新兴
市场
股票
策略
观点
衡量
涨势
五个
问题
20190304
26
How much growth has been(p)repriced?Our Growth Factors suggest EM nmarkets are pri ci ng 50bp of growth accelerati on,whi ch i s a reversal of 1/3 of the decelerati on from last year.We thi nk wi ll be hard to deli ver and suspect growth wi ll li kely“firm”rather than re-accelerate qui ckly.We see modest further upsi de to EM equi ti es(MSCI EM to 1100),and ri sks towards more credi t spread compressi on than our i ni ti al target(EMBI 360bp at end-2019),and modest FX upsi de(medi an 1%spot appreci ati on from current).2019 has been a year of EM“country stories”.The rally has been broad-based nacross asset classes,but we find that country coherence has been hi gh thi s year.That i s,several EMs have outperformed across all asset classes whereas others have underperformed uni formly.I n short,we remai n focused on“pi cki ng the ri ght countri es”rather than broad beta exposure i n EM.Backtesting Val uation:a pro-cycl ical signal.When consi deri ng valuati on nsi gnals usi ng si mple rati os relati ve to hi story(as opposed to deeper fai r value models),we find that i nexpensi ve EMs outperform over ti me apart from drawdowns.EMs that screen as expensi ve on these metri cs tend to be defensi ve(and li kely trade at a premi um because of thei r defensi ve nature);gi ven our constructi ve EM vi ew,we prefer value EMs.Currentl y,on a cross-asset basis,Mexico,Turkey,South Africa,and Russia nscreen as the l east expensive EMs.Brazi l screens as the most expensi ve EM on a cross-asset basi s.I n fixed i ncome,we retai n a top trade recommendati on to recei ve MXN,ZAR,and COP rates vs.low yi elders(whi ch all screen as expensi ve EMs on a relati ve basi s).The hi gh yi elders each have poli ti cal and i di osyncrati c ri sks associ ated wi th them,but our back-test suggests the valuati on di scount warrants a portfoli o-approach and they should outperform as a cluster.Equity Exposure:Focus on Asia,l ess on Brazil.We retai n our vi ew that Brazi l nwi ll underperform EM peers tacti cally unti l reform progress i s more concrete;our Asi a strategi sts remai n posi ti ve on Chi na.Gi ven our vi ew that global growth(and especi ally US)i s li kely to stabi li ze,we find that Tai wan and Korea trade at a di scount relati ve to thei r global betas and see tacti cal catch-up room i n these two global cycli cal markets.Caesar Maasry+1(212)902-8763| Goldman Sachs&Co.LLC Ron Gray+1(212)357-6762| Goldman Sachs&Co.LLCEM Strategy Views Five Q&As on Gauging t he EM Rally4 March 2019|8:44AM ESTI nvestors should consi der thi s report as only a si ngle factor i n maki ng thei r i nvestment deci si on.For Reg AC certi ficati on and other i mportant di sclosures,see the Di sclosure Appendi x,or go to ve Q&As on Gaugi ng the EM Rally (1)Wai ti ng for Growth:how much has been(p)repri ced?Conti nued momentum i n EM assets(equi ty and credi t more than local)and posi ti ve headli nes regardi ng Chi na trade tensi ons have furthered ri si ng opti mi sm for EM exposure i n Q1 2019.Whi le much of posi ti vi ty around EM has focused on a dovi sh turn i n the Fed,we note that US front-end rates(2-year)have barely moved i n February and argue that the recent moves i n EM have been more growth-dri ven(or,rather,dri ven by expectati ons of growth).We are finally begi nni ng to see some si gns of stabi li zati on i n Chi na i n addi ti on to the bottomi ng i n EM ex-Chi na that has been vi si ble si nce September.To be clear these are pretty modest,especi ally relati ve to the run-up i n markets,but i t i s reassuri ng that these are becomi ng more apparent and wi ll need to strengthen for the EM rally not to falter.Below,we refresh our Growth Factors across the US and EM markets and find that the US has re-pri ced about 80bp of growth si nce the Q4 lows and EM roughly 50bp(see Exhi bi t 1).Whi le we do see a turn i n global growth ahead wi th the US the strongest candi date,we do not expect the US to mai ntai n a growth rate near 3%over the course of 2019.For EM,a 50bp accelerati on would reverse roughly 1/3 of the slowi ng of last year,and we are skepti cal the market wi ll conti nue pri ci ng a further re-accelerati on unti l the data are more supporti ve.