Emerging-market debt: Navigating the allocation dilemma

Transcription

1 EATON VANCE TOPIC PAPER MARCH 2017 Emerging-market debt: Navigating the allocation dilemma Michael A. Cirami, CFA Vice President Co-Director Global Income Group Matthew F. Murphy, Jr., CFA, CAIA Vice President Institutional Portfolio Manager Global Income Group Positive fundamental changes in emerging markets over the past years, along with robust economic growth and the proliferation of emerging-market debt (EMD) indices, have helped transform EMD into a mainstream asset class. Growing interest in EMD has led to increased resources dedicated to these assets and expanded access for investors, which, in turn, has added to the complexity of making allocation decisions relating to this asset class. Investors endeavours to blend local currency sovereign debt, corporate debt and off-benchmark country exposures with traditional external (hard currency) sovereign debt investments have, to date, had mixed success. The predominant allocation approach among investors looking for broader access to the EMD opportunity set a purely top-down blend of the three main EMD indices (whether allowing for tactical weighting variations or not) seems to resonate with pension fund boards, but is not without inherent flaws. Eaton Vance believes investors should adopt an active, unconstrained approach to EMD that focuses on country-specific and detailed risk factor-focused analysis across the entire tradable universe. A shift to a risk factor-focused approach requires, on the part of the end investor, a new mindset. For the investment manager, a risk factor-focused approach requires specialist investment expertise and an extensive analytical capability.

2 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 2 Introduction Emerging-market (EM) economies and their capital markets continue to have strategic significance for institutional investors. EM economies today account for around 58% of global GDP and their aggregate growth rate remains way ahead of that for advanced economies. 1 According to International Monetary Fund (IMF) forecasts, EM economies will grow at more than double the pace of advanced economies in 2017 (4.6% versus 1.8%). 2 Indicative of the growing importance of EM economies globally, the Chinese renminbi last year joined the elite ranks of the US dollar, euro, yen and UK pound as an IMF global reserve currency. The fundamentals and credit quality of the EMD market overall are considerably better today compared to what was the case in the 1980s and 1990s. Improvements have come by way of better fiscal management, more effective central bank policies, a strengthened financial framework, robust economic growth, expanded local debt markets (reducing countries vulnerability to external shocks) and a shift away from currency pegs. Looking ahead, EM debt markets are expected to evolve and diversify further as EM nominal GDP increases, capital accounts gradually liberalise, local pension funds continue to develop and new sub-asset classes emerge (e.g., China s municipal bond market). Today, institutional investors are well versed in the arguments typically advanced for investing in emergingmarket debt (EMD); not least, arguments about diversification and yield pickup. However, in reality, interest in this asset class is often tempered by the challenges institutional investors experience in deciding on a correct allocation strategy and the perception that an allocation to this asset class appears to demand a disproportionately large proportion of their risk budget for what is usually a small allocation. With this in mind, in this paper we will endeavour to examine the pros and cons of different allocation approaches to this asset class. What many investors don t know is that EMD investing does not need to be as volatile as the benchmarks suggest. Sources of return within this asset class can be more diversified (e.g., liquidity, manager skill, administrative/ operational), yield and returns can be relatively higher, and correlation with an existing portfolio can be lower. Growing interest in EM debt markets Over the past 15 years, interest in EM debt investing among pension funds and other institutional investors has picked up considerably. Part of this change has had to do with the launch of several EM debt indices since the early 1990s, as shown in the following abridged timeline : JPMorgan Emerging Market Bond Index (EMBI) Government US dollar-denominated debt, originally composed only of Brady bonds; debt whose principal was backed by the US. 1993: JPMorgan Emerging Market Bond Index+ (EMBI+) The first expansion of the index beyond Brady bonds. 1999: JPMorgan Emerging Market Bond Index Global (EMBIG) Government US dollar-denominated debt, with relaxed criteria to include more issuers than in the EMBI+. Currently the most commonly used EMBI benchmark. 2005: JPMorgan Government Bond Index-Emerging Markets (GBI-EM) Local currency sovereign debt. 2007: JPMorgan Corporate Emerging Markets Bond Index (CEMBI) Hard currency corporate debt. 2011: JPMorgan Next Generation Markets Index (NEXGEM) Hard currency debt of lower-rated sovereign issuers. The proliferation in EMD benchmarks has made this asset class more accessible to investors. It has also encouraged interest in not only US dollar-denominated bonds issued by governments in developing economies, but local currency issues, too. Positive trends in emerging economies have played a part. These include better macroeconomic and public debt 1 Source: EconomyWatch.com, Economic Statistics database, 6 January Source: IMF World Economic Outlook, October 2016, page 2. GDP forecasts are on a purchasing power parity basis.

