As GMO looks forward to its 20th year managing emerging debt portfolios, we offer our perspectives on the frequently-asked questions that have come up over the years, including:
• What is meant by emerging debt (external, local, corporate)?
• Why and when to own it: portfolio fit considerations, alpha, and absolute and relative value.
• How to own it: dedicated external, local, or corporate; "blended"; or "multi asset" (including emerging equities). What is meant by emerging debt (external, local, corporate)?
"Emerging debt" is a mixed bag of U.S. dollar, local currency, and corporate debt, and what's considered "emerging" varies across the sub-types and across time.1 Below are the most important features helpful for our discussions about value. Importantly, credit rating is not used as a criterion for any of the types, different from the common distinction in credit markets among high-yield and investment-grade issuers. We focus on the features of the benchmark used most commonly for the type of debt.
Emerging external debt is issued by countries with material default risk, in foreign currency (generally U.S. dollars) under foreign jurisdiction. The relevant index is the J.P. Morgan Emerging Markets Bond Index Global (EMBIG).2 J.P. Morgan defines "emerging" for EMBIG to be low- and middle-income countries as well as those that have restructured their external or local debt during the past 10 years or currently have outstanding restructured external or local debt. Therefore, either because of the income status or default history, it is expected that external debt has material default risk. The EMBIG's average S&P rating is currently BBB-, and there are 46 countries in the index. In addition to sovereign debt, debt issued by companies either 100% government owned or whose debts are 100% government guaranteed is also included ("quasi sovereigns").
Emerging local currency debt is denominated in the local currency of the issuer, regardless of jurisdiction. The relevant family of indexes is the J.P. Morgan Global Bond Index-Emerging (GBI-EM), of which there are three main types based on their investability to foreigners: GBI-EM, GBI-EM Global, GBI-EM Broad. All three are restricted to low- and middle-income countries, as with EMBIG, although there's no criterion related to a country's default history.
The GBI-EM Broad includes all eligible debt (fixed-rate and zero-coupon), regardless of investability to foreigners (16 countries). The GBI-EM Global restricts the universe to only those countries accessible by a majority of foreign investors (14 countries), and the GBI-EM further restricts the universe to that subset freely accessible by foreigners (12 countries).3 Due to the small number of countries, some of which would have large market-capitalization-based weights, investors focus on the "diversified" versions of these, which cap each country at 10%. The average S&P index ratings of the three (diversified) versions currently are: GBI-EM Broad (A-), GBI-EM Global (BBB+), GBI-EM (BBB+).
Emerging corporate debt is issued by corporates domiciled in emerging countries. So far, the relevant index is the J.P. Morgan Corporate Emerging Markets Bond Index Broad Diversified (CEMBIB-D), although this exclusively captures the U.S. dollar (USD) segment (ignoring, as EMBIG does, other major currencies like euros or yen, and completely ignoring local currency corporates). Interestingly, J.P. Morgan departs from its EMBIG/GBI-EM method for defining "emerging countries" in this context, adopting a regional-based approach. In CEMBIB-D's case, companies headquartered in Latin America, Eastern Europe, Middle East, Africa, or Asia ex-Japan are considered eligible, as are those with 100% of their operations there (as long as the bonds are guaranteed by the local operation). Quasi-sovereigns of the variety eligible for EMBIG are specifically excluded, but a company can migrate from one index to the other based on nationalizations/privatizations. The index contains issues from 32 countries spanning AAA-rated Singapore down to single-B rated Venezuela and Jamaica. The average S&P rating is currently BBB.
Why and when to own it: portfolio fit considerations, alpha, and absolute and relative value Portfolio fit
As with any risk asset, the time to own it in risk-seeking portfolios is when its prospective returns adequately compensate you for its risks, taking into account the relationship with other risky assets you already own or may want to buy ("portfolio fit").
We are assuming emerging debt is fair or under-valued, an assumption we'll return to. From a portfolio fit perspective, we observe:
• As shown in Exhibit 1, emerging external debt has a lower statistical correlation than local currency debt with other common portfolio holdings (equities, credit, global bonds). External debt's lower correlation is mostly explained by offsetting responses of its two main drivers (spreads and USD rates) to other risky assets. Spreads generally rise and fall in tandem with other risk assets, but U.S. rates generally move in the opposite direction, muting the asset class's total response to moves in risk assets. Local debt, on the other hand, doesn't have this dampening USD rate exposure as its two main drivers are FX and local bonds. The FX tend to move with risk assets in the way credit spreads do on external, but local emerging bonds' responses have varied over time. In severe crises, they act like risk assets, compounding the correlations; other times they act more like "rates" markets, with a similar dampening effect as the USD rates in external debt. Therefore, if statistical correlation is a factor important in your determination of portfolio fit, it's more likely that external debt will be your answer if you already have exposures to typical risk investments.
