abcGlobal Research

We identify four country-specific and

three global drivers of EM spreads

Continued US recovery is expected to benefit EM in general, defying the excess fear about the US tightening

Turkish spreads appear cheap while Philippines looks rich

A quantitative drill-down of fundamentals The sell-offs in EM external debt (EXD) in the second half of 2013 and early-2014 invite a deep look at the underlying reasons. Why did the fragile five suffer while the CEE happy family escaped unscraped? Would a growing US economy benefit or harm EM in general? At this juncture which countries spreads look cheap and which look rich?

To answer these questions, we investigate what economic and financial variables are the most important drivers of EM EXD and CDS spreads. We find real effective exchange rate (REER), import cover, current account plus FDI as % of GDP, and net public external debt as % of GDP to be most critical country-specific drivers. We also identify VIX, UST volatility (MOVE) and UST 10y yield as the most influential global drivers. Our study confirms the long-run negative relationship between UST 10y yield and EM EXD spreads. However, sharp spikes in volatilities, such as the run-ups in UST volatility last summer, could widen EM spreads dramatically and thus distort this negative relationship.

We econometrically study the effect of US economic growth on the health of EM external sectors. We conclude that most EM countries are likely to benefit from continued US growth.

We conduct a rich-cheap analysis of EM spreads using an econometric technique called cointegration. Our analysis shows Turkish spreads offer value while Philippines appears rich. Russia and Brazil also look cheap, however for Russia we believe caution is warranted given the tension in Ukraine and Russias worsening long-term fundamentals. For Brazil, the cointegration relationship is invalid, implying some other variables may have helped drive the spreads wider.

External Debt Emerging Markets

EM Spread Fundamental ModelDeciphering the key drivers

16 April 2014 Victor Fu EM Strategist HSBC Securities (USA) Inc. +1 212 525 4219 victor.w.fu@us.hsbc.com

View HSBC Global Research at: http://www.research.hsbc.com

Issuer of report: HSBC Securities (USA) Inc

Disclaimer & Disclosures This report must be read with the disclosures and the analyst certifications in the Disclosure appendix, and with the Disclaimer, which forms part of it

VOTE HSBC in the 2014 Institutional Investor LatAm survey Click here to see HSBC's LatAm Team Roster

Request a ballot: http://www.institutionalinvestor.com/research_customerservice.aspx

stripat2Highlight

stripat2Highlight

stripat2Highlight

stripat2Highlight

stripat2Highlight

stripat2Highlight

2

External Debt Emerging Markets 16 April 2014

abc

Motivation of the model The motivation of this model is to uncover the drivers of the ebbs and flows of EM EXD and CDS spreads. 2013 witnessed a U-turn in investor appetite for EM EXD, an unpleasant turn likely brought by the US Feds QE tapering announcement in May. Since then investors have been pulling money out of this asset class, as evidenced by a -5.3% total return registered for EM EXD in 2013. The losses in 2013 were across the board, but more skewed to the fragile five Turkey, Indonesia, Brazil, South Africa and India, whose EXD indices returned -12.6% to -4.3%. Yet, there were also country indices that showed positive returns, such as 3.5% for Hungary. Two questions that could naturally come to the readers mind would be: 1) what were the driving variables that caused the across-the-board poor performance of EM EXD in 2013? 2) What drivers differentiated the performances of individual country indices, e.g., Turkey vs. Hungary?

For the first question, one commonly perceived driver was the widening of US Treasury (UST) yields, with the UST 10y yield rising from c1.6% in early May to above 3% at the end of 2013. However, history shows that higher US 10y yield usually has resulted in tighter EM EXD spreads (see Figure 1). This long-term negative relationship broke down between May and September of 2013 with mounting uncertainty about the timing of the QE tapering. In fact, this

conundrum can be explained by a large spike in UST volatility during the same period, as can be seen in the later shown positive sensitivities of EM spreads to MOVE index. As the tapering uncertainty settled in late 2013, this negative relationship got restored.

Figure 1. EXD spread negatively correlated to UST 10y yield

Source: Thomson Reuters Datastream

The answer to the second question comes from country-specific fundamental variables including current account balance, which has become a household buzzword among EM investors since the tapering talk surfaced, and which has differentiated the fragile five from countries like South Korea, Hungary, etc..

In this report, we strive to provide an in-depth analysis of these two questions. We conduct our analysis for 11 countries Brazil, Colombia, Mexico, Peru, Hungary, Poland, Russia, South Africa, Turkey, Indonesia and Philippines. We

11.522.533.544.555.56

100200300400500600700800900

1000

06 07 08 09 10 11 12 13 14EMBIG spread (bp) UST 10y yield (%, RHS)

Show me the drivers

Special thanks to Dilip Shahani, Head of Asia-Pacific Research, Gordian Kemen, Head of LatAm FI Research, and Di Luo, CEEMEA FI Strategist, for their valuable comments

stripat2Highlight

stripat2Highlight

3

External Debt Emerging Markets 16 April 2014

abc

choose these countries because they all have a relatively large and liquid pool of hard currency bonds and we believe these EXD markets are efficient, i.e., prices reflect all relevant information. We exclude high-beta credits like Argentina, Venezuela and Ukraine as these markets are often driven by hard-to-measure factors such as political noise. We use a linear regression model on about 10 years of data to study that, in addition to UST yields and country current account balance, what other systematic and idiosyncratic variables would drive the evolution of EM EXD and CDS spreads. We then investigate the sensitivities of EM spreads with respect to these drivers. We believe these sensitivity statistics would provide a quantitative link between portfolio allocation and the economic dynamics of EM countries and US. Furthermore, we econometrically conduct a rich-cheap analysis of EM spreads.

Literature review There is a large body of literature studying the drivers of EM EXD/CDS spreads. The drivers investigated can be categorized into two groups country-specific macroeconomic variables and common factors.