3 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 3 management, improved inflation records, improved protection for creditors, strong economic performance and the growth of local institutional investors such as pension funds and insurance companies. For foreign investors, an additional attraction of holding local currency EMD has been the potential for FX gains over time. Governments of developing countries have also warmed to local currency issuance: By shifting a greater portion of their funding to local currency instruments they have been able to reduce their vulnerability to external economic and financial shocks. 3 What should a benchmark be used for? This rapid expansion in EMD benchmarks has been both positive and challenging. On the one hand, it has increased awareness among foreign investors of a larger and more diverse opportunity set offering new risk factors: currency exposure, local rates and corporate credit. On the other hand, it has added to the complexity of decisionmaking in terms of investment strategy, governance, execution, operational issues and cost. So, what should a benchmark be used for? In our view, benchmarks can be useful from a fiduciary perspective (e.g., to monitor manager performance), but should not be used as a reference point for determining an investment approach to this asset class. There are a number of reasons why indices are a poor starting point for allocation decisions. Indices, as discussed in more detail later on, are a poor representation of the true associated opportunity set. An index-referenced allocation also undermines the presumed goal of efficient capture of the best risk-adjusted return opportunities because it typically includes exposure to parts of the index that are relatively less attractive. There are also considerable disparities in investment opportunity within the EMD universe at any given point in time; disparities that argue in favour of a much more surgical approach to this asset class. As an example, consider the year-over-year output growth prospects among developing countries. Brazil (0.5%), Nigeria (0.6%) and South Africa (0.8%) will lag far behind India (7.6%) and China (6.2%) in 2017, according to the IMF s October 2016 World Economic Outlook report. 4 Similarly, if one looks at EM bonds in terms of their component risk factors currency, interest rates, sovereign credit spreads and corporate credit spreads you will see substantial differences across countries. An example of this is shown in Exhibit A, which highlights major differences in nominal interest rates across various countries as at the end of Exhibit A Major differences in five-year nominal interest rates. 25 PERCENT (%) Zambia Ghana Tanzania Egypt Uganda Ukraine Nigeria Kenya Sri Lanka Rwanda Turkey Brazil Dominican Republic Honduras Georgia South Africa Russia Indonesia Mexico Pakistan Colombia India Costa Rica Senegal Cape Verde Serbia Vietnam Barbados Bangladesh Albania Peru Lebanon Philippines Malaysia Mauritius Chile Saudi Arabia Jordan UAE - Abu Dhabi Poland Romania China Qatar Macedonia Morocco Thailand Singapore South Korea Hungary Croatia Bulgaria Taiwan Czech Republic Source: Eaton Vance proprietary data and calculations, as of 31 December Data provided are for informational use only. Past performance is no guarantee of future results. 3 Source: The Evolution of Emerging Market Sovereign Debt: Dramatic Growth in Local Currency Sovereign Debt Is Reducing Emerging Market Financial Vulnerabilities, Moody s Investors Services, published 2 September Source: IMF World Economic Outlook, October 2016, page 2. GDP forecasts are on a purchasing power parity basis.

