A Bullish 2018 EM Outlook... any Questions?

Transcription

1 3 December 27 4:5PM EST A Bullish 28 EM Outlook... any Questions? n n n n n n n A solid outlook, but what is good enough for markets? We expect above-trend growth in 28, but the starting point is already a healthy one. Given markets trade growth acceleration and revisions, note that consensus forecasts are less positive (and below our numbers) on 28 growth than coming into 27. Historically, EM assets withstand up to a 25bp negative revision to consensus GDP without posting negative returns (a low risk in our view). What does a synchronous growth cycle mean for assets? EM and DM had a synchronized slowdown in 2-2 but diverged from The homogenous improvement in growth rates has not led to high return correlation across EM assets (though perhaps it has suppressed volatility). We see plenty of room for EM return dispersion next year, particularly in FX and local rates. Is 28 more of the same for EM Assets? The dynamics of continued strong growth suggest EPS and spread compression will continue to be the driver of equity and credit returns, respectively. However, higher US rates will likely pressure total returns in credit and local rates more so than in 27. Nevertheless, the levels of yields in EM local bonds (~6.2%) and a constructive EM FX outlook (~2% appreciation) should help cushion duration risk in aggregate. Equity - Where is the best growth story? We are tactically bullish on Asian growth, both in trade (KR,TW) and domestic (CH,IN) but for 28 we would focus on where profit margins have room to recover (BR,MX,ID,RUB). Credit - Stick with high yield? We expect high-spread credits to continue to outperform in 28. This strategy underperforms when spreads widen by more than 2bp per month, but we see low risk of persistent spread widening. Local Rates - Any insulation from US rate risks? BRL, INR, COP, PEN, and RUB local rates trade with low beta to US rates; CLP is more sensitive but has a steep curve. Inflation is supportive in these EMs but will be less so for BRL and INR. FX - When to rotate into Value? We prefer global growth exposure in commodity-ems (CLP, PEN, BRL) and Asia Growth FX (KRW, IDR, INR). But Value (MXN, ZAR, TRY) has started to recover. We prefer to trade these FX vs. their local credits in relative value space. Caesar Maasry +(22) caesar.maasry@gs.com & Co. LLC does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. For Reg AC certification and other important disclosures, see the Disclosure Appendix, or go to Analysts employed by non-us affiliates are not registered/qualified as research analysts with FINRA in the U.S.

2 A Bullish 28 EM Outlook... any Questions? Over the past couple weeks, we have published our 28 global outlook reports painting a relatively bullish picture for fundamentals and asset markets heading into the new year. In this report, we respond to the most commonly asked questions by macro clients regarding our outlook, which have generally revolved around two issues: The first centers on downside risks to our view, particularly with regard to growth and/or policy shocks. Second, investors generally characterize the strong performance across risky assets in 27 as beta amidst a highly-synchronized recovery in global growth; looking at 28 they often question if dispersion may pick up and where we have the highest conviction within asset classes. While we find that most investors agree with the relatively upbeat view for 28, most macro clients fall in the camp of reluctant bulls, and in the below report we aim to address some of the prevailing concerns heading into next year. Question - How much growth is good enough for EM assets next year? The overarching theme of our global outlooks has been an emphasis that the global economic expansion has further to run and that asset markets, although stretched in terms of valuation signals, should continue to perform well driven by fundamentals. While clients acknowledge underlying strong global growth dynamics, one key question is how much of this growth is already expected/priced in? As we have long argued, it is growth acceleration not growth rates that drive asset markets. From this perspective, we take a look at Consensus Economics forecasts heading into a new year historically - Exhibit plots the year ahead growth acceleration expected across EM (bars) and the subsequent realized acceleration posted during that year (diamonds). Interestingly, we find that consensus growth (acceleration) expectations are actually lower than in previous years. In each of the past 5 years, economists across the street have signaled between 3 and 6 bp acceleration of EM GDP from one year to the next. As the red diamonds illustrate, however, these expectations proved highly optimistic. Even in 26, as EM assets recovered strongly, expectations were too high starting the year (though the miss was much less significant than during 23-25, and most of that miss was due to a weak Q). It was only this year (27) that marked a veritable acceleration in EM GDP, and EM assets have continued to perform well. Looking into next year, consensus forecasts call for quite a muted EM acceleration (of just 5bp). This is somewhat reassuring, in that it suggests the bar for upside surprise next year is actually somewhat low. Our GDP forecasts call for a 4bp acceleration in EM GDP (GDP-weighted) and 2 bp on an equal-weighted basis (which is how we present the data in the exhibits below). Relative to last year s outlook, consensus views have moderated most in, CE-3,,, and Taiwan but incrementally more positive in, Peru, Colombia. 3 December 27 2