Country-specific drivers Country-specific economic variables, such as external debt-to-GDP ratio and foreign reserves, are deemed critical drivers of EM sovereign spreads in many studies. Reinhart et al (2003) shows that a countrys debt payment history and debt-to-GNP thresholds play a key role in the variations of the countrys sovereign spreads. Fovero and Giavazzi (2005) discovers that Brazilian sovereign bond spreads were strongly correlated with exchange rates and interest rates. Remolona et al (2008) decomposes CDS spreads into expected losses from default and risk premia required by investors as compensation for default risk. The paper finds

that country-specific fundamentals primarily drive sovereign risk. Bellas et al (2010) concludes that in the long term, fundamental variables, such as external debt-to-GDP, interest payments-to-reserves, trade openness, etc., are significant determinants of EM sovereign bond spreads.

Common drivers Global market variables, e.g., VIX, UST yields, drive EM EXD/CDS spreads, according to many studies. Pan and Singleton (2008) suggest that a substantial portion of the co-movement of the CDS spreads of Mexico, Turkey and Korea during some sub-periods of 2001-2006 was induced by changes in investors risk appetite measured by VIX rather than by reassessment of the fundamental strengths of those countries. Longstaff et al (2007) finds that a large sample of CDS spreads for developed and EM countries were more influenced by US equity and high-yield bond markets, and global risk premia than by the local economic factors. Levy-Yeyati and Williams (2010) asserts that UST curve steepening represents an import risk factor for EM spreads.

Driver selection We identified REER, import cover, current account plus FDI as % of GDP, and net public external debt as % of GDP as the most significant country-specific drivers of EM EXD spreads. For common drivers, we found VIX, UST option volatility (measured by the MOVE index), and UST 10y yield to be most influential. Initially we also included in our study variables like fiscal balance as % of GDP, trade openness, and foreign reserves as % of M2, which were utilized in some of the literature. But these variables turned out to perform poorly in terms of explaining the movements of EM EXD/CDS spreads the beta coefficients in the regression have varied signs among countries and the p-values are mostly insignificant. We describe the retained country drivers in order of explanatory power as follows.

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

4

External Debt Emerging Markets 16 April 2014

abc

REER: This is to measure a countrys international competitiveness. An external debt crisis usually comes with or follows a currency crisis. An overvalued REER is likely to cause imports to grow faster than exports, resulting in a large and persistent current account deficit, which has to be funded by external financing. When the external financing environment tightens, an overvalued REER could trigger a sharp sell-off in the nominal exchange rate. This in turn would raise the concern about the countrys external debt serviceability and thus induce a rush-out from the hard currency bond market by investors. Therefore, REER should be one of the most influential fundamental factors for EXD/CDS spreads. A vivid example is the most recent bout of EM stress with pressure concentrated on currencies like TRY, ZAR, BRL, coupled with widening spreads (see Figure 2). Our study also reveals that real and nominal exchange rates are highly correlated historically for most of the EM.

Figure 2. Turkey 5y CDS moves in tandem with TRY

Source: Bloomberg

Import cover (in months): This is a measure of a countrys ability to provide USD, particularly in case of a stress. Studies have showed that measures of reserves, e.g., import cover, M2-to-reserves, are important leading indicators of financial crises. Frankel and Sarvelos (2011), for instance, states that measures of reserves and the real exchange rate are the top two most statistically significant

determinants of crisis incidence in more than half of the 83 papers they surveyed.

Current account balance plus FDI as % of GDP: The importance of this metric in assessing the health of a sovereigns creditworthiness is needless to say just based on its frequent appearance of the word current account in headline news since the Fed unveiled the tapering plan in May 2013. A current account deficit needs to be financed externally. Hence we net it with FDI to measure the sustainability of the deficit as FDI is deemed a more stable type of financing than portfolio inflows.

Net public external debt as % of GDP: This metric is computed as (public external debt foreign reserves) / GDP. It measures not only a countrys leverage ratio but also the ability to sustain that level of leverage. Hence this metric is directly relevant to the valuation of a countrys external debt.

For common drivers, in addition to the two widely studied variables VIX and UST 10y yield, we also include MOVE as we believe UST volatility is an important source of the variation in EM spread (see Figure 3), particularly in case where global liquidity could become tighter.

Figure 3. EM IG spread has responded to MOVE spikes

Source: Bloomberg, Thomson Reuters Datastream

100

150

200

250

300

350

400

1.3

1.5

1.7

1.9

2.1

2.3

2.5

Apr-10 Apr-11 Apr-12 Apr-13 Apr-14TRY TR 5y CDS (bp, RHS)

0

50

100

150

200

250

300

100

200

300

400

500

600

700

06 07 08 09 10 11 12 13 14EMBIG spread (bp) MOVE (RHS)

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

5

External Debt Emerging Markets 16 April 2014

abc

Econometric study Data We collect JPMorgan EM Global Diversified country index stripped spreads from Datastream and 5y CDS spreads from Bloomberg for the selected 11 countries. We take the monthly averages of the daily spread data, which go back to June 2004 (Indonesia and Philippines CDS spreads start from October 2004). VIX, MOVE and UST 10y yield are downloaded from Bloomberg and monthly averaged. For country-specific variables, we obtain the raw data from Datastream and the websites of various central banks or ministries of finance, and then do the calculation when needed. REER and import cover data come at a monthly frequency. Current account, FDI and external debt stock data arrive quarterly for most of the countries and are interpolated to a monthly frequency.

Table 1 shows a snapshot of country-specific variables readings as of 31 March 2014 vs. YE 2012 (for REER it is the return since YE 2012). It can be seen that most countries have experienced a depreciation in REER with an average loss of 6.2%. Note that some REERs, e.g., for Brazil, had enjoyed a decent rally over the past two months. Given REERs highest sensitivity to external shocks among the chosen fundamental indicators, the across-the-board depreciation could largely explain the pressure seen on EM EXD over the past year. For the other three variables, we see an average decrease of 0.3 months in import cover and of 0.1% in current account + FDI as % of GDP, as well as an average improvement of 0.1% in net public external debt as % of GDP. Among the 11 countries, six, six, and seven saw a better reading in the above three variables, respectively. These statistics demonstrate that EM long-term external sector fundamentals in general have been pretty much intact compared to YE 2012, implying the bouts of the broad sell-offs in EM

EXD over the past year might have been more driven by the fear about the QE tapering and the painful short-term adjustments in EM currencies than the worsening of the long-term EM fundamentals. That said, country divergence in the long-term fundamentals does exit, e.g., Hungary has improved all the three long-term variables remarkably while Russia has shown an apparent worsening in those variables. Out of the fragile fives, South Africas long-term fundamentals appear to be most resilient.