(2)“Country Af finity”-measuring country coherence across assets Gi ven the strong repri ci ng of growth ahead of the data actually i mprovi ng,i t follows that we would recommend i nvestors become selecti ve i n EM exposure after the broad market rally.The“Fed repri eve”of Q4 and our beari sh vi ews on the USD have dri ven our preference at the asset class level for EM local bonds si nce November,but we hi ghli ght Exhibit 1:Growth pricing in the US and EM is more optimistic than data woul d impl y 0.0%0.5%1.0%1.5%2.0%2.5%3.0%3.5%4.0%4.5%5.0%5.5%6.0%7580859095100105110Dec-14Mar-15Jun-15Sep-15Dec-15Mar-16Jun-16Sep-16Dec-16Mar-17Jun-17Sep-17Dec-17Mar-18Jun-18Sep-18Dec-18Mar-19USA US Activity(CAI,RHS)US Growth Factor Growth Factors 1.cyc vs.def equities 2.local rate curve(2y vs 5yr)3.Trade-weigted FX 1.52.02.53.03.54.04.55.05.56.06.57.06065707580859095100105110Mar-10Aug-10Jan-11Jun-11Nov-11Apr-12Sep-12Feb-13Jul-13Dec-13May-14Oct-14Mar-15Aug-15Jan-16Jun-16Nov-16Apr-17Sep-17Feb-18Jul-18Dec-18May-19EM Activity(CAI,equal-weighted)EM EM Growth Factor Growth Factors 1.cyc vs.def equities 2.local rate curve(2y vs 10yr)3.Trade-weigted FX Source:Fact Set,Dat ast ream,Goldman Sachs Global Invest ment Research4 March 2019 2Goldman SachsEM Strategy Vi ewsbelow that EM has been tradi ng“country stori es”,and suggest i nvestors consi der i ndi vi dual opportuni ti es goi ng forward.As shown i n the tables below,both January and February were months of strong“country coherence”;that i s,certai n EMs outperformed across all asset classes,whi le others underperformed across all asset classes.I n January,Brazi l,Argenti na,Turkey,Colombi a,Russi a,and South Afri ca(apart from equi ti es)beat thei r peers across FX,credi t,local bonds,and equi ti es whi le I ndi a,Tai wan,Malaysi a,and Poland lagged across all asset classes.I n February,markets reversed wi th Brazi l,Argenti na,and Turkey underperformi ng across all asset classes and Colombi a,Malaysi a,Peru,and Chi na outperformi ng.I n other words,duri ng each month,i nvestors needed to only pi ck the“ri ght country story”to outperform regardless thei r speci fic asset class.Of course,we are not suggesti ng i t i s easy to pi ck the ri ght country each month,but rather,we hi ghli ght that soverei gn ri sk i s often the pri mary dri ver for a gi ven countrys relati ve performance vs.peers rather than the techni cal backdrop of the i ndi vi dual country-asset.Looki ng more structurally at country coherence,we run the average pai rwi se correlati ons of relati ve returns for each asset class wi thi n each EM(e.g.the correlati on of Brazi ls relati ve equi ty return vs.EM peers wi th BRL s relati ve return vs.EM peers)to formulate a basi c i ndi cator of how coherent EM asset markets are behavi ng(see Exhi bi t 3,left-hand si de).I n short,thi s metri c has ri sen over ti me and has always been reasonably posi ti ve over the past 20 years(avg.0.30).Anecdotally,we find that country coherence i s often much hi gher than most i nvestors appreci ate,parti cularly those focused on a speci fic asset class.Often,a gi ven country-assets returns are associ ated wi th speci fic techni cals or flows,and the broader country story goes unnoti ced i n our vi ew.Exhibit 2:“Country stories”general l y drove the outperf ormance/underperf ormance in assets in January and February Performance in JanuaryPerformance in FebruaryFXSov.Loc BondsEquityRanked ReturnsFXSov.Loc BondsEquityRanked Returns(vs.USD)Credit(USD Ret)(local FX)FXCred BndsEQAVG.