Working Paper
China’s Evolving Oil Demand
Slowing Overall Growth, Gasoline Replacing Diesel as Demand
Driver, Refined Product Exports Rising Substantially
Gabriel Collins, J.D.
Baker Botts Fellow in Energy & Environmental Regulatory Affairs
© 2016 by the James A. Baker III Institute for Public Policy of Rice University
This material may be quoted or reproduced without prior permission, provided
appropriate credit is given to the author and the James A. Baker III Institute for
Public Policy.
Wherever feasible, papers are reviewed by outside experts before they are released.
However, the research and views expressed in this paper are those of the individual
researcher(s) and do not necessarily represent the views of the James A. Baker III
Institute for Public Policy.
This paper is a work in progress and has not been submitted for editorial review.
ExecutiveSummary:
•
Three key factors characterize the “new normal” of China’s oil demand: (1) overall
annualoilproductsdemandgrowthwilllikelyslowfromanaverageofnearly5%annually
toonecloserto2.8%betweennowand2020;(2)thecountry’soildemandprofilehas
becomeprimarilygasoline-driven,and(3)oilrefineryovercapacityismakingChinaoneof
theworld’slargestnetexportersofrefinedproducts,atrendlikelytocontinuelookingout
to2020.
•
In2016,China’sincrementalcrudeoildemandwilllikelyincreasebyapproximately360
kbd—roughlytheaveragedailyconsumptionofthePhilippines.
•
China’soildemandisbecomingincreasinglyconsumer-driven.Gasolinehasbecomethe
primemoverofChina’soildemandascarownershiprisesandindustrially-drivendiesel
usedeclines.Risingdomesticairtravelhasalsosignificantlyincreasedkerosenedemand,
althoughthemacroimpactisnotnearlyaslargeasthatofgasoline.
•
Consumer-drivenoildemandislikelytobesignificantlymorevolatileintheshort-term.
Consumer sentiments and behavior evolve and shift much more rapidly than China’s
traditionaloildemandpattern,whichwasunderpinnedbyindustrialdieselfueluseand
wascloselytiedtothe5-YearPlans.Greaterdemand-sidevolatilityinChinastandstohave
profoundeffectsontheglobalcrudeoilandrefinedproductsmarkets—particularlyifthe
Chinesegovernmentdoesnotbecomemoreregularandtransparentinitsdisclosuresof
energydata.Betterdisclosurewouldreducemarketparticipants’needtospeculateasto
thetruestateofsupplyanddemandbalanceswithintheChineseoilecosystem.Thisis
turnwouldhelpsmooththespikesaspricesrespondtomarketfundamentals,benefitting
producersandconsumers.
•
Threecore“wildcards”meritcloseattentionoverthenextfiveyearsfortheirabilityto
influenceChina’soildemandtrajectory:(1)roadcongestionthatmayreducevehicleuse
despite continuing strong sales of new cars, (2) penetration of more efficient gasolinepoweredvehiclesandelectricvehiclesasChina’sgrowingmiddleclasspushesbackagainst
severe air pollution in many cities, and (3) the degree to which an increasingly brittle
Chinese political leadership can craft successful, sustainable solutions to multidimensionaleconomicpolicychallenges.
•
Policymaking missteps that accelerate declines in China’s economic growth rate would
substantiallyenhancetheriskthatpricescontinuetotradeinanarrowband.Currently,
themostlikelyscenarioisthataslowing—butstillrelativelyoil-thirsty—Chineseeconomy
combines with deep global CAPEX cuts and marginal costs of new supply exceeding
$65/bbltoyieldaglobalbenchmarkcrudeoilpricerangeaveraging$60-70/bblbetween
nowand2020.
1
I.
China’s Oil Demand Will Remain Influential Even as the Global Market
RemainsOversupplied
China’seconomicgrowthtrajectoryofthepast15yearsappearstobeataninflectionpoint.As
the country’s high annual oil demand growth rate downshifts to a slower one, market
perceptionsputdownwardpressureonoilprices.Indeed,thecarnageinglobaloilmarketsover
thepast24monthshascomeaboutinpartbecauseChina’sslowinggrowthinitsrealeconomy
andenergyconsumptioncoincidedwitharapidincreaseinunconventionaloilsupplies.China’s
economicslowdown—andthecommensuratedecelerationoftheoildemandgrowthrate—isin
manywaysnatural.Theeconomymustgrowoffamuchhigherbasethanwasthecaseten,five,
oreventwoyearsago.Independentofthebaseeffects,oilintensityperunitofeconomicoutput
isalsodecreasingastheeconomygraduallyfavorsservice,consumer,andlightindustrialactivity
(moreonthisinSectionII,below).
Akeydownsiderisktoglobaloilpricesisthatrecentinternalpoliticaldevelopmentsexacerbate
theriskofaneconomichardlanding,whichcoulddepressChina’soildemandgrowthmuchmore
rapidlythanthestructuraleconomictrendsmentionedabove.Tothatpoint,itbearsnotingthat
China’sleadershipcurrentlygrappleswithadegreeofpolicydisunitynotseeninthePRC’supper
echelonsofgovernmentsinceatleastthelate1980s.1
AmoresustainedChinaoildemandslowdownwouldbeaseriousheadwindforpricesgiventhe
country’sinexorablyincreasingimportanceastheglobaloilmarkets’“consumeroflastresort.”
Chinaaccountedforapproximately43%ofglobalincrementaloildemandgrowthbetween2009
and 2015 (Exhibit 1). This demand growth helped underpin global oil prices despite overall
weaknessintheglobaleconomy.ToputChina’sannualincrementaloildemandgrowthduring
thisperiodintoperspective,China’sincrementalcrudeoildemandin2009grewby275kbd—
roughly the daily consumption of the Philippines and by 1,158 kbd in 2010, roughly the daily
consumptionofTaiwan,oneoftheworld’smostindustrializedeconomies.
1
LinglingWeiandJeremyPage,“DiscordBetweenChina’sTopTwoLeadersSpillsIntotheOpen,”TheWallStreet
Journal,22July2016,http://www.wsj.com/articles/discord-between-chinas-top-two-leaders-spills-into-the-open1469134110
2
Exhibit1:ChineseIncrementalDemandPoweredGlobalOilPricesOverthePastDecade
Leftaxis:incrementaloildemandgrowth(kbd),Rightaxis:China’s%ofglobalincrementaloil
demandgrowth
Source:BP,IEA,Author’sAnalysis
AsChineseoildemandrampedup,itimpactedpricesontwointer-relatedfronts.First,China’s
growing thirst for crude directly increased global crude prices. Second came the demand
increases in commodity-producing (e.g., iron, copper, etc.) countries whose economies were
ridinghighonthebackofrobustChinesedemandforrawmaterials.Indeed,thefiveregional
groupingsthataccountforthemajorityofcommodityexportstoChina(MENA,LatinAmerica,
SoutheastAsia,FormerSovietUnion,andSub-SaharanAfrica),accountedforroughly10.5million
barrelsperdayofoilproductsdemandincreasebetween2000and2015—1.6timestheamount
thatChina’sownoilproductsdemandgrewbyduringthattime.2
Theglobaloilmarketcurrentlylaborsunder:(1)acontinuedsupplyoverhangfrombloatedglobal
crude and refined products inventories and (2) burgeoning production from OPEC members
seekingtogainmarketshareandpayforexpansivemilitarycampaignsanddomesticpolitical
support(SaudiArabiaandIraq)andthoseseekingtorejointheglobaloilmarket(Iran).Russia
2
CalculatedfromBPStatisticalYearbook2016Data,OriginalconceptdrawnfromGabeCollins,“HowExtensively
CouldChina’sSlowdownImpactGlobalOilPrices?,”ChinaOilTrader™,No.1(23May2012),
http://www.chinaoiltrader.com/?p=37
3
also continues to robustly increase production as Russian E&P companies capitalize on lower
servicecosts—roughly80%ofwhicharepricedindevaluedrubles—whileoilexportsarepriced
inmorevaluabledollars,andastheKremlincontinuestoencourageactivityinoneofitsfew
globallycompetitiveindustries.
AreductioninChina’soildemandgrowthratefromroughly5%annuallyoverthepastfiveyears
to2.8%—whichisthisauthor’sbasecaseoutlookforChina’scrudeoildemandgrowthoverthe
nextfiveyears—wouldcastapalloveroilprices.DatafromtheBPStatisticalReviewofWorld
Energy 2016 shows that only five of the 10 largest global oil consuming countries by volume
actuallysawpositivedemandgrowthin2015,despitethelowestaverageoilpricesinadecade.3
These10countriescollectivelyaccountedfor60%ofglobalcrudeoildemandin2015.Thefive
large consumers who experienced growth in demand during 2015 were (in order of demand
growth):China,India,theU.S.,SaudiArabia,andSouthKorea.China’s2015oildemandgrowth
accordingtotheBPdatacameinat766kbd:nearly30%higherthantheincrementalgrowthof
IndiaandtheU.S.combinedandbasicallyequaltothecombinedincrementaldemandgrowthof
theU.S.,India,andSaudiArabia.
