Capital Flows to Emerging Markets

Transcription

1 Capital Flows to Emerging Markets MAY 29, 214 Private capital inflows to emerging markets are benefitting from a supportive global environment. Nonetheless, we have reduced our forecast for aggregate flows in 214, largely due to weaker inflows in 214H1 as the Ukraine crisis is weighing heavily on flows to Russia. Overall flows are projected to decline in 214, in part reflecting falling inflows to China, given the government s efforts to discourage short-term capital inflows. A slowly improving global macro picture and sustained risk appetite are projected to support a gradual pick-up of capital flows on a sequential basis from 214H2 onwards. Capital flows to EMs remain vulnerable to a possible escalation of and contagion from the Ukraine/Russia crisis as well as surprises in the timing, pace and magnitude of eventual Fed policy tightening. EMERGING EUROPE AND CHINA WEIGH ON CAPITAL FLOWS IN AN OTHERWISE SUPPORTIVE ENVIRONMENT The environment for capital flows to EM has become very supportive as the global economy gains momentum (though at a somewhat lower pace than expected in our January report), risk aversion has fallen to very low levels, and U.S. long-term interest rates have eased. In this context, portfolio equity and bond inflows have recovered from their lows at the beginning of the year. FDI flows have also held up well, although bank lending flows have retrenched. However, the Ukraine/Russia crisis has weighed heavily on inflows to EM Europe, particularly on bank lending to Russia. And while inflows to China were stronger than projected in 213, we now anticipate a decline this year, given the government s efforts to discourage short-term capital inflows. On balance, we have revised downwards our annual projections for our sample of 3 EMs by around $47 billion for this year and $27 billion for next year (Chart 1 and Table 1). Relative to EM GDP, capital flows are Chart 1 Emerging Market Private Capital Inflows, Net EM Europe Latin America 14 MENA EM Asia ex. China 12 China Total, Percent of EM GDP IIF Forecast Charles Collyns CHIEF ECONOMIST ccollyns@iif.com Felix Huefner DEPUTY DIRECTOR Global Macroeconomic Analysis fhuefner@iif.com Robin Koepke ECONOMIST Global Macroeconomic Analysis rkoepke@iif.com Saacha Mohammed RESEARCH ASSISTANT Global Macroeconomic Analysis smohammed@iif.com Hung Tran EXECUTIVE MANAGING DIRECTOR htran@iif.com Sonja Gibbs DIRECTOR Capital Markets & Emerging Markets Policy sgibbs@iif.com Emre Tiftik FINANCIAL ECONOMIST Capital Markets & Emerging Markets Policy etiftik@iif.com Table of Contents Global Overview 2 Portfolio Flows Drivers & Risks 7 Emerging Market Valuations 12 EM Bond & Equity Issuance 16 EM Asset Allocation 18 Flows by Region 21 Annex and Tables Source: IIF iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

2 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 2 Table 1 Emerging Market Economies: Capital Flows Capital Inflows Total Inflows, Net: Private Inflows, Net Equity Investment, Net Direct Investment, Net Portfolio Investment, Net Private Creditors, Net Commercial Banks, Net Nonbanks, Net Official Inflows, Net International Financial Institutions Bilateral Creditors Capital Outflows Total Outflows, Net Private Outflows, Net Equity Investment Abroad, Net Resident Lending/Other, Net Reserves (- = Increase) Memo: Net Errors and Omissions Current Account Balance Source: IIF estimated to fall to 3.6% by 215, down from 5% in 212 and less than half the share seen in 27. On a sequential basis, net inflows to EMs are estimated to have dropped sharply in the first half of this year, reflecting the events in Ukraine/Russia and China. The main drop is in the area of bank related flows. Portfolio flows have been more resilient, as suggested by our portfolio flows tracker which indicates a significant recovery since the start of the year (see page 7), consistent with a return of global risk appetite and notwithstanding the limited signs of a turnaround in activity seen in EMs so far. While bond issues have been weaker so far this year than last, possibly reflecting rising concerns about corporate indebtedness, EM equity markets have risen noticeably (see page 12). Net inflows to EMs are estimated to have dropped sharply in the first half of this year Going forward, flows are forecast to increase in the quarters ahead (Chart 2), assuming that the global economy continues to pick up speed, market expectations remain centered on a gradual Fed exit, and that tensions surrounding Ukraine and Russia are contained. In this context, both equity and debt portfolio inflows are envisaged to rise moderately, while other inflows (mainly banking related) should pick up. FDI inflows, which account for around half of all capital flows to EMs, are projected to grow steadily as they benefit in particular from the benign global growth environment (Chart 3). iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

