高盛高華全球經(jīng)濟(jì)周評(píng)1115

《高盛高華全球經(jīng)濟(jì)周評(píng)1115》由會(huì)員分享，可在線閱讀，更多相關(guān)《高盛高華全球經(jīng)濟(jì)周評(píng)1115（12頁(yè)珍藏版）》請(qǐng)?jiān)谘b配圖網(wǎng)上搜索。

1、高盛國(guó)際高盛國(guó)際高盛集團(tuán)INDBRAINDBRABRA2012 年 11 月 15 日Issue No: 12/38全球經(jīng)濟(jì)周評(píng)研究報(bào)告全球制造業(yè)格局改變?nèi)蛑圃鞓I(yè)活動(dòng)的改變導(dǎo)致全球制造業(yè)活動(dòng)在過(guò)去幾年里出現(xiàn)了明顯改變。相對(duì)來(lái)說(shuō)，在這些變化中有的已廣為人知，比如制造業(yè)在 GDP 中的占比下滑，以及全球制造業(yè)活動(dòng)朝向新興經(jīng)濟(jì)體轉(zhuǎn)移等。但是，隨著發(fā)達(dá)經(jīng)濟(jì)體迎接經(jīng)濟(jì)增長(zhǎng)停滯和高失業(yè)率的挑戰(zhàn)，以及新興經(jīng)濟(jì)體尋求更平衡且具可持續(xù)性的發(fā)展道路，重新思考發(fā)展策略的趨勢(shì)越發(fā)明顯。Kamakshya Trivedi+44(20)7051-4005 Jose Ursua(212) 357-2234 高盛集團(tuán)針對(duì)發(fā)

2、展策略的再評(píng)估針對(duì)發(fā)展策略的再評(píng)估我們?cè)噲D闡明與這一再評(píng)估過(guò)程有關(guān)的問(wèn)題。具體而言，我們研究了制造業(yè)附加值的再平衡、探討了危機(jī)以來(lái)制造業(yè)競(jìng)爭(zhēng)力層面（包括勞動(dòng)力成本、生產(chǎn)率以及George Cole+44(20)7552-3779 發(fā)展支持程度）的改變，并研究了一些分析制造業(yè)在長(zhǎng)期發(fā)展及趨同過(guò)程中所扮演角色的最新學(xué)術(shù)報(bào)告。Julian Richers制造業(yè)再平衡勢(shì)頭的巨大潛力我們?cè)诖岁U述的制造業(yè)轉(zhuǎn)變可能會(huì)在相對(duì)較長(zhǎng)的時(shí)間里推進(jìn)，因此對(duì)投資的影響可能更為間接且更具中期屬性。除了制造業(yè)或外商直接投資相關(guān)資金流對(duì)個(gè)股投資造成的影響之外，在這一領(lǐng)域再平衡勢(shì)頭所蘊(yùn)含的潛力將對(duì)整體發(fā)展道路、以及對(duì)匯市和債市

3、等其他資產(chǎn)市場(chǎng)帶來(lái)顯著影響。在過(guò)去十年中，新興經(jīng)濟(jì)體在全球制造業(yè)中的占比飆升制造業(yè)附加值（按 2000 年可比美元匯率計(jì)算；全球占比）(212) 855-0684 多米尼克威爾遜(212) 902-5924 高盛集團(tuán)50% of world40JAPJAPJAP30GERGERGER20100USA2000USA2005USA2010CHNRUS2000CHNRUS2005CHNINDRUS2010N112000N112005N112010G3BRICsN11資料來(lái)源：聯(lián)合國(guó)工業(yè)發(fā)展組織、高盛全球經(jīng)濟(jì)、商品和策略研究投資者不應(yīng)視本報(bào)告為作出投資決策的唯一因素。 有關(guān)分析師的申明和其他重要信息，

4、見(jiàn)信息披露附錄，或參閱 年 11 月 15 日Global Economics WeeklyThe changing shape of global manufacturingThe past few years have seen remarkable changes in global manufacturing activity. Some ofthose changes are relatively well known, such as the declining share of manufacturing incountries GDP and the shift in glob

5、al manufacturing activity towards EMs. But with DMscontending with the challenges of stagnation and high unemployment, and EMs attempting amore balanced and sustainable growth path, there is an increasing tendency to rethink growthstrategies. Related to this reappraisal, a new set of questions has m

