the World ties Trust (NPRT)—was established in 1968 at the time of Bank team to its 2019 annual report that has limited informa- independence in 1968. Similar to Kiribati’s RERF, the trust tion about the composition of the fund and its investments. fund was established to hold revenue from phosphate mining. The team used the financial statements in the annual report The NPRT was composed of four funds, established for sep- and publicly available information about relevant investment arate reasons: the Long–Term Investment Fund, the Land products that Nauru trust fund has been investing in to decon- Owners’ Royalty Trust Fund, the Housing Fund, and the Re- struct the level of its total risk level and its investment costs. habilitation Fund. Whereas formally the NPRT and the state’s Given that the Nauru trust fund is in the early stages of build- budget were separate, in practice there was little distinction ing up its portfolio, we are not in the position to benchmark its (Gould 2014). performance and risk characteristics versus appropriate refer- ence portfolios as it would require a longer time period. At this Because of the fall in phosphate exports in the 1990s, the gov- stage we can only comment that Nauru trust fund is pursuing ernment revenue fell while expenditures continued to increase the growth investment strategy with 85 percent allocation to thereby leading to significant budget deficits. The value of the risk assets and relatively high allocation to private markets at trust fund declined from an estimated $A 1.3 billion in 1990 to 22 percent, as summarized in table 5.8, and that their costs $A 300 million in 2004. The investment strategy of the NPRT are among the highest in our Pacific sample, a close second was to invest phosphate revenue in international property, to Tuvalu’s (as discussed in the next section). > > > T A B L E 5 . 8 . - Implementation of Nauru Trust Funds Asset Allocation, 2018–2018. 2019 2018 Fixed Income 18% 25% Public Equity 60% 60% Private Markets 22% 15% Total 100% 100% Income Assets 15% 21% Risk Assets 85% 79% Source: World Bank estimates based on the annual report and the fund website. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 42 6. Benchmarking of Investment Management Costs Investment expenses, like investment returns, compound over structures and comprehensive accounting of all relevant time, thereby affecting wealth accumulation over the invest- costs, included in their annual reports and presented in ap- ment period. Managing costs are particularly important for pendix C. The team was able to reconstruct the Pacific funds’ small funds. Whereas in the past the small funds lacked pric- asset management fees—a subset of total costs—from sev- ing power compared with their large peers, with the develop- eral sources. Figure 5.4 presents the aggregate asset man- ment of the passive index industry these funds are positioned agement fees for the five Pacific funds. Kiribati’s fee is by far to get similar costs for their passive mandates just as the the smallest since RERF replaced its active mandates with large funds do. This development is one reason for the fast passive benchmark replication strategies. The Tuvalu Trust growth of the passive low-cost mandates that we discuss later Fund’s costs are the highest reflecting the cost of the OBAA in the section. strategy. Asset management fees for the Marshall Islands and the Federated States of Micronesia reflect a combination of Among the Pacific funds only the Marshall Islands and the higher-cost strategies and transition from higher-cost active Federated States of Micronesia funds have transparent cost mandates to low-cost passive mandates. > > > F I G U R E 5 . 4 . - Pacific External Fund Management Costs, 2019 0.70 0.64 0.60 0.60 0.53 0.50 0.38 Percentage 0.40 0.30 0.20 0.08 0.10 0.00 Tuvalu Nauru Micronesia, Fed. Sts Marshall Islands Kiribati Source: World Bank calculations based on data provided by funds. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 43 The cost effectiveness of asset management is not about min- for the two selected global asset managers, resulting in about imizing costs, but about generating value for the portfolio net $A 1 million21 savings in asset management fees annually for of costs. Thus, evaluation of asset management costs should the country. be done considering the broader context of the funds’ invest- ment approach and what net value is being generated through Following the Global Financial Crisis (GFC), the financial higher-cost services. industry has been experiencing significant reduction in as- set management fees for a number of reasons, including (a) One important tool to manage investment management fees greater efficiency in public markets, (b) fee pressures within for Pacific Funds is a public competitive selection of invest- the industry owing to higher competition and lower yield lev- ment managers based on well-defined requirements. In 2015, els, (c) an increasing role of technology and intermediates, as a part of its fund restructuring, Kiribati’s RERF undertook a (d) greater access to market information through multiple plat- public procurement process, or request for proposals (RFP), forms, and (e) a growing awareness of market participants. to replace its two incumbent active asset managers who were More specifically, according to Bloomberg, investors are pay- no longer meeting RERF’s objectives. To realign RERF’s new ing roughly half as much for the same investment strategies requirements with the industry’s best practice, an RFP speci- as they were nearly two decades ago and about a quarter fied detailed investment requirements and was issued to all less than five years ago (Waite, Massa, and Cannon 2019). market participants. RFP criteria included (a) a manager’s figure 5.5 presents the reduction of investment fees over the capacity to provide relevant investment options, (b) demon- decade since GFC for public equity and fixed income invest- strated service capability and commitment to clients’ needs, ment products. (c) experience with similar regional or sovereign wealth funds, and (d) total costs and fees. Formal proposals were received Passive strategies’ share of the investing universe continues from one regional and seven global asset managers, of which to increase at the expense of active strategies, as figure 5.6 four were shortlisted for further due diligence and interviews. demonstrates. Furthermore, demand for active management Two were selected to manage public equity and fixed income has been shifting to come from more sophisticated investors mandates. As a result of the RFP, RERF realized significant that have demonstrated their institutional ability to extract val- reductions of fees from the incumbent managers’ 0.22 per- ue from active management at an appropriate price, as active cent and 0.17 percent down to 0.06 percent and 0.10 percent management has not added value on average. > > > F I G U R E 5 . 5 . - Evolution of Investment Fees (Expense Ratios) Since GFC 2018 average Change since 2008 Type Trend expense ratio (percentage points) Active equity funds 0.76% -0.18 Active bond funds 0.55% -0.10 Index equity ETFs 0.20% -0.09 Index bond ETFs 0.16% -0.03 Index equity funds 0.08% -0.10 Index bond funds 0.07% -0.09 Source: Waite, Massa, and Cannon, 2019. Note: GFC = Global Financial Crisis; ETF = exchange traded fund. 21 This savings amount is based on the $A 750 million RERF assets under management at the time. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 44 > > > F I G U R E 5 . 