3 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 3 Chart 2 IIF EM 3: Capital Inflows by Component IIF Forecast Other Portfolio Debt Portfolio Equity Q1 212Q3 213Q1 213Q3 214Q1 214Q3 215Q1 215Q3 Source: IIF, IMF, National Sources. FDI The regional breakdown of capital inflows shows a wide variety of country-specific dynamics. Flows to EM Europe are projected to fall by 25% between 213 and 215, largely due to much weaker debt inflows to Russia this year on the back of the Ukraine crisis, but also somewhat lower debt flows to Turkey, reflecting heightened political tensions. By contrast, the Africa and Middle East region will see flows rise 2% over the same period, driven by large equity inflows to UAE which offset weaker flows to Nigeria, where terrorism concerns are expected to dampen portfolio inflows. Capital flows to EM Asia are projected to decline only modestly, as lower debt inflows to China are offset by increases elsewhere, notably equity inflows to India and Korea, attracted by reform-minded governments. Flows to Latin America are also expected to stay relatively steady, with higher inflows to Brazil (primarily debt inflows attracted by the rise in interest rates), Chile and Mexico contrasting with lower inflows to the other countries, especially Colombia. Flows to EM Europe are projected to fall by 25% between 213 and 215 Chart 3 Chart 4 Private Capital Inflows to Emerging Markets, equity = FDI + portfolio equity; debt = bank lending and other private creditors 8 6 Equity Inflows IIF Forecast Correlation across EM Exchange Rates average 3-day rolling correlation between percentage change in MSCI EM FX index and USD/local exchange rate.7 Post-28 Crisis Average Debt Inflows Source: IIF Pre-28 Crisis Average Source: Bloomberg, IIF. iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

4 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 4 PULL FACTORS DAMPEN, PUSH FACTORS SUPPORT FLOWS While in the latter part of 213, most of the discussion about capital flows was focused on the timing of Fed exit, attention so far this year has primarily focused on the situation in emerging economies themselves. Thus, pull factors have been more important than push factors as drivers of capital flows (see pages 9-12 for a quantification of these effects). One indicator of this development is the decline in the correlation in movements across EM exchange rates, which suggests that country-specific factors have become more important relative to common drivers (Chart 4). Pull factors have been more important than push factors as drivers of capital flows this year The most important shifts on the pull side have been the Ukraine/Russia crisis and the disruption of carry trade flows to China. But more broadly, we have also continued to shave our projections for EM GDP growth in Mature economies have generally had a weak start to the year, but the underlying momentum remains positive and should lift growth from the second half onwards. Emerging markets will benefit from this uplift, particularly those with close export ties (such as Korea), but overall EMs have shown little sign of a pick up in business confidence in contrast to the mature economies (Chart 5). This has led us to revise downwards our 214 GDP growth forecast for EMs by a cumulative.4ppt relative to our January Capital Flows Report, and.1pp relative to our April Global Economic Monitor. To a large extent, this reflects substantial downward revisions in EM Europe, notably Russia, Turkey and Ukraine, but also somewhat weaker growth in China, South Africa, Thailand and Latin America, notably Argentina, Chile and Mexico (Chart 6). By contrast, we have raised the outlook for India, Korea, Hungary and Czech Republic. While the pull factors for capital flows have thus deteriorated somewhat, push factors have clearly become more supportive. Most importantly, risk aversion in financial markets has declined to very low levels over the past couple of months as reflected for example in the VIX index as well as in options pricing in other asset classes (Chart 7). This has supported EM equities, which have rebounded following weakness earlier in the year, notwithstanding the softer growth outlook (Chart 8). However, from a valuation Risk aversion in financial markets has declined to very low levels over the past couple of months Chart 5 Chart 6 Manufacturing PMI index, breakeven=5 214 GDP Growth Revisions Since January Capital Flows Report percentage points Mature Economies Global Emerging Economies Jul 12 Oct 12 Jan 13 Apr 13 Jul 13 Oct 13 Jan 14 Apr 14 Source: Bloomberg, Datastream, Markit Economics. Source: IIF. iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

5 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 5 Chart 7 VIX Index index Source: Bloomberg. Chart 8 Global Equity Markets index, Jan. 1, 213= U.S. 9 Emerging Markets 8 Jan 13 Jun 13 Nov 13 Apr 14 Source: Bloomberg. perspective, EM equities still generally remain much cheaper than mature market equities, although differences at the country-level remain wide (see page 14). With monetary policy in mature economies expected to remain very easy, U.S. long term interest rates have declined this year (in sharp contrast to the taper tantrum episode of 213, Chart 9), helped by a downward adjustment of expectations for future Fed policy rates from their highs reached last year. This partly reflects the temporary weakness in the U.S. economy at the start of 214, but also rising perceptions that the new normal level of policy interest rates even after policy support is unwound could be substantially lower than in the past. Moreover, while U.S. monetary policy historically has been the main global factor affecting capital flows, expected steps towards further monetary easing in the Euro Area and the Bank of Japan would clearly be supportive for global risk appetite even as the Fed shifts into tightening mode. Chart 9 U.S.:1-yr Treasury Yield percent Source: Datastream. AGGREGATE VIEW MASKS SUBSTANTIAL EM DIVERSITY While pull factors in the aggregate weigh on capital flows, there is substantial crosscountry diversity along several dimensions. Political events are one, albeit important, factor, as is exemplified in the case of Russia and Ukraine. But political factors are also relevant in other countries: for example, the recent coup has clouded the outlook for Thailand further, while the election of a new government in India raises the hope of structural reforms that are expected to attract more inflows. With a number of important elections still scheduled in EMs this year, including in Turkey and Brazil, political developments will remain squarely on the agenda. There is substantial crosscountry diversity along several dimensions On the macroeconomic side, there has been some adjustment of vulnerabilities and policy gaps since last year. Of the countries with large current account deficits in 212, India has been the one where the most adjustment has occurred, contributing to a iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