Table 1. EM country-specific variable data snapshot

Source: Bloomberg, Thomson Reuters Datastream.

Methodology To study what factors drive EM EXD and CDS spreads, we run a linear regression model of each countrys EXD and CDS spreads on a list of global and country-specific variables, which include those discussed in the section of Driver Selection as well as additional candidate variables such as fiscal balance as % of GDP, trade openness, and foreign reserves as % of M2. But these additional variables are later discarded due to their poor explanatory power. We therefore have totally 22 models for 11 countries. The in-sample data span from Jun 2004 (October 2004 for Indonesia and Philippines CDS spreads) to November 2013. All the variables but REER are first differenced (REER return series is used) before they are fed into the regression model since the levels of these variables are found to be nonstationary (or a random walk in plain English) by an Augmented Dickey-Fuller (ADF) test (see Appendix 1). Econometric theory

REER Import cover CA+FDI % GDP Net EXD % GDPCountry Return since YE12 Current YE12 Current YE12 Current YE12BR 1.15% 18.91 20.36 -0.79 0.49 -14.61 -14.07CO -10.93% 8.87 7.61 1.00 0.96 2.14 2.41HU -0.99% 5.31 5.20 3.60 3.16 16.25 23.36ID -3.19% 6.68 7.06 -1.64 -1.18 2.77 1.52MX 0.05% 5.68 5.29 1.10 -1.83 -3.12 -3.03PE -5.00% 18.70 18.68 -0.06 2.62 -19.33 -18.87PH -11.51% 15.15 16.19 3.18 2.98 -13.97 -13.90PL -0.09% 6.13 6.43 -1.35 -2.71 9.95 9.56RU -7.99% 16.05 17.39 0.99 3.72 -3.37 -9.31TR -13.16% 5.09 5.09 -6.03 -4.99 1.06 0.47ZA -16.54% 5.74 5.72 -4.37 -6.45 6.23 6.74

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

6

External Debt Emerging Markets 16 April 2014

abc

says that if the variables included in a linear regression model are nonstationary, the residuals may be nonstationary, which would render the ordinary least square (OLS) estimator inconsistent (this means the estimated coefficients are not reliable and do not converge to the true coefficients even when the data sample size goes to infinity). If on the other hand the residuals are stationary but serially correlated, the OLS estimator would be inefficient and result in distorted coefficient standard errors and R-Squared (see, e.g., Davidson and MacKinnon 2004, Tsay 2010, etc).

The regression results are promising. Table 2 and 3 show the statistics of the regressions using EXD indices and 5y CDS spreads, respectively, as dependent variables. The green color code means the p-value is significant at 95% level while the light blue at 90%. An empty cell in a column indicates the variable is excluded due to the insensible sign of the

coefficient. Notice that the models for most of the countries have a R-Squared (R2 in the tables) of more than 50%, indicating a nice fit, while Durbin-Watson (DW in the tables) statistics unveil that in most of the models residuals are uncorrelated, confirming the validity of the p-value and R-Squared statistics. It is obvious that among country-specific variables, REER is most significant, followed by import cover, while the rest two show a lower degree of significance, which could be due to the nature of low data frequency. As for global variables, VIX is shown to be most influential on EM spreads, followed by UST volatility. Interestingly, UST 10y yield is not as significant as commonly perceived. As expected, it has a negative relationship with EXD spread for all the countries but Indonesia (coefficient: 8.59, p-value: 0.56) with an average coefficient of -25.33. However, its relationship with 5y CDS spread is unpredictable with the coefficients sign varying for different countries.

Table 2. EXD index spread regression coefficients and statistics

Country REER p-value Import cover

p-value CA+FDI % GDP

p-value EXD % GDP

p-value VIX p-value MOVE p-value UST 10y

p-value R2 DW

BR -3.23 0.00 -3.00 0.36 -9.89 0.16 3.54 0.28 2.94 0.00 0.29 0.04 -10.98 0.17 0.69 1.71 CO -1.53 0.08 -31.89 0.00 -7.27 0.27 10.23 0.09 4.65 0.00 0.17 0.32 -9.66 0.32 0.66 1.92 MX -4.62 0.00 -6.63 0.59 -1.94 0.83 2.85 0.38 2.47 0.00 0.08 0.47 -9.06 0.16 0.73 1.77 PE -0.90 0.63 -7.34 0.09 -11.47 0.13 2.44 0.41 3.34 0.00 0.51 0.01 -21.72 0.04 0.58 2.08 HU -14.23 0.00 -24.48 0.11 3.13 0.22 0.99 0.14 0.68 0.00 -55.25 0.00 0.46 1.84 PL -2.68 0.01 -14.72 0.06 -3.50 0.40 1.54 0.54 1.68 0.00 0.60 0.00 -16.28 0.03 0.55 1.90 RU -0.22 0.88 -15.53 0.00 -2.68 0.73 5.80 0.05 2.54 0.00 0.77 0.00 -62.16 0.00 0.62 1.60 TR -3.43 0.00 -33.28 0.00 -13.28 0.06 17.00 0.00 3.76 0.00 0.34 0.05 -18.72 0.05 0.69 1.86 ZA -2.09 0.01 -9.97 0.30 -2.83 0.54 4.09 0.00 0.31 0.14 -6.42 0.56 0.61 1.65 ID -3.33 0.00 -24.96 0.01 -17.48 0.34 1.24 0.91 6.96 0.00 0.07 0.77 0.65 1.95 PH -1.06 0.27 -1.61 0.85 -4.24 0.33 10.77 0.00 3.42 0.00 0.32 0.06 -43.02 0.00 0.61 1.94 Source: HSBC

