《策略焦点：检验股票估值(第三部分)：席勒市盈率的压力测试1105.ppt》由会员分享，可在线阅读，更多相关《策略焦点：检验股票估值(第三部分)：席勒市盈率的压力测试1105.ppt（13页珍藏版）》请在三一办公上搜索。
1、,高盛国际,高盛国际,2012 年 11 月 3 日欧洲策略焦点证券研究报告检验股票估值（第三部分）：席勒市盈率的压力测试目前欧洲股票市盈率的 10 年均值较历史水平有很大的折让。如果将历史作为引导，那么这一折让意味着股票投资者未来的实际回报将较高。然而，投资者往往强调未来增长放缓和结构性下滑行业的权重是估值倍数较低的原因。我们发现，即使对未来增长作出较为保守的假设，股票投资者的实际回报前景依然具有吸引力。目前股市较长期均值仍具有折让，即使在剔除价格最低的行业或国家后也是如此。,目前的低估值隐含未来的高回报基于正常化盈利的估值显示目前欧洲股市的潜在回报具有吸引力。衡量方法之一便是将当前股价用市
2、盈率的 10 年均值相除，且两者皆为实际值（往往被称为格拉汉姆-多德方法或席勒市盈率）。目前欧洲股市的席勒市盈率为 12.1 倍，相对历史均值 18.2 倍存在 34%的折让。从历史的角度看，低估值往往意味着未来的高实际回报。然而，许多投资者认为未来经济增速和盈利增长的放缓是估值较低的主,Matthieu Walterspiler+44(20)7552-3403 彼得欧品海默+44(20)7552-5782 高盛国际,要原因，而不是市场真正具有吸引力。Sharon Bell,CFA,未来增长或更加缓慢，但估值仍显示股票投资者的回报可能具有吸引力盈利增长的确可能低于历史水平。欧洲公司的利润率已与
3、美国同业趋近，不太可能再度上升。然而，即使对未来盈利增长作出较为保守的假设，我们发现股票投资者依然预计可获得年均 6.2%的实际投资回报。估值水平低的情况不限于几个板块或国家虽然金融、公用事业和电信板块正在压低整个欧洲股市的估值水平，但若剔除这些板块，市场的估值水平将仍较历史均值存在折让（只是幅度缩小）。此外，根据行业分类基准（ICB）划分的 19 个超行业中有 16 个的经周期因素调整后市盈率较历史均值存在折让。类似的，尽管欧元区边缘国家部分导致了目前的低估值水平，欧洲核心国家和非欧元市场也较长期均值存在折让。,+44(20)7552-1341 高盛国际Gerald Moser+44(20)
4、7774-5725 高盛国际Christian Mueller-Glissmann,CFA+44(20)7774-1714 christian.mueller-Anders Nielsen+44(20)7552-3000,高盛国际高盛与其研究报告所分析的企业存在业务关系，并且继续寻求发展这些关系。因此，投资者应当考虑到本公司可能存在可能影响本报告客观性的利益冲突，不应视本报告为作出投资决策的唯一因素。有关分析师的申明和其他重要信息，见信息披露附录，或参阅 由非美国附属公司聘用的分析师不是美国 FINRA 的注册/合格研究分析师。,高盛集团,高盛全球经济、商品和策略研究,2,2012 年 11
5、月 3 日,欧洲,Testing equity valuations part 3:Shiller PEThis report is the third and last in our series taking a critical look at how to use valuationin the aftermath of the financial crisis.Strategy Matters:Testing equity valuations part 1:Isthe low bond yield deceiving us?(October 5,2012)analysed how
6、to value equity marketsin the current environment of extremely low yields.Testing equity valuations part 2:Equities vs.bonds(October 16,2012)looked at the impact of the relative yield on equity vs.bond performance in different bond yield environments.In this report,we look at thecurrent low cyclical
7、ly adjusted PE and whether it can be explained by lower future growthor by specific sectors and countries.As we have highlighted in previous research(see:GOAL-Global Strategy Paper:No.4-TheLong Good Buy;the Case for Equities,March 21,2012)the European market looks attractiverelative to its own histo
8、ry on a PE adjusted for the cyclical variations in earnings.One of thesemeasures is computed by dividing the current price by the 10-year average EPS(both of them inreal terms);it is often called the Graham-Dodd or Shiller PE(see Irrational Exuberance,Robert JShiller Princeton University Press 2000)
9、.On this measure of valuation the European market istrading 34%below the long-run average(Exhibit 1).The US market Shiller PE,by contrast,is at21.5x compared with an average over the same time horizon(since 1983)of 22.8x.Exhibit 1:The European market cyclically adjusted PE is significantly below the
10、 long-runaverage40,35,Shiller PE,Average,302520151051983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011Source:Datastream,Goldman Sachs Global ECS Research.In our view,valuation is the most important factor for predicting future long-term returns(seeGlobal Economics Paper 182 F
11、inding Fair Value in Global Equities:Part IIForecasting Returns,高盛全球经济、商品和策略研究,-15,3,2012 年 11 月 3 日,欧洲March 23,2009).Exhibit 2 shows the PE based on 10-year average earnings for the Europeanmarket and the five-year forward total return(in real terms)from an investment made at that time.A high Shill
12、er PE has generally led to negative real returns,while a low valuation has beenrelated to high real returns.Exhibit 2:Low valuation has historically been related to a high forward real returnShiller PE and annualized real total return for the European market,3025,5-year annualisedreal return,2015105