6 . - Shift in Market Demand for Lower-Fee Products 2018 revenue margin (bps) 180 175 Private market Active specialty alternatives 75 equity 70 65 60 Public market 55 alternatives 50 45 40 Active core equity Active core Multi-asset fixed income 35 30 25 Money Active core 20 Passive Active specialty market fixed income 15 other fixed income 10 Passive equity Passive 5 multi asset 0 -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 16.0 2018 net flows, % Passive Alternatives Active Bubble size - 2018 AUM Source: Baghai et al., 2019. Note: AUM = assets under management; bps = basis points. In its analysis of the global asset management industry, McK- gies and only a few top performing funds at the lowest costs insey & Company described a further paradigm shift in asset have seen some inflows. Thus, strategies to optimize invest- management fees during 2018 with significant implications for ment costs deployed by the Marshall Islands and the Feder- the industry going forward: a move from the market where in- ated States of Micronesia funds and complete shift from ac- vestment performance ruled above all else—investors seek- tive to passive mandates by Kiribati’s fund are in line with the ing “good performance at a fair price”—had shifted to “good industry evolving practices. Tuvalu’s fund continues to employ performance at the best price” (Baghai et al. 2019). Overall, high-cost asset managers, despite their consistent underper- investors have been taking their money out of active strate- formance on their mandates since its inception. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 45 6. Caption: Kiribati couple | Source: Author >>> Conclusions Pacific funds’ investment outcomes have been diverse, and our proposed benchmarking ap- proach helps to better understand the key contributing factors to this divergence. A direct comparison of the Pacific funds’ investment returns is challenging because of the funds’ reporting periods and granularity of the available data. Hence, our approach of using refer- ence portfolios calibrated to specific markets and funds’ risk parameters allows us to infer the implicit investment benchmarks for each funds’ investment approach and to compare them with the available investment alternatives specific to each fund. These implicit investment bench- marks allow us to draw general conclusions about whether the fund’s investment approach is in line with the fund’s investment objective and fully uses its investment authorization and its risk parameters. The most significant factor explaining the differences in investment outcomes is the Pacific funds’ investment strategies that their boards pursued to achieve investment objectives. The Pacific funds followed two approaches: (a) a traditional strategic asset allocation approach pursued by Kiribati, the Marshall Islands, and the Federated States of Micronesia and (b) an objective-based asset allocation (OBAA) strategy in Tuvalu. In the traditional strategic asset allocation approach, the boards expressed their tolerance for the overall level of risk through an asset allocation mix that was translated into an investment benchmark. In contrast, in the OBAA strategy, which the Tuvalu Trust Fund approved in 2012, the decision on asset allocation was delegated from the board to the asset managers. The thinking was that the managers were in a better position to anticipate the market movements and deliver superior returns versus if the board pursued the traditional asset allocation approach. OBAA returns since inception are comparable with a conservative portfolio and are lower than comparable returns of its regional peers that have delivered returns in line with balanced to growth investment strategies. In fact, if the TTF board had set a simple and easy to implement benchmark of 50 percent global equities and 50 percent fixed income, it would have delivered 10.6 percent returns versus the OBAA strategy’s 6.1 percent since its inception in 2012. Thus, our analysis revealed that the implicit benchmark for the OBAA strategy is a highly conservative investment strategy that the TTF board could have pursued by implementing a passive portfolio. For example, the TTF board could have pursued the $A Morningstar conservative index or 80 percent fixed income/20 percent global equity, for a fraction of the cost. In contrast, the funds of the Marshall Islands, the Federated States of Micronesia, and Kiribati pursued an explicit invest- ment growth strategy, and these funds’ implicit benchmarks were in line with having 50 percent to 80 percent growth assets. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 47 Another contributing factor to the returns difference was the active risk for the total portfolios (Marshall Islands and Feder- Pacific funds’ different approaches to implementing their in- ated States of Micronesia). In fact, the level of complexity of vestment policies. All examined funds experienced cases of the investment mandates is one of the main factors explaining significant underperformance of their respective benchmarks the differences in investment costs. For the Marshall Islands, (for example, Tuvalu) or suffered outright default (for example, the Federated States of Micronesia, Nauru, and Tuvalu, high- Kiribati). The level of underperformance was directly related er fees are reflective of their investment in alternative asset to the level of the investment mandate’s complexity or degree classes or higher complexity mandates, whereas Kiribati’s low of allowable active risk. Consistent with global experience, costs reflect fully passive mandates. Furthermore, Kiribati was more complex or active mandates require significant in-house able to further reduce its managers’ fees through a competi- skills to implement and monitor. As a result of these experi- tive selection process. As a market practice globally, competi- ences and in line with trends globally, Pacific funds, with the tive selection reduced asset management fees for Kiribati and exception of Tuvalu, have either moved entirely to passive would bring value to the Pacific funds. mandates (Kiribati) or have been reducing the overall level of K E Y TA K E A W AY S Key takeaways for the Pacific funds from this benchmarking study are consistent with best practices for long-term investors: • To ensure long-term sustainability of Pacific funds, the investment strategy should reflect the fund’s investment purpose, which should have in place a well-defined investment governance framework to ensure the strategy is formalized, implemented, and monitored in line with global best practices. • Investment benchmark, representing replicable strategic asset allocation based on the investment policy, typically accounts for 80 percent to 90 percent of the returns and risk of the portfolio. • More complex investment approaches and mandates are costly and require more sophisticated governance and ongoing efforts to oversee and manage these mandates. Boards should ensure that they have the required time, inclination, and knowledge to oversee such mandates and that these ap- proaches add value net of cost to the fund over time • OUTSOURCING IS NO SUBSTITUTE FOR ACCOUNTABILITY: boards have a fiduciary responsibility to the fund’s beneficiaries and are ultimately responsible for investment decisions, including those that are outsourced. As such, they need to ensure ownership of the risks that are being delegated and en- sure that robust processes are followed to select and monitor service providers in line with the fund’s needs. • HIGH-QUALITY GOVERNANCE of the investment process is necessary for the long-term success of in- vestment funds. Statutory governance—that is, clearly defined rules and investment parameters—should be well articulated to provide clarify and accountability to manage the funds. Operational governance— quality of day-to-day decision making and exercising fiduciary responsibility—will directly affect the financial results of the funds. The frameworks should be reviewed periodically by impartial and independent parties to ensure that investment parameters evolve with the evolving market and policy contexts. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 48 7. Caption: Kiribati community | Source: Author >>> References and Other Readings Australian Government Future Fund. 2019. “2018–19 Future Fund Annual Report.” https://www. futurefund.gov.au/about-us/annual-reports. Baghai, Pooneh, Kevin Cho, Onur Erzan, and Ju-Hon Kwek. 2019. “Beyond the Rubicon: North American Asset Management in an Era of Unrelenting Change.” Report, Global Banking and Securities Practice. McKinsey & Company, New York. Beath, Alexander D., and Chris Flynn. 2018. “Real Estate Investment Performance by Invest- ment Implementation Style: The Experience of Large U.S. DB Pension Funds.” CEM Bench- marking, Toronto. https://cembenchmarking.com/ri/insight/19. Beath, Alexander D., and Chris Flynn. 2019. “Asset Allocation and Fund Performance of Defined Benefit Pension Funds in the United States, 1998–2017.” CEM Benchmarking, Toronto. Iverson David, Qing Ding, Charles Hyde, Carly Falconer, Christopher Worthington and Matthieu Raoux. 2020. “2020 Reference Portfolio Review.” New Zealand Super Fund. https://www.nzsu- perfund.nz/assets/Uploads/The-2020-Reference-Portfolio-Review-v2.pdf Brinson, Gary P., L. Randolph Hood, and Gilbert L. Beebower. 1986. “Determinants of Portfolio Performance.” Financial Analysts Journal 42 (4): 39–44. https://www.jstor.org/stable/4478947. Brinson, Gary P., Brian D. Singer, and Gilbert L. Beebower. 1991. “Determinants of Portfolio Performance II: An Update.” Financial Analysts Journal 47 (3): 40–48. https://www.jstor.org/ stable/4479432. Drew, Aaron. 2017. “Improving the Performance of Sovereign Funds in the Pacific.” New Zea- land Institute for Pacific Research, Aukland, NZ. https://apo.org.au/node/104286. Drew, Aaron, and Bart Frijns. 2017. Reference Portfolios for Sovereign Funds in Pacific Island Nations. Aukland, NZ: New Zealand Institute for Pacific Research. http://natlib-primo.hosted. exlibrisgroup.com/NLNZ:NLNZ:NLNZ_ALMA11305584670002836. Drew, Aaron, Bart Frijns, and Ross Fowler. 2016. Assessment Framework for Sovereign Funds in Pacific Island Nations. Aukland, NZ: New Zealand Institute for Pacific Research. http://natlib- primo.hosted.exlibrisgroup.com/NLNZ:NLNZ:NLNZ_ALMA21277983080002836. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 50 Drew, Aaron, and New Zealand Institute for Pacific Research. 2016. The Role of Sovereign Funds in Pacific Island Nations. Aukland, NZ: New Zealand Institute for Pacific Research. http:// natlib-primo.hosted.exlibrisgroup.com/NLNZ:NLNZ:NLNZ_ALMA21277983050002836. Dynkin, Lev, Peter Ferket, Jay Hyman, Erik van Leeuwen, and Wei Wu. 1999. “Value of Security Selection versus Asset Allocation in Credit Markets.” Journal of Portfolio Management 25 (4): 11–27. https://doi.org/10.3905/jpm.1999.319759. Dynkin, Lev, Jay Hyman, and Wei Wu. 2000. “Value of Skill in Security Selection versus As- set Allocation in Credit Markets.” Journal of Portfolio Management 27 (1): 20–41. https://doi. org/10.3905/jpm.2000.319780. FEMM (Finance and Economic Ministers Meeting). 2014. Presentation. 18th Forum Eco- nomic Ministers Meeting in Honiara, Solomons Islands. July 9, 2014. https://www.forumsec. org/2014/07/10/establishing-and-managing-trust-funds-in-the-region-2/ FEMM (Finance and Economic Ministers Meeting). 2019. Decision paper. Internal document. Gould, Martin. 2014. “Managing Manna from Below: Sovereign Wealth Funds and Extractive Industries in the Pacific.” Economic Roundup, Australian Treasury, Canberra. https://treasury. gov.au/publication/economic-roundup-issue-1-2010/economic-roundup-issue-1-2010. Hitt, Jack. 2000. “The Billion-Dollar Shack.” The New York Times Magazine, December 10. https://www.nytimes.com/2000/12/10/magazine/the-billion-dollar-shack.html. Hughes, Helen. 2004. “From Riches to Rags: What Are Nauru’s Options and How Can Australia Help?” Issue Analysis 50 (August). The Centre for Independent Studies, Sydney, Australia. IMF (International Monetary Fund). 2019. World Economic Outlook, October 2019: Global Man- ufacturing Downturn, Rising Trade Barriers. Washington, DC: IMF. Intergenerational Trust Fund for the People of the Republic of Nauru. 2019. “Annual Report 2019.” Internal document. IWG (International Working Group of Sovereign Wealth Funds). 2008. “Sovereign Wealth Funds: Generally Accepted Principles and Practices: ‘Santiago Principles.’” Report. IWG, Washington, DC. https://www.ifswf.org/sites/default/files/santiagoprinciples_0_0.pdf. Kahneman, Daniel. 2003. “Maps of Bounded Rationality: Psychology for Behavior- al Economics.” American Economic Review 93 (5): 1449–75. https://www.jstor.org/ stable/3132137?seq=1#metadata_info_tab_contents. Kiribati. 1983. Public Finance (Revenue Equalization Reserve Fund) Rules. NACUBO-TIAA (National Association of College and University Business Officers–Teachers In- surance and Annuity Association). 2018. 2018 NACUBO-TIAA Studies of Endowments. Wash- ington DC: NACUBO and TIAA. NACUBO-TIAA (National Association of College and University Business Officers–Teachers In- surance and Annuity Association). 2019. 2019 NACUBO-TIAA Studies of Endowments. Wash- ington DC: NACUBO and TIAA. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 51 New Zealand Super Fund. 2020. “Investment Cost Effectiveness Analysis.” NZ Super Fund, Auck- land. December 31, 2020. https://www.nzsuperfund.nz/assets/Uploads/CEM-Survey-2020.pdf. Scott, Brian J., James Balsamo, Kelly McShane, and.Christos Tasopoulos. 2017. ”The Global Case for Strategic Asset Allocation and an Examination of Home Bias.” Report. Vanguard Re- search, Valley Forge, PA. Taleb, Nassim N. 2004, Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets. London: Texere, second edition. Originally published 2001. Trust Fund for the People of the Federated States of Micronesia. 2020. “Fiscal Year 2019 Annual Report.” https://fsmcfatf.com/file-repository/. Trust Fund for the People of the Republic of the Marshall Islands. 2020. “Fiscal Year 2019 Annu- al Report.” https://rmicfatf.com/file-repository/. TTF (Tuvalu Trust Fund). 2019. “Tuvalu Trust Fund Annual Report 2018.” TTF, Funafuti, Tuvalu. Waite, Suzy, Annie Massa, and Christopher Cannon. 2019. “Asset Managers with $74 Trillion on Brink of Historic Shakeout.” Bloomberg, August 8. https://www.bloomberg.com/graphics/2019- asset-management-in-decline/. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 52 PENDIX A.Appendix A. Reference Portfolios: Definition and Construction >>> Appendix A. Reference Portfolios: Definition and Construction Our comparative analysis approach draws on the recommendations from Drew and Frijns (2017) for Pacific funds to use reference portfolios as governance benchmarks. The reference portfolio approach is a method for formulating a portfolio governance benchmark that limits the assets being used to publicly available asset classes that could be cheaply and simply implemented as a passive portfolio. This approach was pioneered by the New Zealand Super Fund in 2010 (Iverson et al. 2020). In 2017 New Zealand Institute for Pacific Research (NZIPR) released a study on the relevance of the reference portfolio approach for Pacific Island Nations (Drew and Frijns 2017). Table A.1 describes the key attributes of reference portfolios. > > > T A B L E A . 