• If, instead, a strategic need to diversify away from the USD (or, via hedging, other major currencies) is important for your portfolio fit, then emerging local currency debt is more likely your answer. External debt doesn't help on this.
• Investors interested in emerging corporate debt need to examine ways they may already have such exposure in their portfolios. Increasingly the corporate bond management practice, whether investment grade or high yield, is going global at least in opportunity set, and a number of emerging corporate issuers find themselves there.
So, if you're looking for something fundamentally different, check to see if you already have it. Another way you may already have exposure to this set of issuers is via your emerging equities holdings, and here the issuer overlap is very high indeed. Box 1 details our observations on this topic.
Alpha
Emerging debt is one of the asset classes in which active managers have most consistently generated alpha over the past two decades. Exhibit 2 shows the longrun returns for median and 5th percentile managers tracked by eVestment as well as the benchmark. We use the external debt asset class to make this case, since it has a longer history for active management. If ever there were an asset class in need of active management, this is it. Of course some of this outperformance is due to deficiencies in the benchmarks for the asset class, which suffer from all of the known ones listed by one of my colleagues in his paper "Bond Benchmark Baloney."
The asset class is stunningly inefficient, likely due to the heterogeneous mix of its active investors. Relatively patriotic (or at least captive) domestic investors drive prices for their own country's bonds to irrational heights, and total-return oriented investors can outperform simply by avoiding them once that happens. "Crossover" buyers (from domestic or global aggregate or, lately, global high yield) of fixed income styles arrive and leave in herds, often with poor timing. In local currency debt, capital controls and other regulations segment the markets. Therefore, sensible, value-oriented investors can buy the bits of the market that are cheap, and avoid those that aren't. We include ourselves in the class of managers that are simply "benchmark aware," regarding tracking error as a senseless indicator of risk.
(For this reason alone, we think ETFs make no sense in this market, because they buy "what's available" rather than "what's cheap." Anyway, emerging debt ETF fees are hardly a bargain relative to the discounts available in equities.4)
This is why GMO's Asset Allocation team includes the expected alpha to its assessment of the attractiveness of emerging debt, since emerging debt alpha has been higher and more persistent.
Absolute and relative value
To determine absolute value, we begin with the prevailing yields on the three sub-classes, divide those up into payment for known risk factors, and then ask ourselves: is it enough?
Box 2 identifies the principal risks disclosed in typical offering documents. When examining absolute value, we're most concerned with the long-term drivers of return, which are credit risk for external and corporate bonds, and currency valuation risk for local currency bonds. Over shorter horizons, which we'll discuss in the upcoming section on relative value, other factors (credit spread volatility for external and corporate debt, currency volatility for local debt, and liquidity for all three) can dominate, but generally these are short-lived.
When examining the absolute value of credit spreads, we need to evaluate whether their current level adequately compensates the investor for the expected losses due to default. Given that neither the sovereign nor the corporate index is defined with respect to credit rating, there's no reason that expected losses should be stable over time. Looking at the average credit rating of EMBIG over time, Exhibit 3 shows that there's been a general upward drift, meaning expected losses have been declining.
However, newer issuers have tended to be lower quality ones (in 2011, for example, single- or double-B rated Jordan, Senegal, Namibia, and Nigeria debuted in EMBIG), so this drift is periodically counteracted through compositional effects. This has also been the case in CEMBIB-D, where recently the "reach for yield" mentality of the market has allowed junkier credits to issue into the index. Therefore, we examine the credit spread offered relative to the expected loss of the index constituents at the time to gauge "excess" spread over expected losses. This is as laborious as it sounds, but we believe it is the correct approach to evaluating credit for its long-term fundamentals.5
What we find for EMBIG is that there has been a lot of variability in the "extra" spread one has received over the expected credit losses. Sometimes it is very high – say, after the Lehman collapse – and sometimes it is low – say, summer of 2007 when our Asset Allocation team was screaming that everything was overpriced. Looking back over history, the average "extra" spread is 300% over and above the EMBIG's expected default loss. Given EMBIG's current composition, we compute the expected default losses to be 68 bps per annum, which means that an average "fair" spread for EMBIG is 4*68 = 272 bps. With EMBIG's current spread near 400 bps, we'd say this represents good value – from a credit perspective.
We haven't made a similar computation for CEMBIB-D, although given the extra risks outlined in Box 2, as conservative credit managers we'd be inclined to say that an extra 100% on top of EMBIG's 300% sounds reasonable. If the extra risks weren't enough to justify this extra margin of safety, we'd instead appeal again to our "portfolio fit" reasoning and say there's an extra hurdle for adding something so fundamentally indistinguishable from emerging equities.