13、0-5-10PE based on 10-year average earnings,5,10,15,20,25,30,35,40,Source:Datastream,Goldman Sachs Global ECS Research.However,as uncertainties surrounding future growth are high,we often find in our conversationswith investors that many doubt the reliability of the signal currently provided by the S
14、hiller PE.The PE computed on trend earnings yields the same valuation signalThe underlying argument for taking the 10-year average EPS is that the cyclical variation inearnings is removed and thus the current price is being compared to the underlying earningspotential.Of course this method is not wi
15、thout drawbacks.First,since EPS tends to trend higherover time,the average of each cycle is above the average of the previous one.Therefore,bycomparing the current price to the trailing average EPS,one might overstate the PE onnormalized earnings by only correcting for cyclical variations and not ta
16、king into account long-term trend growth.Exhibit 3 highlights this point:it shows the evolution of the European marketreal earnings over time and successive 10-year averages.Each 10-year average is above theprevious one.高盛全球经济、商品和策略研究,4,2012 年 11 月 3 日,Exhibit 3:Evolution of Europe real EPS and 10-y
17、ear averages,欧洲,900,Log-scale,Real EPS(indexed 100),Averages,44%above54%above37%above90,1973,1977,1981,1985,1989,1993,1997,2001,2005,2009,Source:Datastream,Goldman Sachs Global ECS Research.As a result,it could be argued that the normalized real EPS derived from averaging thepast 10 years is too low
18、,given that it fails to take into account normal trend growth.However,when looking for over/under valuation signals,we tend to look at the current Shiller PErelative to the average rather than the absolute level of the Shiller PE.Therefore,as long as thetrend growth is unchanged,the valuation signal
19、 given by the Shiller PE is valid.One way to showthis is to compare the Shiller PE with an alternative measure of cyclically adjusted PE.We usethe PE based on trend earnings.Although the level of these two alternative measures isvery different(the average Shiller PE is 18.2x while the average PE on
20、trend earnings is15.2x),the valuation signal is comparable.Exhibit 4 compares the premium/discount totheir respective averages of the Shiller PE and the PE on trend earnings.Exhibit 4:Shiller PE and PE on trend earnings point to the same valuation signal120%100%80%60%40%20%0%-20%,-40%-60%,Shiller PE
21、 Premium/Discount vs.AverageTrend PE Premium/Discount vs.Average,-80%1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011Source:Datastream,Goldman Sachs Global ECS Research.高盛全球经济、商品和策略研究,1.,2.,3.,5,2012 年 11 月 3 日,欧洲,Future growth probably lower but not enough to explain thedi
22、scountOne of the main problems in interpreting the valuation signal from the Shiller PE is theassessment of future growth.Even though the current multiple might be cheap relative to thehistorical average,many investors are sceptical that the rate of growth in future years will becomparable to that a
23、chieved over the last 30 years and therefore believe that the discount of thecurrent multiple to the average is more a reflection of lower future growth than of cheap equities.We find the average real earnings growth achieved by European companies since 1970shas been 3.9%pa,with a large part of that
24、 derived from higher margins.We have discussed in previous research the factors underpinning the rise in margins:(1)shifts inthe industry mix of the market towards higher-margin sectors;and(2)shifts in labour costs ascompanies have outsourced,restructured formerly nationalized industries,sought more
25、 flexiblelabour contracts and moved operations to lower-cost locations(see The margin debate:cyclicalrisks vs.structural gains,April 18,2011).Although it seems hard to envisage that the expansionin margins will continue,we are also sceptical that these gains will be reversed and margins willreturn t
