Emerging Markets Debt: the next 10 years

Transcription

1 Emerging Markets Debt: the next 10 years Paper VII, May 2012 This communication is only for professional investors, professional financial advisers and, at their exclusive discretion, their clients. It is not to be generally distributed to the public.

2 Paper VII: EMD the next 10 years Introduction 2 The changing sovereign debt landscape 4 Future structure of EMD 7 When emerging emerges 10 A word on demand for EMD 13 Conclusion 14 Appendices 15 About Investec Asset Management Investec Asset Management is a specialist provider of active investment products and services to institutional and individual investors. Established in South Africa in 1991, the firm has been built from a small start up into an international business managing US$98.4 billion (as at 31 March 2012) on behalf of third party clients. The business has grown largely organically from domestic roots in Southern Africa to a position where we proudly serve a growing international client base from the Americas, the UK and Continental Europe, Asia, the Middle East, Australia and Africa. We employ 125 investment professionals. The firm is still managed by its founding members, representing continuity and stability throughout the firm s successful growth. The firm seeks to create a profitable partnership between clients, shareholders and employees and our aim is to exceed our clients investment and client service expectations and to manage their money to the highest possible standard. Investec Asset Management is a significant component and independently managed entity within the Investec Group, which is listed in London and Johannesburg.

3 Executive summary Today: Total stock of EMD (emerging market debt local and hard currency) has increased from US$1.6 trillion to US$8.3 trillion over the past 10 years Over the same period, despite increasing its share of global GDP from 27.1% to 39.9%, emerging markets share of total global government debt has fallen from 19.6% to 16.8% By 2022 we expect: Emerging markets share of global government debt to grow to 20.4% in real terms from US$8.3 trillion to US$14.5 trillion The biggest markets will, unsurprisingly, be China (US$4 trillion), India and Brazil (both c. US$2 trillion) Asia will account for over 50% of global EMD from 45% currently. Latin America will shrink from 31% to 26% The maturity of EMD should increase on average by four years The proportion of fixed-coupon debt to increase from 70% to 80% thanks to more stable inflation and currencies Foreign ownership of local currency bonds will continue to grow The JP Morgan local currency index will increase from 14 to 22 countries (Nigeria is most likely next entrant probably within 12 months) The JP Morgan hard currency index could increase from 48 countries to 55 Potential demand for EMD: With an estimated $80 trillion in conventional assets under management, a 1% shift in allocation towards EMD would result in as much as $800 billion in flows Current allocations to EMD are low with some estimates claiming on average 2.2% in US, UK and European pension funds If US pension funds increased their allocation to EMD to the European average of 3.9% it would result in a $370 billion flow Authored by Antoon de Klerk, Grant Webster, Peter Eerdmans Emerging Markets Debt & Currency team The views expressed in this document are those of the authors and reflective of the wider Emerging Markets Debt & Currency team in which they are members. Investec Asset Management is a multi-specialist house and therefore the views expressed may or may not be held by our other specialist investment teams. 1

4 Introduction In this paper, the seventh in our series on emerging market debt (EMD), we consider how we believe the asset class is likely to evolve over the next decade. We assess the economic fundamentals that will drive this development and its implications for countrycomposition, size, structure and sovereign/corporate split of EMD portfolios in five and ten years time. Emerging debt markets, like the economies they are based on, have come a long way over the past ten years (December 2001 December 2011). Significant developments have taken place along several lines: for example, the total stock of debt (public and private/corporate) has increased five-fold from US$1,600 billion to US$8,270 billion; average maturity of local currency debt has increased by 3.5 years (from 5.5 to 9 years) 1 ; the average portion of local currency funding of governments has increased from 62% to 80%; and private sector bonds have grown from 27% to 43% of outstanding emerging market debt 2. These developments have not gone unnoticed by global investors contrast the US$3 billion flows into the asset class in 2005 with the US$80 billion received in 2010 and US$64 billion in In this paper, we consider the future of the asset class, with an emphasis on its supply side dynamics. The specific questions we look to answer are: i. which countries will have the capacity to borrow ii. how much will they be able to borrow iii. in what form (currency, tenor, etc.) are they likely borrow? EMD stock has increased five-fold to US$8.3 trillion over 10 years The share of local currency and private sector EMD has increased Note on methodology and assumptions Forecasting is all about making assumptions. Its value lies in the outcomes produced by different sets of assumptions. To focus on the relevant outcomes we limit the assumptions we make: 1. All amounts are in current (2012) US dollars unless explicitly stated otherwise. That is, we ignore inflation and consider everything in real terms. 2. We make no assumptions regarding future exchange rates. Forecasting exchange rates is a very relevant exercise in and of itself. But here we hold exchange rates steady to highlight changing debt dynamics that might be obscured otherwise. 3. Predominantly we use IMF forecasts out to 2017 as published in its most recent (April 2012) World Economic Outlook. Where we differ or where the relevant series is not forecasted by the IMF we explicitly state our assumptions. 1 Data source: BIS, IAM calculations. EM countries included in calculation; Colombia, Czech Republic, Hong Kong, Indonesia, Peru, Russia, South Africa and Turkey. 2 Data source: BIS, IMF, IAM calculations. Countries included as EM are: Brazil, China, Colombia, Czech Republic, Hong Kong, Hungary, India, Indonesia, Malaysia, Peru, Poland, Russia, South Africa, South Korea, Thailand and Turkey. 3 Data source: EPFR, JP Morgan. 2

5 Figure 1: Evolution of emerging markets debt: December 1995 to December 2011 December 1995 US$ billion 7,000 6,000 5,000 4,000 3,000 Total = US$ billion 2% 13% 8% 7% 21% 1% 2,000 1,000 14% 34% Sovereign ST Corporate ST Total ST Sovereign LT Corporate LT Total LT December 2005 US$ billion 7,000 6,000 5,000 4,000 3,000 2,000 1,000 Total = US$3, billion 13% 0% 8% 7% 7% 41% 23% 1% Sovereign short term domestic Sovereign short term international Sovereign long term domestic Sovereign long term international Private short term domestic Private short term international Private long term domestic Private long term international Sovereign ST Corporate ST Total ST Sovereign LT Corporate LT Total LT December 2011 US$ billion 7,000 6,000 5,000 4,000 3,000 2,000 1,000 Total = US$8, billion 23% 1% 12% 8% 4% 14% 0% 38% Sovereign ST Corporate ST Total ST Sovereign LT Corporate LT Total LT Sovereign ST = Government ST, Corporate ST = Private ST, Sovereign LT = Government LT, Corporate LT = Private LT Source: BIS, IAM calculation EM includes: China, Colombia, Brazil, Hungary, India, Indonesia, Malaysia, Mexico, Philippines, Poland, Russia, South Africa, Thailand and Turkey. May