4 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 4 Allocating to EM debt: The status quo Currently, pension fund boards that allocate to EM debt typically do so with reference to the composition of one or more indices. Mandates are usually either passive, or a variant on passive (e.g., rebalancing index exposures according to an a priori allocation decision), or constrained active, where an active investment manager s ability to deviate from a benchmark is limited. Constraints could, for example, take the form of narrow tracking error limits and limited flexibility to allocate to off-benchmark issues. The prevalence of index-like or index-constrained strategies appears to reflect both the complexity of the asset class and pension fund governance practice. Asset studies that precede portfolio recommendations to the board of directors of a pension fund tend to start with analyses of different indices. Analysis of the return and risk behaviour of an index or indices becomes, almost imperceptibly, the reference point for allocation decisions and, in turn, a benchmark for the management of that allocation. The conflation of an index (or indices) and a benchmark then defines, to a considerable degree, the investment universe for that allocation. EM fixed-income indices afford ready understanding and peace of mind for boards seeking proper execution of their fiduciary duties. Despite flaws in EM debt indices (discussed later on), it is very seldom that pension fund boards approve a mandate with a fully unconstrained benchmark. A further explanation for the preponderance of index-based strategies could be difficulties pension funds have experienced in finding a sufficient lineup of active investment managers having the expertise and capability to undertake detailed country-by-country analysis and to trade off-benchmark exposures in a cost-efficient manner. Admittedly, successful management of the underlying risk factors in this asset class demands rather unique skills. Nonetheless, asset owners who adopt an index-based approach should bear in mind that what they are effectively doing is turning over investment decisions to an index provider, typically an investment bank, which has its own interests at heart. Typically, index providers cater to the largest common denominator; they want the largest investment managers to be able to replicate the index and focus essentially on those countries in which they have business operations. There is often no compelling investment reason why certain countries are excluded from an index. EM local currency indices such as the JP Morgan GBI-EM Global Diversified Index are an example. Local currency markets, which are operationally more difficult to trade than hard currency EM bonds, are poorly covered in the GBI-EM GD Index. Exclusion affects not only smaller markets. Larger markets are not necessarily less volatile, easier to analyse or easier to trade. Flaws in the index-based allocation approach As mentioned previously, we believe index-constrained approaches are not the best choice. As can be seen in Exhibit B, the EM tradable debt universe remains much larger than what is implied by the main indices in this asset class. Limiting the investment scope of a portfolio manager to the constituents of a benchmark index, or a small deviation from those constituents, can hamper portfolio returns over time. The portfolio will be unable, or less able, to capture attractive off-benchmark opportunities and may also be at risk of having to hold meaningful positions in Exhibit B The tradable EM debt universe is much larger than EM indices suggest. Local Sovereign External Sovereign Corporate Loans JPM GBI EM Global Diversified Broad Universe JPM EMBI Global Diversified Broad Universe JPM CEMBI Broad Diversified Broad Universe N/A Broad Universe Countries N/A 40 Market Value $715 bn $2 + tn $420 bn $1 + tn $377 bn $700+ bn N/A $400+ bn Sources: Eaton Vance and JP Morgan as of 31 December 2016.