26、o the long-run average,given that structural change has raised European companiesmargins to the same level as those for US companies.To look at potential earnings growth,we use our economists projection of long-term real GDPgrowth(see Global Economics Paper:208 The BRICs 10 Years On:Halfway Through
27、TheGreat Transformation,December 7,2011)in various regions to project long-term sales growth.We then test different scenarios for margins.No change in margins beyond our explicit forecast horizon of 2013(net income margin of7.3%).We find this would result in 2.7%average real earnings growth between
28、now and2050 compared with the 3.9%achieved historically.Margins decline to their previous peak in 2000(net income margin of 5.5%).Average realearnings growth in this scenario would be 2%.Margins decline to their average level since 1981(net income margin of 3.5%).Average realearnings growth in this
29、scenario would be 0.8%.We find that our base projection for top-line growth would need to expand by 11 bp per year from2013 to 2050 in order for real earnings growth to reach the historical rate of 3.9%.On thattrajectory,margins would be close to 11.6%by 2050 almost 60%higher than their 2011 level.E
30、xhibit 5 details the impact of the different scenarios on real earnings growth.高盛全球经济、商品和策略研究,5,4,6,6,2012 年 11 月 3 日,Exhibit 5:European real earnings growthYear-over-year European real earnings growth projected under different scenariosEarnings YoY Real Growth(constant margin and FX)Earnings YoY Re
31、al Growth(margin declining to 5.5%,constant FX)Earnings YoY Real Growth(margin declining to 3.5%,constant FX)Historical Average3210-1,欧洲,2014E,2018E,2022E,2026E,2030E,2034E,2038E,2042E,2046E,2050E,Source:Goldman Sachs Global ECS Research.While,even with constant margins,we would not expect earnings
32、to grow at the rate achieved inthe last 30 years(2.7%vs.3.9%),we find that market implied growth for the next 40 years issubstantially lower than our most negative scenario.Exhibit 6 shows the 40-year market impliedgrowth over time assuming a constant ERP of 3.5%and a conservative terminal real grow
33、th rateof 1%.Currently it points to an annualized real decline in earnings of 2.5%.Exhibit 6:Evolution of European earnings implied growth840-year implied growth420-2-4,Jan-90,Jan-93,Jan-96,Jan-99,Jan-02,Jan-05,Jan-08,Jan-11,Source:Datastream,Goldman Sachs Global ECS Research.高盛全球经济、商品和策略研究,7,2012 年
34、 11 月 3 日,欧洲,Still solid real return for equity holders on conservative growth assumptionsAs discussed above,although there are reasons to expect earnings growth to be lower than it hasbeen in the past 30 years,the level of real growth currently implied by the market seems too loweven compared to ou
35、r most negative scenario.Yet,one question remains:if growth is lower,what is the fair multiple value towards which the market should converge?One way toanswer this question is to use the justified PE.The justified PE uses a one-stage DDM toderive the theoretical PE of the market given expected growt
36、h,payout ratio and the cost of equity.This PE is not comparable in level terms to the Shiller PE but we can make the comparison with anormal 12-month forward PE.Exhibit 7 shows the discount compared to the justified PE under different assumptions for realearnings growth and with a 44%payout ratio an
37、d 6.2%real cost of equity(in both cases,themarket average since 1989).We find that the current 12-month forward PE(for whichearnings are still below normalized level)would be at fair value with 2.2%annualized realearnings growth.Exhibit 7:Even conservative growth assumptions would leave equity holde