6 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 6 turnaround in investor confidence and strengthening inflows more recently (Chart 1). By contrast, external balances in Turkey and South Africa remain deeply in deficit and are financed to a smaller extent by FDI (see Heat Map, page 8). With investors becoming more discriminating across EMs, steps to raise policy credibility by reducing vulnerabilities and closing policy gaps have an impact on capital inflows, especially portfolio flows. A key adjustment challenge is the need to adapt monetary policy settings. Several EMs have faced inflation well above target levels (often coupled with high credit growth and current account deficits), requiring policy rate hikes to increase real interest rates, which in turn help attract investors. Brazil is most advanced in this regard, having raised policy rates consecutively since spring 213, bringing real interest rates well into positive territory, although expected inflation remains well above the inflation objective (Chart 11). By contrast, real interest rates in South Africa are still negative. These divergences are reflected in the outlook for debt capital inflows: while overall debt inflows are projected to rise by around 1% in Brazil this year, debt inflows to South Africa are envisaged to fall by 15%. Countries with inflation rates below the central bank s target, by contrast, have used the room to cut rates and support growth, benefitting equity capital inflows, for example in some of the Central and Eastern European countries. Steps to raise policy credibility by reducing vulnerabilities and closing policy gaps have an impact on capital inflows DOWNSIDE RISKS CLOUD THE OUTLOOK This baseline scenario is surrounded by risks. On the downside, and, most immediately, an escalation of political tensions primarily in the Ukraine/Russia crisis could have a serious impact on flows, not just on Ukraine/Russia but also on other neighboring EMs like Turkey that have quite close economic ties with Russia through energy and tourism (see pages 23-24). Such an event could also contribute to a general rise in risk aversion. If risk aversion (as measured by the U.S. corporate BBB spread over Treasuries) were to rise back towards its average over 2-present, capital flows to EM could be around An escalation of political tensions primarily in the Ukraine/Russia crisis could have a serious impact on flows Chart 1 Chart 11 Current Account Balances 214 vs. 212 Inflation and Real Interest Rates percent of GDP 3 inflation forecast - inflation target/mid-point of target band (%) 4 Turkey Thailand South 3 CAB/GDP, 212 Poland Africa Russia Brazil Chile 1 Peru Indonesia India Mexico Mexico Romania Colombia real interest rate (%) Indonesia -1Korea India Brazil Czech Poland China Chile -2 Hungary Turkey Philippines South CAB/GDP, Source: Bloomberg, National Sources, IIF. Inflation is IIF's 214-end Africa -6 forecast. Real interest rate is current official interest rate end Source: IIF. inflation forecast. iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

7 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 7 $5 billion lower (see page 11). Also, a shift in market expectations towards significantly faster Fed exit would hurt prospects for EM capital flows. For example, if markets start to expect that policy rates would be hiked at twice the speed currently priced into markets, EM portfolio flows could fall by around $1 billion (see pages 11-12). In the event of an increase in stress levels, countries that have made the least progress in boosting credibility and reducing vulnerabilities since the previous turmoil would be most at risk. The Heat Map of country vulnerabilities on page 8 provides an overview. On the policy side, Turkey is a case in point: with the central bank recently cutting its one-week repo rate on the back of improved sentiment, despite inflation running above target, the country risks being exposed in a case of renewed volatility, particularly if there were to be spillovers from instability in the region. A shift in market expectations towards significantly faster Fed exit would hurt prospects for EM capital flows. Looking at idiosyncratic country factors, flows to China could be dampened further, particularly if the housing correction drives a broader growth slowdown and exchange market conditions remain choppy. At the same time, comparatively cheap equity valuations in EMs pose an upside risk in particular for portfolio flows in an environment where global investors continue to search for yield. Idiosyncratic country factors also have the potential to surprise on the upside; for example impressive structural reform progress by the new government in India could lead to stronger than projected capital inflows. PORTFOLIO FLOWS: TRACKING DRIVERS AND QUANTIFYING RISKS Our recent work on tracking portfolio flows and decomposing their determinants allows us to provide some novel insights about this important source of external financing. Besides painting a more refined picture of recent developments, we also identify several concrete downside risks related to global financial developments, which we quantify with the help of an econometric model. Since there is only limited high-frequency data on EM portfolio flows available, it is difficult to follow the actions of global investors in real time. Among the most useful sources are portfolio flows data released by a limited group of EM countries that provide monthly or even daily portfolio flows estimates. Earlier this year, we launched an EM Portfolio Flows Tracker that aggregates the most timely country-level data sources on flows and incorporates a range of other data to provide a monthly estimate of total nonresident portfolio flows to emerging markets. Our latest estimates through May show a continued pickup in both debt and equity inflows since the start of the year to the highest level since September 212 (Chart12). Our latest portfolio flows estimates through May show a continued pickup. Taking advantage of the monthly country flows database developed for the tracker we have investigated the determinants of portfolio flows during the last several years. We run pooled estimations that allow us to take advantage of recent experience across a range of countries, making it possible to develop more accurate and robust findings iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