Table 3. 5y CDS spread regression coefficients and statistics

Country REER p-value Import cover

p-value CA+FDI % GDP

p-value EXD % GDP

p-value VIX p-value MOVE p-value UST 10y

p-value R2 DW

BR -4.28 0.00 -0.92 0.83 -8.56 0.34 3.40 0.44 2.99 0.00 0.23 0.20 0.58 1.51 CO -2.40 0.00 -31.18 0.00 -12.08 0.06 12.78 0.02 3.72 0.00 0.22 0.18 0.62 1.70 MX -5.20 0.00 -17.78 0.16 -3.54 0.72 12.68 0.00 2.54 0.00 0.23 0.06 0.76 1.67 PE -3.01 0.10 -6.08 0.13 -14.55 0.05 2.11 0.44 3.53 0.00 0.25 0.16 0.51 1.75 HU -14.90 0.00 -12.01 0.38 1.00 0.17 0.81 0.00 -44.56 0.00 0.55 1.78 PL -7.55 0.00 -17.32 0.01 -7.79 0.04 0.93 0.01 0.13 0.28 -14.37 0.04 0.63 1.65 RU -6.41 0.00 -8.63 0.07 11.03 0.00 2.17 0.00 0.67 1.77 TR -3.35 0.00 -32.63 0.09 -9.44 0.24 4.61 0.60 3.83 0.00 0.46 0.01 0.66 2.15 ZA -3.09 0.00 -1.66 0.85 2.67 0.00 0.35 0.03 0.60 1.79 ID -2.72 0.01 -25.20 0.01 -26.21 0.13 5.54 0.60 6.20 0.00 0.49 0.04 0.69 2.11 PH -0.99 0.33 -9.29 0.28 -6.46 0.15 4.99 0.20 3.95 0.00 0.37 0.04 -3.26 0.75 0.59 1.90 Source: HSBC

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

SiddharthHighlight

7

External Debt Emerging Markets 16 April 2014

abc

Hungary and Poland have significant and highly negative coefficients and Philippines has an insignificant negative coefficient. On the other hand, Indonesia has a large positive coefficient of 26.27 with a p-value of 0.06 and other countries coefficients are mildly positive and insignificant with an average of 5.67. The less uniform influence of UST 10y yield on CDS than on EXD could be because that, while the underlying rates of EM USD bonds are UST yields and hence a direct link to UST, EM CDS contracts are often used to express a directional view in cases of risk shocks, which could be related to but not limited to changes in US rates outlook, hence CDS has a more direct link to risk gauges (VIX and MOVE) than to UST yields.

In terms of sensitivities, among all the countries Hungary EXD/CDS spread is most sensitive to REER with one percentage of appreciation inducing a c14bp rally in the spread. Countries with a low import cover show a large sensitivity to this variable, e.g., an increase of one month in import cover could tighten the EXD/CDS spread by c33bp for Turkey. This reminded us of a massive widening in Turkish spreads in autumn 2013 when CBRT kept depleting FX reserves rapidly to defend the currency. Given the high significance and sensitivity plus the monthly data frequency, REER and import cover should be followed most closely, in our view. For global variables, the country most vulnerable to a spike in VIX is Indonesia, followed by Colombia and Turkey, while standing in the opposite are two CEE countries Hungary and Poland. If there is a large rise in UST 10y yield, Hungary is shown to benefit most. However if the UST yield rise comes very fast, resulting in a spike in UST implied volatility, Hungarian spreads are also expected to widen more than peers.

These sensitivity statistics are important to portfolio allocation and risk management. The investor can forecast the future changes in a

countrys EXD/CDS spread by plugging consensus or proprietary estimates of country-specific and global variables into the model. This allows portfolio allocation to be done in an objective way that does not depend on subjective judgments. Of course, judgments can be applied on top of the quantitative analysis if deemed essential. As an application for risk management, faced with a major risk event, the investor can assess the potential spread changes in different possible scenarios of the event.

Long-run anchors and short-run adjustments The chosen explanatory variables can serve different purposes in the model. Among the global variables, UST 10y yield provides a long-run global backdrop, VIX serves as a short-run risk shock in cases of global growth concerns and various crises (e.g., US subprime crisis, eurozone debt crisis, etc.), and MOVE joins VIX but is more specifically related to shocks to UST rates outlook. EM spreads evolve based on the long-run anchor of UST 10y yield while adjusting to the shocks from VIX and MOVE. On the surface these three variables should be highly correlated and one could wonder why all of them are included in the model. Yet a quick study reveals that the correlations are less strong than commonly expected and are time-variant. Table 4 shows the averages of the three-month rolling correlations of UST 10y yield, VIX and MOVE since 2005 are pretty benign. And Figure 4 displays the three rolling correlations series, which are clearly time-variant but mean-reverting. The correlation between UST 10y yield and MOVE has trended higher since mid-2013, which is expected as US rates outlook has taken centre stage during the period. This quick study confirms the merit of keeping all the three global variables as they allow us to value EM spreads in different global environments.

8

External Debt Emerging Markets 16 April 2014

abc

Table 4. Average rolling correlations (3-month window)

UST 10y & VIX UST 10y & MOVE VIX & MOVE

-0.33 0.15 0.17 Source: Bloomberg, HSBC

Figure 4. UST 10y, VIX and MOVE 3m rolling correlations

Source: Bloomberg, HSBC

Likewise, among the country-specific variables, the low-frequency current account plus FDI as % of GDP and net public external debt as % of GDP can serve as a long-run anchor while REER can provide a short-run adjustment. Import cover can function in between. Therefore, the sell-off of EM EXD in the second half of 2013 could be attributed to the short-run adjustment to the shock of UST volatility directly, and indirectly to the rapid REER depreciation, which itself was an adjustment to the shock of UST volatility and also rising US rates. And the rally since early February 2014 could be because the market realized that the short-run adjustments were overdone since EM long-term external sector fundamentals have largely remained unchanged compared to YE2012, as depicted in Table 1.

EM links to the US economy As a US economic recovery appears to have gained a firmer footing, it would be important to get a handle of the impact of US recovery on the EM economies external sectors. We select the Conference Board US Leading Index MoM (LEI) as a proxy of US recovery. To obtain a gauge of

the health of an EM countrys external sector, we apply a statistical procedure called Principal Component Analysis (PCA) to derive a single variable from REER, import cover, current account plus FDI as % of GDP, and net public external debt as % of GDP (the sign is flipped to make the interpretation consistent with other three). That single variable is termed first principal component or PC1 and is expected to explain most of the variations in the four underlying variables. We then regress each economys PC1 on US LEI. Table 5 shows that most countries PC1s respond positively to US LEI with Mexico topping the list, which is natural given Mexican economys dependence on the neighbour. Though the beta coefficients for Peru and Philippines are slightly negative, the p-value statistics are highly insignificant. We also can see that most of the PC1s explain more than 90% of the variations. This analysis reveals that a continuation of US recovery will likely benefit EM external sectors in general. This conclusion is consistent with the previous finding that higher US long yields resulting from US growth should bring down EM EXD spreads in the long run, echoing that last years market fear about the US tapering might have been overplayed.

