天然气行业市场最新消息：全球液化天然气的未来10130220.ppt

,2013 年 2 月 20 日全球天然气市场最新消息全球液化天然气的未来 10 年研究报告并非所有美国液化气项目都将拥有市场,供应增长疲弱将在 2014 年之前继续抑制需求我们延展对未来 10 年全球液化气项目产能和未受限需求的预测，以更好地展现全球液化天然气(LNG)市场周期的潜在变化。首先需注意的是 2014 年之前的需求预测在很大程度上仍受到现有产能限制。尽管名义产能似乎远高于需求，但这可能具有误导性，因为一些生产商出于种种原因无法维持其出口供应。2016-17 年市场将更为疲弱，但需求将最终赶上从 2015 年开始，随着更多液化气产能投产并逐渐达产，全球需求可能重新开始以更适宜的速度增长。假设目前在建的液化气项目将最终建成，则我们认为现货市场将在 2016-2017 年进入下行周期，随后自 2020 年开始进入新的上行周期，原因是以中国和印度为首的非经合组织(OECD)买家扩大了市场份额，帮助吸收了,Samantha Dart+44(20)7552-9350 高盛国际杰夫可瑞(212)357-6801 高盛集团Johan Spetz(212)357-9225 高盛集团,市场中已有的供应。预测时间跨度越长不确定性越大我们目前假设 2016-2020 年期间美国将建成 700 亿立方米/年的液化气产能。我们还假设 2022 年之前加拿大将增加 480 亿立方米/年的液化气新产能，且莫桑比克的 LNG 出口项目将潜在增加 210 亿立方米/年的供应。总体而言，我们的基本假设为 10 年后全球液化气产能将从目前的 4,000 亿立方米/年增至 6,600 亿立方米/年左右。围绕美国 LNG 出口的政治辩论显然对这一预测构成了风险，正如在建液化气项目进度和莫桑比克气田开发进度的不确定性也构成风险一样。何谓供应过剩？重要的是，根据我们目前的 LNG 需求预测，我们认为未来十年中美国液化气产能不应超过 800 亿立方米/年。在我们看来，美国的新增产能将令全球市场在 2020年以后处于供应过剩状态。这不一定影响美国的出口，因为其中很大一部分供应合同已参照基准价格（比如油价甚至亨利港天然气价格）预先签订，但可能会拖累 LNG 现货价格。投资者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,2,2013 年 2 月 20 日,全球,Global LNG:The next 10 yearsWith global LNG markets exceptionally tight currently and a wave of new liquefactioncapacity on the horizon,we extend our global LNG balance table through the next 10years.We maintain our view that global LNG balances will likely transition from 2015into a softer market with the start up and subsequent ramp up of numerousliquefaction projects.However,we highlight that potential demand growth in the next10 years,particularly from Asia,is likely to once again tighten spot markets,creatingthe necessary conditions to accommodate some(but not all)of the large liquefactionprojects currently being proposed in North America.That said,we do not believe themarket needs more than 80 Bcm/y of US liquefaction capacity additions for the nextdecade.Our extended S&D balance points to a bearish cycle in 2016-2017,followed by asubsequent transition into a new bullish cycleWe have extended our expectations on global liquefaction capacity and unconstrained LNGdemand through the next 10 years to better visualize potential changes in the global LNG marketcycle(see Exhibit 1).This means looking beyond what is currently under construction,both on thesupply and demand sides,and making assumptions on potential capacity additions and utilizationrates.As a result,the further out we look,the greater the degree of uncertainty.The first thing to note in Exhibit 1 is that forecast demand through 2014 is still very muchconstrained by available capacity.Although nameplate capacity appears to be well abovedemand,this may be misleading as several producers have not been able to maintain their exportsupply for a variety of reasons.For instance,Egypt,Algeria and Indonesia have faced steepproduction declines and growing domestic demand,while Yemeni production has been frequentlydisrupted by terrorist attacks targeting gas pipelines.More recently,available supplies havebecome even tighter as Shell has declared a force majeure on Nigeria LNG supplies,affectingapproximately 40%of total Nigerian production.As a result,global LNG demand has beenrestrained,forced to remain at relatively stable levels for more than two years,while prices haverisen to accommodate the surge in Japanese and South American demand led by a nuclear anda hydro power crisis,respectively(see our January 28,2013 Global Gas Watch).From 2015,as additional liquefaction capacity is brought online and gradually ramps up,globaldemand may start to grow more comfortably again,as seen in Exhibit 1.The resulting marketbalance can be better visualized,however,by the liquefaction utilization rate implied by oursupply and(unconstrained)demand expectations,as shown in Exhibit 2.The main takeawaysare that,assuming that what is