8 Arpitha Bykere CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 8 HEAT MAP OF SELECTED VULNERABILITY INDICATORS IN EMERGING ECONOMIES Argentina Brazil Bulgaria Chile China Colombia Czech Hungary India Indonesia Korea External Financing Vulnerabilities: Current Account Balance/GDP (%) Net FDI/Current Account Deficit Ratio Reserve Coverage Ratio* Short-Term External Debt/Total External Debt (%) External Debt/GDP (%) REER (% deviation in March 214 from avg.) Domestic Financial Vulnerabilities: Non-Financial Corporate Debt/GDP (%, change between Q3) Households Debt/GDP (%, change between Q3) Real Domestic Credit Growth (% oya, avg.) Foreign Investors' Share in Local Sovereign Debt Market (%, 213-end) Cross-Border Financial Claims/Domestic Credit (%, 212-end) Policy Vulnerablities: end Inflation (vs. Central Bank Target)** 6.5 (4.5±2) 4. (3±1) 1.9 (3.5) 3. (3±1) 1. (2±1) 1.4 (3.) 7.5 (8.) 5.7 (4.5±1) 2.7 ( ) Real Interest Rates*** Fiscal Balance/GDP (%) Gross Government Debt/GDP (%) Elections in 214**** Malaysia Mexico Peru Philippines Poland Romania Russia South Africa Thailand Turkey Ukraine External Financing Vulnerabilities: Current Account Balance/GDP (%) Net FDI/Current Account Deficit Ratio Reserve Coverage Ratio* Short-Term External Debt/Total External Debt (%) External Debt/GDP (%) REER (% deviation in March 214 from avg.) Domestic Financial Vulnerabilities: Non-Financial Corporate Debt/GDP (%, change between Q3) Households Debt/GDP (%, change between Q3) Real Domestic Credit Growth (% oya, avg.) Foreign Investors' Share in Local Sovereign Debt Market (%, 213-end) Cross-Border Financial Claims/Domestic Credit (%, 212-end) Policy Vulnerablities: end Inflation (vs. Central Bank Target)** 3.7 (3±1) 2.8 (2 ±1) 3.5 (4±1) 2.1 (2.5±1) 3.4 (3±1) 7.4 (5.) 6.4 (3-6) 8.5 (5.) 17.4 (4-6) Real Interest Rates*** Fiscal Balance/GDP (%) Gross Government Debt/GDP (%) Elections in 214**** Note: Data refers to 214 estimates unless otherwise specified Vulnerability for each country is shown relative to other countries in the sample and uses the following color code: Red - most vulnerable; pink - vulnerable; light green - less vulnerable; dark green - least vulnerable Colors are assigned according to the following thresholds: For CAB/GDP: Below -3% (most vulnerable), between -3% to (vulnerable), between to +3% (less vulnerable), above +3% (least vulnerable). For reserve coverage ratio: Below.5 (most vulnerable),.5 or above but less than 1. (vulnerable), 1. or above but less than 2. (less vulnerable), 2. or above (least vulnerable). For other variables, the mean and standard deviation are calculated across countries and the colors are assigned based on the following thresholds: Above +1 standard deviation from the mean, between mean and +1 standard deviation, between mean and -1 standard deviation, below -1 standard deviation from the mean The colors for headline indicators (external financial vulnerabilities and policy vulnerabilities) are assigned as follows: The variables under each indicator are given a score based on their color: Dark green=, light green=1, pink=2, red=3. The headline score is then calculated using the IMF methodology applied to 'Overheating Indicators for the G2 Economies' by adding up the score of each variable and dividing it by the maximum possible sum of scores. The headline color is then assigned using the methodology explained above. *FX reserves excl. gold at the end of 213/(current account deficit plus short-term external debt by original maturity plus amortization payments for 214) **The colors are assigned as follows: Inflation above the target band (red); inflation above the mid-point but at or below the upper-end of the target band (pink); inflation between the lower-end and mid-point of the target band (light green); inflation below the target band (dark green) ***Current policy rate - projected 214-end inflation. The colors are assigned as follows: For negative real interest rate: Red if inflation is above the mid-point of the target band; pink if inflation is at or below the mid-point of the target band. Positive real interest rate: Light green if inflation is above the mid-point of the target band; dark green if inflation is at or below the mid-point of the target band ****Includes local, parliamentary & presidential elections Data Sources: IIF for external financing vulnerabilities, real credit growth, inflation and real interest rates; April 214 IMF WEO database (Oct 213 database for Ukraine) for fiscal balance and gross government debt; IIF calculations using BIS, OECD, IMF data for non-financial corporate debt/gdp, household debt/gdp and cross-border financial claims; IIF calculations using National Sources for foreign investors share in government debt market iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