-1-0.8-0.6-0.4-0.2

00.20.40.60.8

1

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014UST 10y & VIX UST 10y & MOVE VIX & MOVE

Table 5. EM fundamental links to US Leading Economic Index

Country Beta to US LEI p-value PC1 weight

BR 0.97 0.00 94.52% CO 0.54 0.04 96.13% MX 1.13 0.00 96.47% PE -0.17 0.49 61.27% HU 0.34 0.05 61.04% PL 0.68 0.00 83.95% RU 0.41 0.19 69.07% TR 0.42 0.10 95.09% ZA 0.81 0.02 95.14% ID 0.55 0.07 97.57% PH -0.11 0.62 90.15% Source: HSBC

9

External Debt Emerging Markets 16 April 2014

abc

Rich-cheap analysis We use an econometric technique called cointegration to conduct rich-cheap analysis. If a set of time series variables are nonstationary but a linear combination of them is stationary, then we say these variables are cointegrated. There are two approaches to the estimation of a cointegration relationship the Enger-Granger two-step method (Enger and Granger 1987) and the Johansen cointegrated vector autoregression method (Johansen 1988). We adopt the former for its accessibility to a broader audience.

Given a set of variables, the Enger-Granger method starts with a linear regression of one variable of the user choice on the others. If the residual series is tested, e.g., by an ADF test, to be stationary, then these variables are declared to be cointegrated, or in plain English, they tend to move together in the long run, albeit they may diverge in the short run. If this regression contains financial market variables, then the short-run divergence can be used to conduct the rich-cheap analysis of one financial variable versus the others. In our study we would like to assess the value of EM EXD and CDS spreads, the natural variables to bring into the analysis would be US corporate BBB index spread and CDX North America Investment Grade (CDX NA IG) index spread, as Figure 5 shows that these two variables have moved in tandem with EM EXD IG index

and a synthetic CDX EM IG index spreads.

Despite the long-run equilibrium relationship between EM and US corporate spreads, the short-run divergence could be due to fundamental reasons. We therefore include the four country-specific variables, along with VIX to control systematic shocks, into the cointegration study. In the end for EXD we regress each countrys EXD spread on US corporate BBB spread, VIX and country-specific variables. And for CDS we replace country EXD spread and US corporate BBB spread with country CDS spread and CDX NA IG spread. All the variables are level series. Like the first difference regression, the in-sample data used to fit the cointegration relationships are from Jun 2004 (October 2004 for Indonesia and Philippines CDS spreads) to November 2013.

The ADF tests of the regression residuals (see Table 6 in Appendix 1) depict that for most countries a cointegration exists. The only exceptions are Brazil for both EXD and CDS and Mexico for EXD.

Appendex 2 and 3 display the cointegration residuals for all the countries for EXD and CDS, respectively. The curves in black are out-of-sample residuals, which are computed based on the in-sample fitted coefficients and on the observations after November 2013. Interestingly, for most countries the residuals appeared to reach extremely high (cheap) levels in November 2013, calling for a broad rally ahead, which did happen after early February 2014. Post the rally, Turkey and Russia EXD spreads still look cheap while Philippines and Peru appear rich. For the 5y CDS spread, Russia and Turkey are on the cheap side while Philippines and Indonesia are on the opposite. For Russia, caution is warranted given the ongoing tension in Ukraine and worsening domestic fundamentals. Brazil EXD and 5y CDS spreads both look cheap, however the residuals are not statistically cointegrated, indicating there may be omitted variables that have helped drive the residuals gradually higher in the long run.

Figure 5. EM EXD and CDS IG spreads move together with US counterparts

Source: Thomson Reuters Datastream, Bloomberg, HSBC

0

100

200

300

400

500

600

700

Apr-07 Apr-08 Apr-09 Apr-10 Apr-11 Apr-12 Apr-13 Apr-14

CDX EM IG CDX IG NAEM EXD IG US Corp BBB

10

External Debt Emerging Markets 16 April 2014

abc

Appendix 1. Augmented Dickey-Fuller test Table 6. Country variables and EXD/CDS cointegration residual ADF test p-values

Country REER CA+FDI % GDP Import cover EXD % GDP EXD Cointegration CDS Cointegration

BR 0.22 0.89 0.53 0.05 0.35 0.16 CO 0.13 0.21 0.24 0.04 0.04 0.02 MX 0.05 0.15 0.96 0.34 0.14 0.01 PE 0.86 0.23 0.47 0.27 0.00 0.00 HU 0.09 0.64 0.64 0.14 0.01 0.01 PL 0.03 0.44 0.74 0.79 0.01 0.03 RU 0.37 0.72 0.12 0.04 0.00 0.00 TR 0.26 0.74 0.20 0.02 0.02 0.00 ZA 0.19 0.13 0.15 0.78 0.00 0.00 ID 0.36 0.78 0.13 0.30 0.00 0.00 PH 0.48 0.11 0.95 0.87 0.00 0.00 Source: HSBC

Note: A p-value smaller than 5%/10% shows that the time series is stationary at a 95%/90% confidence level. We can see that the four country-specific variables are nonstationary in level for most countries. The first-difference series (return series for REER) are tested to be all stationary with the test results omitted here.