HereitisworthwhiletozoominonIndia,whichsomeanalystsviewashavingthepotentialto
offsetChina’sslowingoildemandgrowth.4UsingtheBPdataasabaseline,considerthat6.0%
annualdemandgrowthbetween2016and2020wouldmeanIndianoildemandgrowthaveraging
281 kbd. Raising the average annual demand growth rate to 8.0%--an optimistic assumption
giventhattheIndiangovernmentlikelycannotunleashaninfrastructureandheavyindustrial
boomonthescaleofwhatChinahasdone—wouldraiseaverageIndianoildemandgrowthto
390kbdbetween2016and2020.Forreference,China’sannualincrementaloildemandgrowth
has averaged nearly 475 kbd over the past decade. This suggests that even in a high-growth
scenario,India’sincrementaloildemandwouldstruggletooffsetthecombinationofasignificant
ChinesedemandslowdowncoupledwithcontinuedstructuraldeclineinEurozoneoiluseand
theworldwidephase-inofmorefuel-efficientautomobiles.
China’s massive incremental demand growth contribution and the relative lack of a positive
demandresponsetolowcrudepricesinothermajorconsumingcountriessuggeststhateven
with lower crude prices, other consumers may not be able to compensate for a significant
demand slowdown in China. As such, absent significant lost production from deferred and
cancelledupstreamCAPEXexpenditures,aslowingChinawouldsubstantiallyenhancetherisk
thatpricesof$45-50/bblcouldbecomea“newnormal”fortheglobalBrentandWTIcrudeoil
benchmarks.
3
TheBPStatisticalReviewofWorldEnergycalculatesdemandusingthefollowingformula:Inlanddemandplus
internationalaviationandmarinebunkersandrefineryfuelandloss.Consumptionofbiogasoline(suchas
ethanol),biodieselandderivativesofcoalandnaturalgasarealsoincluded.
4
See,forinstance:AmritaSenandAnupamaSen,“India’sOilDemand:OntheVergeof‘Take-Off’?,”March2016,
TheOxfordInstituteforEnergyStudies,pdfavailablefordownloadat
https://www.oxfordenergy.org/publications/indias-oil-demand/.
4
Atthispoint,perhapsthemostlikelyscenarioisthataslowing—butstillrelativelyoil-thirsty—
ChineseeconomycombineswithdeepglobalCAPEXcutsandmarginalcostsofsupplyofaround
$65/bbltoyieldaglobalbenchmarkcrudeoilpricerangeaveraging$50-65/bblbetweennow
and2020.Thispriceviewdrawsheavilyupongloballiquidssupplycostdatapublishedinlate
2015byRystadEnergy,aswellasrecentdatafromtheIEAshowingthatglobalupstreamoiland
gasinvestmentfell25%in2015,isexpectedtodeclinebyanadditional24%in2016,andcould
continuefallingthrough2017,whichwouldmeananunprecedentedthreeconsecutiveyearsof
industryinvestmentcuts.5
Chineseoilrefineriescurrentlytakein,onaverage,morethan10millionbarrelsperdayofcrude
oil,andduringpeakdemandperiodsinsummer2015rannearly12millionbarrelsperday.These
figurescomefromtheJointOilDataInitiative(“JODI”)andreflecttheintakeofcrudeoilthatis
actually destined to be converted into products.6 The numbers’ reflect the Middle Kingdom’s
heftyandgrowingcrudeoildemandrole:in2000,Chinaaccountedfor5.8%ofglobalcrudeoil
demand.By2015,thatfigureexceeded12%--roughlyoneineveryeightbarrelsusedworldwide.
Againstthisbackdrop,itbecomesclearwhyoilproducers,traders,consumers,andotherparties
with exposure to the global crude oil value chain seek to better understand the potential
trajectories of oil consumption in China over the next three to five years. China’s oil demand
ecosystemisimmenseandcomplicated,andinmanyways,stillquiteopaque.Thestudyadapts
totheserealitiesbyfocusingnotonprovidinganencyclopedicoverview,butrather,identifying
andtheninvestigatingkey“influencefactors”thatarecomingtodriveChina’soildemand,and
arelikelytodosointhenextfivetotenyearsaswell.
Simplywatchingbroadhigh-levelflowssuchasimportsorrefinerycrudeintakeonlytellsapartial
story.Barrelsmayflowintoarefinery—orintostrategicstorage.Barrelsmayflowintoarefinery,
butthenbeprocessedintodieselfuelthatisexported.Ortheymaybeturnedintogasolineto
powerthecountry’smorethan120millionpassengervehicles,thevastmajorityofwhichare
gasoline-fueled.
To tease out the underlying trends, this analysis employs a range of data sources and
techniques—includinglocalChineselanguagemedia,companydata,andofficialstatisticsfrom
5
“GLOBALLIQUIDSCOSTCURVE(OCTOBER2015),”RystadEnergy,
http://www.rystadenergy.com/NewsEvents/PressReleases/global-liquids-supply-cost-curve;SeealsoBenoit
Faucon,“Oil,Gas-FieldInvestmentSettoFallforTwoStraightYears,IEAReportSays,”TheWallStreetJournal,13
September2016,http://www.wsj.com/articles/oil-gas-field-investment-set-to-fall-for-two-straight-years-ieareport-says-1473795274
6
“Refinery Intake,” JODI-Oil World Database - Full Version (Primary Products Table),
http://www.jodidb.org/TableViewer/summary.aspx (accessed 1 August 2016). The caveat matters. In the first 11
monthsof2015,purchasesofcrudeoilslatedtofillChina’sgrowingstrategicpetroleumreserve(“SPR”)boostedthe
country’sapparentcrudeoilintakebymorethan440kbd—roughlytheaveragedailycrudeoiluseofPakistan.This
studyfocuseson“demand”inthesenseofbarrelsofoilthatareactuallybrokendowninarefinery,burned,or
otherwisephysicallytransformed.Assuch,itdoesnotaddressChina’sSPR,whichstores—butdoesnottransform—
crude.
5
thelocal,provincial,andnationallevels.Whereatallpossible,thestudyemploysindependent
data sources, since official Chinese energy statistics frequently fail to provide the granularity,
regularity,orreliabilitynecessarytodrawanalyticallyusefulconclusions.7
II.
KeyMacro-LevelFactorsInfluencingOilDemandinChina
Threethemesdominatethemacro-leveloildemandpictureinChina.First,China’soildemand
growthrateisclearlyslowinginbothpercentageandabsolutevolumeterms.YetevenifChina
movesontoaslowergrowthtrajectory,thebaselinenumbersarenowlargeenoughthatevena
slower annual percentage growth rate can still yield significant absolute oil demand volume
increaseseachyear.
Second,China’soildemandstructureiscurrentlyatatransitionpoint.Transportationdemandis
now transitioning to a gasoline-driven structure as passenger vehicles increasingly supplant
diesel-poweredtrucksandindustrialmachineryasasourceofincrementalfueldemand.Third,
the country’s heavy industrial sector is stagnating and declining as a proportion of national
economicoutput.Asfactoriesrunatlowerutilizationrates,minersdigandtransportlesscoal
andironore,andfewergoodsaretruckedfromfactorytoport,dieselfueldemand—whichis
heavilyleveragedtoindustrialactivity—suffersacommensuratedemandimpact.
Theme1:China’sOilConsumptionIsFundamentallyTransportationDriven
GasolineisnowtakingoverfromdieselfuelasthebaselinedriverofoildemandinChina.Diesel
fueluseisdistinctlyseasonal,withintra-yearvariationsbetweenthespring“trough’andwinter
“peak”ofasmuchas1,500kbd(Exhibit2).Dieselusehasalsostagnatedsincethewinterof
2010/11andcontractedineachofthepasttwoyears,witha3.7%YoYdeclineindieseldemand
during 2015.8 Gasoline demand, in contrast, rose 7% YoY in 2015, mirroring the sharp and
sustainedriseinpassengercarsalesthathascharacterizedmuchofthepastdecadeinChina.9
7
Chineseenergydataarestillsometimesunreliable,asillustratedbythesizeablerevisionsmadein2015tocoal
consumption data dating back to at least 2000. Coal data is plagued by a greater degree of reliability problems
because there are thousands of producers—some of whom operate illegally—and an even large number of
consumerswhocanprocurecoaldirectlyfromtheproducer.Suchisnotthecaseintheoilsector,wherefewerthan
10companiesaccountforthevastmajorityofrefiningcapacityanddownstreamoilproductdistributionandsales,
functionswhichgivedeepinsightintodemand.Intheoilsector,supplychainconcentrationenhancesobservability
andimprovesreliabilityofthedata.Thisisreinforcedbythefactthatoilsmugglinghasbeengreatlyminimizedin
recentyears,makingcustomsdataonimportsandexportsofrefinedproductsmuchmorereliableandaccurate.
8
“ChinaDieselUseSlumpsasNaturalGas,GasolineDemandGains,”Bloomberg,24January2016,
http://www.bloomberg.com/news/articles/2016-01-25/china-diesel-use-slumps-as-natural-gas-gasoline-demandgains
9
Id.
6
Exhibit2:ConsumptionofGasolineandDieselFuelinChina,kbd
Source:JODI,Author’sAnalysis
GranulardatafromresearchersatSinopec,China’slargestoilrefinerbyvolume,showthatdiesel
andgasolinedemandinthecountryoverwhelminglystemsfromtransportapplications.Nearly
70%ofdieseldemandcomesfromon-roadtransportapplications,whilemotorvehicledemand
accounts for nearly all gasoline demand—91% from vehicles with more than two wheels and
7.4%frommotorcycles(Exhibit3).