(vs.USD)Credit(USD Ret)(local FX)FXCred BndsEQAVG.Brazil6.0%3.8%8.3%10.6%26544.3Colombia1.1%0.1%1.7%5.7%113114.0Argentina1.7%10.3%10.4%19.7%121314.3Malaysia0.7%0.6%1.3%1.7%28354.5Turkey2.7%3.5%10.7%14.8%97225.0Peru0.6%0.7%1.0%2.2%37544.8South Africa7.8%5.2%11.7%3.5%121175.3Chile0.3%1.1%0.7%(0.8%)416146.3Colombia4.3%4.3%5.6%8.5%54865.8China0.3%0.4%0.2%4.2%610836.8Russia5.6%3.2%9.0%7.7%49476.0Philippines0.3%0.8%1.7%(1.4%)542166.8Chile5.9%2.6%6.6%5.4%3116138.3India(0.1%)0.7%(0.0%)(0.0%)759118.0Thailand4.0%4.4%4.1%5.2%6310148.3Taiwan(0.2%)(0.3%)(0.3%)4.9%8161029.0Mexico3.4%3.1%5.7%6.1%7107118.8Hungary(0.8%)0.7%(0.5%)0.2%1461210 10.5Indonesia2.8%4.2%4.0%5.5%8511129.0Mexico(1.1%)0.9%0.4%(2.1%)153719 11.0China2.3%1.2%3.3%11.2%101612310.3Czech Rep.(0.4%)(0.3%)(1.3%)1.5%91515611.3Peru1.5%3.2%3.0%7.7%13813810.5Indonesia(0.4%)(0.0%)1.2%(2.0%)1014418 11.5Philippines0.9%2.5%4.9%7.6%15129911.3Korea(0.6%)(0.5%)(0.4%)0.8%121811812.3Hungary1.8%1.6%2.7%3.7%11141416 13.8South Africa(5.2%)1.0%(5.1%)0.5%20220912.8Korea0.3%0.4%(0.3%)9.9%181919515.3Thailand(0.5%)(0.6%)(0.8%)1.2%112013712.8Poland1.1%1.1%1.7%4.1%14171515 15.3Russia(0.6%)0.2%(0.9%)(1.2%)13111415 13.3Czech Rep.0.3%0.4%1.3%6.2%17181710 15.5Brazil(2.6%)0.2%(2.2%)(0.7%)17121613 14.5Malaysia0.9%2.3%1.6%0.4%16131619 16.0Poland(2.1%)0.6%(2.3%)(1.7%)1691717 14.8Taiwan0.1%0.0%0.0%1.7%19201818 18.8Turkey(2.7%)(0.5%)(3.5%)(0.7%)18191812 16.8India(1.9%)1.5%(1.6%)(0.2%)20152020 18.8Argentina(4.6%)(0.4%)(4.0%)(8.6%)19171920 18.8EM2.6%4.4%4.6%7.1%EM(0.9%)1.0%(0.7%)1.8%Source:Dat ast ream,Goldman Sachs Global Invest ment Research4 March 2019 3Goldman SachsEM Strategy Vi ewsI nteresti ngly,country coherence decli ned duri ng the peak-QE years,perhaps as FX returns were moti vated by carry more than growth(whi ch presumably were larger dri vers of equi ty and credi t markets).But over the past few years,country coherence as we measure i t below has returned closer to hi stori cal average levels.Separately,we find that peri ods of EM market drawdowns are associ ated wi th sharp spi kes i n country coherence;we suspect thi s i s due to EM soverei gn vulnerabi li ti es bei ng exposed and dri vi ng certai n markets to underperform si gni ficantly(i.e.,taper tantrum i n 2013,commodi ty exposure i n 2014-15).Thi s was also the case bri efly last year duri ng the sharp October sell-off.(3)How to pick countries-is Val ue the right approach?Gi ven our vi ew that 2019 has been a year of“country stori es”,we suggest i nvestors consi der a broad valuati on si gnal when thi nki ng about speci fic EM country exposures.One method to aggregate the vari ous market data across EMs i s to run z-scores of valuati on metri cs relati ve to a long-term hi stori cal range and to rank the markets.Below,we hi ghli ght our methodology of doi ng so,usi ng equi ty market rati os,FX relati ve to our GS-DEER and GS-FEER models,soverei gn credi t spreads,and real 10-year local bond yi elds(see Exhi bi t 4).From thi s perspecti ve,a number of hi gh-yi elders appear to be the most i nexpensi ve EMs(Mexi co,Turkey,South Afri ca,Russi a,and I ndi a),wi th Brazi l and some low-yi elders appeari ng expensi ve vs.hi story(Hungary,Chi na,Poland,Chi le,and Korea).Exhibit 3:Country coherence tends to be more pronounced during periods of EM pressure(0.30)(0.20)(0.10)0.000.100.200.300.400.500.600.7019891992199519982001200420072010201320162019Country correlation across assets 