11

External Debt Emerging Markets 16 April 2014

abc

Appendix 2. EXD cointegration residual charts Figure 6. Brazil residuals Figure 7. Colombia residuals Figure 8. Mexico residuals

Figure 9. Peru residuals Figure 10. Hungary residuals Figure 11. Poland residuals

Figure 12. Russia residuals Figure 13. Turkey residuals Figure 14. South Africa residuals

Figure 15. Indonesia residuals Figure 16. Philippines residuals

Source for all charts on this page: HSBC

-150

-100

-50

0

50

100

150

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014BR in-sample residual Out-of-sample residual

-100

-50

0

50

100

150

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014CO in-sample residual Out-of-sample residual

-60

-40

-20

0

20

40

60

80

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014MX in-sample residual Out-of-sample residual

-80

-60

-40

-20

0

20

40

60

80

100

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014PE in-sample residual Out-of-sample residual

-200

-150

-100

-50

0

50

100

150

200

250

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014HU in-sample residual Out-of-sample residual

-100-80-60-40-20

020406080

100120

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014PL in-sample residual Out-of-sample residual

-150

-100

-50

0

50

100

150

200

250

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014RU in-sample residual Out-of-sample residual

-100-80-60-40-20

020406080

100

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014TR in-sample residual Out-of-sample residual

-80

-60

-40

-20

0

20

40

60

80

100

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014ZA in-sample residual Out-of-sample residual

-100-80-60-40-20

020406080

100

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014ID in-sample residual Out-of-sample residual

-80

-60

-40

-20

0

20

40

60

80

100

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014PH in-sample residual Out-of-sample residual

12

External Debt Emerging Markets 16 April 2014

abc

Appendix 3. CDS cointegration residual charts Figure 17. Brazil residuals Figure 18. Colombia residuals Figure 19. Mexico residuals

Figure 20. Peru residuals Figure 21. Hungary residuals Figure 22. Poland residuals

Figure 23. Russia residuals Figure 24. Turkey residuals Figure 25. South Africa residuals

Figure 26. Indonesia residuals Figure 27. Philippines residuals

Source for all charts on this page: HSBC

-150

-100

-50

0

50

100

150

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014BR in-sample residual Out-of-sample residual

-150

-100

-50

0

50

100

150

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014CO in-sample residual Out-of-sample residual

-60

-40

-20

0

20

40

60

80

100

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014MX in-sample residual Out-of-sample residual

-80

-60

-40

-20

0

20

40

60

80

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014PE in-sample residual Out-of-sample residual

-250-200-150-100

-500

50100150200250

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014HU in-sample residual Out-of-sample residual

-80-60-40-20

020406080

100120

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014PL in-sample residual Out-of-sample residual

-200

-150

-100

-50

0

50

100

150

200

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014RU in-sample residual Out-of-sample residual

-100-80-60-40-20

020406080

100

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014TR in-sample residual Out-of-sample residual

-100-80-60-40-20

020406080

100

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014ZA in-sample residual Out-of-sample residual

-100

-80

-60

-40

-20

0

20

40

60

80

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014ID in-sample residual Out-of-sample residual

-80

-60

-40

-20

0

20

40

60

80

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014PH in-sample residual Out-of-sample residual

13

External Debt Emerging Markets 16 April 2014

abc

References Bellas, D., M. G. Papaioannou, and I. Petrova, 2010, Determinants of Emerging Market Sovereign Bond Spreads: Fundamentals vs Financial Stress, IMF Working Paper, WP/10/281.

Davidson, R. and MacKinnon, J. (2004) Econometric Theories and Methods, Oxford University Press.

Engle, R. and Granger, C. (1987) Co-integration and Error-correction: Representation, Estimation and Testing, Econometrica 55, 251-276. Favero, C. and Giavazzi, F. (2005) Inflation Targeting and Debt: Lessons from Brazil, in F. Giavazzi, I. Goldfajn and S. Herrera (eds.), Inflation Targeting, Debt and the Brazilian Experience 1999 to 2003, MIT Press. Frankel, J. and Saravelos, G. (2011) Can Leading Indicators Assess Country Vulnerability? Evidence from the 2008-09 Global Financial Crisis, Harvard Kennedy School. Johansen, S. (1988) Statistical Analysis of Cointegrating Vectors. Journal of Economic Dynamics and Control 12:231-54.

Levy-Yeyati, E., and T. Williams, 2010, US Rates and Emerging Markets Spreads, Universidad Torcuato Di Tella, Business School Working Papers 02/2010.

Longstaff, F., Pan, J., Pedersen, L., and Singleton, K. (2007), How Sovereign is Sovereign Credit Risk?, unpublished working paper, UCLA Anderson School, MIT Sloan School, NYU Stern School, and Stanford Graduate School of Business. Pan, J. and Singleton, K. (2008). Default and Recovery Implicit in the Term Structure of Sovereign CDS Spreads, Journal of Finance, 63, 2345-2384. Reinhart, C., Rogoff, K., and Savastano, M. (2003) Debt Intolerance, NBER working paper series. Remolona, E., Scatigna, M., and Wu, E. (2008) The Dynamic Pricing of Sovereign Risk in Emerging Markets: Fundamentals and Risk Aversion, unpublished working paper, Bank for International Settlements and University of New South Wales. Tsay, R. (2010) Analysis of Financial Time Series, Third Edition, Wiley.

14

External Debt Emerging Markets 16 April 2014

abc

Disclosure appendix Analyst Certification The following analyst(s), economist(s), and/or strategist(s) who is(are) primarily responsible for this report, certifies(y) that the opinion(s) on the subject security(ies) or issuer(s) and/or any other views or forecasts expressed herein accurately reflect their personal view(s) and that no part of their compensation was, is or will be directly or indirectly related to the specific recommendation(s) or views contained in this research report: Victor Fu

Important Disclosures This document has been prepared and is being distributed by the Research Department of HSBC and is intended solely for the clients of HSBC and is not for publication to other persons, whether through the press or by other means.

This document is for information purposes only and it should not be regarded as an offer to sell or as a solicitation of an offer to buy the securities or other investment products mentioned in it and/or to participate in any trading strategy. Advice in this document is general and should not be construed as personal advice, given it has been prepared without taking account of the objectives, financial situation or needs of any particular investor. Accordingly, investors should, before acting on the advice, consider the appropriateness of the advice, having regard to their objectives, financial situation and needs. If necessary, seek professional investment and tax advice.

Certain investment products mentioned in this document may not be eligible for sale in some states or countries, and they may not be suitable for all types of investors. Investors should consult with their HSBC representative regarding the suitability of the investment products mentioned in this document and take into account their specific investment objectives, financial situation or particular needs before making a commitment to purchase investment products.