For local currency debt, the primary risk factor is currency risk rather than credit risk. We'll focus here just on currency valuation risk: the risk that the investor will suffer declines in the value of local-currency denominated assets if they are purchased when the currency is overvalued. This is the most directly comparable risk factor to the expected loss due to the credit question in EMBIG and CEMBIB-D. The fact is, it's hard to estimate a currency's fundamental value, although of course we take a stab at it when constructing our local debt portfolio. Luckily, we can appeal to the fundamental relationship between a country's credit spread and its currency level to back it out via relative value to external debt. Next, we cross-reference this with our independent estimate of the currency's fundamental value. So, rather than define an absolute level of yield that is "enough" for local currency debt, we'll make the case relative to external debt.
Also luckily, this is, in fact, the question we're asked most often: which is better to buy right now, external or local? Box 3 highlights the relationship between credit spreads and FX, showing that, at the country level, the relationship is strong, and rarely is there "relative value." In other words, if we can show that credit spreads are cheap (using the expected loss method above), it's likely the case that the currency is cheap, too.
But the comparison is never asked at the country level. Instead it's asked at the asset class level. This is a harder question to answer chiefly because of compositional differences (remember EMBIG has 46 countries and GBI-EMD 12, an "apples to pears" issue). Also, EMBIG contains bonds rather than pure credit instruments (CDS), and bonds themselves can diverge from the same-country CDS (the so-called bond-CDS "basis"). Therefore, to answer the question "which is better, relative value external or local?" you need to know:
• Which kind of bond (external or local) offers the best credit spread when fully hedged of its other risks (USD interest-rate risk for external, and currency and local interest-rate risk for local)? This isolates the credit dimension and is achieved mathematically by factoring the costs of the various hedging dimensions. Exhibit 4 illustrates four ways to take Hungarian credit exposure, for example. Some assumptions will be required, but because the assumptions are used the same way for all of the bonds, they shouldn't have an effect on the estimation of relative value; • Is there a relative value discrepancy between the level of the currency and the level of credit spreads for the country? This can be estimated empirically as shown in Box 3; Hungary's plot is in Figure 3.1. The main assumption is that of mean reversion: that if the FX is exceptionally weak or strong for the current level of credit spreads, the relationship will revert to the mean over time (generally the time horizon here is quite short, usually hours or days, absent a credit event).
Don't be discouraged, though. There are occasional switching opportunities between external and local currency debt that survive the large transactions costs of doing so, usually motivated by differential asset class flows. For the reasons outlined above, these can't be seen as "pure" relative value but rather compositionally-motivated statistical relative value. Exhibit 5 depicts the credit spread of the EMBIG and the FXrebased level of the GBI-EMD. Notice that occasionally there are outliers where mean reversion might be expected (but again, not guaranteed by fundamentals due to the differential compositions).
How to own it: dedicated external, local, or corporate; "blended"; or "multi asset" (including emerging equities)
The next question we're generally asked is whether the best way to own emerging debt is through dedicated funds targeting one of the sub-classes or via "blended" strategies that bundle any or all of the above. We believe that for those investors with the time and ability to manage their asset allocations actively, owning the individual sub-classes makes most sense, because it gives the investor the most freedom to align his investments according to his portfolio fit. Some investors would prefer to give the manager timing authority, and we're fine with that, too. For the reasons just mentioned, though, we temper these investors' expectations with respect to how active such a switching program likely will be.
"Multi-asset" strategies extend the mix to include emerging equities. These are targeted at investors who would like to have more "emerging" exposure but are worried about adding to their emerging equities given the volatility. Based on research by our Asset Allocation team, we believe that such strategies should be able to offer lower overall volatility with similar returns to equities through the judicious use of somewhat less volatile emerging debt.
Summary Emerging debt is a mixed bag of U.S. dollar, local currency, and corporate debt. The dollar debt has a broad group of countries. The local debt has a narrow group of many of the same countries. Corporate debt is essentially "emerging equities lite."
Any or all should be owned in risk-seeking portfolios when they're cheap to their fundamentals in proportion either to their relative value or portfolio fit.
They can be owned in an unbundled fashion, which gives the investor greater flexibility on the portfolio fit angle. Or, they can be owned in various types of pre-arranged bundles, where the manager then has flexibility to focus on relative value in a portfolio-fit agnostic fashion.
Box 1 Emerging Corporates: Chances Are, You Already Own Them!
For investors with long horizons, the fact that the statistical correlation between two assets you might own is high is less relevant the more fundamentally different they are. A classic example is the high statistical correlation that catastrophe bonds have with the S&P when the S&P is declining precipitously. However, nothing about the S&P's decline has raised the likelihood of, say, an earthquake in California, which ought to have a negative impact on a California earthquake-related catastrophe bond. They are fundamentally different assets.
Corporate bonds and equities are fundamentally the same thing, except that corporate bonds are a more senior claim and have term-structure exposure, and only through that fact is their fundamental and statistical relationship diluted. Emerging corporate bonds are no different in this regard.