7
Exhibit3:EstimatedDieselandGasolineDemandinChinabyEndUser,4Q2015
Source:SinopecPresentation,September2015
Theme2:TheOilIntensityLevelofChina’sEconomyHasDeclinedSubstantially
Overthepast20years,China’seconomicoutputhasbecomemuchlessoilintensive.Producing
onebillionRMB’worthofGDPrequired200,000bblofoilin1995,butthatsamebillionRMB
requiredonly64,500bblin2015(Exhibit4).Indollarterms,thismeansthatproducingonebillion
USD’worthofGDPin1995consumednearly1.7millionbarrelsofoil,butin2015,producingthe
samebillionUSD’worthofGDPonlyrequiredabout418thousandbarrelsofcrude.
Overthepastfiveyears,China’soilintensitycurvehasnoticeablyflattened.Thissuggeststhat
thecountryissettlingintoatrendwherebyeachbillionRMBofeconomicoutputwillrequirein
theneighborhoodof60,000to65,000barrelsofcrudeoil.Barringawholesalereturntofixed
asset-led growth—which is unlikely due to rising concerns about pollution and unsustainable
debt burdens incurred from financing underutilized big-ticket infrastructure projects—the
country’s oil intensity is likely to continue gradually trending downward as services and light
industrycometocompriseagreaterportionofeconomicactivity.
8
Exhibit4:OilIntensityofChina’sEconomy
Leftaxis:BarrelsofCrudeOilperBillionRMBofGDP,Rightaxis:BarrelsofCrudeOilPerBillion
USDofGDP
Source:BP,NBSChina,Author’sAnalysis
Theme3:China’sOilDemandLikelyFacesanEconomicTransition
AsChinabecametheworld’smanufacturingfloorinthe1980-2008period,oildemandthatmight
haveotherwiseoccurredelsewhereintheworldinsteadhappenedinChina.Asthecountryloses
manufacturingcompetitivenessanditseconomyslowsoverall,China’soildemandgrowthrate
islikelytodraweverclosertotheglobalbaselinerate—whichsince2005hasaveraged1.2%
annually,accordingtoBP.
AsChinaincreasinglyconfrontsthenecessityofreducingfixed-assetinvestmentasaproportion
oftotaleconomicactivity,oildemandgrowthmayslowinafashionakintothetrajectoriesof
TaiwanandSouthKorea(Exhibit5).Ineachjurisdiction,aperiodof15-20yearsofrapideconomic
expansionandoildemandgrowthwasfollowedbyaperiodinwhichoildemanddeclinedfrom
peaklevels(Taiwan)orgrewmuchmoreslowly(SouthKorea).
9
Exhibit5:OilDemandProfilesofTaiwanandSouthKorea
ThousandBarrelsPerDayofCrudeOilDemand,Leftais:SouthKorea,Rightaxis:Taiwan
Source:BPStatisticalReview,Author’sanalysis
ItisalsoinstructivetoconsidertheoildemandprofileoftheUnitedStates,sinceitsoildemand
ismoredependentontransportation,asopposedtotheAsianTigers,whereindustrialactivity
dominatesoilconsumption.AfteratorridperiodofdemandgrowthafterWorldWarII,U.S.oil
demandtemporarilypeakedintheearly1970swiththefirstOilShock(Exhibit6).Butitthen
grewessentiallyunbrokenforapproximately25yearsuntiltheGlobalFinancialCrisiscrashed
demand.Thelowoilpricessincelate2014havehelpedre-animatecarsales,drivingactivity,and
gasolinedemand,whichhasunderpinnedathirdwaveofU.S.oildemandgrowth.
InSouthKorea,thepastdecadehaswitnessedamilder,smootheroildemandrecovery.The
U.S.,incontrast,hasseenamuchsteeperdemandincreasesincethebleakesttimesoftheGlobal
FinancialCrisisin2008and2009.TheU.S.andSouthKoreanoildemandprofiles,whilevastly
different,illustratethepotentialfora“secondwind”thatre-animatesoildemandgrowthafter
amajoreconomicshockorstructuralchange.UnderascenarioinwhichChina’spolicymakers
“get it right” or other serendipitous events intervene catalyze renewed economic vigor, a
resumption of more robust oil demand growth is an upside risk factor that merits serious
consideration.
10
Exhibit6:OilDemandGrowthProfileofTransportation-LedUnitedStates
ThousandBarrelsPerDay
Source:EIA
Againstthemacrobackdropprovidedabove,thestudywillnowprovideanindepthanalysisof
thethreekeysectorsthatdriveChina’soildemandandhowtheyarelikelytoevolveinthenext
fiveyears.Firstisroadtransportation,thesecondisfueloil(formerlyusedasrefineryfeedstock
andalsousedasmarinefuel)andjetfuel,whilethethirdcomesfromcrudeoil-derivedchemical
feedstocks.
11
III.
HowOilDemandisChanginginKeySectorsoftheChineseEconomy
A. Transportation
China’sownofficialdatasuggestneartermdieselfueldemandmayhavepeakedinatleast13
provinces,whichcollectivelyaccountfornearly60%ofthecountry’sdieselfueluse(Exhibit7).
Thisdevelopmentmattersgreatlybecausedieselaccountedfornearly1/3ofChineseoilproduct
consumption in 2015.10 Furthermore, China’s share of global diesel fuel consumption has
increasedfrom9.6%in2005to13.1%in2015,andChinaaccountedfornearly36%oftheglobal
netincreaseindieselfuelconsumption,accordingtoJODIdata. Sinopec,China’slargestrefiner,
takesinapproximately3.27barrelsofcrudeoilforeachbarrelofdieselfuelproduced.11Because
thevastmajorityofdieselfuelusedinChinaisrefinedinthecountry,thisratiosuggeststhat
conservatively,each100kbdofdieselconsumedrequiresthatrefineriesprocessatleast300kbd
ofcrude.
China’s national diesel consumption declined in both 2014 and 2015. Consecutive years of
contraction,coupledwithweaknessinotherphysicalindicatorsofindustrialactivity,increasingly
suggestthatChina’sdieseldemandarchitectureisstructurallyevolving.Itremainsprematureto
concludethatChina’sdieseldemandhas“permanently”peaked.Buttheevidenceincreasingly
suggeststhatforone-to-fouryearoilpricepathanalysispurposes,theMiddleKingdom’sdiesel
demandisnotcomingback.Indeed,theIEAitselfforecastsacontinueddeclineinChinesediesel
consumptionin2016.12
China’stectonicdieselconsumptionshifthasgivenpausetotheglobaloilmarket.Anditwill
continuetodosoasproducers,traders,andcapitalprovidersadjusttoanewnormalinwhich
sustainedmassiveoildemandgrowthfromChinacannolongerbetakenforgranted.
10
IEAOilMarketReport,13July2016,https://www.iea.org/oilmarketreport/omrpublic/currentreport/#Demand
ChinaPetroleum&ChemicalCorporation,ThirdQuarterlyReport,2015.
http://english.sinopec.com/investor_center/reports/
12
IEAOilMarketReport,13July2016,https://www.iea.org/oilmarketreport/omrpublic/currentreport/#Demand
11
12
Exhibit7:ChineseProvincesWhereDieselFuelDemandAppearstoHaveAlreadyReacheda
Near-TermPeak
Source:NBSChina,ProvincialStatisticalBureaus,Author’sAnalysis
Theprovinceswheredieseldemandappearstohavepeakedfallintothreeprimarygroups.First
aretheexport-focusedprovinces/citiesofBeijing,Fujian,Guangdong,Shanghai,andZhejiang.
TheseareasfacestagnantdemandabroadformanyChinese-madeproductsandtherisingwage
andothercoststhatareunderminingChina’shistoricalexportcompetitiveness.
Second are the rustbelt and coal belt provinces of Inner Mongolia, Liaoning, and Shandong,
whereovercapacityandslowingheavyindustrialactivityarecrimpingdieselfueldemand.Given
thatChinahaslikelyhitPeakCoal,itisdifficulttoimaginedieseldemandrecoveringinthese
provincesinthenextfiveyears.ThirdaretheHinterlandProvinces:Henan,Hubei,andHunan.
The massive economic stimulus efforts of 2008-2010 focused heavily on these areas. China’s
stimuluscreatedunsustainablyhighlevelsoffixedassetinvestmentandconstructionactivityin
these three provinces and after several years of elevated activity, diesel demand began to
decline.
Chinadieselbullscoulddisagreeandsaythattheapparentprovincialdemandpeaksaresimply
“local maxima” as opposed to being definitive structural inflection points. To pre-empt this
13
potentialcriticism,thepresentanalysisdividesthepeakdieselprovincesintotwomaingroups:
(1) those which are “clearly in structural demand decline” and (2) those which remains
“susceptibletostimulus.”
AstimuluspackagesimilarinmagnitudetowhatChinaunleashedin2008-2010(4trillionRMB)
could potentially temporarily revive diesel demand in borderline provinces such as Jiangsu.13
Suchalargeinfrastructure-orientedstimuluseffortcouldverywelltemporarilypropupdemand
for diesel. But it would do so at high cost and introduce risks the Central leadership seeks to
avoid. For one, a large stimulus package would increase systemic risk by further bloating the
country’sbalancesheet.Itwouldalsopressureexistingrealestateandinfrastructureassetsthat
are already seriously debt-laden and underutilized, while creating additional Ghost Cities and
other“zombie”assets.14Finally,itwouldseriouslyundermineBeijing’spolicygoalsofeliminating
overcapacity, reducing pollution, and shifting China’s economy toward a more sustainable
consumption-orientedstructure.