6 The changing sovereign debt landscape As is now widely publicised, emerging market growth is expected to substantially outstrip that of developed markets such that, within the next ten to fifteen years, emerging economies are likely to account for at least half of total global GDP. Already they account for approximately 37% of global GDP. But will we see the same trend in their level of government debt? Certainly over the past 10 years this has not been the case, with emerging market sovereign debt falling as a proportion of total global debt, as highlighted in Figure 2. Figure 2: Emerging markets share of global GDP and global government debt Emerging economies should account for around 50% global GDP in years Over the past 10 years, the EM share of total global government debt has actually fallen % GDP Government Debt Source: IMF, Investec Asset Management calculations May 2012 In trying to measure the future stock of emerging market debt we have relied on two fundamental and widely used estimates: the total size of GDP and debt-to-gdp ratios. Using IMF forecasts of real GDP growth over the next five years 4, as well as expected debtto-gdp ratios, which are in turn based on borrowing requirements within each country, we are able to estimate the potential future stock of government debt. Furthermore, using inflation-adjusted growth rates allows us to draw comparisons across time. We use IMF assumptions for debt and GDP growth to Over this five year period annual GDP growth in emerging and developed markets is forecast at 5.6% and 2.3% respectively. Beyond 2017, the scenario we forecast is not overly conservative, but instead assumes a moderate amount of debt reduction in emerging markets and a slightly faster pace of debt reduction in developed markets. GDP growth from 2017 to 2022 is forecast to be 2.3% for developed markets and 5.2% in emerging markets. As such, we assume that emerging economies will have debt-to-gdp ratios that trend down to 50% (by 2030) where they are above this level in 2017, or remain constant if they are below 50%. For developed markets, we assume that debt-to-gdp will either trend lower to 90% or remain constant after 2017.To put these assumptions into perspective, the IMF s 2012 Fiscal Monitor included an analysis that measured the primary balances required to reduce debt-to-gdp ratios to 40% in the case of emerging economies, and 60% in the case of developed economies by We believe that these targets are unrealistic and would be excessively restrictive on fiscal policy. Our estimates show that by 2022 the global stock of government debt will grow to almost US$70 trillion a 21.5% real increase from 2012 levels, representing an average of US$1.2 trillion in new debt per year. While at first the stock of debt may appear large, the increase is in fact very modest relative to the previous ten years growth. From 2002 to 2012 we estimate that emerging market debt grew by 26.4%, while developed market debt grew by a staggering 70.9%. Clearly the last four years have been exceptional, but even before the 2008 financial crisis developed markets increased their stock of debt by over US$1 trillion per year. 4 Please see the appendix for more details of our methodology and assumptions. Note that where a country has already achieved its target debt-to-gdp ratio by 2017, we assume that it maintains that level for the following 5 years. We have also adjusted the IMF trend GDP growth rates where these levels appear unsustainably high. 4

7 The charts below show the annual stock of global sovereign debt as well as the share made up by emerging markets. We also show the associated debt-to-gdp levels across developed and emerging markets. Due to its size and generally low debt-to-gdp levels, we separate China from the rest of the emerging markets. From the charts it is clear that, despite substantially higher GDP growth, emerging markets will only gradually expand their share of global sovereign debt over the next ten years. Under our assumptions emerging market debt will only increase from 16.2% to 20.4% of total global sovereign debt by By 2022 we expect global government debt to reach US$70 trillion a 21.5% real increase from 2012 Figure 3: Debt-to-GDP ratios across EM and DM 120% 100% Debt-to-GDP ratio 80% 60% 40% 20% Forecast We expect EMD to increase from 16.2% to 20.4% of total global sovereign debt by % EM China DM Source: IMF, Investec Asset Management calculations, May 2012 Figure 4: Stock of global government debt Total debt stock (USD billion) 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 Forecast 25% 20% 15% 10% 5% 0% EM % of Total [RHS] Total EM DM Source: IMF, Investec Asset Management calculations, May 2012 As regards individual issuers, we find unsurprisingly given the size of their economies that China, India and Brazil will hold the largest stock of emerging market debt in The noticeable change will be a relatively large increase in debt by India. This is driven by India s strong economic growth compounded by a high, but falling, level of debtto-gdp. Russia, which is expected to experience lower growth, has the most scope amongst the BRICs to increase its levels of debt relative to GDP and, as such, it moves up the rankings. Unsurprisingly China, India and Brazil will hold the largest stock of EMD by

8 Figure 5: Forecast Emerging market government debt levels, 2012 and 2022 US$ billion 4,500 1,000 4,000 3, , ,500 2, ,500 1, The noticeable change will be a relatively large increase by India China India Brazil Mexico Turkey Poland Russia Thailand Venezuela Indonesia Taiwan Argentina Malaysia Egypt South Africa Israel Colombia Pakistan Czech Republic Philippines Hungary Hong Kong SAR Morocco Nigeria Vietnam Source: IMF, Investec Asset Management calculations May 2012 China s ability to almost double its stock of debt over the next ten years does not come at the cost of increasing leverage. Between 2012 and 2022 we have in fact assumed that China s debt-to-gdp remains fairly constant at around 25% and that the country s economic growth accounts for its ability to increase debt issuance. Outside of the traditional BRICs, the countries with the largest capacity to issue debt over the next ten years will be Mexico (US$200 billion), Thailand (US$136 billion), Malaysia (US$101 billion) and Turkey (US$89 billion). For investors wanting to capture the growth in frontier markets however, it is worth considering those economies which, although small, have significant capacity to grow their debt issuance: Morocco, Nigeria, Vietnam, Tunisia, Kenya, The Dominican Republic and Ghana all have the capacity, we forecast, to increase their stock of debt by at least 60% over the next ten years with issuance ranging between US$12 billion and US$40 billion 5. On a regional basis, the growth in China and India pushes Asia s share of global emerging market sovereign debt to more than 50% by Brazil, though still a major potential source of debt issuance, loses market share as we expect the country s debtto-gdp levels to fall from 65% currently to approximately 53%. We forecast China to double its stock of debt by 2022 without increasing leverage Outside of BRIC, big issuers will be Mexico, Thailand, Malaysia and Turkey Smaller countries: Morocco, Nigeria, Vietnam, Tunisia, Kenya, The Dominican Republic and Ghana can significantly grow issuance Figure 6: Forecast Regional proportions of emerging markets government debt, 2012 and 2022 Middle East and Africa 11% Europe 13% Brazil 18% Latin America 13% 2012 Asia 11% India 13% China 21% Middle East and Africa 10% Europe 11% Brazil 14% Latin America 12% 2022 Asia 11% India 15% China 27% Source: IMF, Investec Asset Management calculations, May For full details on country issuance and assumptions please see the appendix 6