3 Consensus once again forecasts a bullish outlook for and in terms of growth acceleration. Exhibit : Growth forecasts heading into next year are actually muted relative to recent history 2. % Consensus Growth Acceleration 28 vs 27.5 %. %.5 %. % (.5)% (.)% (.5)% (2.)% Consensus Expected Growth Acceleration heading into new year (bars) Realized Growth Acceleration Peru Colombia Taiwan Thailand Korea Czech Republic 27 vs 26 (dots, as of Dec '6) % Source: Consensus Economics, Haver Analytics, Global Investment Research In terms of growth revisions, 27 marked the first full-year since 2 that EM GDP estimates were revised upward (4bp in total), which has been met with strong market performance. Below, we highlight the breaking points across EM assets relative to growth revisions. For EM FX and equities, a growth revision of negative 25bp or worse tends to present an insurmountable obstacle for positive returns. For EM sovereign credit (spread component), the threshold is a bit higher, at roughly 3bp of negative revisions (see Exhibit 2). Perhaps coincidentally, the average revision to EM GDP forecasts over the course of a year (dating back to 2) is -32bp; so perhaps asset markets have been conditioned to expect negative growth surprises relative to consensus starting numbers entering a new year. Exhibit 2: EM assets can withstand up to a -25bp revision to consensus GDP 6% 4% 2% EM GDP Revisions & Asset Returns (quarterly revisions/returns) EM Equity (LHS) EM Credit Spreads (RHS) -5bp -4bp -3bp -2bp -bp -2% -4% -6% EM FX vs. USD (LHS) 25bp negative revision is the "breaking point" Revision to Consensus EM GDP Forecasts -5bp bp or worse or better bp bp 2bp 3bp 4bp 5bp Source: Consensus Economics, Datastream, Global Investment Research 3 December 27 3

4 Another argument brought forth by clients is that incremental growth revisions should matter less when the overall growth picture is healthy (above trend), and that other factors such as positioning and valuation may matter more as the cycle matures. We find some evidence of this, in that the beta of EM asset returns to consensus GDP revisions appears to decline when decently above trend growth, but the coefficient is still quite positive (see Exhibit 3). Rather, we view the very high coefficients on growth surprises when growth is near trend as a signal that markets focus on growth data to determine if the cycle is heading into a proper expansion or back towards contraction. When growth is below trend, it is potentially the direction of growth rates that matter more than incremental growth surprises (think of in 27, which posted growth below expectations but still far better than a dismal 26 and asset markets rallied strongly). Exhibit 3: GDP revisions are most impactful when EM growth is near trend std. dev move of Eq, FX & Credit to bp shift in EM GDP forecasts EM GDP growth vs. average is... Beta of EM Assets to Shifts in Growth Forecasts (grouped by 'lateness of the cycle) Well Below (-75 bp or more) Below (-75bp to -25bp) Near Average Above (25bp to 75bp) Well Above (+75bp or more) Source: Consensus Economics, Datastream, Global Investment Research Question 2 - What does a synchronous growth cycle mean for EM assets? A growing narrative in the market (one that we subscribe to as well) is that most economies in the world are growing in a synchronous fashion - which suggests that the growth cycle has further room to run. Clients often respond with two questions: first, what exactly do we mean by a synchronous cycle, and second, is there a clear implication for asset markets other than a positive view that the cycle may endure? On the first, we have argued that the current synchronous EM expansion (defined as having low dispersion across growth rates) is likely to have longevity as well. Looking at both EM and DM, the cycles are starting to match up as well. Unlike the 2-22 synchronous slowdown through the Euro sovereign crisis (see Exhibit 4, left-hand side), global growth is now improving. In contrast, the period was quite asynchronous, in that DM recovered quickly, whereas EM growth continued to slow. This was an important divergence, because incremental tightening in US monetary policy had a significant impact on EM assets (as EM growth was weak). With the synchronous improvement over the past 8 months, however, we find EM is much more insulated from US monetary policy. 3 December 27 4

5 Within EM, the recent period appears to be quite a homogenous recovery, usually only seen after global recessions (as seen in the early 2s and again in 28). As we show below, almost all the EM economies we cover had accelerating growth rates in late 26/early 27, and that share has come down to roughly 6 (see Exhibit 4, right-hand side). But importantly, these figures are still well above those of the bull market years of EM. Exhibit 4: Growth is in-sync between EM and DM, as well as within the EM complex 9% 8% 7% 6% EM & DM Growth "Synchronicity" synchronous slowdown EM Growth (CAI, LHS) asynchronous cycle synchronous expansion % % % EM "Synchronicity" % of EM economies with accelerating growth very high share of EMs accelerating 5% 4% 3% Dec-9 DM Growth (CAI, RHS) Jun- Dec- Jun- Dec- Jun-2 Dec-2 Jun-3 Dec-3 Jun-4 Dec-4 Jun-5.5%..5%. 3 2 Dec- Dec- Dec-2 ~5 EMs "accelerated" during bull market Dec-3 Dec-4 Dec-5 Dec-6 Dec-7 Dec-8 Dec-9 Dec- Dec- Dec-2 Dec-3 Dec-4 Source: Global Investment Research We do not find a very clear market implication of synchronous growth cycles, apart from the argument we have previously made that this sort of expansion is likely to continue through 28 (and therefore broadly support risky assets). Clients have asked if there is an implication in terms of return dispersion - with the assumption that less fundamental dispersion should lead to great market correlation. Surprisingly, we find that opposite has in fact taken place (see Exhibits 5 and 6). Across EM equities, FX, and credit markets, we find that the average pairwise correlation of returns within each asset class has actually declined in 27. In local rates space, correlations are often quite low to begin with, but even in this asset class, correlations are towards the low end of the distribution. That said, correlations are not always straightforward in their interpretation; a more likely explanation for low correlations is the current regime of low volatility. Although correlations are low, asset prices are generally moving in the same direction (of the 76 EM country-assets we cover across equities, FX, credits, and local rates only 9 have posted negative returns this year). This result is likely driven by the strong, synchronous global growth cycle in that a strong economic underpinning is an important driver of the current low volatility regime. Our core message to investors is that we like EM assets en masse, in particular in equity and credit, but for EM-focused investors, we discuss country-specific stories in the next sections. 3 December 27 5