The value of and the income produced by the investment products mentioned in this document may fluctuate, so that an investor may get back less than originally invested. Certain high-volatility investments can be subject to sudden and large falls in value that could equal or exceed the amount invested. Value and income from investment products may be adversely affected by exchange rates, interest rates, or other factors. Past performance of a particular investment product is not indicative of future results.

HSBC and its affiliates will from time to time sell to and buy from customers the securities/instruments (including derivatives) of companies covered in HSBC Research on a principal or agency basis.

Analysts, economists, and strategists are paid in part by reference to the profitability of HSBC which includes investment banking revenues.

For disclosures in respect of any company mentioned in this report, please see the most recently published report on that company available at www.hsbcnet.com/research.

Additional disclosures 1 This report is dated as at 16 April 2014. 2 All market data included in this report are dated as at close 14 April 2014, unless otherwise indicated in the report. 3 HSBC has procedures in place to identify and manage any potential conflicts of interest that arise in connection with its

Research business. HSBC's analysts and its other staff who are involved in the preparation and dissemination of Research operate and have a management reporting line independent of HSBC's Investment Banking business. Information Barrier procedures are in place between the Investment Banking and Research businesses to ensure that any confidential and/or price sensitive information is handled in an appropriate manner.

15

External Debt Emerging Markets 16 April 2014

abc

Disclaimer * Legal entities as at 8 August 2012 UAE HSBC Bank Middle East Limited, Dubai; HK The Hongkong and Shanghai Banking Corporation Limited, Hong Kong; TW HSBC Securities (Taiwan) Corporation Limited; 'CA' HSBC Bank Canada, Toronto; HSBC Bank, Paris Branch; HSBC France; DE HSBC Trinkaus & Burkhardt AG, Dsseldorf; 000 HSBC Bank (RR), Moscow; IN HSBC Securities and Capital Markets (India) Private Limited, Mumbai; JP HSBC Securities (Japan) Limited, Tokyo; EG HSBC Securities Egypt SAE, Cairo; CN HSBC Investment Bank Asia Limited, Beijing Representative Office; The Hongkong and Shanghai Banking Corporation Limited, Singapore Branch; The Hongkong and Shanghai Banking Corporation Limited, Seoul Securities Branch; The Hongkong and Shanghai Banking Corporation Limited, Seoul Branch; HSBC Securities (South Africa) (Pty) Ltd, Johannesburg; HSBC Bank plc, London, Madrid, Milan, Stockholm, Tel Aviv; US HSBC Securities (USA) Inc, New York; HSBC Yatirim Menkul Degerler AS, Istanbul; HSBC Mxico, SA, Institucin de Banca Mltiple, Grupo Financiero HSBC; HSBC Bank Brasil SA Banco Mltiplo; HSBC Bank Australia Limited; HSBC Bank Argentina SA; HSBC Saudi Arabia Limited; The Hongkong and Shanghai Banking Corporation Limited, New Zealand Branch incorporated in Hong Kong SAR

Issuer of report HSBC Securities (USA) Inc 452 Fifth Avenue, 9th floor HSBC Tower New York, NY 10018, USA Telephone: +1 212 525 5000 Fax: +1 212 525 0354 Website: www.research.hsbc.com

This material was prepared and is being distributed by HSBC Securities (USA) Inc., ("HSI") a member of the HSBC Group, the NYSE and FINRA. This material is for the information of clients of HSI and is not for publication to other persons, whether through the press or by other means. It is based on information from sources, which HSI believes to be reliable but it is not guaranteed as to the accuracy or completeness. This material is not, and should not be construed as, an offer or the solicitation of an offer to buy or sell any securities. The opinions contained within the report are based upon publicly available information at the time of publication and are subject to change without notice. HSI and/or its affiliated companies will buy or sell from customers on a principal basis the securities of the issuer(s) whose securities are recommended in this report. Employees of HSI and its affiliates not involved in the preparation of this report may have positions in the securities mentioned in this report and may from time to time add or dispose of any such securities in a manner different than discussed in this report. Past performance is not necessarily a guide to future performance. The value of any investment or income may go down as well as up and you may not get back the full amount invested. Where an investment is denominated in a currency other than the local currency of the recipient of the research report, changes in the exchange rates may have an adverse effect on the value, price or income of that investment. In case of investments for which there is no recognised market it may be difficult for investors to sell their investments or to obtain reliable information about its value or the extent of the risk to which it is exposed. In the UK this report may only be distributed to persons of a kind described in Article 19(5) of the Financial Services and Markets Act 2000 (Financial Promotion) Order 2005. The protections afforded by the UK regulatory regime are available only to those dealing with a representative of HSBC Bank plc in the UK. In Singapore, this publication is distributed by The Hongkong and Shanghai Banking Corporation Limited, Singapore Branch for the general information of institutional investors or other persons specified in Sections 274 and 304 of the Securities and Futures Act (Chapter 289) (SFA) and accredited investors and other persons in accordance with the conditions specified in Sections 275 and 305 of the SFA. This publication is not a prospectus as defined in the SFA. It may not be further distributed in whole or in part for any purpose. The Hongkong and Shanghai Banking Corporation Limited Singapore Branch is regulated by the Monetary Authority of Singapore. Recipients in Singapore should contact a "Hongkong and Shanghai Banking Corporation Limited, Singapore Branch" representative in respect of any matters arising from, or in connection with this report. In Hong Kong, this document has been distributed by The Hongkong and Shanghai Banking Corporation Limited in the conduct of its Hong Kong regulated business for the information of its institutional and professional customers; it is not intended for and should not be distributed to retail customers in Hong Kong. The Hongkong and Shanghai Banking Corporation Limited makes no representations that the products or services mentioned in this document are available to persons in Hong Kong or are necessarily suitable for any particular person or appropriate in accordance with local law. All inquiries by such recipients must be directed to The Hongkong and Shanghai Banking Corporation Limited. In Korea, this publication is distributed by either The Hongkong and Shanghai Banking Corporation Limited, Seoul Securities Branch ("HBAP SLS") or The Hongkong and Shanghai Banking Corporation Limited, Seoul Branch ("HBAP SEL") for the general information of professional investors specified in Article 9 of the Financial Investment Services and Capital Markets Act (FSCMA). This publication is not a prospectus as defined in the FSCMA. It may not be further distributed in whole or in part for any purpose. Both HBAP SLS and HBAP SEL are regulated by the Financial Services Commission and the Financial Supervisory Service of Korea. In Australia, this publication has been distributed by The Hongkong and Shanghai Banking Corporation Limited (ABN 65 117 925 970, AFSL 301737) for the general information of its wholesale customers (as defined in the Corporations Act 2001). Where distributed to retail customers, this research is distributed by HSBC Bank Australia Limited (AFSL No. 232595). These respective entities make no representations that the products or services mentioned in this document are available to persons in Australia or are necessarily suitable for any particular person or appropriate in accordance with local law. No consideration has been given to the particular investment objectives, financial situation or particular needs of any recipient. This publication is distributed in New Zealand by The Hongkong and Shanghai Banking Corporation Limited, New Zealand Branch incorporated in Hong Kong SAR. In Canada, this document has been distributed by HSBC Bank Canada and/or its affiliates. Where this document contains market updates/overviews, or similar materials (collectively deemed Commentary in Canadaalthough other affiliate jurisdictions may term Commentary as either macro-research or research), the Commentary is not an offer to sell, or a solicitation of an offer to sell or subscribe for, any financial product or instrument (including, without limitation, any currencies, securities, commodities or other financial instruments). Copyright 2014, HSBC Securities (USA) Inc, ALL RIGHTS RESERVED. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, on any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of HSBC Securities (USA) Inc. MICA (P) 118/04/2013, MICA (P) 068/04/2013 and MICA (P) 077/01/2014