9 Future structure of EMD The exact structure of a given bond market is greatly influenced by the economic history and current macro policies of its host economy. Past debt and financial crises, in particular, leave their mark on the willingness of lenders to take certain risks. For several Latin American countries (Brazil and , Mexico 1990, Peru 1989) with painful experiences of hyper-inflation, local investors have traditionally been loath to take on inflation and interest rate risk and consequently these governments have issued relatively more inflation-linked and floating rate bonds. However, as a country builds a track record of macro stability, investors become more comfortable to take on these risks and the proportion of non-vanilla fixed interest rate bonds decreases. This development is further supported by increasing participation of offshore investors, who typically make their entrance into new markets via fixed interest bonds. As evidenced by the credit crisis of 2008/09 when the issuance of corporate bonds in the developed world all but dried up and aggressive fiscal stimulus meant very large issuance of sovereign bonds the structure of mature bond markets is not static. Financial innovation and unique developmental paths also make forecasting future emerging market bond market structures, particularly at an aggregate level, very challenging. Balancing practicality with relevance, we identify potential broad market developments and highlight specific emerging markets in which these are likely to have a significant influence. As well as having a relatively large bond market relative to GDP, there are a number of broad characteristics that can be clearly identified in mature bond markets. We discuss each of these characteristics below with reference to the reasons behind their development, and also the prospects for this trend to occur in key emerging markets over the next 10 years Sovereign/corporate bond mix Mature debt market characteristic: stable ratio between the amount of sovereign and corporate debt; Dynamic at work: private /corporate bond market development follows that of public (sovereign) with some lag as domestic and offshore investors tend to price credit yields off sovereign yield curves; 10-year emerging markets view: increasing issuance of private bonds vis-à-vis public bonds visible especially in countries, such as Brazil and South Africa, which have already come some way in establishing private bond markets. It is foreseeable that in these markets the private bond markets could be of at least equal size to sovereign markets in the ten years to come. A good track record allows greater issuance of fixed interest bonds, attracting foreign investors Increasing issuance of private bonds can be expected in many emerging markets Figure 7a: Sovereign-corporate bond mix and per capita GDP Private debt securities as % of Total South Africa Forecast GDP p.c. USD '000 (real 2011) 1992 Dec-96 Dec-98 Dec-01 Dec-03 Dec-06 Dec-08 Dec-11 Dec-13 Dec-16 Dec-18 Dec-21 Corporate all: South Africa p.c. GDP: South Africa 6 See Braun and Briones (2006): The development of bond markets around the world for a more detailed discussion. 7

10 Figure 7b: Private debt securities sovereign-corporate bond mix and per capita GDP Private debt securities as % of Total Brazil Forecast GDP p.c. USD '000 (real 2011) 1992 Dec-96 Dec-98 Dec-01 Dec-03 Dec-06 Dec-08 Dec-11 Dec-13 Dec-16 Dec-18 Dec-21 Corporate all: Brazil p.c. GDP: Brazil Source: IMF for GDP growth & GDP p.c. data. BIS for debt data.; IAM calculations, May 2012 Forecasts: We use the relationship between p.c. GDP growth and private debt as a % of total debt in developed markets. 2. Bond maturity Mature debt market characteristic: both sovereign and corporate bond yield curves have long maturities; Dynamic at work: improved credit ratings and decreasing macroeconomic volatility allow governments and corporates to issue debt at longer maturities; 10-year emerging markets view: continuing extension of the yield curve within and across emerging markets. On average we see emerging markets adding another four years to their yield curves. We expect on average emerging markets to add four years to their yield curve Figure 8 Average maturity of outstanding bonds Forecast Average duration of outstanding debt Korea Poland Turkey Colombia Industrial Countries Source: BIS, IAM calculations, May 2012 Forecasts: extrapolation of the annual rate of increase of the individual country but slowing down at 0.1 years per year Local/hard currency bond mix Mature debt market characteristic: Eurodollar bonds represent a very small portion of total bonds outstanding for governments as well as a decreasing portion of private bonds; Dynamic at work: Borrowers would rather repay debt and interest in the same currency as their income. As international lenders gain more trust in the credibility of monetary and fiscal policies, they become increasingly willing to lend in local currencies; 10-year emerging markets view: local currency bond issuance and share of total outstanding debt to dominate that of hard currency in an increasing number of emerging markets. Local currency bond issuance to dominate in an increasing number of emerging markets

11 Figure 9: Local currency component of sovereign debt outstanding 100 Forecast % Indonesia Peru Philippines Russia Turkey Ukraine Korea Source: IMF, BIS, IAM calculations, May 2012 Forecasts: We calculate the relationship between growth in per capita income and the local currency component in outstanding government debt. This relationship is used along with IMF p.c. growth forecasts and the starting point of individual countries to calculate forecasts. 4. Fixed interest/inflation-linked mix Mature debt market characteristic: stable ratio between inflation-linked and fixed interest bonds with the latter dominating the mix; Dynamic at work: growing track records of macro stability give investors enough comfort to take on inflation risks; 10-year emerging markets view: increasing issuance of plain-vanilla fixed interest bonds. We see EMs on aggregate moving towards the current developed market structure 80% fixed, 10% floating, 5% inflation-linked and 5% exchange rate linked. Figure 10: Emerging markets average % total bonds outstanding by bond type 100 Forecast Fixed Floating Inflation-linked Developed Markets 2010 Source: IMF, BIS, IAM calculations, May Forecasts: We use current developed bond market characteristic (DM 2010) to forecast where EM on average will evolve to over the next 10 years. DM includes: Australia, Canada, Italy, Netherlands, Spain, United Kingdom, United States, Japan and Finland 9