1. DriversofChina’sDieselDemandSlowdown
ThedemandslowdowninChina’sprimedieselconsumingregionsreflectsabroaderslowdown
in industrial activity. A number of industrial activity metrics corroborate the diesel demand
decline reflected in official data. Foremost is electricity demand, which is driven primarily by
industrialuse. Powerconsumptionbeganplateauingin2013andhasseendemandgrowthslow
meaningfullyeachyearsince(Exhibit8).15
13
“China's4trillionyuanstimulustoboosteconomy,domesticdemand,”Xinhua,9November2008,
http://news.xinhuanet.com/english/2008-11/09/content_10331324.htm
14
Thephenomenonofso-called“GhostCities”thatarebuilttohouselargenumbersofpeople,butremain
virtuallyuninhabitedhasproliferatedastherealestateboominChinarunsupagainstfundamentaldemand
constraints.Multiplemediasourceshavedocumentedtheexistenceofthesecities,withextensivecoverage
beginninginthe2013timeframe.Considerforinstance“China'sGhostTowns:StrollingtheThames(VICEonHBO
Ep.#6Extended),”https://www.youtube.com/watch?v=qPjGWcM3Awcand“China'sGhostCities-60Minutes
Australia,”https://www.youtube.com/watch?v=F1ZsVUB_-qg.(lastaccessedon27August2016).
15
GabrielB.CollinsandAndrewS.Erickson,“China’sStockMarketBoom:BuckleUpforaWildRide!”China
SignPost™(洞察中国)89(14June2015).http://www.chinasignpost.com/2015/06/14/chinas-stock-market-boombuckle-up-for-a-wild-ride/
14
Exhibit8:China’sElectricityConsumptionSlowdownTracksDecliningDieselFuelUse
Source:NBSChina
Electricitydemandisahigh-confidencedatasetforseveralreasons.First,China’spowersupply
comes from a small set of generators and grid operators, allowing throughput to be easily
tracked.Second,unlikeoil,coal,andotherenergycommodities,thegenerationandtransmission
ofelectricitycorrelatecloselywithactualusebecauseelectronscannotbestoredeconomically
onascalelargeenoughtoinducepotentialinventorycalculationproblemsofthetypewhich
plague analysis of other commodity markets in China. Third, electricity use is pervasive
throughouttheindustrialsector,sincevirtuallyeverymanufacturingactivityhasmultiplesteps
for which electricity is an input. For these reasons, the steady and significant slowdown in
electricity demand across the Chinese economy suggests that the official diesel fuel demand
numbersareprobablysufficientlyaccuratetogiveconfidenceinthetrendstheyareshowing.
Second,otherphysicalindicatorsofindustrialactivityinChinahavealsoflat-linedoverthepast
18months.Asidefromelectricity,sulfuricacidandcausticsodaalsopervadeabroadrangeof
industrial processes, making demand for them a lead indicator for the health of an industrial
economy.Tothatpoint,alkalis(foremostamongthemcausticsoda)arekeyfoundationsforthe
AmericanChemistryCouncil’sChemicalActivityBarometer,anindexbasedoncomparingmore
than60years’ofdatacorrelatingdemandforkeyindustrialinputchemicalswithcyclesinthe
overallUSeconomy.16Chinesedataonlyshowmonthlyproductionofsulfuricacidandcaustic
16
“ChemicalActivityBarometer,”AmericanChemistryCouncil,
https://www.americanchemistry.com/jobs/cab/Barometer.pdf
15
soda, not actual “consumption.” Nonetheless, because the country is largely self-sufficient in
both chemicals and a relatively small share of total output is exported, the production data
provideanacceptableproxyforlocalusage.
Since late summer 2014, production of each chemical has flat-lined (Exhibit 9). Sulfuric acid
productionis“choppy”inpartbecauseasubstantialportionofChina’sacidoutputisabyproduct
of oil refining and metal smelting. The lack of growth in caustic soda output is especially
noteworthybecausefromearly2014onwards,thecountryhasfacedasignificantovercapacity
situation, which would normally prompt increased exports.17 The lack of such a development
suggestsweakdemandabroad(Australia,foroneisakeymarketforChinese-madecausticsoda),
which in turn further weighs on Chinese domestic industrial activity that drives diesel fuel
demand.
Exhibit9:ProductionofSulfuricAcidandCausticSodainChina,MillionTonnesperMonth
Source:NBSChina
17
“2015 Caustic and Sulfuric Acid Market Analysis: Balancing faces Many Challenges” (2015年烧碱市场分析：
仍将面临多重挑战), Tianjin Tiankai Chemical Industries Import and Export Corporation,
http://tiankaichem.com/a/xinwendongtai/2/259.html
16
2. WhyChina’sDieselFuelDemandLikelyWillNotRecovertoItsPriorPeakintheNext
FiveYears—IfEver
Reason one is that China has likely passed “Peak Truck.” The CEO of Cummins stated in the
company’sfourthquarter2015earningscallthatwithregardtothenumberoftruckssoldinthe
country,“it’snotclearthatwe'lleverhitthatnumberagain.”18Cumminsdataalsoshowthat
truckenginesalesinChinadeclinedsharplyfrommorethanonemillionunitsin2013to750,000
unitsin2015,withsalesforecasttodeclineafurther4%year-on-yearin2016.Thecompany’s
statementmeritcloseattentionbecauseitsuppliesnearlyoneineveryfivedieselenginesgoing
into heavy and medium duty trucks manufactured in China, giving it deep visibility into the
market(Cummins2015AnalystDayPresentation).Slowingtruckenginesalesstronglysuggest
that truck replacement rates are declining as the slower pace of infrastructure activity and
manufacturingreducesthemilesdriveneachyearbymuchofthetruckfleetinChina.
Heavytrucksalesreflectthistrend,astheboomperiodthatsawnewvehiclesalesrisefrom237
thousandheavy-dutytrucksin2005,toapeakofmorethanonemillionin2010(Exhibit10).New
heavytrucksaleshavenowfallenbackto2008levels,with551thousandvehiclessoldin2015,
accordingtoindustrydata.ThefactthatChina’snationwidedieselfuelconsumptiontaperedoff
in2014and2015despitenearly1.3millionnewheavytrucksenteringthefleetsuggeststrucking
activityhasslowedsubstantiallyasindustrialactivitylevelsdecline.
18
Cummins4Q2015EarningsCallTranscript,SeekingAlpha,http://seekingalpha.com/article/3867056-cumminscmi-n-thomas-linebarger-q4-2015-results-earnings-call-transcript?part=single
17
Exhibit10:ChinaNewHeavyTruckSalesvs.NationalDieselFuelDemand
Thousandtrucks,lefty-axis;Thousandbarrelsperdayofdieselfuel,righty-axis.
Source:CAAM,NBSChina,TruckIndustryResearch,Ltd.
China’struckfleetisalsomovingtofavorsmallervehiclesthatuselessfuel.Between2005and
2014,“lightdutytrucks”steadilytookfleetsharefrommediumdutytrucks,whileheavytrucks’
fleet share stagnated (Exhibit 11). Between 2009 and 2014, heavy trucks’ share of the fleet
hovered at approximately 25%, while light trucks have steadily gained share over the past
decade,climbingfrom51%in2005,to57%in2010,tomorethan65%in2014,accordingtoNBS
data.
Chineseheavytruckstypicallyuseasmallerdisplacementengine(e.g.6to10liters)thanthose
usedbytruckssoldintheU.S.andEuropeanmarkets(10to15L).19Nonetheless,theirengines
aremanytimeslargerthanthoseusedinthelightdutytrucksthathavegrabbedfleetsharefrom
heavy and medium-duty machines. For instance, a typical light duty diesel-powered truck in
China has an engine displacement of 2.0-to-2.5 liters—as little as 1/5 that of a heavy duty
19
GabeCollinsandAndrewErickson,“CountingtheBarrels:Heavytrucksalesareabetterbarometerthancar
salesofChina’soildemandgrowth,”ChinaSignPost™(洞察中国),No.20(31January2011).
http://www.chinasignpost.com/2011/01/31/counting-the-barrels-heavy-truck-sales-are-a-better-barometer-ofchinas-oil-demand-growth-than-car-sales/
18
vehicle.20Furthermore,thelightdutytrucksarelesslikelytobemakingthelonger-distancehauls
typicallymadebylargervehicles,meaningthattheynotonlyuselessfuelperkilometerthan
theirlargercousins,butarealsodrivingfewerkilometerspervehicleoveragiventimeperiod.
Each of these dynamics reduces diesel fuel consumption from the trucking sector, which is
China’sprimarydieselfueldemanddriveratpresent.