Think of emerging markets corporate bonds as a hybrid instrument combining issuer-specific default risk and USD term structure risk. Unbundled this way, it's clear that the issuer-specific risks ought to be compared with other like risks, which in this case would be emerging equities, while USD term-structure risk can be evaluated relative to U.S. interest-rate swaps or U.S. Treasuries.
From a portfolio fit perspective, the issuerspecific risk ought to have at least one of the following properties: (1) it's the better value relative to its equity counterpart; or (2) it's diversifying given that it exists in circumstances where public equity doesn't exist for that issuer. On the first issue, we're often amused that the comparable presented for corporate bonds in general (not just emerging ones) is Treasuries or perhaps other forms of debt (loans, for example) rather than equities. Our hypothesis is that corporate bonds allow investors to "sneak in" more equity risk when they're constrained at the policy level from doing so.
As value-oriented investors, we find this uncompelling from an asset-allocation perspective: buy what's cheap. On the second issue, we took a peek at the composition by country, by industry, and by issuer of the CEMBIB-D and the MSCI-EM equity index. This is what we found.
By country (Figure 1.1) the name overlap is high, although CEMBIB-D caps country weights and so is a bit less lumpy. BRICs plus Mexico and Korea comprise 40 and 65%, respectively. By industry (Figure 1.2)
– allowing for the different industry classifications mysteriously propagated by the debt vs. equity underwriters – the overlap is also high. Banks (boy, do the underwriters like to underwrite their colleagues' issues) are the biggest in both. Among the larger countries in both indexes, the overlap by individual issuer (Figure 1.3) is amazingly high, and we note with some alarm that perpetual bonds are included in CEMBIB-D. Perps are about as close to equity as we bond folks can get. We've also noted in Figure 1.3 the bonds included in investmentgrade and high yield indexes. Chances are, you have it!
Continue reading.
1 In fact, some on the sell-side, trying to distinguish "emerging" from the more fiscally challenged "developed" or "advanced" country counterparts, are promoting a less pejorative re-branding. Instead of "emerging," these are "growth" markets rather than "mature" markets. "Frontier" markets are being rebranded "Next Generation" markets in an attempt to remove the Wild West image. Regardless of the branding, we view these as "risk" assets not "risk-free" or even "relatively less risky than mature country debt" markets. Lucky for us, however, it's generally been the case that investors in "emerging" have been more than compensated for the risks entailed.
2 J.P. Morgan separately quotes a EUR EMBIG and prior to May 2002 had included local law instruments. The asset class has its origins in the bonds issued out of restructured bank loans from the 1980s and 1990s ("Brady bonds"), although today Brady bonds are only 5% of the EMBIG market value.
3 The local debt index these days contains just fixed-rate (including zero-coupon) bonds, although in the past it had included Chilean inflation-linked bonds. Barclays produces a global inflation-linked index, including emerging issuers Chile, Korea, South Africa, Mexico, Brazil, Colombia, Poland, and Turkey, although Brazil dominates the index by a wide margin. In local debt there is also a host of very interesting other instruments: floating-rate notes, foreigncurrency linked, etc., that are not currently captured by the index.
4 Another weakness of the emerging debt ETF is its need to provide a high level of liquidity to buyers of the product. As you will see later in Box 2, Figure 2.2, emerging debt at best has poor liquidity, and sometimes it is truly dreadful. No investment vehicle can make an illiquid underlying market liquid, although it can try by using, as emerging debt ETFs do, restricted benchmarks that include only the most liquid (or, in the case of local debt, most accessible) bonds. Investors are likely to be surprised when the discount to NAV rises sharply during a market disruption. Instead, funds ought to carry purchase and redemption fees (paid to the fund shareholders) to highlight to investors the need to be long-term oriented in the asset class. After all, separate accounts pay these implicitly when they buy assets at the offer and sell at the bid. Figure 2.2 shows the large and sometimes very large transactions costs inherent in emerging debt markets.
5 Fortunately for us, because we don’t believe in the portfolio fit of corporates, we’ve saved ourselves some labor on this score!
Disclaimer: The views expressed are the views of Tina Vandersteel through the period ending 5/29/12 and are subject to change at anytime based on market and other conditions. This is not an offer or solicitation for the purchase or sale of any security. The article may contain some forward looking statements. There can be no guarantee that any forward looking statement will be realized. GMO undertakes no obligation to publicly update forward looking statements, whether as a result of new information, future events or otherwise. Statements concerning financial market trends are based on current market conditions, which will fluctuate. References to securities and/or issuers are for illustrative purposes only. References made to securities or issuers are not representative of all of the securities purchased, sold or recommended for advisory clients, and it should not be assumed that the investment in the securities was or will be profitable. There is no guarantee that these investment strategies will work under all market conditions, and each investor should evaluate the suitability of their investments for the long term, especially during periods of downturns in the markets.