6 Exhibit 5: Equity and FX correlation has declined Exhibit 6: Fixed income market correlation is also declining Avg. Pairwise Correl. across EMs (6-mo correl of weekly returns) EM FX Equity falling correlation Dec-4 Jul-5 Feb-6 Sep-6 Apr-7 Nov-7 Jun-8 Jan-9 Aug-9 Mar- Oct- May- Dec- Jul-2 Feb-3 Sep-3 Apr-4 Nov-4 Jun-5 Jan Avg. Pairwise Correl. across EMs (6-mo correl of weekly spread/rate chg) Credit Local Rates falling correlation Dec-4 Jul-5 Feb-6 Sep-6 Apr-7 Nov-7 Jun-8 Jan-9 Aug-9 Mar- Oct- May- Dec- Jul-2 Feb-3 Sep-3 Apr-4 Nov-4 Jun-5 Jan-6 Source: FactSet, Global Investment Research Source: Datastream, Global Investment Research Question 3 - Is 28 shaping up to be more of the same for EM Assets? From a macro standpoint, our outlook for 28 is similar to that of the 2H of 27 - strong growth but an incremental tightening of core interest rates. Clients often then suggest we are calling for more of the same as we enter next year. In aggregate, this is largely true vis-a-vis EM assets: we expect further gains in 28 but not at the same pace as earlier in 27. To frame this view, we run quickly through the asset classes from a total return composition stand point relative to past cycles and highlight where we see things changing at the margin. Equity: Even after a strong run in 26, EM equities have posted their strongest post-gfc return this year, which has been driven by all components (see Exhibit 7). The clear shift this year was a significant improvement in EPS growth - last year s gains were driven almost wholly by P/E expansion. Looking into 28, we do expect more of the same in that EM EPS will likely be the main driver of return (~2% vs. the ~2 gain in 27). We see limited scope for P/E multiples to expand as US rates rise, but as historical data show, EM P/Es do tend to continue to rise somewhat when EPS growth remains healthy. The bull market years of were driven by both strong EPS growth and continued P/E expansion, for example, with modest boosts from FX and dividends. 3 December 27 6

7 Exhibit 7: EPS growth has driven equity performance in 27, similar to bull years Equities PE EPS FX DIV Total Return Source: FactSet, I/B/E/S, Global Investment Research Credit: This year, EM sovereign credit has posted a strong return, albeit marginally less than 26 s solid return (see Exhibit 8). Compositionally, 27 has been buoyed by US rates staying effectively flat point-to-point, but as spreads tightened significantly last year, both carry and the space for incremental spread tightening has been more limited. Heading into 28, EM credit appears similar to the 25 setup: EMBI spreads hovered at 34bp in early 25 (vs. 32bp currently) and US rates rose ~4bp during the year (we forecast ~6bp rise in US -year rates next year). However, the level of starting yield of EM USD cash bonds was much higher back then at ~7.5% vs. 5.5% currently. That carry component gave much more cushion to digest potential spread widening or US rate rises. For 28, we expect total returns of EM sovereign credit to be below those of 26 and 27, more likely in the ballpark of 5% (which represents a 35bp spread tightening from current levels and a ~6bp rise in US rates). Exhibit 8: We expect lower total returns in credit in 28 with less scope for further spread compression and duration risk moving against cash bonds 4% 2% 8% 6% 4% 2% -2% -4% -6% -8% Credit Spread Change Duration Risk Carry Total Return Source: Datastream, Global Investment Research 3 December 27 7

8 EM FX: Despite the market narrative that EM FX has been the disappointing asset class of EM this year, we find that EM FX returns are similar to those of the bull market years of Nominal carry remains well above levels of 25-27, but even spot appreciation has been a major factor in EM FX performance this year (see Exhibit 9). Looking into 28, we forecast an average spot appreciation of 2% and current nominal carry is 2.4%, suggesting roughly 4.5% returns vs. the USD next year. This would be quite similar to 26 s EM FX performance. Exhibit 9: FX has performed strongly this year from carry and spot appreciation 5% 5% FX FX Spot Yield Total Return -5% - -5% Source: FactSet, Global Investment Research EM Local Rates: EM local bonds have tended to perform with EM FX historically (though 25 was a year in which EM local bonds shrugged off FX weakness), which sets a somewhat positive view for 28. However, our view of rising core rates is likely to pressure local duration as well. Nevertheless, local yields and FX shifts have generally been the more important factor for total returns than pure duration exposure (see Exhibit ). At the GBI-EM level, nominal yields are currently 6.3%, which is in-line with 5-year average yield levels. If local rates rise in-line with our above-consensus forecasts of US -yr rate rises (though GBI duration is closer to 5 years), this would wipe out roughly /3 of the yield return. Given our views of modest EM FX appreciation, our base-case signals ~6% GBI-EM total returns for next year, which is once again more similar to the return profile of 25 than more of the same (i.e. 26 or 27). 3 December 27 8