5 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 5 markets and securities whose fundamentals or relative valuations are deteriorating. Another drawback is the tendency of index-based approaches to diminish the often touted diversification benefits of investing in EMD. This assertion has its basis in Eaton Vance research conducted in February It shows that investors seeking exposure to local currency debt would likely experience higher correlations between local currencies if they used the JPM GBI EM Global Diversified Index as the basis of their local debt allocation rather than an index-unconstrained approach. This research looked at rolling 36-month median cross correlation figures over a 10-year period to 30 November 2016 for all local currencies in the EMD universe. It focused specifically on the following currency pairs: Currencies in the JPM GBI EM Global Diversified Index benchmark. Currencies outside of the benchmark (109 currencies). Currencies inside the benchmark versus those outside the benchmark. The findings of this research are shown in Exhibit C. They highlight the point that while, as one might expect, correlations change over time, the power of diversification can be far more pronounced when index constraints are removed. The shift to blended allocations As awareness about the breadth of opportunities in the EM debt universe has grown, so, too, has awareness about the potential benefits of a more inclusive approach (i.e., complementing hard currency sovereign exposures with positions in local currency, corporate and frontiermarket debt). In principle, a broader, more inclusive approach would likely deliver diversification benefits and relatively better risk-adjusted returns over time. Blended allocations take a number of forms. One of these entails an allocation at the board level to separately managed individual sleeves, with each sleeve treated as either a strategic or opportunistic exposure. Another entails delegation to a single investment manager operating within an agreed blended allocation framework. Among large US public pension plans, those that allocate to EM debt (usually below 5% of total plan assets) have sought to incorporate local currency debt in different ways. 5 Some plans treat it as an opportunistic exposure, some leave currency exposures unhedged as part of their policy on foreign investments, some fully hedge back to US dollars and some will partially hedge subject to an upper US dollar exposure limit for their overall fixed-income portfolio. Current blended approaches: A big-picture, top-down theme Notwithstanding the above-mentioned variations, the common feature of most blended approaches currently is their embodiment of a big-picture, top-down allocation approach. Exhibit C Local currency benchmark offers less currency diversification. 0.8 Benchmark Off Benchmark Benchmark versus off-benchmark 0.7 CORRELATION '05 '06 '07 '08 '09 '10 '11 '12 '13 '14 '15 '16 Source: Eaton Vance as of 30 November The chart shows median rolling 36-month cross correlations for (1) currency pairs within the JPM GBI EM Global Diversified Index; (2) currency pairs outside this index and: (3) index and nonindex currency pairs. The median figures in this analysis closely approximate the averages. 5 Source: World Bank Treasury Report, Emerging Markets Local Currency Debt and Foreign Investors, November 2014.

6 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 6 A top-down blended approach, in our view, can represent a step forward for investors who want to capture a much broader opportunity set. According to research by Danish investment consulting firm Kirstein A/S, a static, equally weighted allocation to the three key subasset classes (hard currency sovereigns, as represented by the JP Morgan EMBI Global Diversified Index; local currency sovereigns, as represented by the JP Morgan GBI-EM Global Diversified Index and hard currency corporates, as represented by the JP Morgan CEMBI Broad Diversified Index) has the potential to deliver around 60% of the total annualised return that could be achieved if an investor were able to pick the best-performing subasset class in each calendar year. 6 The Kirstein A/S research is illustrated in Exhibit D. The three bars on the far right show annualised returns for the best calendar-year performances (January 2015 to July 2016), the worst calendar-year performances and a static, equally weighted blend of the three key subasset classes class over this period. Potentially, a purely top-down, index-based blend represents a conservative approach from a fiduciary standpoint. However, from an investment standpoint, it is not a defensive approach. It also has several drawbacks: The allocation is usually reviewed only once a year by the board. In reality, decisions made annually at a board level do not match the speed at which investment conditions and relative valuations change within this asset class. For example, allocation decisions made at the beginning of 2016 save in instances where the investment manager had been afforded meaningful discretion were unlikely to have been revisited immediately following Donald Trump s November US presidential election victory, even though his win raised the prospect of a stronger US dollar, upward pressure on Treasury yields and the possibility that the US Federal Reserve may raise US overnight interest rates at a pace faster than previously envisaged. Blending key EM debt asset classes via a big-picture, top-down approach typically entails a focus on key indices (i.e., EMBI GD, GBI-EM GD and CEMBI BD) with the result that, even allowing for tactical overweights or underweights, underlying allocations share the limitations inherent in these indices. Allocating across index-like exposures also raises other issues. Consider the following: Country risk factors are major drivers of asset performance, particularly for emerging-market countries. 7 Unfortunately, country representation across Exhibit D Static, index-based blend: A potentially conservative allocation approach. 40 ANNUAL RETURN (%) Top performing Worst Performing Blend YTD 11% 7% 2% Average Source: Kirstein A/S as of 31 July Exhibit shows the best and worst returns among the three sleeves (hard, local and corporate bonds) over the past decade, as well as the returns for a blended approach consisting of an equal-weighted allocation to each sleeve. Return figures for each sleeve are based on commonly used indices: the JP Morgan EMBI Global Diversified (hard currency sovereign debt), the JP Morgan GBI-EM Global Diversified (local currency government debt) and the JP Morgan CEMBI Broad Diversified (corporate debt). 6 Source: Emerging-market debt investing: A Nordic perspective, Kirstein A/S and Eaton Vance, November Sources: Baldacci, Gupta, & Mati, 2008; Beck, 2001; Eaton Vance, 2015; Heston & Rouwenhorst,1995; Rowland & Torres, 2004; Serra, 2000; Stocker, in press.