currently under construction on the liquefaction sidegets built,spot markets can transition from 2015 into a bearish cycle following thestart up of new liquefaction capacity,with excess supply expected to develop in themarket in 2016-2017,as these projects ramp up supplies.We expect the global LNGmarket to subsequently move towards another bullish cycle from 2020,as demandgrowth re-balances the market.To be clear,our unconstrained demand expectations shown in Exhibit 1 and underlying theutilization rate path in Exhibit 2 are not necessarily what will be realized in the market.Theyrepresent the volumes we think each region would normally absorb given their expectedregasification capacity and their need to utilize that capacity.If,however,more supplies areavailable than what global markets would ordinarily consume,spot prices will likely be underpressure,so that these volumes can be accommodated which is what we expect to see in the2016-2017 period.If,alternatively,global supplies are lower than the expected unconstrainedglobal demand,actual demand would then be restricted by that supply availability,likely lendingsupport to spot prices,much like what we observe in the market currently.In our base case,webelieve a similar process may take place again from 2020.高盛全球经济、商品和策略研究,50,3,2013 年 2 月 20 日Exhibit 1:Global LNG demand will likely continue to berestrained by supply through 2014Bcm/y,全球Exhibit 2:Our base case for utilization rates suggests abearish cycle in 2016-2017 followed by a tightening12-month moving average,in%,750650,Realized,Expected,88%86%,Declining available capacityrestrained global demand:bullish,Steady demand growth eventuallytightens the balance:bullish(assuming 70 Bcm/y of US LNGexport capacity gets built),550,84%,450350250,Global LNG(nameplate)capacity,82%80%78%,Financial crises weighedon demand:bearish,Additional liquefaction capacitysoftens the market:bearish,150,Global LNGdemand,76%Dec-05 May-07 Oct-08,Mar-10 Aug-11,Jan-13,Jun-14 Nov-15 Apr-17,Sep-18 Feb-20,Jul-21,Dec-22,Jan-05 May-06 Sep-07 Jan-09 May-10 Sep-11 Jan-13 May-14 Sep-15 Jan-17 May-18 Sep-19 Jan-21 May-22 Sep-23,Global liquefaction utilization rateAssuming 80 Bcm/y from the US,Historical average(2005-2012)Assuming 90 Bcm/y from the US,Source:Waterborne Energy,company data,GS Global ECS Research,Source:Waterborne Energy,company data,GS Global ECS Research.,estimates.We discuss below the main points regarding Exhibit 2 and the key assumptions driving ourconclusions:Declining utilization rates can be bullish or bearish.The key is understanding thedrivers behind it.If it is caused by lower supply availability restraining demand,like in thepast year,it is bullish for the market.If it is driven by lower(unconstrained)demand,likein 2009,or increased available capacity,as expected from 2015,it is bearish.Persistence of supply availability issues may permanently lower the marketutilization rate.Although the historical average for liquefaction utilization rate is 82%,the 12-month moving average shown in Exhibit 2 suggests this has somewhat declinedmore recently.We believe the supply availability issues in places like Egypt,Algeria andIndonesia are likely to continue,with Egypt taking steps to start importing LNG later thisyear and Indonesia having inaugurated its first importing terminal last year.Start up does not always mean ramp up.Although the bulk of the liquefactioncapacity additions are expected in 2015,a real softening of the market is only likely oncethese projects start ramping up their volumes,which may roll into 2016.The further out we look,the greater the uncertainty.Our base case assumes that70 Bcm/y of liquefaction capacity will get built in the United States between 2016 and2020.We have also assumed 48 