Exhibit12:ChinaTruckFleetbyTruckClass
Percentageoftotalfleet
Source:NBSChina,Author’sAnalysis
Reason two: increasing concerns over diesel engine emissions. Volkswagen’s diesel exhaust
emissionsscandalcrystallizedconcernsthathavebeenbuildingforsometimeconcerningdiesel
engineexhaustemissions.Mostpointedly,datasuggestthatevenifdieselengineshavenotbeen
riggedtogameemissionstests,theystillposesignificanthealthrisksduetotheirpropensityto
emit small particulates and nitrogen oxides. Chinese policymakers are almost certainly taking
note of the air quality problems afflicting many European capitals following many countries’
strategicdecisioninthe1990stoincreasedieselengineuseinpassengercarsasawayofreducing
carbon dioxide emissions. France, for instance, began moving in mid-2015 to retract diesel
20
“TakingaComprehensiveLookat2.5LiterDieselEngines,”(结构参数⼀起看细说那些2.5升柴油机),
http://www.360che.com/tech/160314/52811_all.html
19
friendlypoliciesandreportedlyplanstobandieselvehiclesmanufacturedbefore2011fromParis
by2020.21
WorldHealthOrganizationdataonPM2.5particulateemissionssupportEuropeanauthorities’
concerns.ThelatestannualreadingsshowedthatParishadanannualmeanPM2.5levelof17
microgramspercubicmeterofair(Exhibit13).Thisfigureisnearly24%higherthanthereading
forcar-chokedHouston,whichhasnearly2.5timestherateofpercapitavehicleownershipas
Paris and also has substantial petrochemical and manufacturing activity within the same air
pollution sampling zone.22 The key difference is that most passenger vehicles in Houston are
gasoline-powered,whiletheoverallFrenchpassengercarfleetisnearly80%diesel-propelled.
LessCO2,butmuchmorePM2.5permiledriven.
Exhibit13:DieselExhaustEmissionsCauseSeriousParticulatePollution
MicrogramsofPM2.5particulatespercubicmeterofair,2014
Source:WHO,Author’sAnalysis
21
“AutomakersslamFrenchdecisiononcleandiesels,”AutomotiveNewsEurope,26June2015,
http://europe.autonews.com/article/20150626/ANE/150629983/automakers-slam-french-decision-on-cleandiesels
22
HarrisCountyVehicleRegistrationData,http://www.txdmv.gov/reports-and-data/publicationsreports/cat_view/13-publications/25-reports-data/65-veh;
http://www.harriscountytx.gov/CmpDocuments/74/Budget/FY16%20Population%20Study.pdf;Nombrede
VoituresenFrance,http://www.statistiques-mondiales.com/france_voitures.htm(lastaccessedon30July2016)
20
Beijing,China’scarownershipcapital,exemplifieswhyChinesepolicymakerswilllikelymaintain
a jaundiced eye toward any attempts to increase the use of diesel-powered vehicles in the
country.Gasoline-poweredcarscurrentlydominatepassengercarsales(andintheoverallfleet).
Despitethefueleconomybenefitsofsmall,moderndieselengines,policymakersarelikelyto
shun them due to emissions concerns. Reinforcing the likely role of autos as a driver of air
pollutionproblems,since2009,Beijinghascutitscoalconsumptioninhalf,yetPM2.5pollution
asmeasuredbytheUSEmbassyremainssevereandhasdeclinedmuchmoreslowlythanwould
beexpectedgivenplummetinglocalcoaluse(Exhibit14).
Exhibit14:Beijing’sParticulatePollutionProblemsRemainDespiteFallingCoalUse
PM2.5levelinmicrogramspercubicmeterofaironleftverticalaxis,coaluseinmilliontonnes
peryearonrightverticalaxis
Source:USStateDepartment,Author’sAnalysis
There are two possible explanations for Beijing’s persistent PM2.5 pollution problems: (1)
pollutionfromsurroundingprovincescontinuestodriftintoBeijingor(2)motorvehicletailpipes
areinfactthecoresourceofBeijing’sPM2.5emissionsproblems.Thelikelyansweristhatboth
areresponsible,butthatthedataaffirmmotorvehicles’responsibilityforasubstantialportion
ofBeijing’sPM2.5problems.
Officials in Hebei Province, which is China’s largest steel producer and surrounds the Beijing
municipality, say they are eliminating millions of tonnes of output capacity across emissions21
intensiveheavyindustrialsectorsincludingsteel,cement,andplateglass.23Yetproductiondata
forhigh-emissionscommoditiesshowamoremixedpicture.LocaldatashowthatHebei’scement
production declined nearly 15% YoY in 2015, but steel products output rose 5.5% YoY.24
Meanwhile,NationalBureauofStatisticsdatashowthatHebei’selectricityproductiondeclined
veryslightlyYoYin2015,butralliedsignificantlyinthepastfourmonthsoftheyear.
Uncertainty aside, aggregate heavy industrial activity in Hebei is likely on the low side of the
normalrangethathasdefinedthepastseveralyears.Moreover,Beijinghasachieveddramatic
reductionsinlocalcoalburning,whichhasbeenreplacedtoalargedegreebycleannaturalgas.
Such a backdrop suggests that if industry were the primary culprit, PM2.5 levels should have
declinedmoresignificantlythantheyhave.Yetthedatainsteadappeartovalidatetheviewthat
motorvehicleemissionsareindeedakeysourceofPM2.5emissionsinBeijing.
Eventhoughthebulkofthefleetthereisgasoline-powered,thefactthatairpollutionremainsa
persistentproblemanddieselvehiclesclearlydriveseriousairqualityissuesinEuropeancities
suggestspolicymakershereandelsewhereinChinawillrestrictthesaleanduseofdieselvehicles
moreseverely.Indeed,theEuropeancitiesassessedabovehavelittleheavyindustrialactivity
anddonotburncoalforpower,yethaveseriousparticulateemissionsproblems,forwhichdiesel
vehiclesappeartobetheprimaryculprit.
B. Transportation Theme 2: Rising Gasoline Demand Means China’s Crude Oil Use is
DecouplingfromtheIndustrialEconomy
GasolinehastakenoverasthemaindriverofincrementaloilproductsdemandgrowthinChina.
This ties China’s crude oil appetite more tightly to consumer trends, as opposed to industrial
growth. In the Chinese market, it takes approximately 19 passenger cars to equal the annual
crudeoildemandcreatedbyoneheavy-dutydieseltruck.25In2015,551,000newheavytrucks
23
“Pollution control slashes GDP growth in Hebei,” China Daily, 8 January 2015,
http://www.chinadaily.com.cn/china/2015-01/08/content_19276430.htm
24
“Hebei 2015 Economic and Social Development Report” (河北省2015年国民经济和社会发展统计公报发布),
Great Wall Network, 1 March 2016, http://news.hebei.com.cn/system/2016/03/01/016694836_01.shtml
25
Thisestimateusesthefollowingassumptions:(1)Fueluse—EachheavytruckinChinaconsumes44litersof
dieselfuelper100km(BenSharpeandRachelMuncrief,“LiteratureReview:Real-WorldFuelConsumptionOf
Heavy-DutyVehiclesInTheUnitedStates,China,AndTheEuropeanUnion,”TheICCT,WhitePaper,January2015,
http://www.theicct.org/sites/default/files/publications/ICCT_HDV_FC_lit-review_20150209.pdf)whileChina’s
passengercarfleethadanaveragefueleconomyof7.12liters/100kmin2015(“ChinaPassengerVehicleFuel
Consumption,”2015DevelopmentAnnualReport,TheInnovationCenterforEnergyandTransportation
December2015,http://www.icet.org.cn/english/admin/upload/2015120241280305.pdf.)(2)Annualvehiclekm
travelled—52,000kmperyearforheavytrucks(H.Huoetal./EnergyPolicy43(2012)6–16(10),
http://www.tsinghua.edu.cn/publish/ess/7835/20120719144509070242457/[8]%20Huo_EP_2012.pdf)and13,000
km/yearforpassengercars(He,K.,Huo,H.,Zhang,Q.,He,D.,An,F.,Wang,M.,Walsh,M.,2005,“OilConsumption
andCO2EmissionsinChina’sRoadTransport:CurrentStatus,FutureTrends,andPolicyImplications,”Energy
Policy33(2005):1499–1507).Dieselandgasolineyieldsperunitofcrudeoilrunthroughrefineryarederivedfrom
Sinopec’s2015Form20-Freport.
22
weresoldinChina,versus21.2millionnewpassengercars—a38:1ratio,meaningthatpassenger
carsandtheirgasolinethirstarenowdecisivelydrivingincrementaloildemandgrowthinChina.
For the next three years, each million new passenger cars sold in China will likely create
approximately10kbdofincrementalgasolinedemand.Thisisworthapproximately3.8barrels
ofcrudethroughputina“typical”Chineserefinery.26Beyond2018-19,higherscrappingratesand
technologicalchangesinvehiclepowertrainsthatincreasefuelefficiencywilllikelyreducethe
incrementalfueldemandimpactofeachnewcarsoldinChina.Therisingimportanceofgasoline
demandandthedeclineindieselusereflectabroadstructuralshift:crudeoilisrapidlybecoming
aconsumergoodinChina,notanindustrialcommodityasitwasduringthehalcyonyearsof
burgeoningdieselfueldemand.
Between2003and2012,China’scrudeoildemandusedtobedrivenprimarilybydieselfuelfrom
thetrucks,bulldozers,andgeneratorsthatpoweredthecountry’sindustrialboom.Yetasheavy
industrialactivityslowedandgasolinebecameamoreimportantsourceofoildemandgrowth,
crudeoilstartedbecomingmoreofaconsumergood,asopposedtoanindustrialone.Thisshift
willprofoundlyinfluenceglobalcrudeoilandrefinedproductsmarkets.