12 5. Local/foreign ownership mix Mature debt market characteristic: A high level of foreign participation by foreign investors in both sovereign and corporate bond markets; Dynamic at work: Economic growth, macro and political stability and market liquidity attract international investor interest as new markets present opportunities for diversification and potentially higher returns; 10-year emerging markets view: we see the current trend of increasing foreign participation in emerging bond markets continuing at a steady pace although not quite reaching developed market levels over the next ten years. Increasing foreign participation in EMD to continue at a steady pace Figure 11: Foreign ownership of sovereign securities 80% Forecast 60% 40% 20% Malaysia Mexico Poland Thailand EM Average DM average Source: IMF, BIS, World Bank, IAM calculations, May Forecast: Using the current EM average of 17.3% we calculate the annual rate of increase required for this level to reach 30% - a level which we see as commensurate with the investment opportunity and macro prudence for EM on average. DM includes: Australia, Canada, Italy, Netherlands, Spain and United Kingdom. Note that the United States and Japan are explicitly excluded due to their non-typical role as investment destinations. EM includes: Brazil, Colombia, Hungary, Indonesia, Malaysia, Mexico, Philippines, Poland, South Africa, Thailand and Turkey. When emerging emerges Inherent in their classification as emerging is a long-term issue for emerging market investors: what happens when these markets are no longer emerging but arrive at the point when they have actually emerged? As yields fall (commensurate with reduced risk) and correlations with developed market assets increase, the appeal of investing in emerging market debt decreases. But what is often overlooked is the fact that as one country emerges, another is waiting to fill the gap, and so we expect that the universe of investible emerging markets, due to its dynamism, will maintain the attraction of higher yields and lower correlations over time. To analyse how the universe of emerging market debt may change over time, we have taken the rather bold step of trying to forecast the constituents of JP Morgan s emerging market benchmarks: the EMBI Global Diversified (a benchmark of hard currency denominated emerging market debt) and the GBI-EM Global Diversified (a benchmark of local currency denominated emerging market debt). To be considered for inclusion, an emerging economy is required to be classified as Lower-Middle Income by the World Bank for two consecutive years, and also have a sufficiently liquid debt market. Using forecasts for real income growth and our modelling of the development of each country s local currency denominated debt market (described above), we believe JP Morgan s GBI-EM Global Diversified index will include 13 new countries by One of the most likely candidates to be included is Nigeria. The country has been classified as Lower-Middle Income for a number of years already and the local debt market is well developed. As the debt market becomes What happens when these markets are no longer emerging? As one country emerges, another is waiting to fill the gap JP Morgan s GBI-EM GD index could include 13 new countries by 2022: Nigeria soon, India and China depending on foreign restrictions 10

13 more accessible to international investors, with increasing liquidity being provided by international banks, we expect Nigeria to be included imminently. Other countries which we expect to qualify include Romania, Zambia, Ghana (all by 2017), Serbia and Vietnam. It is also likely that both India and China could be included when restrictions on foreign investors are lifted. While many countries meeting the income status classification will be included by virtue of the size and liquidity of their debt markets, others, such as Zambia and Ghana, may well be included because they issue large, liquid global bonds denominated in their own local currencies. This is not an unusual phenomenon as already Chile, Brazil and Russia have issued such bonds. These instruments trade on international exchanges and give foreign investors access to the local currency debt and in turn this helps the local debt market develop an efficiently priced yield curve. Figure 12: Income per capita with local currency benchmark cut-offs (refer to appendix for detail on methodology) Income per capita (USD) 20,000 18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 Forecast China Ghana India Ukraine Vietnam Nigeria Romania Philippines Brazil Chile Colombia Egypt Hungary Indonesia Malaysia Mexico Poland Russia South Africa Lower middle income level High income level Source: IMF, World Bank, UN Population Division, IAM calculations, May 2012 The countries most likely to leave the index are Poland and Hungary, which have been classified as High Income for a number of years now. Technically a country should be excluded once it has been classified as High Income for five consecutive years. However, this remains at the discretion of JP Morgan and it seems unlikely that they would remove a country while it is still perceived as emerging with a high dispersion of income levels across the population. Other potentially significant exits by 2022 could include Turkey and Russia, both of which will breach the level of High Income based on our calculations. Once again, though, we believe that a number of qualitative factors may overrule the technical measures for exclusion. The JP Morgan EMBI Global Diversified (hard currency EMD) index will also experience significant changes over the coming ten years as an additional 13 countries could potentially be included, taking the total number of issuers to around 55. Based purely on the technical index rules we believe that seven countries could exit the benchmark. These include Croatia and Lithuania, as well as the same countries expected to leave the local currency benchmark. It is relatively easy to predict upcoming inclusions based on which countries are planning to issue dollar debt. Recently it was announced that Azerbaijan will be joining the benchmark. This could soon be followed by Mongolia. Other countries likely to enter the benchmark include Angola and Zambia, both of which are looking to issue inaugural dollar bonds. Based on their expected levels of income and issuance potential, other countries that we believe will have the ability to join the benchmark by 2022 include Kenya, Tanzania, Botswana, Paraguay, Nicaragua, Albania and Armenia. Most likely to leave are Poland and Hungary JP Morgan s hard currency index could include 13 new countries by 2022, and 7 could leave 11