9 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 9 about the drivers of portfolio flows than would be possible with only aggregate data (see Box 1, page 1). The key results of our modelling are: 1. EM portfolio flows are highly sensitive to shifts in global risk aversion. Our preferred proxy of global risk is the U.S. BBB-rated corporate bond spread over Treasuries (Chart13). We estimate that an increase in this spread by 1 basis points is associated with a reduction in EM portfolio flows of $71 bn (or.41% of EM GDP) during the same month. All regression estimates apply to our group of 3 EMs excluding China and are reported in 214 dollars (see Box 1 for details). 2. A 1 percent increase in EM stock market prices is associated with increased portfolio inflows in the range of $9 to $22 billion (or.5% to.13% of EM GDP) during the same month. The lower bound of this range is based on a country-level estimate of the relationship between the local stock market prices and portfolio flows. The upper bound adds the impact of EM-wide stock market movements to this, which may be partly driven by global financial developments and partly reflect spillover effects from other EM economies. 3. A 1 basis point upward shift in expectations of future U.S. policy interest rates is estimated to reduce non-resident flows by $56 bn (or.33% of EM GDP) during the same month. An upward shift is estimated to have about twice the impact on portfolio flows as a downward shift, suggesting that emerging markets suffer more from Fed tightening surprises than they benefit from Fed easing surprises. 4. Portfolio bond and equity flows seem to behave in broadly similar ways. One difference is that equity flows are positively autocorrelated, indicating momentum in investor behavior. Also, debt flows seem to respond more strongly to reassessments in the outlook for Fed policy and to changes in global risk appetite. The regression results for equity and bond flows are reported in the annex on page 29. EM portfolio flows are highly sensitive to shifts in global risk aversion Upward shifts in expectations of future U.S. policy rates tend to reduce inflows. The described effects refer to the contemporaneous impact of a change in the underlying driver, for example the effect of a risk shock on portfolio flows in the same month when risk aversion increases. It seems plausible that some of the impact of a change in risk aversion would play out in subsequent months. To test this hypothesis, we include a lag of the respective independent variable in the regression. We find tentative evidence that some of the impact of changes in Fed policy expectations plays out over time, but the evidence is not statistically robust. These empirical results have some direct implications for interpreting portfolio flow movements during the past 12 months. In terms of domestic ( pull ) and external ( push ), the push variables in the model include global risk and U.S. monetary policy, while the pull side is captured by developments in the local stock market. Based on this distinction, we find that the EM sell-off in June last year was primarily driven by external developments (a push shock ), while the volatility in early 214 stemmed from a pull shock arising from idiosyncratic developments in a range of emerging market countries (Chart 14). Most recently, in May 214 both external and domestic developments seem to have turned supportive for EM portfolio flows. The EM sell-off in June last year was primarily driven by a push-shock. iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

10 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 1 BOX 1: WHAT DRIVES INTERNATIONAL PORTFOLIO FLOWS? We estimate a simple econometric model of the drivers of portfolio flows across a range of EM economies using monthly data. The dependent variable is total non-resident portfolio flows to a pool of 11 EM economies, representing about 62% of total portfolio flows to our group of 3 EMs. 1 The pooled model allows us to get an aggregate estimate of the impact of the various drivers based on the experiences of the individual countries. We include country-specific fixed effects to allow for structural differences across countries. The model is estimated with portfolio flows expressed in percent of the recipient country s GDP (which makes it possible to pool the regressions across countries and obtain an aggregate estimate). To allow for an intuitive interpretation, we report the regression results in billions of 214 dollars by multiplying the estimated coefficients by our projection for 214 EM nominal GDP. 2 The sample period is from January 21 to February 214. Our baseline regression is as follows: Yt = -7.7 * Riskt +.9 * Country_SMt +1.3 * EM_SMt -56. * Tighter_MPt 26. * Easier_MP (.) (.57) (.9) (.1) (.44) P-values are reported in parentheses. A value below.1 indicates significance at the 1% level etc. Riskt is the change in the average BBB spread in percentage points, Country_SM is the percent change in the country s stock market index, EM_SMt is the percent change in the MSCI EM stock market index. The last two variables both refer to the change in the interest rate implied by fed funds futures contracts 34 months into the future in percentage points. We use a dummy=1 for months when rate expectations shifted upward ( Tighter_MP ) and another dummy=1 for months when rate expectations shifted downward. Further details on the role of U.S. monetary policy in driving EM portfolio flows are discussed in a recent IIF Working Paper entitled Fed Policy Expectations and Portfolio Flows to EMs. Functional forms are based on standard stationarity tests for all variables (Augmented Dickey-Fuller tests). The adjusted R 2 for the model is.19. All explanatory variables are significant at least at the 1 percent level and have the expected signs. We also tested a number of alternative model specifications, particularly with respect to the country-specific variables. For example, we used EM purchasing manager indices, consensus growth forecasts, and an economic surprise index as alternatives to the stock market variables. We did not find statistically robust evidence that these variables were drivers of EM portfolio flows. We also included a lagged dependent variable to allow for momentumtype effects, but the estimated coefficient was not statistically significant. 1 The countries are Brazil, Bulgaria, Chile, Czech Republic, India, Indonesia, Korea, Mexico, Poland, South Africa, and Turkey. 2 We exclude China from the calculation of EM GDP because (1) China accounts for almost 4% of EM GDP but only about 17% of EM portfolio flows and (2) Chinese portfolio inflows appear to be behave differently than inflows to most other EMs in our group. QUANTIFYING RISKS OF PORTFOLIO FLOWS REVERSALS Our modelling also enables us to quantify several risk scenarios for the outlook for portfolio flows. Present financial conditions provide a benign backdrop for EM investments: current levels of risk aversion are quite low, while markets appear to have digested the ongoing Fed tapering process with little strain. A less benign implication is that there is significant potential for deterioration in the global financial environment that could drive another pull-back in EM flows: There is significant potential for deterioration in the global financial environment iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