6、oved to the forefront ofinvestors minds: how far has the migration of manufacturing from DMs to EMs extended? Arethere any signs that the declines in manufacturing in DMs are stabilising? How have the forcesthat determine the manufacturing competitiveness of location evolved and, in particular, how

7、havethey changed since the global financial crisis?These questions are difficult to answer definitively, not least because it is hard to find extensiveand consistent data on many of the dimensions of interest. Still, in this Weekly, we attempt toshed light on some of these issues. In particular, we

8、explore the rebalancing in manufacturingvalue added in detail, highlighting what has changed, discuss the shifts in aspects ofmanufacturing competitiveness since the crisis, and also review new academic work thatunderlines the role of manufacturing activity in long-run growth processes.Our main find

9、ings are as follows:Manufacturing has rebalanced across DMs and EMs over the past decade or so, withthe former losing weight in global value added and exports in favour of the latter(especially China). EMs have gained market share in various branches of industry,especially but not exclusively those

10、associated with low-cost products.The race for manufacturing competitiveness has involved various combinations of lowerlabour costs, higher productivity and the creation of better growth environments acrosscountries. In some EMs, manufacturing wages have begun to increase, although theyare still at

11、relatively low levels. But sustaining competition will require increasedinvestment in human capital.Looking ahead, convergence dynamics will likely support manufacturing growth in EMs,while expenditure on research and development (R&D) is likely to continue to boostinnovation and productivity in DMs

12、. These two forces partly offset each other, but notentirely.The asset pricing implications of the manufacturing rebalancing story are more subtlethan those from short-run cyclical fluctuations. But markets may reward the continuationof some medium-run processes that benefit structural winners, at t

13、he country level, insingle-stock space or in FX, for example.Manufacturing has rebalanced across DMs and EMsSince the early 1980s, manufacturing has declined as a share of global output from almost 25%to close to 16% (while the share of services has risen from 57% of GDP to 70%). In the US, forexamp

14、le, it has fallen from 22% of GDP to about 13%, and similar declines have occurred in otherDMs. Manufacturing activity in EMs is still thriving, although it also has fallen as a share ofGDPin China, for example, it has fallen from 40% to 30%. The fact that manufacturings sharehas also declined in EM

15、salthough at a slower ratereveals that non-manufacturing sectors inthese countries have also grown rapidly in recent decades. Manufacturing also continues toemploy a large number of people. Its share of total employment has halved in the US but itremains high in other DMs. In EMs, the declines have

16、been smaller and manufacturingemployment has grown strongly in China, where workers have been transferred from low-value高盛全球經(jīng)濟(jì)、商品和策略研究ApparelBasicmetalsLeather/footwearElectricalmach.TextilesNon-metalmineralOthertransport.TobaccoRubber/plasticsFurnitureChemicalsPaperMotorvechiclesMach.&equip.Food/be

17、veragesWood(ex.Furnit.)Refinedoilprdcts.FabricatedmetalsOffice/ComputersMedicalinstr.Printing/publishingTV/Comm.equip.JAPINDRUSINDRUSRUS32012 年 11 月 15 日Global Economics Weeklyactivities in the primary sector to factories in urban centres (see Asia Economics Analyst 12/16,China: Shifting to a lower

18、gear).But beyond those trends within countries, which are relatively well known, manufacturing haschanged across countries in important ways. For example, China and other EMs increased theirshare in manufacturing global value added by a large fraction during the last decade (Exhibit 1).The BRICs (ma

19、inly China) have gained almost 10ppt compared with a loss of about 8ppt in theG3 (half of it in Japan). India has doubled its share, albeit from very low levels, while the N-11group has expanded it by 1.5ppt. In terms of manufacturing exports, of the DMs only Germanysshare has gained a touch, but Ja

20、pans share fell 3ppt and the US lost almost 5ppt. On the back ofthese losses, Chinas share has almost tripled, Indias has more than doubled and the N-11s hasrisen about 1ppt. The magnitude of Chinas expansion to cover 13% of global manufacturedexports is impressive (it started the decade with a shar

21、e of 4.5% and ended it with 4.5ppt morethan the US).EMs have made gains over DMs in virtually all sub-sectors of industry, and not necessarily inareas that represent the largest shares of their own composition of manufactured goods. Exhibit 2shows a breakdown of the top 22 branches of industry and E

22、Ms gains in the global market sharefrom 2000 to 2010. The highest gains have occurred in apparel, basic metals andleather/footwear, which saw advances of more than 30ppt. Electrical machinery, tobacco,furniture, chemicals, motor vehicles and food/beverages, saw gains higher than 15ppt, and evenin mo