14 The development of emerging market debt, in terms of depth, size and liquidity, has led these markets to attract a broader range of investors, including cross-over investors who are less benchmark orientated. This is especially the case as many developed debt markets have come to behave far more like risk assets since the 2008 financial crisis and subsequent European fallout. As a consequence, we believe that there is a strong possibility that in the near future, new benchmarks will be created to incorporate a combination of high-quality developed and emerging market debt. These global benchmarks will down-weight heavily indebted countries in favour of higher yielding, growth-orientated nations with more favourable fiscal outlooks. New benchmarks could be created, combining highquality developed and EMD Figure 13: Potential inclusions and exclusions to the JP Morgan GBI-EM Global Diversified local currency benchmark Previously/currently included Exits New Entrants China India Kenya Lebanon Morocco Serbia Ukraine Vietnam Ghana Nigeria Romania Zambia Ghana Nigeria Romania Zambia Philippines Philippines Philippines Brazil Brazil Brazil Brazil Chile Chile Chile Chile Colombia Colombia Colombia Colombia Egypt Egypt Egypt Egypt Hungary Hungary Hungary Hungary Indonesia Indonesia Indonesia Indonesia Malaysia Malaysia Malaysia Malaysia Mexico Mexico Mexico Mexico Peru Peru Peru Peru Poland Poland Poland Poland Russia Russia Russia Russia South Africa South Africa South Africa South Africa Thailand Thailand Thailand Thailand Turkey Turkey Turkey Turkey countries 14 Countries, 1 exit, 1 new entry since Countries, 2 exits and 5 new entries since Countries, 5 exits and 13 new entries since Source: Investec Asset Management

15 A word on demand for emerging markets debt The ability for global funds to absorb future emerging market debt issuance is of critical importance in assessing the potential impact on prices. While an analysis of the demand-side dynamics of emerging market debt deserves a full white paper in itself, we are able to offer some insights here. The total size of conventional (i.e. pension, insurance and investment) funds was estimated at US$77 trillion at the end of In addition to this, there is an estimated US$6 trillion in sovereign wealth funds and central banks foreign exchange reserves 8. Clearly from the size of these funds even a small shift in allocation towards emerging market debt would result in a large flow of capital: for each 1% change in allocation we could expect as much as US$800 billion in emerging market bond flows (bearing in mind we forecast emerging market sovereign debt issuance of US$500 billion p.a. for the next ten years). Current allocations to emerging market assets and emerging market debt in particular are low. Morgan Stanley 9 estimates that average allocation to emerging market debt among US, UK and European pension funds stands at only 2.2%. In the US, the largest pension fund market, the allocation is only 1.7% and only 37% of pension funds have exposure to emerging market debt (against 65% that have exposure to emerging market equities). Japan is expected to have even lower allocations due to the fact their allocation to emerging markets (including equity) as a proportion of total cross-country portfolio allocation is only 4%. Were US pension funds to increase their allocation to emerging market debt by 1%, this would represent a flow of approximately US$170 billion, while an increase up to the European average allocation of 3.9% would result in a flow of over US$370 billion. But even at these levels the allocation to emerging market debt is very low relative both to the asset class s contribution to the total global investable universe and to the levels often cited as most beneficial from a diversification perspective (which range very broadly from a 10-30% 10 allocation to emerging market assets depending on the assumptions used). Of course any proportional increase in holdings by foreign investors has to be met by a proportional decrease in holdings by domestic investors, including the local banking and pension fund sectors. It is this complicated supply-demand dynamic that will have implications for the future pricing of emerging market debt. A 1% shift in allocation towards EMD would result in as much as $800 billion in flows Current allocations to EMD are low: on average 2.2% in US, UK and European pension funds US pension funds increasing their EMD allocations to European average would result in a $370 billion flow 7 Morgan Stanley research, OECD, IMF, McKinsey, cityuk 8 While the total AUM of central banks and sovereign wealth funds is estimated at $10-$15 trillion, in reality only a small of these assets are mobilised in the form of investable assets. Morgan Stanley research suggests the total size of investible assets held by central banks and sovereign wealth funds at $6 trillion 9 Morgan Stanley, EM Profile, Small Fish in a Big Pond, May Sources include Mercer, NEPC, P&I, Council of Institutional Investors 10 Morgan Stanley research found an optimal allocation, based on Telser s Safety First Criterion, to be 30%, including 13.26% to hedged local currency debt and 16.06% to sovereign hard currency debt. We use this only as an indication of the range of potential optimal allocations. 13

16 Conclusion Over the past ten years emerging markets have accounted for an increasingly large proportion of global GDP. However, over the same period, their proportion of global sovereign debt has fallen. We expect these debt dynamics to reverse over the next ten years as developed markets undergo an extensive period of de-leveraging while relatively strong emerging market growth continues. We forecast the total global stock of government debt to increase, in real terms, by 21.5% to almost US$70 trillion. Emerging markets will account for 20.4% of this debt, from 16.2% presently as they issue almost US$500 billion per year. While at first the stock of debt may appear large, the increase is in fact modest relative to the previous ten years growth: from 2002 to 2012 we estimate that emerging market debt grew by 26.4%, while developed market debt grew by a staggering 70.9%. Over the next ten years we expect that China will double its stock of government debt to over US$4 trillion. Together with India and Brazil these markets will account for more than 50% of the debt of what is currently classified as emerging markets. However, this increase in debt will not be delivered by an increase in leverage on the contrary our assumptions are that both emerging and developed markets decrease their levels of debt-to-gdp. Instead this increase is founded upon strong economic growth. As well as a relative increase in the size of emerging market debt, we also expect the composition of that debt to change significantly. In line with the historic development trends of mature debt markets, our forecasts show that the average maturity of emerging market sovereign debt will increase by four years, while the proportion of floating and inflation-linked debt will reduce. We also foresee a decrease in the proportion of hard currency denominated debt and hence an increase in the proportion of debt held by foreign investors. At the same time, issuance of corporate debt should increase dramatically as the private sector will be able to price their debt off local currency sovereign yield curves. The development of these markets will be matched by appropriate changes to the traditional emerging market debt benchmarks. By 2022 we believe that JP Morgan s local currency benchmark will grow from 14 countries to 22 while the bank s hard currency benchmark will increase from 48 to 55 issuers (including the potential exits of seven countries). There is also much scope for international investors to increase their exposure to emerging market debt over the next ten years. Currently the average allocation to emerging market debt amongst US, UK and European pension funds stands at only 2.2%. Given that, globally, the total size of assets under management currently stands at approximately US$77 trillion, a 1% shift in allocation towards emerging market debt would result in US$800 billion in flows. Emerging markets have substantial capacity to issue debt over the next ten years while still maintaining sustainable levels of leverage. As developed markets experience a difficult period of deleveraging we believe that global sovereign debt markets will undergo a significant re-ordering over the next ten years. In line with their sustainable growth dynamics we foresee emerging markets continuing to offer attractive debt investment opportunities. 14