11 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 11 Chart 12 Chart 13 IIF Portfolio Flows Tracker: Total Portfolio Inflows to EMs U.S. BBB-Rated Corporate Bond Spread over Treasuries, IIF group of 3 EMs, dashed lines = prelim. estimates percent per annum 5 Prior Estimates 8 4 through April Latest Estimates Average -2 through May May 13 Aug 13 Nov 13 Feb 14 May 14 Source: IIF Source: Bank of America Merrill Lynch. The BBB spread, our preferred risk indicator, is currently about 75 basis points below its long-term average of 2.3 pp (calculated from 2-214, Chart 13). According to our econometric estimations, a normalization of global risk appetite back to this level would be associated with a $53 billion reduction in portfolio equity and bond flows. A more severe increase in the BBB spread to 4 pp (about half the level observed during the global financial crisis) would trigger a reduction in portfolio inflows of about $184 billion. For comparison, the total annual value of net non-resident portfolio inflows to EMs during amounted to about $3 billion. A normalization of global risk appetite would be associated with a large reduction in portfolio flows. Another risk scenario relates to the outlook for U.S. monetary policy. As of late May 214, markets are pricing in a first Fed rate hike around August 215, followed by subsequent rate hikes at a pace of about 25 basis points per quarter. Under this scenario, the federal funds target rate would rise about 2 basis points over the next 3 years, reaching 2.25% in May 217 (our model uses Fed funds expectations 3 years out). By contrast, in the mid-2s tightening cycle rates were raised at twice that Chart 14 Chart 15 Estimated Contributions to EM Portfolio Flows Federal Funds Target Rates During Fed Tightening Cycles, push = global risk and Fed, pull = EM stock markets percent, t in months, market expectations estimated from fed funds futures and Eurodollar contracts t= is Month of First Rate Hike: Feb First Rate Hike Push Effect Pull Effect -5 Jan 13 May 13 Sep 13 Jan 14 May 14 Source: IIF. iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved First Rate Hike: June 24 First Rate Hike Expected in mid-215 t-6 t t+6 t+12 t+18 t+24 Source: Bloomberg, Datastream, IIF.

12 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 12 Chart 16 Federal Funds Target Rate percent per annum Path Expected Under Risk Scenario Path Consistent with Market Expectations as of May 2, Source: Bloomberg, CME Group, IIF. speed, by about 5 basis points per quarter (Chart 15). If investors were to come to expect that the federal funds rate would be increased at this faster pace again in the current cycle, markets would quickly price in a U.S. policy rate as high as 4.25% by May 217 in the futures curve, an additional increase of 2 basis points (Chart 16). According to our estimations, if market expectations were to shift abruptly to this new trajectory, this could trigger a reduction in EM portfolio flows in the order of $112 billion. Note that this scenario does not assume an earlier time for the first rate hike than currently priced in. Given that the first rate hike is likely only about months away, we see the main downside risk not so much in the timing of the first rate hike, but rather in the pace of subsequent rate hikes especially since there is a risk of a sharp reassessment in the outlook for rate hikes once the Fed sets a precedent for the interval between two rate hikes. Finally, note that the second risk scenario is not independent of the first one. If Fed policy expectations were to change adversely, this would likely reduce risk appetite, reinforcing the adverse impact on emerging markets. A shift in expectations to a more normal pace of Fed exit could trigger a retrenchment. A CLOSER LOOK AT WHAT DRIVES EMERGING MARKET VALUATIONS With pull factors tending to dominate push factors, portfolio flows this year suggest increasing investor discrimination across countries. This pattern is reflected in intraemerging market correlations in price movements. Taking the BRIC countries as an example, correlations for both bond and equity markets have been lower this year than during the second half of 213 as markets adjusted to the prospect of Fed tapering (Chart 17). Broader correlations across emerging markets, e.g. for exchange rates (see Chart 4, page 4) tell a similar story Further insight into investor discrimination can be found by looking at how markets price different risks into valuations a key consideration in asset allocation decisions and thus another factor impacting capital flows. Paul Ticu pticu@iif.com Despite a benign global risk backdrop, valuation of emerging market equities is at very low levels iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