abc

Rates EMEA Bert Lourenco Head of Rates Research, EMEA +44 20 7991 1352 bert.lourenco@hsbcib.com Subhrajit Banerjee +44 20 7991 6851 subhrajit.banerjee@hsbcib.com Theologis Chapsalis +44 20 7992 3706 theologis.chapsalis@hsbcib.com Wilson Chin, CFA +44 20 7991 5983 wilson.chin@hsbcib.com Di Luo +44 20 7991 6753 di.luo@hsbcib.com Chris Attfield +44 20 7991 2133 christopher.attfield@hsbcib.com Sebastian von Koss +49 211 910 3391 sebastian.von.koss@hsbc.de Frank Will +49 211 910 2157 frank.will@hsbc.de Asia Andr de Silva, CFA Head of Rates Research, Asia-Pacific +852 2822 2217 andre.de.silva@hsbcib.com Pin-ru Tan +852 2822 4665 pinrutan@hsbc.com.hk Himanshu Malik +852 3941 7006 himanshu1malik@hsbc.com.hk Dayeon Hong +852 3941 7009 dayeonhong@hsbc.com.hk Americas Larry Dyer +1 212 525 0924 lawrence.j.dyer@us.hsbc.com Jae Yang +1 212 525 0861 jae.yang@us.hsbc.com Bertrand Delgado +1 212 525 0745 bertrand.j.delgado@us.hsbc.com Gordian Kemen Head of Latin America Fixed Income Research +1 212 525 2593 gordian.x.kemen@us.hsbc.com Victor Fu +1 212 525 4219 victor.w.fu@us.hsbc.com Alejandro Mrtinez-Cruz +52 55 5721 2380 alejandro.martinezcr@hsbc.com.mx Aaron T Gifford +1 212 525 3277 aaron.t.gifford@us.hsbc.com

Credit EMEA Lior Jassur Head of Credit Research, EMEA +44 20 7991 5632 lior.jassur@hsbcib.com Dominic Kini +44 20 7991 5599 dominic.kini@hsbcib.com Laura Maedler +44 20 7991 1402 laura.maedler@hsbcib.com Anna Schena +44 20 7991 5919 anna.schena@hsbcib.com Pavel Simacek, CFA +44 20 7992 3714 pavel.simacek@hsbcib.com Reza-ul Karim +44 20 7992 3703 reza-ul.karim@hsbcib.com Raffaele Semonella +971 4423 6554 raffaele.semonella@hsbcib.com Ivan Zubo +44 20 7991 5975 ivan.zubo@hsbcib.com Jordan Cant +44 20 7991 5475 jordan.cant@hsbcib.com Asia Dilip Shahani Head of Global Research, Asia-Pacific +852 2822 4520 dilipshahani@hsbc.com.hk Zhiming Zhang +852 2822 4523 zhimingzhang@hsbc.com.hk Devendran Mahendran +852 2822 4521 devendran@hsbc.com.hk Philip Wickham +65 6658 0618 philipwickham@hsbc.com.sg Keith Chan +852 2822 4522 keithkfchan@hsbc.com.hk Louisa Lam +852 2822 4527 louisamclam@hsbc.com.hk Yi Hu +852 2996 6539 yi.hu@hsbc.com.hk Helen Huang +852 2996 6585 helendhuang@hsbc.com.hk Crystal Zhao +852 2996 6514 crystalmzhao@hsbc.com.hk Kelly Fu +852 3941 7066 kellyyfu@hsbc.com.hk Lan Lan +852 3941 7186 lanlan@hsbc.com.hk

Christopher Li +852 2822 3232 christopherbli@hsbc.com.hk

Americas Sarah R Leshner Head of LatAm Corporate Credit Research +1 212 525 3231 sarah.r.leshner@us.hsbc.com Sean Glickenhaus +1 212 525 4131 sean.x.glickenhaus@us.hsbc.com

Global Fixed Income Research Team

Steven Major, CFA Global Head of Fixed Income Research +44 20 7991 5980 steven.j.major@hsbcib.com

Front Page (Page View)EM Spread Fundamental ModelShow me the driversMotivation of the modelLiterature reviewCountry-specific driversCommon drivers

Driver selectionEconometric studyDataMethodologyLong-run anchors and short-run adjustments

EM links to the US economyRich-cheap analysisAppendix 1. Augmented Dickey-Fuller testAppendix 2. EXD cointegration residual chartsAppendix 3. CDS cointegration residual chartsReferences

Disclosure appendixAnalyst CertificationImportant DisclosuresAdditional disclosures

Disclaimer

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description