9 Exhibit : Local rates face a tougher climb in 28, but stable FX should keep returns positive 2 5% 5% -5% - -5% Rates FX Spot Yield Duration Total Return Source: Datastream, Global Investment Research In the below section, we detail the major theme in questions asked by clients in response to our generally positive views across the EM asset spectrum. Question 4 - Where is the best growth story across EM Equity? Our core message this year has been to be long of EM equities en masse. A synchronized improvement in EM growth dynamics has strong historical precedent in suggesting that the entire EM equity complex should outperform DM equities. As we show in Exhibit below, for most of the period, almost all EMs outperformed the DM benchmark in tandem. For 27, roughly 6 of EM equity markets have outperformed (the aggregate has outperformed by a solid this year buoyed by large equity markets primarily in Asia). While we expect broad EM outperformance to continue in 28, we expect differentiation to increase as the cycle matures. In recent years, we have found that relative EPS strength has been the primary axis of differentiation for EM equities, more so than relative valuation for example. Exhibit 2 details that 27 once again has been a year where cross-sectional EM equity returns have been driven by relative earnings power. 3 December 27 9

10 Exhibit : Most EMs are outperforming DM together Exhibit 2: EPS growth is the clear driver of relative equity returns almost all EMs outperformed Share of EM Equity Markets Outperforming DM Price Return (local FX) CNY Earnings Growth has been the major driver of returns 4 PEN KRW 3 PLN TRY EM ZAR INR 2 CLP PHP BRL HUF IDR TWD COP THB CZK MXN MYR 27 EPS Change RUB (forward NTM integer) Source: FactSet, Global Investment Research Source: FactSet, I/B/E/S, Global Investment Research With broad improvement in global growth as articulated above, we think EPS growth will be boosted most in economies with significant operating leverage. One relatively simple way to measure potential leverage is to compare profit margins with history. Below, we compared the current level of profit margins vs. the average of 2-2 (a period when EM growth was at a similar level to our 28 forecasts). While we like Asian equities from a tactical basis, with regional overweights in, Korea, and, as well as Taiwan relative to the broad EM average, we see relatively limited room for further margin expansion in these EMs. For the full year of 28, we maintain a positive view on and would look to rotate into,, and (particularly the domestic exposure in banks and retail there). faces a number of potential political headwinds both domestically (election) and from abroad (US trade policy), and recent growth data have been disappointing, which have weighed heavily on the equity market (worst EM equity market since July). Soft data have been slightly more encouraging (consumer confidence) and to the extent these surveys spill into hard data, we see scope for Mexican equities to regain much of their lost ground in December 27

11 Exhibit 3: There is more room for EPS upside in EMs where margins are still low vs. history Current Profit Margin vs. 2-2 Average less operating leverage -6 more "room to run" -8 Colombia Thailand Korea Taiwan Source: FactSet, I/B/E/S, Global Investment Research Question 5 - Will high yielders continue to outperform in EM credit? Active-managed EM credit funds have posted solid returns this year in excess of the EMBI benchmark, and their performance has been fueled by a beta rally in which high-yielding credits have tightened the most. Given the relatively stretched valuation of EM credit (at least relative to recent history) heading into 28, credit managers are acutely focused on the risk of spread widening and how best to position for the new year. As we show in Exhibit 4, a simple strategy of going long credits with above-median spreads vs. those with below-median spreads has outperformed consistently since early 25. In general, this is a pro-risk trade and correlates (inversely) with overall EM credit spreads. From a top-down perspective, we expect this trend to continue in 28 given our constructive view on EM credit spreads. 3 December 27

12 Exhibit 4: High-spread credits have consistently outperformed low-spread EMs bp 6 High vs. Low Spread Credit (total return) bp 4 2bp 3bp EMBI Spread (inverted, RHS) 4bp 5bp Dec-4 Jun-5 Dec-5 Jun-6 Dec-6 Jun-7 Dec-7 Jun-8 Dec-8 Jun-9 Dec-9 Jun- Dec- Jun- Dec- Jun-2 Dec-2 Jun-3 Dec-3 Jun-4 Dec-4 Jun-5 Jun-8 Dec-8 Source: Datastream, Global Investment Research This relationship is quite consistent across the entire EM credit complex - as we illustrate below, there is a quite linear relationship between starting level of spread and subsequent total return across sovereign bonds (see Exhibit 5). There a few notable outliers, including Lebanon,, and, all of which have had significant political headlines recently. But importantly, this relationship of high-spread credits outperforming has been true even since May, a period when overall EM credit spreads have been roughly flat. To analyze the risks of running a high-spread strategy, we show in Exhibit 6 the average return of the strategy dependent upon overall EM credit spread changes. On average since 24, we find that high-spread credits tend to underperform in months where aggregate EM spreads widen by 2bp or more (an event which occurred in early November this year but has quickly reversed). In short, it takes a relatively significant shock to EM credit spreads for this strategy not to work (shocks such as the global financial crisis, the Euro sovereign crisis, and the commodity unwind in 24 had lasting impact to this strategy). Looking out to next year, most clients view a hawkish Fed as a potential risk to EM - but this strategy outperformed during the taper tantrum, so we view that risk as limited here. 3 December 27 2