23、re high-end branchessuch as medical instruments and computersgains are close to10ppt.The structure of the value added by sub-sector shows that gains in global share have not alwayscoincided with the highest shares within the aggregate of EMs. For example, although the sharegains in apparel are close

24、 to 35ppt, this sector represents only about 3% of value added in thegroup. On the other hand, gains in TVs and communication equipment have been more modest,at close to 5ppt, while their share in the groups value added is close to 10%. This inverserelationship does not hold in every sector, but it

25、reflects that some of the largest share gains ofEMs correspond to relatively cheaper goods, while making inroads in higher-end products hasbeen more challengingalthough it has certainly occurred too.Exhibit 1: EMs share in global manufacturing spikedduring the last decadeValue added in manuf. (in co

26、nstant 2000 US$; % of world)Exhibit 2: As they overtook DMs in various sub-sectorsManuf. value added (Diff. in EMs % of worlds total 2010 vs.2000, and intra-EM shares in 2010)504030% of worldJAPGERGERJAPGER403530252015pptWorld share gained by EMs 2000-2010Share of branch within EMs in 2010 (RHS)%141

27、2108620104100USAUSAUSACHNBRACHNBRACHNINDBRAN11N11N115020200020052010200020052010200020052010G3BRICsN11Source: UNIDO; GS Global ECS Research.高盛全球經(jīng)濟(jì)、商品和策略研究Source: UNIDO; GS Global ECS Research.NORSWZBELDENSWEGERFINAUTNEDAUSFRAIRECANUSAITAJAPUKGSPAGRENZLISRSGP KORARGPORCZRSVKBRAHUNTWNPOLMEXPHL42012 年

28、11 月 15 日Global Economics WeeklyThe race for manufacturing competitiveness: Lower costs, higherproductivity and growth-friendlinessMarket forces related to international competition and the search for higher profit margins havedriven these manufacturing rebalancing patterns. Three key components of

29、these dynamics are (i)labour costs, (ii) labour productivity and (iii) an environment conducive to growth.The financial crisis and its aftermath witnessed noticeable movements in unit labour costs. Butlonger-run trends over the past decade reflect deeper forces stemming from labour-marketcompetitive

30、ness. The abundance of labour has given EMs an important edge over DMs. Exhibit 3compares hourly compensation costs in manufacturing in 1997 vs. 2010 in Dollar terms for alarge set of countries. EMs have the lowest labour costs, but not necessarily the lowest increasesin proportion of starting value

31、s. Given that labour costs in the US rose by close to 50%, largeincreases in EMs reflectin addition to FX appreciationa relative catching up of manufacturingwages over time. Even China and India have seen important increases in hourly earnings, buttheir levels are still only a fraction of hourly ear

32、nings in the USabout 10% in China and 5% inIndiaalthough trends clearly point upwards (Exhibit 4). Increases are likely to persist, as therelative abundance of lower-cost labour from the first cohorts of manufacturing workers isgradually depleted in those countries.As adjustments in labour costs tak

33、e place, the other two factors that define countriescompetitiveness will continue to evolve. On the one hand, labour productivity has increasedmarkedly in places such as the US (at a yearly rate of almost 5.4% since the early 2000s). This islower than the rates seen in places such as Korea, the Czec

34、h Republic and Taiwan (at 6% ormore), but higher than in many other DMs.It is reasonable to expect competitiveness in terms of countries growth environments to progresstoo. We matched labour productivity growth in OECD countries for which data are available withthe corresponding yearly changes in th

35、eir Growth Environment Scores from 1998 onwards (seeGlobal Economics Paper 211, Our 2011 GES: A Sharper Signal for Growth). The simplecorrelation pictured by a scatter plot reveals the positive association between the two (Exhibit 5).As we have stressed in our GES research, the scores pick up microe

36、conomic andmacroeconomic factors that improve (or impede) countries ability toExhibit 3: EMs compete through lower labour costsHourly compensation costs in manufacturing (US$)Exhibit 4: China and India are special cases of low butrising costsHourly earnings of manuf. workers (varying defs.; % of US)

37、70605040US$US20101997109876% USChina (urban units)China (towns and villages)India (all employees)India (prod. workers)302010054321020022003200420052006200720082010eSource: BLS; GS Global ECS Research.高盛全球經(jīng)濟(jì)、商品和策略研究Source: BLS; GS Global ECS Research.ChangeinGES(idx.pts.)crisis52012 年 11 月 15 日Exhibi