13 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 13 Ample global liquidity and dovish forward guidance from central banks have left market gauges of risk at very low levels in other words, risk appetite is at multi-year highs (see the May and June IIF Capital Market Monitors). This appetite has been evident across a range of asset classes, including in particular flows to mature market equities and highyield bonds, which were barely dented during the taper tantrum. However, portfolio flows to emerging economies have benefited less from the more benign global risk backdrop (Chart 18), and this is reflected in valuations. While forward price-earnings ratios for developed market equities (near 14) are at their highest since 27, EM equities trade at a forward P/E of just over 1 the lowest since 25. A number of factors play a role in investors perceptions of the riskiness of emerging markets as an asset class, and hence how they are valued. At present, two key barometers are concern about the lack of momentum in EM economic growth (with prospects for China a particular focus see page 4) and the buildup in corporate debt, which has implications both for the relatively small EM corporate bond market and for EM sovereign debt markets. These factors have both general affecting emerging markets as an asset class and country specific dimensions. Ample global liquidity and dovish forward guidance from central banks have left market gauges of risk at very low levels. Concerns about EM growth momentum have intensified over the past year, but in fact the differential between emerging (EM-3) and mature (G-7) economies has shrunk dramatically over the past several years, dropping from percentage points in 28 to under 2.5 percentage points this year (Chart 19). This narrowing growth premium is clearly reflected in the long-running de-rating of emerging market equities relative to mature market equities (Chart 2). This is consistent with the downshift in overall earnings growth in emerging markets, from rates close to 2% in to the negative growth seen in After two years of expecting a pickup in earnings, forward earnings estimates now suggest negative growth in as well (Chart 21). Chart 17 Chart 18 Correlation Chart XX in BRIC Equity and Bond Markets 6-day rolling correlations BRIC Equities (MSCI) Source: MSCI, JP Morgan, IIF calculations H to date BRIC Sovereign Bonds (EMBIG) Risk Chart Appetite XX and Mutual Fund Flows to Emerging Markets BAML market risk index, inverted, 3mmav EM Debt and Equity Mutual Fund Flows, Net -2 Risk Appetite Index Source: BAML, IIF iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

14 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 14 Chart 19 Emerging -Developed Market Growth Differential percentage points, measured as the difference in oya growth Source: IIF Chart 2 Relative Valuation: Emerging vs. Mature Market Equities MSCI indices, price to 12M forward earnings estimates; dotted line = period average 1.5 EM Equities More Expensive EM Equities Cheaper Source: MSCI, IIF calculations In the same way that aggregate EM growth differentials with mature economies can help explain the relative rating of emerging market equities as an asset class, economic growth rates of individual countries can help explain some of the marked differences in equity market valuation across countries. Plotting expected 214 growth against a country s price-earnings ratio underscores the premium that investors currently place on growth. Countries like Mexico, Malaysia and Philippines, for example, enjoy both relatively high expected growth rates and high valuations. In contrast, Brazil, which has entered a fourth consecutive year of below-trend growth and is expected to see growth of only about 1.5-2% this year, is valued at just 15 times earnings well below the ratios of Mexico or Philippines. Our estimates suggest that growth prospects can explain about a third of the variation in individual countries forward price-earnings ratios, and about 2% of the variation in trailing price-earnings ratios. Chart 21 Emerging Markets: Actual and Forward Earnings Growth percent, oya, period averages Trailing Forward April Source: MSCI, IIF calculations Chart 22 Emerging Market Growth Projections for 214 and Trailing Price-Earnings Ratio 3 HUN PER 25 MEX PHL CHL 2 BRL ZAF KOR MYS IND THA 15 PLN COL IDN CZE CHN 1 TUR ARG y = x R² =.199 RUS Real GDP Growth Projections for 214, percent Source: MSCI, IIF. Trailing Price-toEarnings Ratio iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

15 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 15 Another source of heterogeneity in EM valuations can be found in the varying degrees of buildup of corporate debt across countries. In the ultra-low rate environment since the 28-9 financial crisis, emerging market borrowers particularly private sector non -financial and financial corporates have taken advantage of strong investor interest to come to market (see pages 18-2). This surge in issuance has helped make EM corporate debt a much more established asset class: capitalization of the benchmark J.P. Morgan CEMBI index rose to $825 billion at the end of Q1 214 almost six times its capitalization at the end of 29. Greater depth in this market certainly has made it more attractive to a broader range of investors. However, the buildup in emerging market corporate debt has raised investor concerns about the riskiness of this sector as well as vulnerabilities stemming from corporate indebtedness for individual EM countries. Since 27, overall EM corporate debt has risen from 55% to nearly 8% of GDP a big jump, and one which puts the EM average on a par with many mature markets. The EM average is heavily affected by China, where corporate debt is now nearly 15% of GDP, up from about 95% in 27 (Chart 22). One way to assess the impact of debt buildup on valuations is to look at the growth of corporate debt (as a percentage of GDP), plotted against the market s assessment of the health of corporate balance sheets the price to book ratio. The striking decline in the overall EM price-to-book ratio from about 2.5X in 27 to around 1.5X at present (Chart 23) is in part a reflection of investors concern about the buildup in EM corporate debt. On an individual level, countries that have had a larger buildup in debt (e.g. Hungary, Turkey, Brazil, Korea and most notably China) tend to have lower price-to-book ratios: our estimates suggest that perhaps a quarter of the variation in current price-to-book ratios can be explained by individual countries growth in corporate debt-to-gdp ratios since 27 (Chart 24 and 25). Higher corporate debt levels are another source of pressure on EM equity valuations Chart 23 Emerging Market Equities: Price-to -Book Ratio times, MSCI MXEF index Source: MSCI Chart 24 Chart 25 Emerging Markets: Non-Financial Corporate Debt EM Corporate Debt and Price-to-Book Ratios percent of GDP 4 China IDN Hungary Korea 3 ZAF Singapore Q1 27 IND Thailand MEX Russia Q3 213 India 2 TUR Czech Blue and red, vertical PLN BRA Brazil lines are weighted EM Turkey debt-to-gdp ratios for 1 Poland 27Q1 and 213Q3, KOR Saudi Arabia respectively. HUN RUS South Africa Indonesia Mexico Source: IMF, national sources, IIF calculations Book-to-Price Ratios y = -.3x R² =.2661 CHN Percentage point change in non-financial corporate debtto-gdp between 7Q3-13Q3 Source: IMF, MSCI, national sources, IIF calculations iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