13 Exhibit 5: Starting spread clearly explains performance in 27 across EM credits Exhibit 6: A high vs. low yield strategy tends to back-fire when spreads widen by more than 2bp in a month Spread as 7bp of May-7 Apart from Lebanon, and... Ecuador 6bp Higher Spread Credits outperform Ukraine 5bp Egypt 4bp Argentina Lebanon EMBI Sri Lanka Dominican 3bp Kazakhstan 2bp Colombia Vietnam Panama Peru bp Total Return since May bp -2% 2% 4% 6% 8% 2% 2.5% 2..5%..5%. (.5%) (.) (.5%) (2.) Return of High Yielding Credits vs. Low Yielding Credits -4 bp or better -mo Change in EM Spread or worse Source: Datastream, Global Investment Research Source: Datastream, Global Investment Research As mentioned, not all high-spread credits performed well this year. Of the major EM economies, and stand out as underperformers in credit space. While we acknowledge that buying last year s underperformers has tended to work in EM recently, we would rather gain exposure to relative value trades within these markets than go long the credits. Specifically, we find that the FX of these two economies has larger underperformed the credit, and would take the other side of the recent trend. We include in this group, not because the credit has been a notable outlier, but rather because it is also an EM in which the FX has significantly lagged credit. As we show in Exhibit 7 (right-hand side), the FX to credit carry differential has increased to historical highs in these three EMs. There is also historical precedent to highlight that FX can outperform credit during an EM bull market (which happened from 26-27), and we view the current entry point as attractive from a recent technical perspective. Exhibit 7: For these 3 EMs with idiosyncratic risks, we prefer FX to credit Long MXN, ZAR, TRY vs. Short credit in,, (Short credit = Long US bonds, short EM USD bonds) Long FX vs. Credit in,, 27: FX pricing more 'idiosyncratic' risk than credit FX vs. Credit Carry Differential (2mo carry of FX vs. EMBI spread) 3 Dec-4 Total Return Jun-5 Dec-5 Jun-6 Dec-6 Jun-7 Dec FX Bull Market Jun-8 Dec-8 Jun-9 Dec-9 Jun- Dec Total Return Mar (.) (2.) (3.) (4.) Source: Datastream, FactSet, Global Investment Research Question 6 - Can any EM local rates markets withstand the pressure of core rate rises? While the views on core rates are somewhat mixed across macro clients, there is little pushback against our view that rates should generally rise in 28 driven by at least 3 December 27 3

14 two Fed hikes (we forecast 4 for calendar 28). In aggregate, EM local rates move similar to US rates, but there is significant variation across the various economies. In Exhibit 8, we rank EMs by the beta of local rates to US rates (data since 24). As shown,,,, and have the highest beta to US rates with reasonably high correlation as well. On the other side,,, Colombia, and Peru have exhibited very low correlation with US rates. As we wrote in the section above, we do expect EM local bond returns to be lower in 28 than this year, but the magnitude of US rate rises is not enough to fully reverse the relatively high levels of yield in local bonds (EM average of 5.% at the 5-year tenor). Looking across local markets, we would focus on EMs with a relatively low sensitivity to US core rate shifts and a relatively high starting point of yield.,,, Colombia, Peru, and fit these metrics quite well. We would also consider local inflation dynamics, which are becoming less positive in places like and next year (see Exhibit 9). We are more constructive on CLP and PEN than COP, but with rate curves relatively steep in all three economies vs. the EM average (5 year vs. year), and a strong disinflationary impulse in Colombia next year, we view these markets as having attractive local bonds. Our economists have flagged the potential for 5-year rates to decline in, but we expect the FX to weaken towards 6. next year. Exhibit 8: Higher US rates will likely pressure EM duration, but some local rates have a low correlation with the core rate complex (.2) Colombia correl. (dots) Peru US Rate Sensitivity on Local EM Rates (5 year rates) Taiwan Czech Rep. Thailand Korea beta (bars) 5% Local EM Yield vs. 4% 5y 3% Yield US-rate Sensitivity TRY 2% % BRL 9% ZAR 8% RUB 7% INR MXN 6% COP IDR 5% PHP PEN CLP 4% CNY MYR 3% KRW 2% PLN THB Beta to % CZK HUF US Rates TWD (.2) Source: Datastream, Global Investment Research 3 December 27 4

15 Exhibit 9: We see a differentiated picture across EM inflation dynamics 28 vs. 27 Inflation (GS forecast) Colombia -2pp -pp pp pp 2pp Source: Global Investment Research Question 7 - What bucket of EM FX do we see the best upside? Unlike EM equity and credit, where the asset classes have posed consistently strong returns since early 26, EM FX has been a bumpier ride over the past year. Investors are already switched on to the importance of picking the right spots in EM FX, and we have outlined our preferences in two Top Trades (link), going long BRL, CLP, and PEN as well as long KRW, INR, and IDR vs. SGD and JPY. Macro investors have generally agreed with the exposure in Asia tactically, but have focused on the value currencies of MXN, ZAR, and TRY, which have been the biggest underperformers during 2H 27. As shown below, we find it useful to group EM FX into various clusters as a way to think about their macro profiles. Below, we show the carry and value statistics of a Value cluster (MXN, ZAR, TRY), a Oil-sensitive cluster (COP, RUB), and Metals cluster (CLP, PEN), and a Global Trade cluster (KRW, TWD, IDR, THB). Value-FX is clearly the least expensive based on our GSDEER and GSFEER metrics, to the tune of 2 undervaluation and offers a high nominal carry of ~8%. The Metals cluster is also undervalued by roughly 5% with low carry (~%). Global Trade and Oil-FX both screen as overvalued on our metrics. Our preference for some of the Metals and Global Trade FX is more on their macro sensitivities than valuation per se (though it helps with CLP and PEN), and we have shown how the FX pairs in our Top Trades display equity-centric characteristics and perform strongly during periods of rising global growth (link). 3 December 27 5