1 . - Key Attributes of a Reference Portfolio Characteristic Description The reference portfolio should contain public asset classes where the Be a simple and low cost portfolio that fund can replicate the exposure using low-cost passive management. This could be implemented passively makes the reference portfolio a realistic and viable option The benchmark indexes should represent broad market representation Be diversified of the asset classes, ensuring a high level of diversification for the overall reference portfolio Reflect an appropriate risk level for the A reference portfolio should be constructed with a risk profile in line with Fund, given its purpose the fund’s purpose The reference portfolio needs to be considered from a home market Relevant to its home investor perspective. This may include domestic legislative constraints, the inclusion/exclusion of domestic assets, and currency hedging needs. The reference portfolio should reflect a long-term perspective or A long run concept expectations of the risk and returns of asset classes and the resulting portfolio. Source: Drew and Frijns, 2017. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 54 Reference portfolios can serve two purposes. First, given their • Because all Pacific Funds often use external managers, design, they can be passively implemented at a low cost. How- reference portfolio allows to quantify the value-add of ever, they can also be used as a policy benchmark to analyze these managers, net of fees, against this benchmark, par- the value added or subtracted by active managers. Tradition- ticularly if they are given active management mandates. ally benchmarks may include measures that are not directly • It provides a long-run construct and removes undue influ- investable—such as the use of the S&P + 3 percent as a pri- ences to formulate a benchmark that is overly prejudiced vate equity benchmark or the use of the consumer price index by short-term tactical or other considerations. (CPI) +3 percent for objective-based asset allocation (OBAA) • To simplify Pacific funds administration and optimize op- strategies used for Tuvalu Trust Fund (TTF). Because refer- erational costs through the use of widely available trans- ence portfolios are limited to investable public asset classes, parent market indices, this approach makes fund perfor- they represent a credible investment alternative. mance measurement and monitoring much easier. A reference portfolio is a simplified strategic asset allocation process that imposes straightforward limits on the asset class- es being used, such that it could be implemented as a simple, Building Reference cost-effective, and passively managed fund. Portfolios for Pacific Funds Particularly relevant for Pacific funds, the advantages of using the reference portfolio approach are as follows: To construct reference portfolios for the Pacific funds we used • It represents an easily replicable passive portfolio tailored widely diversified indices of risk assets and income assets to the fund’s individual risk tolerance parameters, which defined in the table A.2. Risk assets have higher long-term could be used as a default investment strategy that would growth, but greater volatility of returns in the short term. In- not require significant resources to implement and man- come assets have lower, but more steady, long-term growth age on ongoing basis. and a lower likelihood of loss in the short term. > > > T A B L E A . 2 . - Definition of Income and Risk Assets Building block Description Examples • Investment-grade government and corporate bonds Income assets Defensive assets that have lower, but more steady long-term growth • Cash • Equities Risk assets Assets with higher return, but also higher risk • Public real estate • Subinvestment grade government and corporate bonds Source: World Bank. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 55 We create two different reference portfolio types. For risk assets, this reference portfolio approach creates a diversified risk basket. Similar to the first approach, the core • Type 1: Passive 50 percent Global Equity/50 percent of the basket remains global equities and the Bloomberg Bar- Global Fixed Income clays Global Aggregate is used for the income asset. In ad- dition, three other asset classes are added to the portfolio to The first, and most simple reference portfolio is constructed increase diversification and improve risk adjusted returns. of 50 percent global equities as the risk asset, and 50 percent global fixed income as the income asset. Portfolio returns are Diversified Risk Basket includes the following asset classes: calculated with monthly rebalancing. For global fixed income, • Global equities (70 percent): The core of the basket we use the Bloomberg Barclays Global Aggregate, which is a is MSCI World, to provide a globally diversified exposure globally diversified index of investment grade government and to equities. corporate bonds. This index is highly diversified and represents • Public real estate (10 percent): According to CEM Bench- a proxy for the entire global investment grade fixed income uni- marking (Beath and Flynn 2018), listed equity real estate verse. For the risk assets we use MSCI World,22 which is an eq- investment trusts (REITs) have a .91 correlation with pri- uity index that includes companies from 23 developed nations. vate real estate and result in higher after-fee returns. We use the FTSE EPRA Nariet Global REITs Index. The passive 50 percent global equity/50 percent global fixed • US small-cap equities (10 percent): According to CEM income reference portfolio allows for a relatively high level of Benchmarking data, US small-cap equities have a .89 global diversification with simple to understand and implement correlation with private equity returns,23 and higher net- cost-effective structure, which serves as a useful baseline. This of-fee returns for funds under US$2 billion. We use the simple structure gives an investable portfolio option that does Russell 2000 index. not require a significant level of technical expertise to maintain • Emerging market US$ debt (10 percent): We add US$- and monitor, while providing exposure to key market factors denominated bonds from emerging market governments that will drive portfolio returns over its investment horizon. because the asset class has relatively high risk-adjusted returns and a lower level of correlation compared with the • Type 2: Diversified Risk Basket Covering Full Range of other risk assets in the basket. We use the JP Morgan EM- Risk Level Exposure BIG Index, which includes the US$-denominated bonds of over 60 emerging market governments and wholly owned A second approach uses a larger mix of risk assets to increase government enterprises. diversification to improve risk-adjusted returns and to provide public proxies for commonly used private market assets. Table A.3 presents correlation of asset classes included in the These reference portfolios offer a more complicated structure reference portfolio that highlights the source of risk diversifica- than type 1 described previously requiring slightly more tech- tion: risky and income assets have exhibited negative correla- nical capacity to develop, maintain, and monitor while provid- tion of monthly returns over the past 25 years. ing higher risk adjusted returns or lower downside risk. > > > T A B L E A . 