17 Appendix 1: A note on methodology We base our estimates of future debt capacity on real GDP and sustainable levels of debt-to-gdp. Using these two inputs we are able back-out the approximate sustainable level of debt for each country for each year. Using real increases in GDP also allows us to draw comparisons on debt levels across time without having to account for the effects of inflation. Two important inputs into our calculations are the long-run, sustainable levels of debtto-gdp and real GDP growth. From 2012 to 2017 we rely on IMF forecasts for GDP growth and debt-to-gdp. Beyond 2017 we follow the IMF s lead and assume that each country has reached its long-run trend level of growth. However, for some countries where we believe the 2017 levels are unsustainable we adjust these levels down. These countries include Brazil, Cambodia, China, India, Indonesia, Iraq, Mexico, Nigeria, Peru, Poland, Russia, Thailand and Turkey. China is a special case too where we adjust the IMF s real growth forecasts down over the whole ten year period. From 2012 to 2017 the IMF forecasts growth of 8.65% which we believe, as an average annual rate, is too high. We adjust this down to an average annual rate of 7.6% and then, from 2017 to 2022, smooth growth down to a long-run trend growth rate of 6%. We also rely on IMF forecasts for levels of debt-to-gdp between 2012 and Beyond 2017 we assume that countries which have debt below the long-run sustainable rate will keep their debt-to-gdp levels constant. Those that are above the long-run sustainable level will trend their debt down to reach those levels by For emerging markets the long run level is 50% and for developed markets this level is 90%. These debt-to-gdp levels are above those highlighted by the IMF in their 2012 Fiscal Monitor report of 40% for emerging markets and 60% for developed markets. Once again China is a special case as we believe that the IMF assumption for debt reduction from 25.8% of GDP in 2011 to 10.1% in 2017 is too aggressive. We assume that China gradually reduces its level of debt to 25% by To forecast gross national income (GNI) per capita we assume that GNI grows at the real rate of GDP between 2012 and We then adjust this for population growth where we use the UN Population Division s 2010 Medium-fertility Variant population forecasts which run over the full period of our analysis. As we are using real increases in GNI we believe that it is suitable to keep the World Bank s income classification levels constant over time. This allows us to assess which countries will change classification which determines their eligibility for the various JP Morgan indices. 15

18 Appendix 2: Table of country details EMERGING MARKETS GDP (USD billion, 2012 prices) Real GDP Growth Debt-to-GDP Government Debt (USD billion) LT Trend LT Trend Albania % 2.5% 2.5% 61.7% 66.9% 59.9% 50.0% Angola % 6.2% 6.2% 23.8% 13.0% 13.0% 13.0% Argentina % 4.5% 4.5% 43.3% 36.9% 36.9% 36.9% Armenia % 4.0% 4.0% 37.6% 40.0% 40.0% 40.0% Azerbaijan % 3.1% 3.1% 11.2% 9.5% 9.5% 9.5% Bangladesh % 7.3% 7.3% 0.0% 0.0% 0.0% 0.0% Belarus % 5.0% 5.0% 37.3% 24.6% 24.6% 24.6% Belize % 2.5% 2.5% 78.1% 71.3% 62.4% 50.0% Bolivia % 5.0% 5.0% 31.5% 25.3% 25.3% 25.3% Bosnia and Herzegovina % 3.5% 3.5% 43.4% 35.4% 35.4% 35.4% Botswana % 4.3% 4.3% 16.2% 8.1% 8.1% 8.1% Brazil 2,568 3,137 3, % 3.7% 3.5% 65.1% 56.7% 53.9% 50.0% 1,672 1,780 2,032 Bulgaria % 4.5% 4.5% 21.3% 11.7% 11.7% 11.7% Cambodia % 7.3% 7.0% 28.6% 26.4% 26.4% 26.4% Chile % 4.5% 4.5% 10.1% 6.6% 6.6% 6.6% China 7,899 11,373 15, % 6.4% 6.0% 25.8% 25.6% 25.4% 25.0% 2,038 2,909 3,941 Colombia % 4.5% 4.5% 32.3% 31.7% 31.7% 31.7% Costa Rica % 4.5% 4.5% 32.4% 37.5% 37.5% 37.5% Côte d'ivoire % 6.7% 6.7% 62.6% 51.0% 50.6% 50.0% Croatia % 2.5% 2.5% 52.8% 63.1% 57.6% 50.0% Czech Republic % 3.5% 3.5% 43.9% 47.1% 47.1% 47.1% Dominican Republic % 5.0% 5.0% 29.9% 30.9% 30.9% 30.9% Ecuador % 3.4% 3.4% 18.3% 22.4% 22.4% 22.4% Egypt % 6.5% 6.5% 79.2% 60.4% 56.1% 50.0% El Salvador % 3.5% 3.5% 50.0% 44.5% 44.5% 44.5% Fiji % 2.3% 2.3% 53.5% 46.2% 46.2% 46.2% Gabon % 2.3% 2.3% 17.3% 12.4% 12.4% 12.4% Georgia % 5.5% 5.5% 32.7% 26.0% 26.0% 26.0% Ghana % 5.7% 5.7% 42.1% 40.3% 40.3% 40.3% Guatemala % 3.5% 3.5% 24.8% 25.5% 25.5% 25.5% Honduras % 4.0% 4.0% 31.2% 28.7% 28.7% 28.7% Hong Kong SAR % 4.4% 4.4% 33.2% 27.8% 27.8% 27.8% Hungary % 2.2% 2.2% 76.3% 71.9% 62.8% 50.0% Iceland % 2.9% 2.9% 97.3% 81.7% 81.7% 81.7% India 1,791 2,593 3, % 6.8% 6.0% 67.6% 64.6% 58.5% 50.0% 1,210 1,676 2,112 Indonesia 897 1,253 1, % 6.4% 6.0% 23.2% 15.5% 15.5% 15.5% Iraq % 7.1% 6.0% 31.3% 12.8% 12.8% 12.8% Israel % 3.5% 3.5% 74.0% 65.8% 59.2% 50.0% Jamaica % 1.5% 1.5% 145.9% 162.8% 115.8% 50.0% Jordan % 4.4% 4.4% 71.9% 67.5% 60.2% 50.0% Kazakhstan % 6.4% 6.4% 9.6% 2.9% 2.9% 2.9% Kenya % 6.5% 6.5% 46.6% 45.2% 45.2% 45.2% Korea 1,156 1,405 1, % 4.0% 4.0% 32.9% 23.1% 23.1% 23.1% Latvia % 4.0% 4.0% 39.1% 33.2% 33.2% 33.2% Lebanon % 4.0% 4.0% 135.3% 135.5% 99.9% 50.0% Lithuania % 3.9% 3.9% 40.9% 36.9% 36.9% 36.9% Malaysia % 5.0% 5.0% 53.1% 57.1% 54.1% 50.0% Mauritius % 4.2% 4.2% 51.3% 45.6% 45.6% 45.6% Mexico 1,196 1,419 1, % 3.1% 3.0% 42.9% 43.1% 43.1% 43.1% Moldova % 5.5% 5.5% 22.5% 15.8% 15.8% 15.8% Mongolia % 9.1% 9.1% 0.0% 0.0% 0.0% 0.0% Montenegro % 2.2% 2.2% 48.9% 58.0% 54.7% 50.0% Morocco % 5.9% 5.9% 56.0% 53.9% 52.2% 50.0% Namibia % 4.4% 4.4% 25.3% 32.1% 32.1% 32.1% Nicaragua % 4.0% 4.0% 68.7% 51.2% 50.7% 50.0% Nigeria % 6.2% 6.0% 18.3% 18.5% 18.5% 18.5%