16 Exhibit 2: Value FX comes with relatively high carry Exhibit 2: We like growth-sensitive FX for their macro affinity more than valuation reasons Nominal 2-mo Carry Overvaluation vs. GSDEER & FEER 5% Global Trade Value vs. Carry Oils % -5% - -5% Rest of EM Metals Value. Value (MXN, ZAR, TRY) Oils (COP, RUB) Rest of EM Metals (CLP, PEN) Global Trade (KRW, TWD, IDR, THB) -2-25% 2-mo Carry -2% -% % 2% 3% 4% 5% 6% 7% 8% 9% Source: FactSet, Global Investment Research Source: FactSet, Global Investment Research Looking at recent performance trends, we find that commodity-sensitive FX have followed somewhat similar paths recently, though Oil-FX outperformed much more meaningfully in H (see Exhibit 22). Going forward, we expect oil prices to remain relatively flat, but have more optimistic forecasts on Copper, driven by continued strong demand - which in turn should keep the CLP and PEN on appreciating paths into 28. In Exhibit 23, we contrast the Global Trade FX with Value FX. Again there is a similar trajectory for most of the year, but a significant divergence opened up in September and October, where Value underperformed. All three Value FX pairs have dealt with idiosyncratic risks this year and we find it hard to argue that political headlines will wane anytime soon (only in is there a clear catalyst in December that could possibly cool the political volatility). As we have often learned in emerging markets in recent years, no news can be good news for inexpensive assets - and a muddle-through scenario in any of these economies can lead to significant gains in the local FX. As mentioned above, we would prefer to trade these currencies relative to their local credits as a way to hedge the idiosyncratic exposure. But directionally, we believe FX appreciation in 28 is more likely than further weakness. 3 December 27 6

17 Exhibit 22: Commodity-sensitive EM FX has performed strongly in 27 Exhibit 23: Trade-FX has done well recently, with Value starting to reverse recent weakness 2 8 Cumulative Total Return of FX vs. USD "Metals-FX" (CLP, PEN) "Oil-FX" (COP, RUB) Cumulative Total Return of FX vs. USD "Value-FX" (MXN, ZAR, TRY) "Trade-FX" (KRW, IDR, THB, TWD) right-axis Jan-7 May-7 Jul-7 Nov-7 Jan-7 May-7 Jul-7 Nov-7 Source: FactSet, Global Investment Research Source: FactSet, Global Investment Research 3 December 27 7

18 Appendix Macro Forecasts and Activity macro forecasts for major EMs and world aggregates Market GDP growth Inflation Policy Rates FX Forecast (USD) 26 27E 28E 26 27E 28E Current 27E 28E Current 3-m 6-m 2-m Colombia Czech Rep Korea Peru Taiwan Thailand EM DM World Source: Haver Analytics, FactSet, Global Investment Research Current Activity Indicators (3-month smoothed) Current activity indicators across EMs Jan-4 % 3-mo smoothed EM CAI (Avg. of 9 major EM CAIs) Apr-4 Jul-4 Oct-4 Jan-5 Apr-5 Jul-5 DM CAI (Avg. of US, Euro Area, and Japan CAIs) Oct-5 Jan-6 Jul-6 Jan-7 3-mo smoothed Jul Momentum (last 3-mo vs prior 3-mo chg in CAI, %) BRL PHP PEN COP KRW ZAR CLP PLN RUB IDR CNY MXN CZK TWD MYR THB HUF -3. Growth (Current CAI vs. -4. Consensus long-term -5. growth, %) INR TRY Source: Global Investment Research Source: Consensus Economics, Global Investment Research Financial Conditions Index EM inflation and policy monitor EM FCI 2.5pp 2.pp.5pp Inflation Change (now vs. 3-mo ago) Inflation & Policy Monitor INR Cutting No Chg. Hiking TRY Jan-3 Jul-3 Jan-4 Jul-4 DM FCI Jan-5 Jul-5 Jan-6 Easier financial conditions Jul-6 Jan-7 Jul-7 Jan-8.pp.5pp.pp (.5pp) THB PLN COP CZK CNY HUF MYR TWD PHP ZAR CLP KRW BRL IDR MXN (.pp) PEN RUB Inflation Rate (.5pp) (current, yoy) % 2% 3% 4% 5% 6% 7% 8% 9% % 2% Source: Global Investment Research Source: Haver Analytics, Global Investment Research 3 December 27 8

19 Central Bank Policy and Inflation Czech Rep Tight Loose real policy rate (%) BRL RUB CNY CLP IDR TWD KRW PEN INR THB RON PHP PLN Policy Rate Inflation (%) Level (%) Market Pricing GS Forecasts Recent Policy Changes Curr 2m fwd 3m ch Current Real 3mo 6mo 2mo 3mo 6mo 2mo m 2m 3m 3-6m Colombia Czech Rep Korea Peru Romania Taiwan Thailand Current (vs 3yr average, Bars) Thailand Korea HUF CZK Financial Conditions Inflation vs Policy ZAR MYR COP TRY - current inflation minus target (%) MXN 3mo Change (RHS, Dots) Source: Haver Analytics, Bloomberg, Global Investment Research current inflation (%) (bps) KRW RUB.5 current inflation vs 3mo ago (pp change) mo Market Implied Policy Changes Market implied 3 Korea Czech Rep Inflation Momentum IDR Thailand MXN CZK BRL HUF PLN CLP CNY TWD ZAR COP MYR PHP THB Taiwan INR RON GS Forecasts Colombia TRY 3 December 27 9