3 . - Correlation of Monthly Returns for Income Assets and Risk Assets, December 1995–April 2020. R - Global Agg 0.12 R - EM Sovereigns 0.03 0.33 Public Real Estate -0.05 0.20 0.61 US Small-Cap Equities -0.04 -0.10 0.51 0.71 R - Global High Yield -0.07 0.17 0.81 0.73 0.67 EM Equities -0.05 -0.04 0.68 0.69 0.71 0.77 Global Equities -0.03 -0.03 0.58 0.75 0.83 0.75 0.82 FI - EM Public Real US Small-Cap FI - Global Cash (US$) FI - Global Agg Sovereigns Estate High Yeld BM Equities Source: World Bank. Note: Agg = aggregate; EM = emerging market; FI = fixed income; HY = high yield. Reference portfolio returns are calculated using monthly rebalancing. 22 Another option would be to use the MSCI ACWI (All Country World Index), which also includes Emerging Markets equities. In practice, MSCI World and MSCI ACWI have had nearly identical return profiles. 23 CEM Benchmarking’s methodology accounts for reporting lags for illiquid assets in private markets. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 56 Scaling the approach to different risk and return tolerances An advantage of this approach is that it can easily be scaled to different risk tolerances by combining different levels of income assets and the diversified risk basket. At the margins the diversified risk 0 percent portfolio is made up entirely of income assets basket (Bloomberg Barclays Global Aggregate), and the diversified risk 100 percent portfolio is made up entirely of the diversified risk basket. Iterations in between are made up of mixes of the returns of these two baskets, rebalanced monthly. For example, as illustrated below, the diversified risk 60 percent portfolio contains a 60 percent weight for the diversified risk basket, and a 40 percent weight for the income assets basket. Tables A.4 and A.5 present asset composition of these reference portfolio for different risk levels and their historic risk and returns characteristics. > > > T A B L E A . 4 . - Asset Composition of Reference Portfolios for Different Risk Levels Diversified Diversified Diversified Diversified Diversified Diversified Risk 0% Risk 20% Risk 40% Risk 60% Risk 80% Risk 100% Diversified Risk Basket % 0% 20% 40% 60% 80% 100% Overall Mix Income Basket % 100% 80% 60% 40% 20% 0% FI - Global Agg 100% 80% 60% 40% 20% 0% Global Equities 0% 14% 28% 42% 56% 70% Public Real Estate 0% 2% 4% 6% 8% 10% Asset Class Weights US Small Cap Stocks 0% 2% 4% 6% 8% 10% EM US$ Sovereign Debt 0% 2% 4% 6% 8% 10% Total 100% 100% 100% 100% 100% 100% Inccome Assets 100% 100% 100% 100% 100% 100% Income Basket FI - Global Agg 100% 100% 100% 100% 100% 100% Diversified Risk Basket 100% 100% 100% 100% 100% 100% Global Equities 70% 70% 70% 70% 70% 70% Risk Basket Public Real Estate 10% 10% 10% 10% 10% 10% US Small Cap Stocks 10% 10% 10% 10% 10% 10% EM US$ Sovereign Debt 10% 10% 10% 10% 10% 10% Source: World Bank. Note: Agg = aggregate; EM = emerging market; FI = fixed income. > > > T A B L E A . 5 . - Risk and Return Characteristics for Income Assets, Risk Assets, and Reference Portfolios, December 1995–April 2020. Average Frequency of Average Minimum Maximum Asset Category annualized negative 12m risk 12m return 12m return returns return Cash (US$) Income Assets 3% 1% 0% 7% 0% FI - Global Agg Income Assets 5% 3% -1% 12% 2% Global Equities Risk Assets 6% 15% -47% 54% 26% Public Real Estate Risk Assets 7% 18% -62% 89% 21% US Small-Cap Equities Risk Assets 8% 20% -42% 64% 27% FI - EM Sovereigns Risk Assets 9% 11% -28% 43% 18% Passive 50% FI/50% Equity Reference Portfolios 6% 8% -25% 29% 18% Diversified Risk 20% Reference Portfolios 6% 4% -9% 16% 4% Diversified Risk 40% Reference Portfolios 6% 6% -19% 25% 12% Diversified Risk 60% Reference Portfolios 7% 9% -29% 35% 18% Diversified Risk 80% Reference Portfolios 7% 12% -38% 45% 23% Diversified Risk 100% Reference Portfolios 7% 14% -45% 56% 25% Source: World Bank. Note: Agg = aggregate; EM = emerging market; FI = fixed income. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 57 Calculating US$ and $A Returns for Reference Portfolios Because our objective it to use these reference portfolios to benchmark Pacific funds with both US$- and $A-base currencies, we need to adjust returns of the reference portfolios to $A when using for comparing returns of Kiribati’s Reserve Equalization Rev- enue Fund (RERF) and Tuvalu’s TTF. To calculate the returns in $A, we use the US$ returns for each index and translate these into $A using the $A currency returns for the month. The exception to this is the Bloomberg Barclays Global Aggregate, where we use the $A-hedged returns directly from Barclays as summarized in table A.6. > > > T A B L E A . 6 . - Treatment of $A Returns for Reference Portfolios Name Index Description Source $A Returns FI - Global AGG Barclays Global Aggregate Global IG bonds Barclays Uses the $A Hedged return provided by Barclays US Small-Cap Equities Russell 2000 US Small Cap Bloomberg Calculate $A returns from US$ returns Emerging Market US$ FI - EM Sovereigns JPM EMBIG Bloomberg Calculate $A returns from US$ returns Sovereign Bonds FTSE EPRA Nareit Global Public Real Estate Public Real Estate Bloomberg Calculate $A returns from US$ returns REITs Index Global Equities MSCI World Global Equities Bloomberg Calculate $A returns from US$ returns Source: World Bank. Note: $A = Australian dollar; Agg = aggregate; BBG = Bloomberg; EM = emerging market; FI = fixed income; FTSE EPRA = Financial Times Stock Exchange Euro- pean Public Real Estate; IG = investment grade; JPM EMBIG = JPMorgan Emerging Market Bond Index Global; m = month; REIT = real estate investment trusts. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 58 PENDIX B.Appendix B. Morningstar $A Multi-Sector Market Indices Key Benefits Multi-Sector Fund Performance Benchmarks These indices are available in Morningstar DirectTM, • Enable more accurate Morningstar calculates and publishes proxy market and on Morningstar’s proprietary fund profiles. and appropriate benchmarks for multi-sector funds. These Morningstar comparisons and Multi-Sector Market Indices incorporate weighted averages How They’re contrasts between the of existing market indices. Calculated funds in each multi- We use identical calculations to determine daily returns sector Morningstar These indices enable investors and their advisers to make for the five indices. We begin the calculation process by Category accurate and appropriate comparisons between multi-sector breaking each index into nine broad asset classes. We • These indices are funds and provide a meaningful performance benchmark for then assign a weighting to these asset classes on the basis available in Morningstar the five Morningstar multi-sector fund categories. of the category’s average asset DirectTM, and allocation, and then apply a widely-used market index to on Morningstar’s These five Multi-Sector Market Indices are: each broad asset class. Finally, we add these asset classes proprietary fund • Morningstar Aus Msec Conservative TR $A together to construct the daily return for each Multi-Sector profiles. (XIUSA04GIY) (ticker 95244) Market Index. Each category’s average asset allocation is • Morningstar Aus Msec Moderate TR $A updated twice-yearly (on 1 March and 1 September) and (XIUSA04GIZ) (ticker 95245) the market indices for each asset class are updated daily. • Morningstar Aus Msec Balanced TR $A (XIUSA04GJ0) (ticker 95246) • Morningstar Aus Msec Growth TR $A (XIUSA04GJ1) (ticker 95247) • Morningstar Aus Msec Aggressive TR $A (XIUSA04GJ2) (ticker 95248) C ontact Us Asset Class Components and Market Indices Contact Morningstar Asset Class