China’s gasoline demand upshifted dramatically in recent years, as passenger car sales
skyrocketed. In early 2012, the ratio of gasoline consumption to diesel consumption in China
begin to rise noticeably (Exhibit 15). This accelerated dramatically in 2015 as gasoline
consumption continued to grow while diesel demand contracted slightly. The ratio currently
stands at approximately 0.94 barrels of gasoline consumed per barrel of diesel. Notably, in
January,February,andMarchof2016,Chinaactuallyconsumedmoregasolinethandieselfuel,
adevelopmentthatlikelyportendsagasoline-dominatedfutureintheChinarefinedproducts
market.
26
SinopecisChina'slargestoilrefinerbyvolume.Sinopecrefinedapproximately236milliontonnesofcrude(1.726
billionbblat7.3bbl/tonne)in2015andproducedapproximately54milliontonnesofgasoline(0.458billionbblat
8.5bbl/tonne).Thisratioofapproximately0.265bblofgasolineperbblofcrudesuggeststhatthe"typical'Chinese
refineryrunsabout3.8barrelsofcrudeperbarrelofgasolineproduced.
23
Exhibit15:RatioofGasolinetoDieselDemandinChina
Source:JODI,Author’sAnalysis
GasolinehasalsosteadilygobbledupshareinChina’stotalpetroleumliquidfuelsandtotaloil
productsmarkets(Exhibit16).Amongpetroleum-derivedliquidfuels,gasoline’ssharerosefrom
26%inJanuary2004tomorethan40%byMarch2016,withthegrowthheavilybiasedtoward
thepastfouryears.Notethatduringthissametimeframe,gasolinegainedsharewhileoverall
petroleum-derivedliquidfuelsdemandroseby68%.Thefactthatgasolinegainedamajorchunk
of market share amidst rapid overall oil products demand growth shows just how dramatic
China’s gasoline transformation has been. The analysis uses a 12-month moving average to
smoothseasonaldemandspikesandmakelong-termtrendsmorevisuallyapparent.
24
Exhibit16:GasolineasaPercentageofTotalOilProductsandLiquidFuelDemandinChina
Source:JODI,Author’sAnalysis
1. PassengerCarSalesAreDrivingGasolineDemand
Gasoline-ledoildemandgrowthstemsfromyearsofrobustpassengercarsalesgrowthatarate
andscalenotseensincetheautomobilefirsttooktheUnitedStatesmarketbystormintheearly
20thCentury.In2009,whentheChinapersonalcarboomtookoff,thecountrysold10.3million
units, a big jump from the 6.7 million cars sold in 2008, according to data from the China
Association of Automobile Manufacturers (CAAM). By 2015, 21.2 million passenger cars were
soldinChina—significantlymorethanthe17.5millionlightcarsandtruckssoldintheUS,long
theworld’sdominantpassengervehiclemarket.27ScarcelyanypassengercarsinChinaaredieselpowered.28
Newcarsalesnumberscloselytracktherapidincreaseinpassenger-haulingvehicles’proportion
oftheoverallfleetinChina.Becausethecarsthathaveenteredthefleetduringtherecentcar
buyingspreestillhavemanyyearsofserviceleftinthem,scrappingratesremainlow,andthe
relationshipbetweennewcarssoldandthenumberofcarsenteringthevehiclefleetonanet
basisismuchclosertoone-to-onethanitwouldbeinamorematuremarket.
Supporting that thesis, the absolute and relative growth rates in the number of passengercarryingvehicleshasrisensharplysince2009,whichcorrelatesstronglywithnewpassengercar
sales(Exhibit17).Theratioofnewpassengercarsalestonetadditionsinthepassenger-carrying
vehiclefleetbetween2009and2013hoveredbetween0.98and1.11,whichsuggeststhaton
thewhole,morethan9ofevery10newcarssoldrepresentedanetadditiontothepassenger
vehiclefleet.
27
MikeSpector,JeffBennett,andJohnD.Stoll,“U.S.CarSalesSetRecordin2015,”TheWallStreetJournal,5
January2016,http://www.wsj.com/articles/u-s-car-sales-poised-for-their-best-month-ever-1451999939.
28
“WhyVolkswagen’sProblemsDon’tIncludeChina,”ChinaRealTimeReport,WallStreetJournal,28September
2015,http://blogs.wsj.com/chinarealtime/2015/09/28/why-volkswagens-problems-dont-include-china/.
25
Exhibit17:PassengerCarryingVehiclesvs.CargoCarryingVehiclesinChina’sCivilianVehicle
Fleet
Millionvehicles
Source:NBSChina,Author’sAnalysis
2. China’sGasolineDemandOutlookFor2016-2018
Threefundamentalfactorsdrivegasolineuse:vehiclefleetsize,howthefleetisdriven,andwhat
typeofpowertrainoptionsareavailable.Industry,government,andprivateanalystsholdarange
ofviewsonChina’sgasolinedemandoutlook.Atthebullishendofthespectrum,CreditSuisse
believesthatChina’scrudeoildemandwillgrowata4%annualratein2016and2017(i.e.,more
than415kbdeachyear).29ThebankbelievesgasolinedemandwillbetheprimemoverofChinese
oil demand growth, with its baseline forecast projecting gasoline demand will increase by
between280kbdand360kbdinboth2016and2017.30ChinesestateoilgiantCNPC’sin-house
teamforecastsa4.3%increaseincrudeoildemandin2016.31
29
“ConnectionsSeries:Oiland…ChineseConsumers,”CreditSuisse,19November2015,GlobalEquityResearch.
Ibid.
31
“ChinaOilDemandtoGrow4.3Percentin2016:CNPCResearch,”Reuters,26January2016,
http://www.cnbc.com/2016/01/26/china-oil-demand-to-grow-43-percent-in-2016-cnpc-research.html.
30
26
Credit Suisse was much more bullish than the November 2015 IEA Oil Market Report—the
“consensusview”—thatcalledforChinesegasolinedemandtorisebyonly110kbdin2016.32
The IEA analytical staff became much more bullish in late 2015, as the December 2015 OMR
basically doubled the agency’s 2016 China gasoline demand forecast. Subsequent IEA OMRs
continuedincreasingtheoutlook,whichbyAprilstoodat228kbd.33ButbeginninginMay2016,
theIEAbegindramaticallytrimmingitsChinagasolinedemandforecastasitrealizedthatmany
barrels of gasoline formerly counted as demand were actually being exported and that real
demandwasinfactnotnearlyasrobustaswidelythought(Exhibit18).
Exhibit18:IEA’s2016ChinaGasolineDemandForecastTrimmedDramatically
Projected2016gasolinedemandgrowth,kbd
Source:CreditSuisse,IEAOilMarketReport,Author’sAnalysis
Toputdemandestimatesinperspective,considertheadditionalimpliedgasolinedemandper
millionnewcarssoldinChina.Essentially,becausenewcarsinChinacorrelatetonetpassenger
vehiclefleetgrowthathighratesbecausescrappageremainslow(seeExhibit18,above),dividing
annual incremental gasoline consumption growth by the number of new cars sold yields an
32
33
IEAOilMarketReport,13November2015,https://www.iea.org/oilmarketreport/reports/2015/1115/#Demand
IEAOilMarketReport,16April2016,https://www.iea.org/oilmarketreport/reports/2016/0416/#Demand
27
approximate“ruleofthumb”forhowmanybarrelsofgasolineperdayeachmillionnewcarssold
inChinaarelikelytoconsume.34
Thepastfiveyearsofcarsalesandgasolineconsumptionlevelsshowsthateachmillionnewcars
soldyieldsanaverageof11kbdinadditionalgasolinedemand(Exhibit19).Thecurvehastrended
downwardandtheIEAdatasuggest2016couldseerealgasolinedemandgrowthinChinasharply
slow.Basedonthesedata,theauthorprojectsthatChinesegasolinedemandwillgrowby75-to120kbdin2016,2017,and2018.Thehighgrowthcasewouldseegasolinedemandgrowby180to-200kbdineachofthoseyears.AtSinopec’s2015refinerygasolineyieldrateof26.5%,Chinese
refinerswouldneedtoprocessapproximately280-to-450kbdofcrudeoiltoproducethebase
casegasolinevolumeand680-to-750kbdofcrudeoilunderthehighcasescenario.35
Exhibit19:ChinaImpliedGasolineDemandGrowthPerMillionNewCarsSold
Source:CAAM,NBSChina,Author’sAnalysis
34
Fororiginalstudy,seeGabrielB.CollinsandAndrewS.Erickson,“GasolineAlley:HowMuchGasolineDemand
AreEachMillionCarsSoldinChinaWorth?”ChinaSignPost™(洞察中国)82(3November2014),
http://www.chinasignpost.com/2014/11/03/gasoline-alley-how-much-gasoline-demand-are-each-million-carssold-in-china-worth/.
35
Sinopec2015Form20-F,
http://english.sinopec.com/download_center/reports/2015/20160425/download/2016042502.PDF.
28
3. GasolineDemandOutlookandImplications
GasolinedemandinChinaislikelytoincreaseby75-to-120kbdin2016and2017.Beyondthat
point, the confluence of vehicle powertrain innovation and other factors—particularly
congestiononChineseroadways—willlikelybegintotrimincrementalgasolinedemandgrowth.