19 EMERGING MARKETS Foreign Currency Debt (% of total) Gross National Income per Capita (USD) JPM Local Currency Benchmark incl./excl. JPM Hard Currency Benchmark incl./excl Albania 42.9% 41.3% 39.7% 4,030 4,459 5,004 IN Angola 41.5% 37.7% 33.9% 4,250 5,093 6,083 IN Argentina 60.0% 49.1% 38.1% 9,613 11,369 13,632 BM Armenia 86.2% 77.5% 68.9% 3,449 4,152 5,049 IN Azerbaijan 46.4% 43.1% 39.7% 5,364 5,808 6,477 IN Bangladesh 44.8% 43.6% 42.3% 767 1,008 1,359 Belarus 89.4% 75.4% 61.3% 6,502 8,225 10,674 BM Belize 84.0% 77.6% 71.1% 3,856 3,963 4,105 BM Bolivia 63.9% 59.2% 54.5% 1,935 2,286 2,705 Bosnia and Herzegovina 63.0% 56.7% 50.4% Botswana 86.2% 73.0% 59.8% 7,122 8,449 10,019 IN Brazil 7.0% 6.0% 5.0% 9,767 11,472 13,348 BM BM Bulgaria 75.1% 62.5% 49.9% 6,520 7,929 10,250 BM Cambodia 98.0% 88.6% 79.1% 826 1,104 1,486 Chile 20.1% 12.6% 5.0% 10,982 13,150 15,852 BM OUT BM OUT China 0.7% 0.7% 0.7% 5,003 7,074 9,572 IN BM Colombia 26.4% 23.7% 21.0% 5,951 6,970 8,243 BM BM Costa Rica 18.3% 15.5% 12.7% 7,170 8,360 9,903 IN Côte d'ivoire 0.0% 0.0% 0.0% 1,145 1,410 1,756 BM Croatia 70.4% 59.7% 49.0% 13,864 15,383 17,619 BM OUT Czech Republic 18.3% 11.7% 5.0% 18,083 20,973 24,743 Dominican Republic 72.1% 63.5% 54.9% 5,354 6,420 7,800 BM Ecuador 100.0% 90.1% 80.1% 4,218 4,729 5,292 BM Egypt 15.0% 15.0% 15.0% 2,416 2,917 3,734 IN BM El Salvador 100.0% 89.5% 79.0% 3,456 3,916 4,489 BM Fiji 14.6% 14.6% 14.6% 3,693 3,956 4,333 Gabon 0.0% 0.0% 0.0% 8,224 8,468 8,678 BM Georgia 90.2% 78.1% 66.0% 3,084 4,162 5,645 BM Ghana 0.0% 0.0% 0.0% 1,475 1,798 2,141 IN BM Guatemala 53.0% 50.9% 48.9% 2,787 2,907 3,076 IN Honduras 0.0% 0.0% 1,926 2,114 2,362 Hong Kong SAR 0.0% 0.0% 0.0% 34,590 40,566 47,952 Hungary 44.3% 37.6% 30.8% 13,124 14,671 16,485 BM OUT BM OUT Iceland 30.2% 17.6% 5.0% 33,596 36,136 39,612 India 4.0% 4.0% 4.0% 1,416 1,925 2,535 IN Indonesia 36.8% 33.1% 29.3% 2,767 3,689 4,839 BM BM Iraq 0.0% 0.0% 0.0% 2,686 3,762 4,609 BM Israel 22.8% 14.9% 7.1% 28,169 31,247 34,559 Jamaica 41.6% 40.8% 39.9% 4,886 5,115 5,435 BM Jordan 31.8% 31.5% 31.2% 4,378 4,825 5,543 BM Kazakhstan 21.2% 13.1% 5.0% 8,440 10,868 14,207 BM Kenya 0.0% 0.0% 0.0% 848 1,007 1,222 IN IN Korea 1.5% 1.5% 1.5% 21,166 25,290 30,345 Latvia 78.5% 61.8% 45.2% 12,625 15,199 18,867 Lebanon 42.0% 35.8% 29.6% 9,145 10,753 12,734 IN BM Lithuania 72.0% 55.3% 38.5% 12,550 15,302 18,900 BM OUT Malaysia 3.7% 3.7% 3.7% 8,250 9,733 11,599 BM BM Mauritius 20.2% 16.4% 12.6% 8,372 10,021 12,112 Mexico 23.5% 21.5% 19.5% 9,393 10,564 11,788 BM BM Moldova 79.1% 71.1% 63.1% 2,024 2,688 3,612 Mongolia 71.5% 59.7% 48.0% 2,491 3,775 5,475 IN Montenegro 15.5% 15.5% 15.5% 6,905 7,588 8,458 Morocco 24.0% 23.6% 23.1% 3,021 3,709 4,750 IN BM Namibia 26.8% 25.2% 23.6% 4,695 5,360 6,185 BM Nicaragua 87.1% 79.9% 72.7% 1,171 1,331 1,530 IN Nigeria 0.0% 0.0% 0.0% 1,343 1,625 1,944 IN BM 17