Market Index Client Services on +61 2 9276 4446 Australian Shares M&P/ASX 300 TR or email International Shares MSCI World Ex Australia NR $A helpdesk.au@morningstar.com Australian Listed Property S&P/ASX 300 A-REIT TR International Listed Property FTSE EPRA/NAREIT Developed NR Hdg $A Unlisted Property S&P/ASX 300 A-REIT TR Australian Fixed Interest Bloomberg AusBond Composite 0+Y TR $A International Fixed Interest BarCap Global Aggregate TR Hdg $A Australian Cash RBA Bank accepted Bills 90 Days Multi-Sector Index Asset Allocation as at 31 December 2018 Asset Class Conservative Moderate Balanced Growth Aggressive Australian Shares 6 .11 11.44 21.95 29.03 36.73 International Shares 6.23 13.10 24.04 32.06 45.65 Australian Listed Property 1.47 2.00 2.29 2.78 2.24 International Listed Property 0.67 1.55 1.65 2.39 2.35 Unlisted Property 0.63 1.09 1.24 2.71 0.89 Australian Fixed Interest 26.41 28.63 20.46 12.37 3.82 International Fixed Interest 26.55 16.69 13.61 8.85 3.14 Australian Cash 31.92 25.49 14.75 9.80 5.18 © 2018 Morningstar, Inc. All rights reserved. Neither Morningstar, its affiliates, nor the content providers guarantee the data or content contained herein to be accu- rate, complete or timely nor will they have any liability for its use or distribution. Any general advice or ‘class service’ have been prepared by Morningstar Australasia Pty Ltd (ABN: 95 090 665 544, AFSL: 240892) and/or Morningstar Research Ltd, subsidiaries of Morningstar, Inc, without reference to your objectives, financial situation or needs. Refer to our Financial Services Guide (FSG) for more information at www.morningstar.com.au/s/fsg.pdf. You should consider the advice in light of these matters and if applicable, the relevant Product Disclosure Statement before making any decision to invest. Our publications, ratings and products should be viewed as an additional investment resource, not as your sole source of information. Past performance does not necessarily indicate a financial product’s future performance. To obtain advice tailored to your situation, contact a professional financial adviser. Some material is copyright and published under licence from ASX Operations Pty Ltd ACN 004 523 782. Source: Morningstar, https://www.morningstar.com/content/dam/marketing/apac/au/pdfs/Legal/MultiSector_IndicesAU_2018-06.pdf. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 60 PENDIX C.Appendix C. Total Costs and Asset Managers Fees for Federated States of Micronesia and Marshall Islands Trust Funds Information in the tables in appendix C comes from the FY19 Annual Reports of the Marshall Islands and the Federated States of Micronesia (Trust Fund for the People of the Republic of the Marshall Islands 2020; Trust Fund for the People of the Federated States of Micronesia 2020). > > > TA B L E C .1 . - Federated States of Micronesia Trust Fund’s Investment and Administrative Costs, FY18 and FY19 FY18 FY19 % Change Investment Expenses $5,013,172 $4,930,063 -1.7% Custodian 141,639 149,990 5.9% Investment Adviser 253,851 1,119,368 374.6% Money Manager 1,005,409 1,214,399 20.8% Money Manager* 3,630,273 2,446,206 -32.6% Investment Adviser - Indirectly Billed 1,201,403 0 -100.0% Managers - Directly Billed 1,283,651 944,975 -26.4% Managers - at NAV (estimated) 1,145,219 1,501,331 31.% % of Net Position 0.79% 0.72% -9.3% Administrative Expenses $172,844 $171,894 -0.5% Executive Administrator 110,460 110,898 0.4% Audit Fees 44,833 46,163 3.0% Accounting Fees 5,484 5,983 9.1% Legal Fees $940 $940 0.0% Miscellaneous Fees** 10,676 7,784 -27.1% % of Net Position 0.03% 0.02% -8.2% Total expenses $5,186,016 $5,101,957 -1.6% Total % of Net Position 0.82% 0.74% -9.2% Note: Amounts rounded to nearest dollar. Souce: FY18 and FY19 Audited Annual Financial Statements and Fund Data Collection. *Money Manager fees directly subtracted from individual fund asset values. Note change in investment expense categories based on revised fee accounting and invoicing to further breakdown fees between Investment Adviser and Money Managers and improve transparency. > > > TA B L E C . 2 . - Marshall Islands Trust Fund’s Investment and Administrative Costs, FY18 and FY19 FY18 FY19 % Change Investment Expenses $1,563,073 $1,670,636 6.9% Custodian - (122) 0.0% Investment Adviser 118,045 122,088 3.4% Money Manager* 1,445,028 1,548,671 7,2% % of Net Position 0.39% 0.38% -1.0% Administrative Expenses $159,380 $153,090 -3.9% Executive Administrator 108,366 109,556 1.1% Audit Fees 41,204 35,000 -15.1% Accounting Fees 3,980 2,411 -39.4% Legal Fees 559 127 -77.3% Miscellaneous Fees** 4,330 5,053 16.7% % of Net Position 0.04% 0.04% -11.1% Total Dollar Expenses $1,722,452 $1,823,726 5.9% Total % of Net Position 0.43% 0.42% -2.0% Souce: FY18 and FY19 Audited Annual Financial Statements and Fund Data Collection. Note: Custodian line includes a $122.10 fee reimbursement made by State Street, the Fund’s former Custodian, in FY19 for a FY17 revised fee calculation. Note: Amounts rounded to nearest dollar. *Money Manager fees directly subtracted from individual fund asset values. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 62 > > > T A B L E C . 3 . - Federated States of Micronesia Trust Fund’s Asset Managers’ Fees, FY18 and FY19 Portfolio Assets Total Expense Ratio in Basis Points Managed Assets Equity Domestic SSgA S&P 500 Index 0.6 SSgA Russell 2000 Index 2.7 International Mercer Non-US Core Equity 42 SSgA Emerging Markets Index 12 Global Acadian Global Managed Volatility 20 Fixed Income Mercer Opportunistic Fixed Income 44 Mercer Core Bond 18 Alternatives Mercer Hedge Fund Investors 150 Mercer Private Investment Partners III 241 Mercer Private Investment Partners IV 250 Mercer Private Investment Partners V 235 Non-Discretionary Assets Real State Prudential Real Estate Investors 93 Alternatives HarbourVest Partners VIII Venture Fund 265 HarbourVest Partners VIII Buyout Fund 265 HarbourVest Partners VIII Mezzanine & Distressed Fund 265 HarbourVest Partners VIII Inti Private Equity Fund 265 Portfolio Advisors Private Equity Fund IV 270 Mercer Managed Asset Fee (per tiered fee schedule) 18 Mercer Non-Discretionary Fee 0 Total Fee Rate 69 Souce: Trust Fund for the People of the Federated States of Micronesia, “FY2019 Annual Report,” 2020. Note: Total fees are an estimate based on 9/30/19 asset values and assumptions around application of fees for the alternative allocations. Does not include carried interest. Total fees will vary over time based on the fee schedule. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 63 > > > T A B L E C . 4 . - Marshall Islands Trust Fund’s Asset Managers’ Fees, FY18 and FY19 Portfolio Assets Total Expense Ratio in % Equity Domestic Vanguard® Total Stock Market Index Fund Institutional Shares 0.030 Vanguard® PRIMECAP Fund AdmiralTM Shares 0.310 Vanguard® WindsorTM Fund AdmiralTM Shares 0.210 International Vanguard® Total International Stock Index Fund Institutional Shares 0.080 Vanguard® International Growth Fund AdmiralTM Shares 0.320 Vanguard® International Value Fund 0.380 Fixed Income Vanguard® Total Bond Market Index Fund Institutional Shares 0.035 Vanguard® Short-Term Investment-Grade Fund Institutional Shares 0.100 Vanguard® Inter-Term Investment Grade Fund AdmiralTM Shares 0.070 Alternatives Mercer Private Equity Partners III 2.41 Mercer Private Equity Partners IV 2,50 Investment Advisory Fee Total 0.39 Souce: Trust Fund for the People of the Republic of the Marshall Islands, “Annual Report FY2019,” 2020. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 64 PENDIX D.Appendix D. Do Active Managers Add Value Net of Fees Over Time? >>> Appendix D. Do Active Managers Add Value Net of Fees Over Time? The aim of active managers, over time, is to dynamically allocate assets between risk assets and defensive assets to capture greater upside and buffer downside risk, whether on a specific benchmark or on the total return. If a manager is successful in delivering (excess) returns and does it consistently, this manager has demonstrated a skill in that particular investment strategy. The role of investors is (a) to select managers that have demonstrated persistent skill, (b) to monitor their performance on ongoing basis, (c) to evaluate whether the managers continue to meet their agreed performance metrics, and (d) to terminate the manager if the manager does not meet the metrics. There is extensive academic literature on fundamental challenges for investment managers to persistently beat the market after fees. In 2019, Bloomberg reported that only 17.9 percent of US large-cap equity funds outperformed the S&P 500, and only 19.8 percent European equity funds beat their benchmark (figure D.1). Even if some managers demonstrate excess performance under specific conditions, persistence of outperformance is extremely low as quantified in a study by the World Bank Treasury. For the sample of 86 active equity fund managers in the top quartile of performance from 1999–2004, there is very low persistence of outperformance over the subsequent 10 years. More managers migrated to the bottom quartile of performance in the subsequent period than remained in the top quartile. By the third period (2009–2014), only 3 of the initial 86 managers remained in the top quartile (figure D2). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 66 > > > F I G U R E D . 1 . - Percentage of Actively Managed Funds That Outperformed the Market US large-cap funds 17.9% 82.1% Europe equity funds 19.8% 80.2% Source: Bloomberg. Note: Based on five-year returns through December 31, 2018, relative to the S&P 500 and S&P Europe 350. > > > F I G U R E D . 2 . - Migration of Top Quartile Managers Over Investment Horizon 1999—2004 2004—2009 2009—2014 86 17 3 15 5 27 3 27 6 Source: World Bank calculations. There are three potentially overlapping explanations for this • Behavioral biases mean that managers can have nega- phenomenon. The first reason is context-specific skill. An ac- tive skill. Behavioral economics has catalogued a series tive manager may, in fact, have real skill to profitably navigate a of cognitive biases, such as loss aversion, that may lead particular market environment. When that specific market envi- investors to make choices that are actively worse than ronment changes, the advantage that the manager has may go simply making a random choice (Kahneman 2003). Man- away. The second reason is market efficiency. The third expla- agers may have negative skill that actively destroys value. nation is that the outperformance may be due to luck, not skill. • Fooled by randomness. An active investment manager who has a positive track record may be skilled or may Practical implications when evaluating whether to select ac- simply be lucky (Taleb 2004). Given the probability distri- tive mandates include the following: bution, it is very difficult to know whether modest levels of • Active management returns are only a fraction of the outperformance are due to luck or to skill. overall returns. While the expectation is that skilled ac- • Cost of monitoring active managers. A fund employing tive managers can add value over time, these returns are active managers will need to spend money and resourc- modest in size compared with the magnitude of the over- es managing their active managers and replacing them all returns because of setting the (benchmark) risk level if they persistently underperform. In contrast, a passive as illustrated on figure 4.2 and because the managers mandate has high certainty to replicate an index’s return lack persistency in their performance results. (with a small level of tracking error) for a low cost. • (High) fees erode value from active management. The • It is better to pursue active managers in markets that higher the level of a manager’s fees, the higher the are inefficient (for example, private markets in general) level of skill necessary to generate net returns above wherein the average manager can still likely add value. the benchmark. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 67 PENDIX E.Appendix E. Comparison with Endowment Model >>> Appendix E. Comparison with Endowment Model The five funds selected for the study are designed as perpetual funds with a common objective of balancing distribution of their income for today’s spending while preserving equity across gen- erations. University endowments funded by financial gifts from alumni or other donors play a vital role in helping to support academic institutions’ long-term missions and goals.24 Trustees, chief business officers, and other decision-makers have a fiduciary duty to develop investment strate- gies and spending policies that best position their endowments to meet the long-term objectives of their institutions. For most, that means achieving two goals: (a) ensuring that the endowment provides appropriate resources for current operations and (b) preserving, on an inflation-adjust- ed basis, the purchasing power of the endowment in support of future generations. As a result, some Pacific funds, Kiribati and Tuvalu in particular, have been exploring whether they should explicitly adopt an endowment model. These funds encountered some challenges when considering this alternative and how to operationalize it. For example, the most typical examples of an endowment model are sophisticated institutions with large endowments, such as Harvard University (~US$37 billion), Yale University (~US$27 billion), Stanford University (~US$25 billion) and Princeton University (~US$23 billion). Whereas these endowments have certainly been the leaders in the field of endowment investment, a significant number of univer- sities and colleges have much smaller endowments; many of them have endowments that are comparable in size or even smaller than Pacific funds, thereby making them a relevant compara- tor to learn from. This section presents the key results from the 2018 study of US University and Colleges endowments undertaken by National Association of College and University Business Officers (NACUBO)25 and Teachers Insurance and Annuity Association of America (TIAA)26 that is based on the analysis of 802 institutions managing endowments with a range of assets under management from US$25 million to more than US$1 billion, making it the largest and most com- prehensive study of its kind (NACUBO-TIAA 2018). 24 For further information on this topic, see NACUBO-TIAA (2018; 2019). 25 NACUBO is a membership organization representing more than 1,900 institutions, www.nacubo.org. 26 TIAA is a Fortune 100 financial services organization that is the provider of financial services in academic, research, medical, cultural, and government fields www.tiaa.org. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 69 The data collected by the survey are used to evaluate endow- purposes, the study offers a wealth of insights into key prac- ment investment returns, asset allocation, and investment tices employed by endowments and on the range of invest- strategies, as well as governance and management issues. ment outcomes across these institutions. Table E.1 presents The data collected by the study are an important indicator of the summary of investment returns over different investment the extent to which campus leaders are managing their en- horizons (1, 3, 5, and 10 years) for portfolios of different sizes. dowed funds in ways that achieve institutional goals. For our > > > T A B L E E . 