38、t 5: GES are related to labour productivity growthGlobal Economics WeeklyExhibit 6: Disparities in competitiveness around theLabour productivity gwth. & change in GES (1998-2009; OECD)Average yearly changes (2001-06 vs. 2007-10)0.6Labor costs(%)Productivity(%)GES(Idx. points)Export value(%)0.4AUSBef

39、.14.6Aft.4.7Bef.1.8Aft.1.7Bef.0.14Aft.-0.10Bef.0.0Aft.-0.20.2Non-EuroCAN 10.6JAP -3.70.59.11.56.00.51.2-0.020.34-0.55-0.38-0.3-0.4-0.9-0.20-0.2-0.4-0.6All economyManufacturingDMsEuroDMsUSAUKGFINGERFRASPAITA-2.76.83.35.48.011.612.20.2-5.52.93.60.70.53.18.55.18.24.43.70.5-0.23.50.4-2.5-3.20.02.1-2.7-0

40、.050.26-0.320.250.200.58-0.03-0.46-0.41-0.36-0.13-0.15-0.48-0.16-0.5-0.4-2.1-0.3-0.3-0.70.4-1.30.1-0.5-0.10.1-0.4-1.2-0.8-10-5 0 5Labor productivity growth (%)1015EMsCZRKORMEX10.28.8na0.0-5.6-1.410.29.4na4.54.3-0.60.440.110.53-0.080.60-0.04na0.3-0.4na-0.10.5TWN-3.2-4.49.97.90.50-0.29nanaSource: OECD

41、; GS Global ECS Research.Source: BLS; INEGI; UN; WDI; GS Global ECS Research.achieve their growth potential. It is intuitive that superior growth environmentscharacterised bystronger macro fundamentals, higher human capital and enhanced easiness of doing business,among other factorsare those where l

42、abour productivity gains have been higher.Exhibit 6 shows the average changes in the three competitiveness variablesunit labour costs,output per worker and GESfor three sets of countries: non-Euro area DMs, Euro area DMs andEMs. The crisis brought widespread reductions in labour costs, but they were

43、 more pronounced inEMs than in most of the DMs. The European periphery shows very large increases in costs in therun-up to the crisis, which were not matched by high productivity gains (at least not like in the USor the European core). EMs did much better on this dimension, and also in terms of GES

44、changes,before and during the crisis.The last two columns in Exhibit 6 show changes in the value of each countrys exports. Thismetric captures the quality of export baskets from low- to high-value-added items withinmanufacturing goods, and it reflects an economys position in the export value chain (

45、see GlobalEconomics Weekly 12/21, Quantifying the Value Chain in Global Exports). Changes in unitlabour costs sometimes reflect movements up and down this value chain, but not always. Indeed,the large increases in labour costs observed for the set of Euro area countries in our samplecoincided with i

46、mportant losses in the value of their export basketand this development is a keyexplanation for the unsustainable external positions in some of the periphery countries. Incontrast, countries such as Korea saw increases in labour costs that were more than offset byproductivity gains and, in addition,

47、 coincided with a movement up the export value chain.Into the future: Convergence behind EMs and R&D behind DMsThe global trends we described above were affected by the financial crisis and its aftermath, butnot substantially (see, for example, US Economic Analyst 12/27, Another Sluggish Summer).Alt

48、hough manufacturing regained space from services, unit labour costs declined and productivityslumped temporarily around 2009, the broader patterns obey fundamental forces more thancyclical headwinds, as the previous exhibits show. In particular, two of those forces are likely toshape the future of m

49、anufacturing.First, we expect convergence dynamics to continue to support higher rates ofmanufacturing growth in EMs. The latest five-year periods since 1995 show that the高盛全球經(jīng)濟(jì)、商品和策略研究62012 年 11 月 15 日Global Economics Weeklydynamism of this sector has come to rest mostly in that groupwith recent gr

50、owth ratesof around 7% vs. the 2% observed for DMs (Exhibit 7). As the Box describes, unlikeother areas of the economy, manufacturing exhibits absolute convergence in the sensethat, even without adjusting for country characteristics, those that are farther back intheir development processes (the maj

51、ority of EMs) actually tend to grow faster thanthose at more advanced stages (most DMs).Second, expenditure on research and development will likely continue to boostinnovation and productivity in DMs. We assembled data on per capita R&D expendituresin industry for a sample of about 30 countries, whi