Bcm/y of liquefaction capacity will be added in Canada,as well as 21 Bcm/y potentially out of a Mozambique LNG export project by 2022.Onnet,our base case assumes global liquefaction capacity rises to approximately 660Bcm/y in 10 years,from 400 Bcm/y currently.The political debate concerning US LNGexports obviously poses a risk to these numbers,as well as the uncertainty regardingthe timing of development of the Mozambique field.Effectively,even what is alreadyunder construction is subject to delays,as exemplified by the Angola LNG project(7Bcm/y),which has been delayed for over a year.Uncertainty also on the demand side.We currently assume that global LNGdemand growth will be led by non-OECD buyers,which will likely double their marketshare to 48%in 10 years and easily compensate for the expected decline in JapaneseLNG demand as their nuclear generation capacity restarts(see Exhibits 3 and 4).However,the potential exploration and development of shale gas in China poses a risk高盛全球经济、商品和策略研究,4,2013 年 2 月 20 日,全球to this view,as it could reduce the need for imports into China.Similarly,uncertaintyregarding new regasification facilities beyond those currently under construction alsoposes a risk to our base case,as it could restrain imported volumes,even if(unconstrained)demand continues to grow.For the moment,we remain relativelyaggressive regarding our Chinese demand growth assumptions,but will continue tomonitor Chinas progress on shale gas exploration and its plans for future regasificationcapacity.,Exhibit 3:We believe non-OECD Asia will drive globalLNG demand growth in the next 10 yearsExpected global LNG demand in Bcm/y600,Exhibit 4:leading non-OECD buyers to double theirmarket share to 48%by 2023Global LNG demand in Bcm/y600,500400300200100,500400300200100,OECDNon-OECD,Realized,Expected,0,2013,2014,2015,2016,2017,2018,2019,2020,2021,2022,2023,0Dec-06,Aug-08,Apr-10,Dec-11,Aug-13,Apr-15,Dec-16,Aug-18,Apr-20,Dec-21,Aug-23,Asia,Europe,Middle East,Americas,Source:GS Global ECS Research estimates.,Source:Waterborne Energy,GS Global ECS Research estimates.,On net,we believe that steady growth in global LNG demand will not only be able toeventually absorb all of the incremental supply currently under construction and expected inthe market in 2015 and 2016,but will also be able to accommodate an additional 140 Bcm/yof supplies from other liquefaction projects,likely including a few(but not all)from the UnitedStates(see Exhibit 5).This relatively tight market outlook for 2020 and beyond increases thelikelihood that at least some of the proposed US liquefaction projects enter additional supplyagreements in the coming years,which could help justify the economics of such projects.This is especially the case as the estimated breakeven cost for North American LNG projectsare relatively well positioned against competing projects elsewhere(see Exhibit 6).As aresult,the main source of uncertainty regarding the development of such projects remainspolitical,with no guarantees available on the volume of US LNG exports that the USDepartment of Energy will ultimately approve.Importantly,given our current LNG demand growth expectations,we do not believe themarket needs more than 80 Bcm/y of US liquefaction capacity for the next decade.AdditionalUS builds would,in our view,leave global markets in surplus beyond 2020.While this maynot necessarily pose a problem for US export facilities with a large share of suppliescontracted forward under a benchmarked price,such as oil or even Henry Hub,it may weighton spot LNG market prices.