1 . - FY19 Net Investment Returns for Different Size Endowments Over $500 Over $250 Over $100 Over $50 Over $25 Total Over $25 million million— $1 million—$500 million—$250 million—$100 million—$50 institutions $1 billion and under billion million million million million 1-year (FY19) 5.3% 5.9% 5.1% 5.0% 5.1% 4.9% 5.5% 5.8% 3-year 8.7% 9.6% 8.9% 8.9% 8.5% 8.3% 8.3% 8.3% 5-year 5.2% 6.1% 5.1% 5.3% 5.0% 4.9% 4.9% 5.5% 10-year 8.4% 9.0% 8.5% 8.4% 8.3% 8.2% 8.4% 7.7% Source: NACUBO-TIAA 2018, 2019). Consistent with the long-term nature of endowments’ investment portfolios, their annual returns fluctuate from year to year depend- ing on the market environment (figure E.1). Consistent with their long-term investment horizon, the endowments exhibit negative annual returns during times of market downturns. However, over a longer investment horizon (10 years) these portfolios benefit from being invested in the full range of investment choices (figure E.2). > > > F I G U R E E . 1 . - Average Annual Endowment Returns, FY10–FY19 20% 19.2% 15% 15.5% 11.9% 11.7% 12.2% 10% 8.2% 5.3% 5% 2.4% 0% -0.3% -1.9% -5% FY10 FY11 FY12 FY13 FY14 FY15 FY16 FY17 FY18 FY19 Source: NACUBO-TIAA 2018, 2019. > > > F I G U R E E . 2 . - Average 10-year Endowment Returns, FY10–FY19 10% 7.7% 7.7% 7.4% 7.4% 7.4% 7.3% 7.2% 7.1% 7.0% 7.2% 8% 6% 4% 2% 3.4% 5.6% 6.2% 7.1% 7.1% 6.3% 5.0% 4.6% 5.8% 8.4% 0% FY10 FY11 FY12 FY13 FY14 FY15 FY16 FY17 FY18 FY19 Source: NACUBO-TIAA 2018, 2019. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 70 As discussed throughout the note, the relative performance The largest difference across endowments of different sizes is of investment institutions depends on many variables, includ- their allocation to alternative strategies with a more significant ing their appetite for risk, liquidity preferences, asset alloca- increase in allocation to the alternatives for the larger endow- tion, and portfolio implementation, among other factors. Fur- ments: endowments of more than US$1 billion allocate about thermore, as we discussed in relation to the Pacific funds, the 58 percent to these strategies, endowments of US$101 mil- absolute size of institutional portfolios also has a noticeable lion to US$250 million allocate about 27 percent, whereas the effect on how these funds seek to achieve their investment ones below US$25 million about 11 percent. Such differenti- goals. Because university endowments cover a broad range ated approach to alternative strategies by smaller funds is not of portfolio sizes—from under US$25 million to into tens of surprising as returns on these strategies depends significantly billions of dollars—it is illustrative how these endowments allo- on the size of the allocation. On average endowments enjoyed cate across different asset classes and, in particular, between 8.3 percent return on their alternative investments, return for risky and income (defensive) assets. this asset class was 11.2 percent for endowments of more than US$1 billion and only 4.5 percent for the smallest funds, As table E.2 illustrates, allocation to risky assets such as US as detailed in table E.3. From the same table, larger funds equities, non-US equities and alternative strategies, increase enjoy quite a bit higher returns (9.7 percent) than the entire for larger size endowments. Thus, endowments of more than universe of endowments on average (8.2 percent) with returns US$1 billion hold about 90 percent in risky assets, endowments ranging from 7.6 percent for the smallest endowments to 8.7 of US$101 million to US$250 million hold about 80 percent percent for endowments up to US$1 billion. and endowments of below US$25 million about 70 percent. > > > T A B L E E . 2 . - FY18 Asset Allocation for Different Size Endowments US$501 US$251— US$101— US$51— US$25— Under Total Over million— US$500 US$250 US$100 US$50 US$25 institutions US$1 billion US$1 billion million million million million million Responding institutions 800 104 84 88 195 154 103 72 US. equities 16% 13% 22% 24% 31% 34% 39% 45% Non-US. equities 20% 19% 22% 22% 22% 22% 18% 15% Fixed income 8% 7% 10% 12% 15% 19% 22% 24% Alternative strategies 53% 58% 41% 38% 27% 22% 16% 11% Short-term securities/cash/other 3% 3% 5% 4% 5% 3% 5% 5% Note: Amounts rounded to nearest dollar. Souce: FY18 and FY19 Audited Annual Financial Statements and Fund Data Collection. *Money Manager fees directly subtracted from individual fund asset values. Note change in investment expense categories based on revised fee accounting and invoicing to further breakdown fees between Investment Adviser and Money Managers and improve transparency. > > > T A B L E E . 3 . - FY18 Asset Class Investment Returns for Different Size Endowments US$501 US$251— US$101— US$51— US$25— Under Total Over million— US$500 US$250 US$100 US$50 US$25 institutions US$1 billion US$1 billion million million million million million Responding institutions 800 104 84 88 195 154 103 72 Percentage of institutions reporting 100.0% 13.0% 10.5% 11.0% 24.4% 19.3% 12.9% 9.0% Total net return 8.2% 9.7% 10.5% 11.0% 24.4% 19.3% 12.9% 9.0% US. equities 13.6% 13.0% 14.3% 13.5% 13.8% 13.3% 13.9% 12.8% Non-US. equities 6.8% 8.1% 7.0% 7.0% 6.6% 6.1% 6.8% 6.7% Fixed income 0.5% 1.3% 0.4% 0.7% 0.1% 0.5% 0.5% 0.3% Alternative strategies 8.3% 11.2% 9.9% 9.0% 9.0% 7.1% 5.6% 4.5% Source: NACUBO-TIAA 2018, 2019. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 71 This study also provided insights into how individual differ- cohort with 9 percent difference between top and bottom 5th ences within each institution managing their respective en- percentiles: 12.6 percent versus 3.5 percent). These differenc- dowment affect investment outcomes. More specifically, table es could be caused by both market factors (that is, different E.4 presents the range of investment outcomes in different exposure to risky assets owing to differences in risk prefer- size levels. Thus, on the aggregate, the top quartile of endow- ences or liquidity requirements) and by skill factors (that is, ments returned 9.1 percent, whereas the bottom quartile’s re- different levels of success in implementing particular invest- turns were 2 percent lower at 7.1 percent in FY18. The differ- ment strategies owing to a number of factors, not least be- ence is greater between top 5th percentile that generated 11.6 cause larger funds typically enjoy more financial and technical percent versus the bottom 5th percentile that generated 5.6 resources). Information presented in the study does not allow percent. This difference appears to be broadly similar across for disaggregation of these factors. all endowment-size levels (with the exception for the smallest > > > T A B L E E . 4 . - FY18 Variability of Investment Returns for Different Size Endowments US$501 US$251— US$101— US$51— US$25— Under Total Over million— US$500 US$250 US$100 US$50 US$25 institutions US$1 billion US$1 billion million million million million million Total institutions 802 104 85 88 195 154 103 73 75th percentile 9.1% 10.9% 9.4% 9.1% 8.8% 8.6% 8.7% 8.6% 50th percentile (median) 8.0% 9.6% 8.6% 8.1% 7.7% 7.6% 7.8% 7.9% 25th percentile 7.1% 8.4% 7.6% 7.5% 7.1% 7.1% 6.4% 6.3% Percentiles 95th percentile 11.6% 12.9% 12.4% 11.7% 9.9% 9.9% 10.2% 12.6% 5th percentile 5.6% 6.7% 6.0% 6.6% 5.5% 5.5% 5.1% 3.5% Source: NACUBO-TIAA 2018, 2019. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 72