16 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 16 MODEST DECLINE IN EM BOND AND EQUITY ISSUANCE IN 214 BUT SIGNIFICANT VARIATION BY SECTOR AND COUNTRY Emerging market bond and equity issuance slowed in the first five months of 214 relative to 213. While 213 as a whole saw over $8 billion in EM bond issuance (above the robust $75 billion average annual pace of 21-12), issuance in the first half of the year was particularly heavy. For 214 year to date, EM bond issuance has been almost 25% below that of the same period last year, mainly reflecting a 5% drop in issuance by non-financial EM corporates (Charts 26 and 27). While tensions surrounding the Russia/ Ukraine conflict likely weighed on issuance in the first months of the year, concerns about rapidly growing levels of corporate debt may also be playing a role (see pages 12 and 15). Anxiety surrounding the Chinese government s March decision to allow the first corporate bond default in recent history has evidently helped cool Chinese issuance non -financial corporate offerings are down 65% from the same period in 213 (Chart 28). Non -financial corporates in India and Brazil have also issued less bonds relative to the same period last year down by 6% and 3%, respectively while Russian non-financial corporate issuance, unsurprisingly, came to an effective halt (down over 9% from the same period last year). Though year-to-date issuance by financial corporates has been broadly at the same levels as last year, there have been significant differences across countries. The concern about China s first corporate default did not have much of an effect on issuance by the financial sector: Chinese financial corporate issuance this year has been 35% above last year s level. Korean financial sector issuance too has been robust this year, 7% higher than the same period last year (Chart 29). In contrast, financial corporates in Russia, Turkey, and India have reduced their bond issuance significantly. In contrast to non-financial corporates, sovereign bond issuance has been solid in 214, up 25% from last year s level China, Poland, Romania, Mexico and Turkey have all seen significant issuance this year. Around 6% of all bonds issued (sovereign and corporate) have been in local currency, slightly below the 213 average (Chart 3 and 31). Chart 26 Chart 27 Emerging Markets: Bond Issuance by Sector Emerging Markets Bond Issuance, by Country, monthly Non-financial, monthly 11 Financial 11 Russia Brazil India 1 Government 1 Mexico Others Korea 9 9 China Jan 13 May 13 Sep 13 Jan 14 May 14 Source: ThomsonONE, IIF calculations Jan 13 May 13 Sep 13 Jan 14 May 14 Source: ThomsonONE, IIF calculations iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.

17 CAPITAL FLOWS TO EMERGING MARKETS MAY 29, 214 page 17 Chart 28 China: Corporate Bond Issuance, monthly 55 Financial 5 Non-Financial Jan 13 May 13 Sep 13 Jan 14 May 14 Source: ThomsonONE, IIF calculations Chart 29 Financial Corporate Bond Issuance, by Country, monthly 6 China Korea 5 Brazil India 4 Malaysia Others Jan 13 May 13 Sep 13 Jan 14 May 14 Source: ThomsonONE, IIF calculations Equity issuance in emerging markets has also been below 213 levels. Excluding initial public offerings, EM corporates have raised some $43 billion year to date, around 45% less than the same period last year (Chart 32). With valuations for EM equities well below their long-run average (see pages 12 and 15), EM companies raised only $13 billion from 125 IPOs this year (Chart 33), lagging behind last year s pace. STILL VERY MUCH IN THE GAME: MUTUAL FUND INVESTORS AND EM ASSET ALLOCATION Despite marked swings during stress episodes (the 28 global financial crisis, the 211 Euro Area sovereign debt crisis and the Fed tapering shock), the asset allocation of Chart 3 Chart 31 Emerging Markets Bond Issuance, by Currency Bond Issuance: Share of Local Currency Bonds, monthly percent of total issuance 12 Foreign Currency Middle East/Africa 1 Local Currency 8 6 EM Europe ytd 4 Latin America Source: ThomsonONE, IIF. includes sovereign and corporate debt EM Asia Source: ThomsonONE, IIF. iif.com Copyright 214. The Institute of International Finance, Inc. All rights reserved.