高盛全球经济、商品和策略研究,Gaspricerequired(US$/mcf)foreconomicbreakeven,5,2013 年 2 月 20 日Exhibit 5:We believe global LNG demand growth canaccommodate sizeable additions to liquefaction capacityEstimated year-on-year growth in Bcm/y60,全球Exhibit 6:North American projects a relatively wellpositioned from a cost perspectiveBreakeven price of non-producing and recently onstream LNGprojects17EvansShoalPlutoLNG1,50,16,ArrowLNG,BrassLNGAbadiLNG,APLNG,40,15,GreaterSunriseLNG,GladstoneLNG AngolaLNGShtokman QueenslandCurtisLNGBrowseLNG,302010,141312,Ichthys,LeviathanLNGGreaterGorgonWheatstoneLNGTanzaniaLNGKitimatLNGLNGCanadaYamalLNGNigeriaLNGTrain7,Prelude,SabinePass,MozambiqueLNG,0,2014,2015,2016,2017,2018,2019,2020,2021,2022,2023,11,0,PNGLNG,5,000,10,000,15,000,20,000,25,000,30,000,35,000,40,000,Demand growth,Liquefaction capacity growth,LNG production(mcf/d),Source:Company data,GS Global ECS Research estimates.Exhibit 7:Global LNG balance tableAnnual forecast through 2023 in Bcm/y,unless otherwise noted,Source:Company data,Goldman Sachs Research estimates.,2013,2014,2015,2016,2017,2018,2019,2020,2021,2022,2023,OECD AsiaNon-OECD AsiaChinaIndiaTaiwanThailandIndonesiaSingaporeMalaysiaEuropeMediterraneanNorthwestMiddle EastDubaiKuwaitIsraelOther EMEANorth AmericaSouth AmericaArgentinaBrazilChileTotal demand*Liquefaction capacityImplied utilization rate,16673272317211164333172310131152433441081.4%,16389362717233259302973311121261534341981.9%,163115483417364365293693422121472538047180.6%,1671315437174964672938114625111582540750480.9%,17114358401751075692940146728101783643253081.6%,17515563441951185712942157811091993745554783.1%,179167674820612867329441889112822114747858781.4%,1831787152216139676304719810113822114749960982.0%,1881887456236149679314919810114823114851962982.5%,192199776124615968232501989114824114953865881.9%,19621180652671610786335321911115824114956066484.4%,*Please note that our demand assumptions are constrained by available supply through 2014.From 2015,the demand numbers shown above should beinterpreted as unconstrained demand,while actual consumed volumes will ultimately be determined by the equilibrium between supply and demand atthe time.Source:GS Global ECS Research estimates.高盛全球经济、商品和策略研究,6,2013 年 2 月 20 日Current trading recommendations,全球,Current trades,First recommended,Initial value,Current Value,Currentprofit/(loss)1,The Commodity Carry Basket:Crude,Corn and Base(CCB)Long the S&P GSCI Petroleum,Corn and Copper total return indices,short the S&P GSCI F3 Aluminium total return index,equally weighted,December 5,2012-2013-2014 Outlook,100.00,100.29,0.29,Long NYMEX natural gas one-by-two call spreadLong one Jul-13 NYMEX natural gas$3.85/mmBtu call,short two Jul-13 NYMEX natural gas$4.70/mmBtu calls,November 11,2012-Natural Gas Watch,$0.12/mmBtu,$0.07/mmBtu,($0.05/mmBtu),Long Jun-13 NYMEX WTI crude vs.short Jun-13 ICE Brent crudeBuy 1 Jun-13 NYMEX WTI crude,sell 1 Jun-13 ICE Brent,August 21,2012-Energy Weekly,($12.33/bbl),($18.62/bbl),($6.29/bbl),Long S&P GSCI Brent crude oil total return indexLong S&P GSCI Brent crude oil total return index at initial index value of 1,174.26,August 21,2012-Energy Weekly,1,174.26,1,261.33,(3.35%),Rolled from a long September 2012 NYMEX WTI Crude Oil position on 21-Aug-12,carrying forward a potential loss of 10.77%Long GoldBuy April 2013 COMEX Gold,sell$1,850/toz Apr-13 call,buy$1,575/toz Apr-13 put,October 11,2010-Precious Metals,$1,717.5/toz,$1,623.7/toz,$224.0/toz,Rolled from a long Dec-12 COMEX Gold future position on 4-Dec-12 with a potential gain of$317.8/tozAs of close on February 15,2013.Inclusive of all previous rolling profits/losses.Source:Goldman Sachs Global ECS Research.高盛全球经济、商品和策略研究,4,0.1,1,2,3,4,7,2013 年 2 月 20 日Price actions,volatilities and forecasts,全球,Prices and monthlychanges1,Volatilities(%)and monthly changes2,Historical Prices,Price Forecasts3,units,17 Feb Change Implied2 Change Realized2 Change 3Q 11 4Q 11 1Q 12 2Q 12 3Q 12 4Q