• What is meant by emerging debt (external, local, corporate)?
• Why and when to own it: portfolio fit considerations, alpha, and absolute and relative value.
• How to own it: dedicated external, local, or corporate; "blended"; or "multi asset" (including emerging equities). What is meant by emerging debt (external, local, corporate)?
"Emerging debt" is a mixed bag of U.S. dollar, local currency, and corporate debt, and what's considered "emerging" varies across the sub-types and across time.1 Below are the most important features helpful for our discussions about value. Importantly, credit rating is not used as a criterion for any of the types, different from the common distinction in credit markets among high-yield and investment-grade issuers. We focus on the features of the benchmark used most commonly for the type of debt.
Emerging external debt is issued by countries with material default risk, in foreign currency (generally U.S. dollars) under foreign jurisdiction. The relevant index is the J.P. Morgan Emerging Markets Bond Index Global (EMBIG).2 J.P. Morgan defines "emerging" for EMBIG to be low- and middle-income countries as well as those that have restructured their external or local debt during the past 10 years or currently have outstanding restructured external or local debt. Therefore, either because of the income status or default history, it is expected that external debt has material default risk. The EMBIG's average S&P rating is currently BBB-, and there are 46 countries in the index. In addition to sovereign debt, debt issued by companies either 100% government owned or whose debts are 100% government guaranteed is also included ("quasi sovereigns").
Emerging local currency debt is denominated in the local currency of the issuer, regardless of jurisdiction. The relevant family of indexes is the J.P. Morgan Global Bond Index-Emerging (GBI-EM), of which there are three main types based on their investability to foreigners: GBI-EM, GBI-EM Global, GBI-EM Broad. All three are restricted to low- and middle-income countries, as with EMBIG, although there's no criterion related to a country's default history.
The GBI-EM Broad includes all eligible debt (fixed-rate and zero-coupon), regardless of investability to foreigners (16 countries). The GBI-EM Global restricts the universe to only those countries accessible by a majority of foreign investors (14 countries), and the GBI-EM further restricts the universe to that subset freely accessible by foreigners (12 countries).3 Due to the small number of countries, some of which would have large market-capitalization-based weights, investors focus on the "diversified" versions of these, which cap each country at 10%. The average S&P index ratings of the three (diversified) versions currently are: GBI-EM Broad (A-), GBI-EM Global (BBB+), GBI-EM (BBB+).
Emerging corporate debt is issued by corporates domiciled in emerging countries. So far, the relevant index is the J.P. Morgan Corporate Emerging Markets Bond Index Broad Diversified (CEMBIB-D), although this exclusively captures the U.S. dollar (USD) segment (ignoring, as EMBIG does, other major currencies like euros or yen, and completely ignoring local currency corporates). Interestingly, J.P. Morgan departs from its EMBIG/GBI-EM method for defining "emerging countries" in this context, adopting a regional-based approach. In CEMBIB-D's case, companies headquartered in Latin America, Eastern Europe, Middle East, Africa, or Asia ex-Japan are considered eligible, as are those with 100% of their operations there (as long as the bonds are guaranteed by the local operation). Quasi-sovereigns of the variety eligible for EMBIG are specifically excluded, but a company can migrate from one index to the other based on nationalizations/privatizations. The index contains issues from 32 countries spanning AAA-rated Singapore down to single-B rated Venezuela and Jamaica. The average S&P rating is currently BBB.
Why and when to own it: portfolio fit considerations, alpha, and absolute and relative value Portfolio fit
As with any risk asset, the time to own it in risk-seeking portfolios is when its prospective returns adequately compensate you for its risks, taking into account the relationship with other risky assets you already own or may want to buy ("portfolio fit").
We are assuming emerging debt is fair or under-valued, an assumption we'll return to. From a portfolio fit perspective, we observe:
• As shown in Exhibit 1, emerging external debt has a lower statistical correlation than local currency debt with other common portfolio holdings (equities, credit, global bonds). External debt's lower correlation is mostly explained by offsetting responses of its two main drivers (spreads and USD rates) to other risky assets. Spreads generally rise and fall in tandem with other risk assets, but U.S. rates generally move in the opposite direction, muting the asset class's total response to moves in risk assets. Local debt, on the other hand, doesn't have this dampening USD rate exposure as its two main drivers are FX and local bonds. The FX tend to move with risk assets in the way credit spreads do on external, but local emerging bonds' responses have varied over time. In severe crises, they act like risk assets, compounding the correlations; other times they act more like "rates" markets, with a similar dampening effect as the USD rates in external debt. Therefore, if statistical correlation is a factor important in your determination of portfolio fit, it's more likely that external debt will be your answer if you already have exposures to typical risk investments.