20 EM Macro Slice performance vs. MSCI EM vs. MSCI EM External Demand 3 25 vs. MSCI EM Commodity Cyclicals vs. MSCI EM Banks Domestic Cyclicals 75 Defensives EM valuation and earnings growth MSCI EM Valuation P/E P/B D/Y Earnings Growth (NTM) (LTM) (LTM) 26 27E 28E (4.3) Korea (7.6) (3.3) Taiwan Thailand EM (Local) EM (USD) % 2 5% 5% -5% - -5% Aug- Feb- 2 Aug- MSCI EM (Consensus EPS growth revisions in USD) Feb-2 Aug-2 22 Feb-3 Aug-3 23 Feb-4 Aug-4 24 Feb-5 Aug *Earnings growth based on consensus estimates in local currency EM cross asset valuation.6.4 Average Valuation Z-Scores across Equity, FX, Credit, and local Bonds.2. (.2) (.4) (.6) (.) +ive z-score = overvaluation, -ive z-score = undervaluation Equity valuation includes P/E, P/B, D/Y in z-score Real bond yields using forward inflation expectations HUF BRL PLN PHP PEN CZK INR IDR KRW COP THB TWD MYR RUB MXN CNY CLP TRY ZAR EM Philippine Czech ReKorea Peru Thailand Taiwan Colombia EM Equity PE (NTM).2x 3.x 9.2x 2.7x 5.2x 9.2x 2.3x 4.2x 8.8x 5.9x 5.3x 4.2x 3.3x 5.7x 6.4x 6.3x 3.6x 8.2x 6.2x 2.7x FX GS DEER (2%) +5% (6%) (9%) (4%) +9% (8%) (7%) +% +5% (%) (2%) (2%) (2) +5% (25%) (9%) (46%) (44%) (%) Credit Spread bp 233bp bp 48bp bp 34bp bp 69bp bp 9bp 8bp 232bp 45bp 39bp 286bp 288bp Bonds Real Rates -.5% 5.%.5%.7% -.4%.5% -.7% 2.7% 2.8% 3..% - 2.7%.3% 3.4% 3.% % 4.8%.9% Equity Z-Score (.7) (.4) (.3) (.22) (.82) (.42).67 (.63) FX Z-Score (.9) (.3).3 (.33) (.4) (.8) (.) (.73). (.38) (.52) (.7) (.9) (.) (.5) (2.4) (.38) (2.49) (.67) (.88) Credit Z-Score (.2) (.37).9 (.7).58 Bonds Z-Score (.) (.64) (.35).33 (.62).3 (.4).68 AVG Z-Score (.33) (.42) (.67) (.67) (.72) (.78).26 Source: FactSet, Datastream, Global Investment Research 3 December 27 2

21 Flows and Positioning EM portfolio investment and exchange-reported foreign flows 5 EM balance of payment Quarterly foreign portfolio investment ($Bil, latest = Q2-27) 6 4 EM exchange-reported foreign inflows ($Bil, Cumulative) Foreign Inflows EM Equity Flow EM Debt Flow Jan-5 Mar-5 May-5 Jul-5 Sep-5 Nov-5 Jan-6 Bond Equity May-6 Equity:,,, Korea,, S. Africa, Taiwan, Thailand, Bond:,, Korea,, S. Africa, Thailand, Jul-6 Nov-6 Jan-7 May-7 Jul-7 Nov-7 Jan-8 Mar-8 Source: Haver Analytics, Bloomberg, Global Investment Research Cumulative flows into EM mutual funds 7 6 $bn Cumulative flows into EM funds 7 6 Active vs. Passive: Cumulative flows into EM funds ($bn) Fund Flows EM credit funds (hard currency bond funds) EM equity funds EM bond funds (local currency bond funds) Passive, equity funds Active, fixed-income funds Passive, fixed-income funds Active, equity funds -2-3 Source: EPFR, Global Investment Research Cross-asset mutual fund positioning Fund Positioning 7% 6% 5% 4% 3% 2% % -% -2% -3% -4% -5% -6% Cross-Asset Country Allocation relative to MSCI EM, EMBI Sov., GBI-EM Equity Funds Credit Funds Loc. Bond Funds Average Peru Taiwan Thailand Korea Colombia Peru Taiwan Korea Czech Rep. Colombia Thailand Czech Rep. Chg in average positioning from 6-mo ago Source: EPFR, Global Investment Research 3 December 27 2

22 FX Performance Return vs USD Forecasts GS Fair Value FX Carry Spot w% 3m% YTD% 3m 6m 2m DEER FEER % overval yr fwd real y ZAR MXN PLN MYR RUB PHP Czech Rep CZK Romania RON INR HUF Korea KRW Argentina ARS Thailand THB Peru PEN CNY Taiwan TWD BRL Colombia COP IDR Egypt EGP TRY CLP % 4% 3% 2% % -% -2% -3% Current (Bars) Real FX Carry (2mo) %-ile vs History (RHS, Dots) % overvalued vs FEER Current Level vs GS Fair Value Metrics TRY ZAR MXN CNY PHP MYR TWD COP CLP HUF PEN RUB CZK INR THB IDR Expensive KRW -4% -5% TRY MXN RUB ZAR BRL IDR INR COP CNY PEN PHP THB MYR CLP PLN KRW TWD CZK HUF -4 BRL Inexpensive % overvalued vs -6 DEER % 8% 7% 6% 5% 4% 3% 2% % -% -2% -3% -4% % Appreciation Forecasted 3mo 2mo BRL CLP TRY IDR INR RUB COP TWD PEN HUF THB KRW CNY MYR CZK RON MXN PLN PHP ZAR ARS TRY 3mo Return vs USD (%) ARS EGP CNY RUB RON INR IDR COP HUF CLP BRL MXN ZAR MYR KRW CZK PHP THB PLN TWD PEN Source: FactSet, Global Investment Research 3 December 27 22