4006008001000120014000.000.100.200.300.400.500.602010201120122013201420152016201720182019MSCI EM(RHS,Inverted)Source:Fact Set,Dat ast ream,Goldman Sachs Global Invest ment Research4 March 2019 4Goldman SachsEM Strategy Vi ews We back-test thi s cross-asset valuati on si gnal to gi ve a sense of when thi s methodology may be helpful,and find that there i s a strong si gnal over the medi um term that i nexpensi ve EMs tend to outperform expensi ve markets(see Exhi bi t 5,back-test i s based on monthly valuati on readi ngs,wi th no look-ahead bi as).However,di ggi ng a bi t deeper,we find somewhat counter-i ntui ti ve results,i n that thi s method of valuati on actually works best duri ng ri sk-on peri ods although we di d see moderate outperformance duri ng the EM bear market from 2013-2016.Perhaps thi s i s due to vulnerable EMs often screeni ng as the least expensi ve markets(whereas defensi ve EMs screen expensi vely)and broad ri sk-on peri ods tend to compress such premi a.Exhibit 4:Mexico,Turkey,South Af rica and India appear most attractive on a country story val ue basis BRLHUFCNYPLNCLPKRWPHPPENTWDTHBCOPIDRMYRCZKINRRUBZARTRYMXNEMEquityPE(NTM)12.8x9.8x10.9x11.8x15.6x9.3x16.8x14.4x13.4x14.4x10.4x15.2x16.0 x13.9x17.6x5.7x13.5x6.8x13.5x11.4xFXvs GS FV(1%)(3%)(1%)(6%)(2%)+10%+4%(2%)+1%+6%(5%)+7%(2%)+4%+2%(3%)(10%)(41%)(8%)(3%)CreditSpread161bp87bp52bp67bp46bp-64bp73bp-113bp109bp71bp-128bp134bp183bp316bp132bp351bpBondsReal 10yr5.1%-0.1%1.0%0.7%1.2%0.3%2.6%3.5%-1.5%3.4%4.7%2.3%-0.2%4.1%3.8%4.8%-2.9%4.9%2.3%EquityZ-Score0.770.13(0.12)0.15(0.61)(0.91)0.300.13(0.52)0.51(0.61)0.20(0.17)(0.22)0.01(0.90)(0.12)(1.63)(1.33)(0.24)FXZ-Score(0.11)(1.17)0.75(1.27)0.140.760.750.290.820.58(0.53)0.230.43(0.67)0.28(0.23)(0.15)(2.35)(0.31)(0.09)CreditZ-Score0.381.071.110.651.06-0.950.99-0.600.710.91-0.830.660.26(1.04)0.060.13BondsZ-Score1.041.67(0.10)2.080.861.01(0.87)(0.35)-(0.88)0.56(1.71)(1.78)0.41(1.90)(1.15)(1.71)2.92(2.18)0.22AVGZ-Score0.520.420.410.400.360.290.280.270.150.070.00(0.14)(0.15)(0.16)(0.20)(0.41)(0.43)(0.53)(0.94)0.00(1.00)(0.80)(0.60)(0.40)(0.20)0.000.200.400.60Average Valuation Z-Scores across Equity,FX,Credit,and local Bonds+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 Source:Fact Set,Dat ast ream,Goldman Sachs Global Invest ment Research4 March 2019 5Goldman SachsEM Strategy Vi ews(4)We l ike“Val ue”in Fixed Income,and adopt a portf ol io approach given the idiosyncratic risks Looki ng at the recent hi story of the“i nexpensi ve”EMs li sted above,we find qui te a pro-cycli cal performance profile over the course of the past year.These markets(Mexi co,South Afri ca,Turkey,I ndi a,and Russi a)outperformed i n early 2018 across fixed i ncome and equi ty,but si gni ficantly underperformed duri ng the bear market last year.Si nce October,however,they have outperformed across the board(I ndi a less so year-to-date).More recently,over the past three weeks,the equi ti es i n these markets have underperformed whereas the local bonds are sti ll towards thei r relati ve hi ghs(see Exhi bi t 6).I n part thi s i s due to the strong performance of Chi na(whi ch i s heavi er i n the equi ty benchmark)and we recommend looki ng more at Asi an markets tacti cally i n equi ty below.But for value-i nvestors seeki ng a bi t more upsi de i n EM,we remai ned focused on the hi gh yi elders i n fixed i ncome space;we retai n a top trade recommendati on i n recei vi ng MXN,ZAR,and COP rates over low yi elders(PLN,HUF,KRW,and THB,whi ch screen as expensi ve EMs i n our cross-asset table above apart from THB).The hi gh-yi elders each carry si gni ficant i di osyncrati c ri sks(electi ons i n ZAR and I NR,a new admi ni strati on i n MXN,and geo-poli ti cal headli nes li kely to conti nue regardi ng TRY,RUB,and I