52、ch we later split into manufacturingand non-manufacturing (Exhibit 8). The US and other DMs dominate this picture.Germany and the US spend almost the same amount on manufacturing R&D, althoughthe latter also devotes an important amount of resources to non-manufacturing sectors.Beyond these intra-DM

53、differences, it is clear from the comparison that EMs spend avery small fraction of DMs expenditure. Although technology transfers or replication mayallow some EMs to leap-frog in some areas, the true boost from innovation will accrueto the DMs where the research originates. This effect partly offse

54、ts the convergenceforces benefiting EMs by making it more difficult for them to catch up. But it alsoincentivises EMs to make investments in human capital accumulation and R&D.Is there a convergence boost in manufacturing that makes it special?The focus on manufacturing in recent policy debates rais

55、es thequestion of whether it is special in some sense. Does thecomposition of growthi.e., growth in the manufacturing versusservices or other sectorsmatter as long as overall GDP itselfgrows? Recent research suggests that the answer to that maybe a qualified yes, because manufacturing may provide a

56、wayfor poorer countries to reach a path of high growth andconvergence.In a nutshell, absolute convergence refers to the implicationfrom neoclassical growth models that less developed countrieswill grow faster. The problem is that when it comes to overallGDP per capita growth, this prediction is not

57、validated in thedata. In contrast, conditional convergence, which qualifies thatstatement by adjusting for country characteristics, has foundthe frontier (say the US) would receive a productivity growth boostof almost 2ppt, and one that was only a quarter as productivewould grow 4ppt faster. On aver

58、age, post-1995 productivitygrowth in the US has been almost 5%. At roughly 20% of thislevel during the late 2000s, Chinas boost from convergencewould have been an additional 5ppt, which takes it to 10% growth,roughly the value shown in Exhibit 7.The extent to which these convergence boosts in manufa

59、cturingpass through to the overall economy depend on the share of thissector in total output. In Rodriks model, GDP per worker is theaverage of labour productivity in the manufacturing ( ) and non-manufacturing sectors ( ), weighted by the manufacturing shareof employment (). The economy wide growth

60、 rate ( ) will begiven by:wide acceptance. In simple terms, there is a tendency for poorercountries to grow faster if they also have substantial levels oflnln,education, sound institutions, etc. What is surprising is thatwhere is the long-run growth rate of the economy, is therecent papers have foun

61、d that, in the case of the manufacturingconvergence coefficient,is the productivity of the frontiersector, absolute convergence does hold (see, for example, DaniRodriks Unconditional Convergence in Manufacturing, 2012).So even without conditioning by country characteristics,manufacturing in poorer c

62、ountries tends to grow faster. It is inthis sense that manufacturing is special and may provide aboost to growth.Rodriks empirical results centre on a convergence rate of 2.9%,meaning that a country that is half as productive as高盛全球經(jīng)濟(jì)、商品和策略研究country and is a reallocation term. For our purposes, the

63、middleterm is of interest. It shows that the convergence boost (whichincreases with the relative productivity shortfall) is scaled by theshare of employment in manufacturing. As we mentioned,countries such as China, with a large number of manufacturingworkers, are expected to benefit more from the c

64、onvergenceboost. But as it develops, and the productivity shortfall narrows,the boost will gradually fade, as it has happened in relatively moreadvanced EMs.ROMCHLMEXPOL SVKTURGRECHNRUSHUNNZLPORITASPACZRCANUKGNEDFRAIREBELNORAUSICEGERAUTSGP KORJAPSWZUSAFIN ISR72012 年 11 月 15 日Exhibit 7: Dynamism of m

65、anufacturing growth now restsmostly in EMsAverage real growth of manufacturing value addedGlobal Economics WeeklyExhibit 8: The bulk of manufacturing R&D still takes placein DMsIndustrial R&D expenditures (2005 PPP cst. US$; per capita)1412108%950850750650$US percapitaManufacturingNon-manufacturing5

66、506424503502500-2-41995-2000ChinaEMsWorldDMs2000-20052005-201015050-50Source: UNIDO; GS Global ECS Research.Source: OECD; Goldman Sachs Global ECS Research.Sketching an investment outlook for manufacturing progressThe shifts in manufacturing that we have described are likely to evolve over a relatively longperiod of time, and any investment implications are therefore more indirect and medium-term innature. Markets may be surprised by the continuation of some developments and the interruptionin o