• If, instead, a strategic need to diversify away from the USD (or, via hedging, other major currencies) is important for your portfolio fit, then emerging local currency debt is more likely your answer. External debt doesn't help on this.
• Investors interested in emerging corporate debt need to examine ways they may already have such exposure in their portfolios. Increasingly the corporate bond management practice, whether investment grade or high yield, is going global at least in opportunity set, and a number of emerging corporate issuers find themselves there.
So, if you're looking for something fundamentally different, check to see if you already have it. Another way you may already have exposure to this set of issuers is via your emerging equities holdings, and here the issuer overlap is very high indeed. Box 1 details our observations on this topic.
Alpha
Emerging debt is one of the asset classes in which active managers have most consistently generated alpha over the past two decades. Exhibit 2 shows the longrun returns for median and 5th percentile managers tracked by eVestment as well as the benchmark. We use the external debt asset class to make this case, since it has a longer history for active management. If ever there were an asset class in need of active management, this is it. Of course some of this outperformance is due to deficiencies in the benchmarks for the asset class, which suffer from all of the known ones listed by one of my colleagues in his paper "Bond Benchmark Baloney."
The asset class is stunningly inefficient, likely due to the heterogeneous mix of its active investors. Relatively patriotic (or at least captive) domestic investors drive prices for their own country's bonds to irrational heights, and total-return oriented investors can outperform simply by avoiding them once that happens. "Crossover" buyers (from domestic or global aggregate or, lately, global high yield) of fixed income styles arrive and leave in herds, often with poor timing. In local currency debt, capital controls and other regulations segment the markets. Therefore, sensible, value-oriented investors can buy the bits of the market that are cheap, and avoid those that aren't. We include ourselves in the class of managers that are simply "benchmark aware," regarding tracking error as a senseless indicator of risk.
(For this reason alone, we think ETFs make no sense in this market, because they buy "what's available" rather than "what's cheap." Anyway, emerging debt ETF fees are hardly a bargain relative to the discounts available in equities.4)
This is why GMO's Asset Allocation team includes the expected alpha to its assessment of the attractiveness of emerging debt, since emerging debt alpha has been higher and more persistent.
Absolute and relative value
To determine absolute value, we begin with the prevailing yields on the three sub-classes, divide those up into payment for known risk factors, and then ask ourselves: is it enough?
Box 2 identifies the principal risks disclosed in typical offering documents. When examining absolute value, we're most concerned with the long-term drivers of return, which are credit risk for external and corporate bonds, and currency valuation risk for local currency bonds. Over shorter horizons, which we'll discuss in the upcoming section on relative value, other factors (credit spread volatility for external and corporate debt, currency volatility for local debt, and liquidity for all three) can dominate, but generally these are short-lived.
When examining the absolute value of credit spreads, we need to evaluate whether their current level adequately compensates the investor for the expected losses due to default. Given that neither the sovereign nor the corporate index is defined with respect to credit rating, there's no reason that expected losses should be stable over time. Looking at the average credit rating of EMBIG over time, Exhibit 3 shows that there's been a general upward drift, meaning expected losses have been declining.
However, newer issuers have tended to be lower quality ones (in 2011, for example, single- or double-B rated Jordan, Senegal, Namibia, and Nigeria debuted in EMBIG), so this drift is periodically counteracted through compositional effects. This has also been the case in CEMBIB-D, where recently the "reach for yield" mentality of the market has allowed junkier credits to issue into the index. Therefore, we examine the credit spread offered relative to the expected loss of the index constituents at the time to gauge "excess" spread over expected losses. This is as laborious as it sounds, but we believe it is the correct approach to evaluating credit for its long-term fundamentals.5
What we find for EMBIG is that there has been a lot of variability in the "extra" spread one has received over the expected credit losses. Sometimes it is very high – say, after the Lehman collapse – and sometimes it is low – say, summer of 2007 when our Asset Allocation team was screaming that everything was overpriced. Looking back over history, the average "extra" spread is 300% over and above the EMBIG's expected default loss. Given EMBIG's current composition, we compute the expected default losses to be 68 bps per annum, which means that an average "fair" spread for EMBIG is 4*68 = 272 bps. With EMBIG's current spread near 400 bps, we'd say this represents good value – from a credit perspective.
We haven't made a similar computation for CEMBIB-D, although given the extra risks outlined in Box 2, as conservative credit managers we'd be inclined to say that an extra 100% on top of EMBIG's 300% sounds reasonable. If the extra risks weren't enough to justify this extra margin of safety, we'd instead appeal again to our "portfolio fit" reasoning and say there's an extra hurdle for adding something so fundamentally indistinguishable from emerging equities.