7 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 7 the popular indices varies greatly. For example, India is in the CEMBI, but not in the local index. Allocation across asset types impacts country allocations. Switching index-like exposures entails a meaningful shift in duration. GBI EM GD has a duration of around 4.5 years, while EMBI-GD has a duration of around seven years. Switches between external sovereigns and external corporates (assuming index-like exposures) will result in a change in US duration. Few investors are fully aware of the extent of developed-market risk within EM indices. (A bottom-up risk factor approach, however, would be very mindful of this.) When a country (e.g., Russia) is prominent in all three indices, the absolute level of risk concentration may be greater than optimal in the combined weightings of the three indices. Institutional investors also run the risk that submanagers may all overweight the same country. Further, there are often also important relative value distinctions between local currency and dollardenominated debt, and between sovereign and corporate issues. Such distinctions fall outside the scope of purely top-down asset allocation. Admittedly, as a matter of practicality, many pension fund guidelines call for the inclusion of a benchmark. In such cases, where investors want a blended approach, using a blended benchmark of something like 50% GBI-EM, 25% EMBI and 25% CEMBI which reflects the debt outstanding in the broader universe could be an option. However, we believe a benchmark should function largely as a reference for portfolio performance. Allocation strategies based principally on overweighting and underweighting common EM indices are, in our view, not optimal for investing in this sector. The future of blended: Active, unconstrained, risk factor-based Unconstrained investing offers access to a much broader range of investment opportunities, as shown in Exhibit E. At Eaton Vance, we prefer an index-unconstrained approach that focuses on country-level macroeconomic and political research across the entire investment opportunity set along with bottom-up analysis of specific risk factors. Bonds are instruments that can be broken down into their component risk factors: currency, interest rates, sovereign credit spreads and corporate credit spreads. Disaggregating and evaluating such idiosyncratic risk factors at the country level is, we believe, an approach that can be a consistent source of alpha. Exhibit E EM economic and demographic potential outside of the GBI-EM. GDP Population 1.2 billion $8.6 trillion $20.4 trillion 4.9 billion GBI-EM ex-gbi-em Source: International Monetary Fund, World Economic Outlook Database, October GDP and population data are for 2015.