20 EMERGING MARKETS GDP (USD billion, 2012 prices) Real GDP Growth Debt-to-GDP Government Debt (USD billion) LT Trend LT Trend Pakistan % 3.5% 3.5% 61.7% 53.2% 51.9% 50.0% Panama % 5.1% 5.1% 36.0% 25.4% 25.4% 25.4% Papua New Guinea % 4.9% 4.9% 0.0% 0.0% 0.0% 0.0% Paraguay % 4.7% 4.7% 13.5% 8.3% 8.3% 8.3% Peru % 5.4% 5.0% 20.7% 17.6% 17.6% 17.6% Philippines % 5.0% 5.0% 40.1% 33.2% 33.2% 33.2% Poland % 3.6% 3.5% 55.7% 48.7% 48.7% 48.7% Romania % 4.0% 4.0% 34.2% 27.4% 27.4% 27.4% Russia 1,925 2,330 2, % 3.3% 3.0% 8.4% 11.0% 11.0% 11.0% Senegal % 5.4% 5.4% 43.7% 46.9% 46.9% 46.9% Serbia % 3.5% 3.5% 53.7% 47.9% 47.9% 47.9% South Africa % 3.7% 3.7% 40.0% 36.4% 36.4% 36.4% Sri Lanka % 6.5% 6.5% 0.0% 0.0% 0.0% 0.0% St. Vincent and the Grenadines Taiwan Province of China % 3.5% 3.5% 70.4% 60.4% 56.0% 50.0% % 5.0% 5.0% 42.5% 33.9% 33.9% 33.9% Tanzania % 7.0% 7.0% 47.7% 50.1% 50.0% 50.0% Thailand 5.3% 4.4% 4.0% 44.4% 51.7% 51.0% 50.0% The Bahamas % 2.5% 2.5% 49.9% 0.0% 0.0% 0.0% Trinidad and Tobago % 2.6% 2.6% 37.3% 47.7% 47.7% 47.7% Tunisia % 6.7% 6.7% 43.5% 45.7% 45.7% 45.7% Turkey , % 4.2% 4.0% 36.0% 31.4% 31.4% 31.4% Uganda % 7.0% 7.0% 29.4% 26.2% 26.2% 26.2% Ukraine % 3.5% 3.5% 35.9% 33.2% 33.2% 33.2% Uruguay % 4.0% 4.0% 49.0% 36.9% 36.9% 36.9% Venezuela % 2.9% 2.9% 51.6% 76.8% 65.6% 50.0% Vietnam % 7.5% 7.5% 37.3% 32.3% 32.3% 32.3% Zambia % 7.7% 7.7% 26.6% 17.9% 17.9% 17.9% DEVELOPED MARKETS LT Trend LT Trend Australia 1,533 1,821 2, % 3.5% 3.5% 24.0% 14.9% 14.9% 14.9% Austria % 1.8% 1.8% 73.9% 69.2% 69.2% 69.2% Belgium % 1.7% 1.7% 99.1% 87.5% 87.5% 87.5% Canada 1,773 1,986 2, % 2.2% 2.2% 84.7% 73.6% 73.6% 73.6% 1,501 1,461 1,632 Denmark % 1.8% 1.8% 51.3% 46.1% 46.1% 46.1% Finland % 2.0% 2.0% 51.6% 54.3% 54.3% 54.3% France 2,790 3,041 3, % 2.0% 2.0% 89.0% 84.6% 84.6% 84.6% 2,483 2,573 2,842 Germany 3,599 3,842 4, % 1.3% 1.3% 78.9% 71.1% 71.1% 71.1% 2,838 2,731 2,910 Greece % 2.9% 2.9% 153.2% 136.8% 117.3% 90.0% Iceland % 2.9% 2.9% 97.3% 81.7% 81.7% 81.7% Ireland % 2.9% 2.9% 113.1% 109.2% 101.2% 90.0% Italy 2,157 2,236 2, % 1.2% 1.2% 123.4% 118.9% 106.9% 90.0% 2,661 2,659 2,536 Japan 5,989 6,409 6, % 1.1% 1.1% 235.8% 256.6% 233.0% 200.0% 14,124 16,446 15,804 Korea 1,156 1,405 1, % 4.0% 4.0% 32.9% 23.1% 23.1% 23.1% Luxembourg % 3.1% 3.1% 23.8% 35.6% 35.6% 35.6% Netherlands % 1.9% 1.9% 70.1% 78.6% 78.6% 78.6% New Zealand % 2.3% 2.3% 36.0% 31.5% 31.5% 31.5% Norway % 2.1% 2.1% 49.6% 49.6% 49.6% 49.6% Portugal % 1.5% 1.5% 112.4% 109.2% 101.2% 90.0% Singapore % 4.0% 4.0% 98.0% 89.0% 89.0% 89.0% Spain 1,466 1,564 1, % 1.8% 1.8% 79.0% 91.9% 91.1% 90.0% 1,159 1,437 1,561 Sweden % 2.4% 2.4% 35.5% 20.4% 20.4% 20.4% Switzerland % 1.9% 1.9% 48.9% 43.9% 43.9% 43.9% United Kingdom 2,437 2,764 3, % 2.6% 2.5% 88.4% 86.8% 86.8% 86.8% 2,154 2,398 2,731 United States 15,412 17,943 20, % 2.8% 2.5% 106.6% 113.0% 103.4% 90.0% 16,429 20,275 21,338 18