ThisisimportantbecauseChina’sdieseldemandhasverylikelyalreadyreachedapermanent
peak,whichstandstomakegasolinethekeydriverofadditionalcrudeoildemand.36
ThechairmanofSinopec,China’slargestoilrefiner,believesthecountry’sgasolinedemandwill
peakaround2025,whileExxonMobil’slatestChinaenergyoutlookforecaststhatpassengercar
fueldemand(i.e.primarilygasoline)willleveloffaround2030.37Asgasolinedemandcontinues
tode-couplefromheavyindustrialindicators(Exhibit20)andcomprisealargershareofChina’s
overalloilproductsdemand,thingslikepersonalairtraveldataandspendingonmealsoutside
thehomewilllikelybecomebetterleadingindicatorsofgasolinedemandthanthetraditional
measuressuchaselectricitythatwouldhavehistoricallymorecloselytrackedChineseoilproduct
consumption.
36
GabrielCollins,“ChinaPeakDieselPosesaSeriousChallengetoSaudiArabia,MayHelpForceOPECProduction
Cut,”ChinaSignPost™(洞察中国)94(5April2016).http://www.chinasignpost.com/2016/04/05/china-peakdiesel-poses-a-serious-challenge-to-saudi-arabia-may-help-force-opec-production-cut/.
37
TimothyCoulter,“China’sFuelDemandtoPeakSoonerThanOilGiantsExpect,”Bloomberg,1April2015,
http://www.bloomberg.com/news/articles/2015-04-01/china-s-fuel-demand-to-peak-sooner-than-oil-giantsexpect;“TheOutlookforEnergy:AViewto2040,”ExxonMobil,24,
http://cdn.exxonmobil.com/~/media/global/files/outlook-for-energy/2016/2016-outlook-for-energy.pdf.
29
Exhibit20:ChinaGasolineDemandvs.ConsumerandIndustrialActivityIndicators
Index,April2014=100
Source:NBSChina,Author’sAnalysis
4. GasolineLikelytoRemainChina’sDominantCarFuelUntilAtLeast2025,ButFleet
isBecomingMuchMoreEfficient
GasolinewillbeChina’sdominantpassengercarfuelforthenextdecade,andlikelybeyond,due
togasoline’ssuperiorpowerdensityandlogisticaladvantagesrelativetopureelectriccarsand
competing liquid fuels. But the gasoline will be burned more efficiently. Gasoline motors are
beingdownsizedandturbochargedand/orlinkedwithelectricpropulsionsystemstoformmore
efficient gasoline-electric drivetrains that wring out significantly more miles per gallon while
delivering high performance that makes driving fun. Lower kilometers driven per vehicle,
drivetrain technology improvements, and a gradual increase in scrapping rates mean that by
2018,salesofnewcarsinChinawilllikelyneedtobeatleast10-to-15%higherthantheywerein
2015simplytosustainthe2015incrementalgasolinedemandgrowthrateofapproximately200
kbd.
DespitethehighconfidenceexpressedbytheChinagasolinebulls,threecorestructuralfactors
deserveclosescrutinygiventheirdeepinfluenceoverhowcarsare—andwillbeabletobe—
driveninChinanowandincomingyears.Thesefactorsare:(1)trafficproblems,(2)potential
changes in car owner sentiment and behavior driven by congestion and other issues, and (3)
30
improvements in fuel efficiency from rising penetration of more advanced drivetrain
technologies.Theinterplayofthesefactorsraisestheriskthatgasolinedemandgrowthineach
ofthenextthreeyearscouldfallbelowthebasecaseof75-to-120kbdthattheauthorcurrently
forecasts.
A. TrafficTroubles
China’s grinding gridlock continues to worsen in many cities as car sales outpace highway
construction.Itisnotyetfullyclearwhatthe“saturation”pointisforvehicleownership,butthat
tippingpointisultimatelymuchmorelikelytobedrivenbyalackofroadspacethanbyanyper
capita car ownership threshold. The data from Beijing—China’s most mature car market—
suggest that at around 200 cars per km of roadway, large Chinese cities will begin to see
ownershipratesleveloff(Exhibit21).
Exhibit21:PrivatePassengerCarsPerKMofRoadintheBeijingandChengduMetroAreas
Source:LocalStatisticalBureaus,NBSChina
Urban consumers dominate the Chinese car market thus far, and so the national car density
numbersaremisleading,becausetheyfactorinruralareaswheremanyresidentsstillcannot
affordcars.AstheusedcarmarketgrowsinChina,itispossiblethatruralareascouldbecomea
moresignificantsourceoffueldemand.Butatthispointandforthenextfiveyears,theproper
31
focuswillbeoncities,asTier1and2marketsreachsaturationandcarsellersincreasinglylook
toTier3-6marketstopropupsales.38
Tobesure,Beijinghasimposedsubstantialrestrictionsoncarpurchases—forinstance,setting
anannualquotaofcarsthatcanbesoldandissuinglicenseplatesviaalottery—aswellasactual
useofprivatepassengervehiclesinthecity.39Yetthetrendofroadwaysaturationplateauingin
the200carsperkmrangeislikelyfundamentallyrootedinphysicalconstraints,asopposedto
administrativeones.
Suchfactorsincludelowroaddensityrelativetopopulatedarea.Forinstance,aSeptember2013
studybySingapore’sLandTransportAuthorityrevealsthatBeijinghadonly1.7kmofroadway
persquarekilometeroflandarea.40ContrastthiswithSeoulandTokyo,whichhad13.5km/km2
ofroadand19.1km/km2,respectively.41NewYorkhas12.9kmofroadpersquarekilometer.42
HarrisCounty,Texashas263passengervehiclesperkmofroad—30%higherthanBeijing—but
aroaddensityofapproximately2.8km/km^2,or65%higherthanBeijing’s.43
B. DataShowWorld-ClassCarSales…AndWorld-ClassCongestion
GranulardatarevealjusthowdeepChina’strafficcongestionproblemsare.TomTom,aleading
globalnavigationalaidmanufacturer,usesanhistoricaldatabasewith14trillionhistoricaltravel
timemeasurementsthatiscontinuallyupdatedandmeasurescongestionontheroadnetworks
of 295 cities in 50 countries around the world.44 The index is expressed in the form of a
percentagenumberthatreflectstheincreaseintraveltimeascomparedtoanuncongested“free
flow”trafficsituation.
TomTom’s2016TrafficIndexyieldstwoimportantinsights.First,China’slargecitiesarealready
amongsomeoftheworld’sworstintermsoftrafficcongestion.Forinstance,oftheworld’s25
most-congestedcitiesin2015accordingtotheTomTomdata,fivewerelocatedinChina(Exhibit
22).
38
GabrielB.CollinsandAndrewS.Erickson,“GasolineAlley:HowMuchGasolineDemandAreEachMillionCars
SoldinChinaWorth?”ChinaSignPost™(洞察中国)82(3November2014),
http://www.chinasignpost.com/2014/11/03/gasoline-alley-how-much-gasoline-demand-are-each-million-carssold-in-china-worth/
39
“BeijingSaidtoExpandRestrictionsonVehicleUsetoFightSmog,”Bloomberg,8January2016,
http://www.bloomberg.com/news/articles/2016-01-08/beijing-said-to-expand-restrictions-on-vehicle-use-to-fightsmog
40
PanDi,“KeyTransportStatisticsofWorldCities,”Journeys,September2013,
http://www.lta.gov.sg/ltaacademy/doc/13Sep105-Pan_KeyTransportStatistics.pdf
41
Ibid.
42
Ibid.
43
TexasDMVRegistrationDatabyCounty,CY2014,http://www.txdmv.gov/reports-and-data/cat_view/13publications/25-reports-data/65-vehicle-titles-registration/229-number-of-vehicles-registered;“HarrisCounty
Profile,”TheCountyInformationProgram,TexasAssociationofCounties,
http://www.txcip.org/tac/census/profile.php?FIPS=48201(accessedon20May2016)
44
“TomTomTrafficIndex,MeasuringCongestionWorldwide,”2015,http://www.tomtom.com/en_gb/trafficindex/
32
Exhibit22:TomTomGlobalCitiesRoadCongestionRatings
Increaseintraveltimecomparedtoafreeflowroadconditionwithnocongestion
Source:TomTom,Author’sAnalysis
Second, congestion worsened in 14 of the 22 Chinese cities on the list, suggesting that the
country’strafficstruggleswilllikelybecomemoreseriousdespiteaslowingeconomyascarsales
outpaceurbanroadconstruction.TheTomTomdatareadlikea“who’swho”oflargeandgrowing
Chinese car markets. Key cities include: Beijing, Chengdu, Chongqing, Guangzhou, Nanjing,
Tianjin,andWuhan,aswellasemergingcarmarketssuchasChangsha,Ningbo,Quanzhou,and
Shijiazhuang.Inshort,congestionissuesarenotonlyplaguingthepowerhousecoastalvehicle
markets—theyarealsoaffectingkeyinlanddemandhubs.
Addingmorecolor,21ofthe22Chinesecitiesonthelisthadworseaveragecongestionreadings
thanthecityofHouston,Texas,whichislegendaryforitstrafficissues.The10mostcongested
ChinesecitiesintheTomTomdatasetaccountedforroughlyoneineveryfivepassengervehicles
ownedinChinaatyear-end2014,somanyChinesedriversarealreadyinthethickofserious
trafficjamsthatarenotlikelytoimproveinthenextfewyears.