For local currency debt, the primary risk factor is currency risk rather than credit risk. We'll focus here just on currency valuation risk: the risk that the investor will suffer declines in the value of local-currency denominated assets if they are purchased when the currency is overvalued. This is the most directly comparable risk factor to the expected loss due to the credit question in EMBIG and CEMBIB-D. The fact is, it's hard to estimate a currency's fundamental value, although of course we take a stab at it when constructing our local debt portfolio. Luckily, we can appeal to the fundamental relationship between a country's credit spread and its currency level to back it out via relative value to external debt. Next, we cross-reference this with our independent estimate of the currency's fundamental value. So, rather than define an absolute level of yield that is "enough" for local currency debt, we'll make the case relative to external debt.
Also luckily, this is, in fact, the question we're asked most often: which is better to buy right now, external or local? Box 3 highlights the relationship between credit spreads and FX, showing that, at the country level, the relationship is strong, and rarely is there "relative value." In other words, if we can show that credit spreads are cheap (using the expected loss method above), it's likely the case that the currency is cheap, too.
But the comparison is never asked at the country level. Instead it's asked at the asset class level. This is a harder question to answer chiefly because of compositional differences (remember EMBIG has 46 countries and GBI-EMD 12, an "apples to pears" issue). Also, EMBIG contains bonds rather than pure credit instruments (CDS), and bonds themselves can diverge from the same-country CDS (the so-called bond-CDS "basis"). Therefore, to answer the question "which is better, relative value external or local?" you need to know:
• Which kind of bond (external or local) offers the best credit spread when fully hedged of its other risks (USD interest-rate risk for external, and currency and local interest-rate risk for local)? This isolates the credit dimension and is achieved mathematically by factoring the costs of the various hedging dimensions. Exhibit 4 illustrates four ways to take Hungarian credit exposure, for example. Some assumptions will be required, but because the assumptions are used the same way for all of the bonds, they shouldn't have an effect on the estimation of relative value; • Is there a relative value discrepancy between the level of the currency and the level of credit spreads for the country? This can be estimated empirically as shown in Box 3; Hungary's plot is in Figure 3.1. The main assumption is that of mean reversion: that if the FX is exceptionally weak or strong for the current level of credit spreads, the relationship will revert to the mean over time (generally the time horizon here is quite short, usually hours or days, absent a credit event).
Don't be discouraged, though. There are occasional switching opportunities between external and local currency debt that survive the large transactions costs of doing so, usually motivated by differential asset class flows. For the reasons outlined above, these can't be seen as "pure" relative value but rather compositionally-motivated statistical relative value. Exhibit 5 depicts the credit spread of the EMBIG and the FXrebased level of the GBI-EMD. Notice that occasionally there are outliers where mean reversion might be expected (but again, not guaranteed by fundamentals due to the differential compositions).
How to own it: dedicated external, local, or corporate; "blended"; or "multi asset" (including emerging equities)
The next question we're generally asked is whether the best way to own emerging debt is through dedicated funds targeting one of the sub-classes or via "blended" strategies that bundle any or all of the above. We believe that for those investors with the time and ability to manage their asset allocations actively, owning the individual sub-classes makes most sense, because it gives the investor the most freedom to align his investments according to his portfolio fit. Some investors would prefer to give the manager timing authority, and we're fine with that, too. For the reasons just mentioned, though, we temper these investors' expectations with respect to how active such a switching program likely will be.
"Multi-asset" strategies extend the mix to include emerging equities. These are targeted at investors who would like to have more "emerging" exposure but are worried about adding to their emerging equities given the volatility. Based on research by our Asset Allocation team, we believe that such strategies should be able to offer lower overall volatility with similar returns to equities through the judicious use of somewhat less volatile emerging debt.
Summary Emerging debt is a mixed bag of U.S. dollar, local currency, and corporate debt. The dollar debt has a broad group of countries. The local debt has a narrow group of many of the same countries. Corporate debt is essentially "emerging equities lite."
Any or all should be owned in risk-seeking portfolios when they're cheap to their fundamentals in proportion either to their relative value or portfolio fit.
They can be owned in an unbundled fashion, which gives the investor greater flexibility on the portfolio fit angle. Or, they can be owned in various types of pre-arranged bundles, where the manager then has flexibility to focus on relative value in a portfolio-fit agnostic fashion.
Box 1 Emerging Corporates: Chances Are, You Already Own Them!
For investors with long horizons, the fact that the statistical correlation between two assets you might own is high is less relevant the more fundamentally different they are. A classic example is the high statistical correlation that catastrophe bonds have with the S&P when the S&P is declining precipitously. However, nothing about the S&P's decline has raised the likelihood of, say, an earthquake in California, which ought to have a negative impact on a California earthquake-related catastrophe bond. They are fundamentally different assets.
Corporate bonds and equities are fundamentally the same thing, except that corporate bonds are a more senior claim and have term-structure exposure, and only through that fact is their fundamental and statistical relationship diluted. Emerging corporate bonds are no different in this regard.