38、rs earning a 6.2%real cost of equityJustified PE under different earnings growth scenario,GrowthScenarioHistoricalrealearningsgrowthrateConstantmarginsandGDPgrowthprojectionsfortoplineMarginsdeclineto5.5%andGDPgrowthprojectionsfortoplineGrowthrequiredforPEtobefairgivenrealcostfoequityMarginsdeclinet
39、o3.5%andGDPgrowthprojectionsfortopline,Payoutratio(%)4444444444,RealCostofEquity(%)6.26.26.26.26.2,RealGrowth(%)3.92.72.22.00.8,JustifiedPE*18.712.411.010.38.0,Current12monthforwardPE11.011.011.011.011.0,Premium/Discount(%)41.011.50.06.837.0,*JustifiedPEisderivedfromaonestageDDM.Justifiedpe=payoutra
40、tio/(CostofequityGrowth)Source:Goldman Sachs Global ECS Research.The reassuring conclusion we came to is that,even on a conservative assumption of 2.2%real earnings growth(which would make the current multiple fair relative to a one-stageDDM derived PE),an equity investor would expect to earn the co
41、st of equity(6.2%in realterms).高盛全球经济、商品和策略研究,35,25,20,8,2012 年 11 月 3 日,欧洲,Low Shiller PE:not just driven by a few sectors or countriesAnother argument we hear when discussing the attractive valuation of the European market isthat the current valuation is driven by a few sectors facing structural d
42、ifficulties and that,excludingthem,the market is actually on a much higher multiple.The sectors most often cited areFinancials,Telecoms and Utilities,which account for almost 30%of the STOXX Europe 600weight.However,we find that,even excluding these three sectors,the market remains on anattractive c
43、yclically adjusted PE.Exhibit 8 shows the Shiller PE for the European marketexcluding Financials,Utilities and Telecoms.Exhibit 8:Market remains attractive even excluding Financials,Utilities and Telecoms40Shiller PE ex Fin,Utilities andTelecomsShiller PE30AverageAverage ex Fin,Utilities andTelecoms
44、1510583 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11Source:Datastream,Goldman Sachs Global ECS Research.Although excluding these sectors raises the overall valuation for the remainder of the market(theShiller PE moves from 12.1x to 15.5x),as they are generally
45、on lower multiples than the broadermarket,the market excluding these sectors is still on a 15%discount relative to its own history(including these sectors,the European market is on a 34%discount to the average since 1983).In addition,there are some difficulties in pointing to sectors as facing struc
46、tural issues.At the endof the 1990s few investors would have considered Telecoms to be facing structural issues(although stretched valuation would prove a structural issue for the sector investment returns),while Basic Resources which in the next decade would benefit from the rapid development ofBRI
47、Cs countries was one of the few sectors trading at a discount to its average Shiller PE.Another way to assess how broad-based inexpensive valuation is across the market is to look atindividual sector multiples rather than the aggregate multiple.Exhibit 9 shows over time theproportion of sectors for
48、which the Shiller PE is below the average Shiller PE.We find that 16 ofthe 19 ICB super sectors(84%)are currently trading below their long-run average cyclicallyadjusted PE.This proportion is large compared to the longer history(the long-run average isaround 50%).高盛全球经济、商品和策略研究,100,9,2012 年 11 月 3 日
49、,Exhibit 9:84%of European sectors are trading below their average Shiller PEProportion of sector with Shiller PE average Shiller PE9080706050403020100,欧洲,Jan-90,Jan-93,Jan-96,Jan-99,Jan-02,Jan-05,Jan-08,Jan-11,Source:Datastream,Goldman Sachs Global ECS Research.Exhibit 10 provides the detail for eac
50、h of the 19 super sectors.As discussed above,only threesectors are trading above their average Shiller PE:Chemicals,Food&Beverages,and PersonalHousehold Goods.Exhibit 10:Only three sectors trade above their average Shiller PE,SectorChemicalsFood&BevPersonalandHHGoodsIndustrialGds&SvsAutos&PartsHealt