23 Equities Performance Equities Performance (%) EPS (NTM %ch) Local USD P/E P/B D/Y Local USD wgt w 3m ytd w 3m ytd NTM LTM NTM m 3m 2m m 3m 2m Czech Rep.2% Peru.4% % Colombia.4% Thailand 2.2% % % % % % % % % % % Taiwan.4% Korea 5.6% ### 58.7.% MSCI EM Thailand Current (Diamonds) Thailand PE Ratios th-th percentile (Bars) Czech Rep Peru Taiwan Colombia Combined Valuation Metrics Korea Utilities -2-3 Materials EPS (USD) 3mo % Chg mo Performance (%) MSCI Sector Performance and Earnings Telecom Indu Finls Con Stap 3mo Change in EPS (Local Currency) EM Energy Con Disc Info Tech Health Peru Korea Average Z-Score of PBK, PE and Div Yld Korea Taiwan Thailand Peru Colombia Czech Rep Source: FactSet, I/B/E/S, Global Investment Research 3 December 27 23

24 Performance of Select EMBI Countries Returns (%) Spread Returns (%) Spread Wgt wk 3m YTD YTD Level %ile/hist wk 3m YTD Wgt YTD Level %ile/hist 3.9% % Europe % % Ecuador.9% % Mid. East % % Ukraine 2.6% % Africa % % Kazakhstan 2.7% % Latam % % 3.7% % Asia % Colombia % 2.9% % A % % 5.2% BBB % % Croatia.7% BB % % Dominican Rep. 2.4% B % % 2.7% % 2.7% % IG % % Argentina 3.4% HY % % Peru 3.% % 2.7% % EMBI Composites Uruguay 2.3% % Plus % Panama 2.9% Global % 2.8% % Global Div % Sri Lanka % 3.4% % 3.4% % 4.3% % 4.3% % Lebanon 2.7% % Egypt % Venezuela.5% Green = Cheap, Red = Expensive relative to history Current EMBI Spread by Country Percentile Rank vs. History (RHS) Venezuela Lebanon Ecuador Ukraine Egypt Argentina Dominican Rep. Sri Lanka Kazakhstan Colombia Uruguay Peru Panama Croatia Spread (bps) AAA AA+ AA Spreads vs Sovereign Ratings A+ A CNY CLP A- BBB+ TRY ZAR DOP MXN BRL KZT COP RUB IDR PEN UYU MYR PAB HUF HRK PHP PLN LKR ECU ARS LBP UAH EGP Regional Relative Performance vs EMBI Asia Latam Ratings Relative Performance vs EMBI B Africa 3 Europe BB Mid East 97 BBB A 88 Jan-7 May-7 Jul-7 Nov-7 Jan-7 May-7 un-7 Jul-7 BBB Nov-7 BB+ BB AA- BBB- BB- B+ B B- CCC+ CCC Source: Datastream, Global Investment Research 3 December 27 24

25 Performance of Select GBI-EM Countries Returns (%) 5y Nominal Yield 5y Real Yield Slope (5y-y) wgt wk 3m YTD Curr % % vs hist Curr % vs hist Curr 7.8% % % % % Czech Rep % % % Peru 2.8% %.8 22% % % % Thailand 8.% %.7 58% % % % % % % % % % 3.3 % Colombia 6.9% % 2.5 9% %.8 % % % % %.5 58% MENA 7.8% % 3.8 Europe 35.2% %.9 43% Green = Cheap Asia 23.7% % Red = Expensive Latin America 33.4% % % (as %-ile vs history) GBI-EM % % 5y Yield (%) TRY Rates Level vs Slope BRL ZAR 8 MXN RUB INR IDR 6 COP PEN PHP CLP 4 MYR CNY 2 THB PLN CZK HUF Slope (5y-y, bps) Local Risk Premium Local Rates less FX Debt Rate (5y) Colombia Peru Nov-6 Regional Relative Performance vs GBI-EM Jan-7 Colombia Africa Real Rates Asia May-7 Peru 5y-2mo fwd infl exp (%) Jul-7 LATAM Europe Thailand Korea Czech Rep Nov-7 Source: Datastream, FactSet, Global Investment Research 3 December 27 25

26 Emerging Market Yield Curves m Current 3m 6m y 2y 5y 7y 9y Czech Republic m 6m y 2y 5y 7y 9y y 2y 5y 7y 9y Colombia 3m 6m y 2y 5y 7y 9y m 3m 6m y 2y 5y 7y 9y Korea 6m y 2y 5y 7y y m 3m 6m y 2y 5y 7y 9y Taiwan m 3m 6m y 2y 5y 7y 9y m 6m y 2y 5y 7y 9y 3m 6m y 2y 5y 7y 9y m 3m 6m y 2y 5y 7y 9y Thailand 3m 6m y 2y 5y 7y 9y 6. 3m 6m y 2y 5y 7y 9y m 3m 6m y 2y 5y 7y 9y m 6m y 2y 5y 7y 9y m 6m y 2y 5y 7y 9y m 3m 6m y 2y 5y 7y 9y Peru 3m 6m y 2y 5y 7y 9y 3m 6m y 2y 5y 7y 9y EM 3m 6m y 2y 5y 7y 9y Source: Datastream, Global Investment Research 3 December 27 26