Think of emerging markets corporate bonds as a hybrid instrument combining issuer-specific default risk and USD term structure risk. Unbundled this way, it's clear that the issuer-specific risks ought to be compared with other like risks, which in this case would be emerging equities, while USD term-structure risk can be evaluated relative to U.S. interest-rate swaps or U.S. Treasuries.
From a portfolio fit perspective, the issuerspecific risk ought to have at least one of the following properties: (1) it's the better value relative to its equity counterpart; or (2) it's diversifying given that it exists in circumstances where public equity doesn't exist for that issuer. On the first issue, we're often amused that the comparable presented for corporate bonds in general (not just emerging ones) is Treasuries or perhaps other forms of debt (loans, for example) rather than equities. Our hypothesis is that corporate bonds allow investors to "sneak in" more equity risk when they're constrained at the policy level from doing so.
As value-oriented investors, we find this uncompelling from an asset-allocation perspective: buy what's cheap. On the second issue, we took a peek at the composition by country, by industry, and by issuer of the CEMBIB-D and the MSCI-EM equity index. This is what we found.
By country (Figure 1.1) the name overlap is high, although CEMBIB-D caps country weights and so is a bit less lumpy. BRICs plus Mexico and Korea comprise 40 and 65%, respectively. By industry (Figure 1.2)
– allowing for the different industry classifications mysteriously propagated by the debt vs. equity underwriters – the overlap is also high. Banks (boy, do the underwriters like to underwrite their colleagues' issues) are the biggest in both. Among the larger countries in both indexes, the overlap by individual issuer (Figure 1.3) is amazingly high, and we note with some alarm that perpetual bonds are included in CEMBIB-D. Perps are about as close to equity as we bond folks can get. We've also noted in Figure 1.3 the bonds included in investmentgrade and high yield indexes. Chances are, you have it!
Continue reading.
1 In fact, some on the sell-side, trying to distinguish "emerging" from the more fiscally challenged "developed" or "advanced" country counterparts, are promoting a less pejorative re-branding. Instead of "emerging," these are "growth" markets rather than "mature" markets. "Frontier" markets are being rebranded "Next Generation" markets in an attempt to remove the Wild West image. Regardless of the branding, we view these as "risk" assets not "risk-free" or even "relatively less risky than mature country debt" markets. Lucky for us, however, it's generally been the case that investors in "emerging" have been more than compensated for the risks entailed.
2 J.P. Morgan separately quotes a EUR EMBIG and prior to May 2002 had included local law instruments. The asset class has its origins in the bonds issued out of restructured bank loans from the 1980s and 1990s ("Brady bonds"), although today Brady bonds are only 5% of the EMBIG market value.
3 The local debt index these days contains just fixed-rate (including zero-coupon) bonds, although in the past it had included Chilean inflation-linked bonds. Barclays produces a global inflation-linked index, including emerging issuers Chile, Korea, South Africa, Mexico, Brazil, Colombia, Poland, and Turkey, although Brazil dominates the index by a wide margin. In local debt there is also a host of very interesting other instruments: floating-rate notes, foreigncurrency linked, etc., that are not currently captured by the index.
4 Another weakness of the emerging debt ETF is its need to provide a high level of liquidity to buyers of the product. As you will see later in Box 2, Figure 2.2, emerging debt at best has poor liquidity, and sometimes it is truly dreadful. No investment vehicle can make an illiquid underlying market liquid, although it can try by using, as emerging debt ETFs do, restricted benchmarks that include only the most liquid (or, in the case of local debt, most accessible) bonds. Investors are likely to be surprised when the discount to NAV rises sharply during a market disruption. Instead, funds ought to carry purchase and redemption fees (paid to the fund shareholders) to highlight to investors the need to be long-term oriented in the asset class. After all, separate accounts pay these implicitly when they buy assets at the offer and sell at the bid. Figure 2.2 shows the large and sometimes very large transactions costs inherent in emerging debt markets.
5 Fortunately for us, because we don’t believe in the portfolio fit of corporates, we’ve saved ourselves some labor on this score!
Disclaimer: The views expressed are the views of Tina Vandersteel through the period ending 5/29/12 and are subject to change at anytime based on market and other conditions. This is not an offer or solicitation for the purchase or sale of any security. The article may contain some forward looking statements. There can be no guarantee that any forward looking statement will be realized. GMO undertakes no obligation to publicly update forward looking statements, whether as a result of new information, future events or otherwise. Statements concerning financial market trends are based on current market conditions, which will fluctuate. References to securities and/or issuers are for illustrative purposes only. References made to securities or issuers are not representative of all of the securities purchased, sold or recommended for advisory clients, and it should not be assumed that the investment in the securities was or will be profitable. There is no guarantee that these investment strategies will work under all market conditions, and each investor should evaluate the suitability of their investments for the long term, especially during periods of downturns in the markets.