8 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 8 Exhibit F provides a very simplified example of how country-specific risk factor analysis can be used to shape portfolio positioning. Unconstrained blended don t try this at home Investment management of an unconstrained, blended mandate is a demanding undertaking. It requires extensive investment research resources, a robust operational infrastructure, a dedicated EMD trading capability and strong risk management. Research. The value add of research in emerging markets is significant. Fewer analysts cover this universe compared to say, the S&P 500 Index universe. Fundamental research, particularly at the country level, is important in order to identify both the most attractive opportunities available and assess downside risk. An example of the latter is Mozambique. At the time of this writing (February 2017), the country, once considered one of Africa s potential success stories given its large natural gas reserves, was seeking a new IMF funding programme after missing a coupon payment in January 2017 on about US$727 million worth of eurobonds issued in April For private creditors, the response by the IMF is significant: The country must first secure new private credit arrangements so that its debt would no longer be deemed distressed. Increasingly, the IMF is becoming reluctant to bail out private creditors of sovereign debt. For investment managers, this means there is now an even greater onus on them to undertake in-depth, fundamental research and due diligence before committing capital to a particular country. Part of this research needs to focus on continually assessing governance standards and government credibility. Ghana is a case in point. The country is currently under an IMF programme with a stipulation for no buildup of arrears. Notwithstanding this stipulation, in January 2017, previously undisclosed spending arrears totalling US$1.6 billion came to light. These arrears could double the size of the country s budget deficit, more or less. Custody, account and trading infrastructure. A very strong operational infrastructure is essential in order to access the full range of local markets. Investment managers need a well-developed, sophisticated trading infrastructure and the ability to manage different accounts across more than 100 countries. Recordkeeping and document-processing functions, in particular, need to be well-resourced in order to cope with the sheer volume of documentation that needs to be created, tracked and kept up-to-date. Failure to keep documentation up-to-date can be costly. An example here is the losses some EMD investors in Nigeria have encountered as a result of not keeping their Certificate of Capital Importation (CCI) disclosures up-to-date and clearly aligned with activity in their banking and trading accounts. To get money into or out of Nigeria, investors need to be able to reconcile every trade with their CCI statement showing how much was initially brought into the country. This may sound straightforward, but because investors are Exhibit F Relative value among various risk factors. Country-specific outlook: Russia (January 2017) Low Conviction Neutral High Conviction Currency Local Interest Rates Sovereign Credit Corporate Credit Source: Eaton Vance Management as of 31 January This outlook is provided for informational use only. It should not be considered investment advice.

9 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 9 not required to repatriate the proceeds when they exit one investment and enter another, this often proves difficult to do. In our experience, few investors make the effort to ensure that their CCIs are fully up-to-date at all times. Trading expertise. Trading over-the-counter instruments in emerging markets can be very challenging. It requires a well-resourced, experienced trading capability focused solely on trading EMD. Without a sufficiently resourced and experienced trading team, the investment manager can end up with money stuck in countries like India or Nigeria, suffer poor FX rates on restricted currencies left with a custodian to trade or be forced to pay taxes that could have been avoided. These losses can be substantial, easily exceeding the circa basis points paid to a suitably skilled third-party investment manager. Risk management. Investment managers need to be fully abreast of the legal framework in which they invest. Unlike in many other asset classes, regulations relating to market access, taxation and trading restrictions change frequently and often suddenly. Managers also need to be fully apprised of relevant trading conventions (e.g., International Swaps and Derivatives Association protocols) and know where the sources of liquidity are. They need to understand and proactively manage implicit developed-market risk within their EMD portfolio. They also need continually to be on the lookout for potential fraud in enterprises in which they have invested. Cognisant of these resource demands, Eaton Vance s global income team employs 30 investment professionals, six EM debt traders and five trading assistants across Boston, London and Singapore. We trade our own FX. The advantage of this is that we avoid the risk of expensive FX trades that can arise when trades are left in the hands of a custodian. Our trading capability also affords us the ability to structure trades advantageously. For example, in Colombia, we put our trades on a payment system that avoids transaction taxes. Other considerations Investors considering moving to internal management of their EM debt allocation might want to be aware of the fact that the pool of EM debt trading expertise is relatively small. Being able to attract sufficient trading talent dedicated to a particular pension fund may prove to be a challenge. Internal management also potentially raises governance challenges around transparency and investment performance. Traditional approaches to management of local currency debt have, to date, disappointed a number of investors. Matching or surpassing the performance of the local currency indices has proved difficult, particularly in up markets. The JPMorgan GBI-EM Global Diversified Index is a gross index it doesn t reflect the drag on returns of taxes paid by real-world investors who want to track its performance. These taxes will be different for different investors depending on factors such as how the fund is structured, which instruments are used to execute a trade and where the fund is domiciled. Investors might want to re-evaluate the cost benefit of how they have traditionally approached local currency debt versus an optimised unconstrained approach. Can an unconstrained, blended approach deliver added value? Clearly, pension funds will need to do their own due diligence on this question. That said, we believe Eaton Vance s track record in blended unconstrained mandates points to the viability of such an approach. Our Emerging Market Debt Opportunities Strategy, which has been running for more than three years, has been able to generate alpha and excess return with lower-thanbenchmark volatility. Counterintuitive as it may sound, a more flexible approach has the potential to offer both better returns and lower volatility. Conclusion EM debt offers a unique source of income and, for portfolios comprising mainly developed-market fixed income, diversification. However, successful investment in this complex asset class requires specialist expertise and a keen awareness of its many idiosyncrasies and potential pitfalls. This is particularly true for investors seeking to move to a blended approach. At Eaton Vance, we believe

10 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 10 a blended, index-based approach represents a defensive, straightforward option that boards may find reasonably easy to approve and oversee. Nonetheless, the inherent flaws in indices point to the potential advantage of a more sophisticated flexible approach one seeking to evaluate risk factors in each country across the entire tradable universe. For investors, adopting such an approach requires a new mindset and becoming comfortable with a portfolio exhibiting a substantial tracking error to a performance benchmark. It presupposes extensive due diligence, but also recognition of how much money traditional approaches currently leave on the table.

11 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 11 Index Definitions The JP Morgan Emerging Markets Bond Index Global Diversified (EMBI GD) is a broad index of hard currency sovereign bonds issued by a selection of emerging-market countries. Like the EMBI Global Index, it includes US dollar-denominated Brady bonds, eurobonds, traded loans and local market debt instruments issued by sovereign and quasi-sovereign entities. However, it differs from EMBI Global in that it limits the weights of countries with larger debt stocks. The JP Morgan Government Bond Index-Emerging Markets Global Diversified Index (GBI-EM GD) is an unmanaged index of local currency bonds with maturities of more than one year issued by emerging-markets governments. Like the GBI-EM Global Index, it includes only those countries that are directly accessible by most of the international investor base. Unlike GBI-EM Global, GBI-EM Global Diversified limits the weights of index countries with larger debt stocks. The JP Morgan Corporate Emerging Markets Bond Index Broad Diversified (CEMBI BD) is a market capitalisation-weighted index that tracks hard currency (US$ -denominated) corporate bonds issued by emerging-markets entities. The JP Morgan Emerging Markets Bond Index Plus (EMBI+) Index is a market cap-weighted index that measures US dollar-denominated Brady Bonds, eurobonds and traded loans issued by sovereign entities. Unless otherwise stated, index returns do not reflect the effect of any applicable sales charges, commissions, expenses, taxes or leverage, as applicable. It is not possible to invest directly in an index. Historical performance of the index illustrates market trends and does not represent the past or future performance.

12 EATON VANCE TOPIC PAPER MARCH 2017 EMERGING-MARKET DEBT INVESTING 12 About Eaton Vance Eaton Vance is a leading global asset manager whose history dates to With offices in North America, Europe, Asia and Australia, Eaton Vance and its affiliates offer individuals and institutions a broad array of investment strategies and wealth management solutions. The Company s long record of providing exemplary service, timely innovation and attractive returns through a variety of market conditions has made Eaton Vance the investment manager of choice for many of today s most discerning investors. For more information about Eaton Vance, visit eatonvance.com. About EVMI Eaton Vance Management (International) Limited (EVMI) is a subsidiary of Eaton Vance Management, a leading U.S. asset management organisation, and markets internationally the investment capabilities of Eaton Vance Management and its affiliates, including Parametric. EVMI has been based in London since This material does not constitute an offer or solicitation to invest in any Eaton Vance fund and/or products. Forecasts may not be attained. Past performance is no guarantee of future results. This material is communicated by Eaton Vance Management (International) Limited, which is authorised and regulated in the United Kingdom by the Financial Conduct Authority and located at 125 Old Broad Street, London, EC2N 1AR, United Kingdom, Tel. +44 (0) EVMI is a wholly owned subsidiary of Eaton Vance Management (EVM). EVM is an investment advisor registered with the United States Securities and Exchange Commission (SEC) and is a wholly owned subsidiary of Eaton Vance Corp. (EVC). The nonvoting common stock of EVC, parent company of EVM, is publicly traded on the NYSE under the symbol EV. For purposes of this material, Eaton Vance or the Company is defined as all three entities operating under the Eaton Vance brand. EVMI markets the services of the following strategic affiliates: Parametric Portfolio Associates LLC (Parametric) is an investment advisor registered with the SEC and is a majority-owned subsidiary of EVC and Hexavest, which is an investment advisor based in Montreal, Canada, registered with the SEC in the United States and which has a strategic partnership with Eaton Vance, who owns 49% of the stock of Hexavest Inc. In Singapore, EVMI has a wholly owned subsidiary, namely Eaton Vance Management International (Asia) Pte. Ltd. (EVMIA), 8 Marina View, Asia Square Tower 1, #07 05, Singapore , which holds a Capital Markets License under the Securities and Futures Act of Singapore (CMS ), is an exempt Financial Adviser pursuant to the Financial Adviser Act Section 23(1)(d) and is regulated by the Monetary Authority of Singapore. This document is to be distributed to Accredited Investors ONLY (as defined in the Securities and Futures Act, Chapter 289 of Singapore). In Australia, EVMI is exempt from the requirement to hold an Australian financial services license under the Corporations Act in respect of the provision of financial services to wholesale clients as defined in the Corporations Act 2001 (Cth) and the Australian Securities and Investments Commission s Class Order 03/1099. EVMI is registered as a Discretionary Investment Manager in South Korea pursuant to Article 18 of Financial Investment Services and Capital Markets Act of South Korea. Eaton Vance Management (International) Limited utilises a third-party organisation in the Middle East, Wise Capital (Middle East) Limited (Wise Capital), to promote the investment capabilities of Eaton Vance to institutional investors. For these services, Wise Capital is paid a fee based upon the assets that Eaton Vance provides investment advice to following these introductions. This document does not constitute an offer to sell or the solicitation of an offer to buy any Securities/Notes/Fund Units/Services referred to expressly or impliedly in this document in the People s Republic of China (excluding Hong Kong, Macau and Taiwan, the PRC ) to any person to whom it is unlawful to make the offer or solicitation in the PRC. The document may not be provided, sold, distributed or delivered, or provided or sold or distributed or delivered to any person for forwarding or resale or redelivery, in any such case directly or indirectly, in the People s Republic of China (the PRC, excluding Hong Kong, Macau and Taiwan) in contravention of any applicable laws. Eaton Vance Management (International) Limited is licensed by the United Kingdom Financial Conduct Authority to engage in the investment management business and hereby operates in Japan under Article 58-2, and Article 61, Paragraph 1 of the Financial Instruments and Exchange Act of Japan. Accordingly, services provided by Eaton Vance Management (International) Limited are available to Japanese investors only to the extent permitted under Article 58-2 and Article 61, Paragraph 1. The views expressed in this material are those of the authors and are current only through the date stated at the top of this page. These views are subject to change at any time based upon market or other conditions, and Eaton Vance disclaims any responsibility to update such views. These views may not be relied upon as investment advice and, because investment decisions for Eaton Vance are based on many factors, may not be relied upon as an indication of trading intent on behalf of any Eaton Vance fund. This material may contain statements that are not historical facts, referred to as forward-looking statements. A fund s future results may differ significantly from those stated in forward-looking statements, depending on factors such as changes in securities or financial markets or general economic conditions, the volume of sales and purchases of fund shares, the continuation of advisory, administrative and service contracts, and other risks. This material is for professional clients only Eaton Vance Management (International) Limited 125 Old Broad Street, London, EC2N 1AR, United Kingdom Telephone: +44 (0) internationalenquiries@eatonvance.com