33
C. HowMightDriverandCarBuyerSentimentsandBehaviorbeChanginginChina?
Chinagasolinemarketanalystsfacethecriticallyimportantquestion:howmuchdoChinesecar
owners actually drive? The question is a complicated one in practice because the Chinese
government does not report vehicle miles travelled (“VMT”) data as, for instance, the US
DepartmentofTransportationdoes.45ResearchersinChinahaveworkedtobridgethisgap,using
twoprimarymethodologies.
ThefirstentailsusingregularlyreportedtotalfreightorpassengertrafficvolumetoestimateVMT
pervehicle.46ThesecondmethodestimatedtherelationshipbetweenVMTandvehicleageand
appliedthistothevehicleagedistributionoftheChinesevehiclefleetinaparticularyear.47An
analysisfromArgonneNationalLaboratorystatesthatinthe2001-2004period,theannualVMT
ofcarsinChina’scitieswasbetween24,000and27,000km.48
This number has likely fallen significantly in recent years, because the early estimates were
inflatedbythelargeproportionoftaxisinChina’scarfleet—16.4%ofvehiclesin2002and10.1%
in2004.49TheArgonneresearchersprojectthatby2030,theaverageVMTofChinesepassenger
carswillbe13,000km—afiguremorecloselyalignedwithactualdatafromJapanandWestern
Europeinthelate1990s.50
PrivatelyownedcarsarecomingtodominateChina’spassengercarfleetandannualVMTpercar
continuestoregresstowardthe10,000-13,000kmlevel.Trafficcongestionlikethatillustratedin
the TomTom data could prompt drivers to use their vehicles less, a potentially deeply
consequential factor for China’s future gasoline consumption trajectory. A 2015 study by
consultancyBainandCompanyshowedthatnearlyathirdofcurrentcarownerssurveyedmight
getridoftheirvehiclesiftrafficcongestionbecame“significantlyworse,”whilesuchacongestion
levelwouldpromptnearly30%ofprospectivecarbuyerstorefrainfrompurchasingavehicle.51
Baincanvassedareasonablylargesamplesize,surveyingmorethan2,100respondentsinsixTier1andTier-2cities.
45
C.Y.CynthiaLin,Jieyin(Jean)Zeng,“TheElasticityofDemandforGasolineinChina,”DepartmentofAgricultural
andResourceEconomics,UCDavis,http://www.des.ucdavis.edu/faculty/lin/China_gasoline_elasticity_paper.pdf
(25)
46
He,K.,Huo,H.,Zhang,Q.,He,D.,An,F.,Wang,M.,Walsh,M.,2005.OilconsumptionandCO2emissionsin
China’sroadtransport:currentstatus,futuretrends,andpolicyimplications.EnergyPolicy,33(2005)1499–1507
47
Lin,X.,Tang,D.,Ding,Y.,Yin,H.,Ji,J.,2009.StudyontheDistributionofVehicleMileageTravelledinChina.
ResearchofEnvironmentalSciences,22(3).;SeealsoOliveret.al,“In-useVehicleEmissionsinChina:
BeijingStudy,”BelferCenterforScienceandInternationalAffairs,HarvardUniversity,March2009,
http://belfercenter.ksg.harvard.edu/files/2009-05-ETIP-Oliver-et-al-2.pdf
48
M.Wang,D.He,H.Huo,andL.Johnson,“ProjectionofChineseMotorVehicleGrowth,OilDemand,andCO2
Emissionsthrough2050,”December2006,ArgonneNationalLaboratory,https://greet.es.anl.gov/files/rwdz78ca
49
Id.
50
Id.
51
“ChinaNewMobilityStudy:2015,”Bain&Company,
http://www.bain.com/Images/BAIN_DIGEST_China_new_mobility.pdf
34
The survey’s question prompt relies on drivers’ subjective perceptions of how bad traffic is
relative to the current conditions used as a baseline. But despite that potential weakness, it
suggeststhatcontinuedgridlockcould:slownewvehiclesales,promptcurrentownerstodrive
less,andinaworstcase,inducecarownersinChina’slargestmetroareastoabandonpersonal
caruseandownershipentirely.Thiswouldbedeeplyproblematicforoilmarketssincethecore
metrosareChina’smajorcentersofgasolinedemand.
Exhibit23:PrimaryBearishFactorsThatCouldReduceCarOwnershipand/orSalesinChina
TrafficCongestion
AverageGasolinePrice
Wouldstopowningacar
Wouldrefrainfrombuyingacar
CarOwners
ProspectiveBuyers
SignificantlyWorse
32%
29%
ModeratelyWorse
16%
13%
DoesNotGetWorse
11%
5%
9.5RMB/l($5.46/gal)
31%
27%
8.5RMB/l($4.88/gal)
13%
9%
7.5RMB/l($4.31/gal)
11%
5%
Source:BainandCompany,Author’sAnalysis
Trafficproblemsarealsomuchmorelikelytoaffectcaruseandsales(andthus,gasolinedemand)
thanistheotherkeybearishfactor:gasolineprices.Congestionaffectsdrivershereandnowand
itimpactsthemeverysingletimetheydriveanywherebeyondtheirimmediateareaofresidence.
Gasolineprices,incontrastonlycausepainwhenadriverhastofillup—somethingthatlikely
happens once per week or less. Secondly, changing global oil market conditions make it very
unlikelythatgasolinepricesinChinawillrisehighenoughtonegativelyimpactcarownershipand
sales.
Bain’s survey data suggests that the “pain point” for current and prospective Chinese drivers
beginswhengasolinepricesclimbto7.5RMBaliterorhigher($4.31/gallon).Thesharpestpain
doesnothituntilpricesclimbabove8.5RMB/l($4.88/gallon).Thesethresholdsaresurprisingly
high—particularly from an American perspective—and reflect the fact that Chinese vehicle
owners are generally wealthier than average and that passenger cars are still a luxury good,
ratherthanthedailynecessitythattheyhavebecomeinotherpartsoftheworld.
ToassesstheriskthatgasolinedemandmightposetocarsalesandownershipratesinChina,the
authorplottedChina’sbaselineretailgasolinepricefor#93fuelagainstthedailyspotpriceof
Brentcrude,thebroadinternationalcrudeoilpricebenchmark.Theresultsuggeststhatitwould
takeanoilpricespikeakintothatexperiencedin2008todriveChineseretailgasolinepricesto
levelslikelytoadverselyaffectsalesandownershipofpersonalcars.
Forinstance,retailgasolinepricesinChinahavehistoricallynotapproachedthe7.5RMB/lmark
unlessBrentcrudewastradingatcloseto$120/bbl.Thetruepainthresholdof9.5RMBperliter
ofgasolinewouldlikelyrequireBrentcrudetotradecloseto$160perbarrel(Exhibit24).The
35
authorishighlycognizantthatatrulyseveregeopoliticalshock—suchaslarge-scaledisruption
ofoiltransitthroughtheStraitofHormuzbymilitaryconflict—couldtriggerpricespikesofsuch
magnitude.Yetsucheventsalsolieveryfaroutontheprobabilitycurve.
Exhibit24:ChinaRetailGasolinePricesvs.BrentCrudeDailySpotPrice
Leftaxis:Brentcrudeprice(USD/barrel),Rightaxis:Retail#90gasolineprice(RMB/liter)
Source:EIA,NDRC,ChineseMedia
The Bain survey data suggest that traffic congestion and gasoline prices will—by a significant
margin—bethestrongestinfluencesoncarpurchaseandusedecisionsinChina.Ofthethree
“secondaryfactors”surveyed,onlyone—taxi/ridesharingavailability—comescloseintermsof
potentialinfluence.Iftaxiavailabilitywere“easy,”18%ofrespondentswouldconsidergetting
ridoftheircars,while13%wouldconsiderdoingsoifitwere“easy”torentacar.
Twocoreconclusionsflowfromthisdata:first,thereisindeedsignificantmarketspaceforthe
growth of car rentals and ride sharing in China, which would explain Uber’s intense drive to
competeintheChinamarket.52Second,thereisasignificantgapintheresponsesofcarowners
andprospectivecarbuyerstothetaxiandrentalcaravailabilityquestions,aswellasthequestion
52
LeslieHook,“Saudiwealthfundtakes$3.5bnUberstake,”FinancialTimes,1June2016,
http://www.ft.com/intl/cms/s/0/a7e31c58-282c-11e6-8b18-91555f2f4fde.html#axzz4B1qWnLw7
36
aimedatgaugingwillingnesstousepublictransportinlieuofacar.Thoserespondentswhohad
notyetpurchasedacarwerenotnearlyasinfluencedbyrental/taxiavailability.
ChinaSignPost’spriorresearchsuggestsapossibleexplanation:prospectivecarbuyersseekcars
notjustforsocialstatusreasons,butforpracticalonesaswell.53Forapersonwhoseekstocarry
groceries,travelinmorecomfortthanpublictransportoffers,andhavepersonaltransportation
availableatatimeandplaceoftheirchoice,apersonalcarfitsthebill.Butoncesuchanindividual
buysthecarandbeginstofacetheburdensofcarownershipinChina—forinstance,extreme
difficultiesfindingparkingspace—theinitialallurefadesandabitofbuyer’sremorsekicksin.54
Exhibit25:SecondaryBearishFactorsThatCouldImpactCarSalesandUseinChina
Wouldstopowningacar
Wouldrefrainfrombuyingacar
CarOwners
ProspectiveBuyers
TaxiAvailability
Easy
Moderate
Difficult
18%
11%
11%
7%
5%
5%
CarRentalAccessibility
Easy
Moderate
Difficult
13%
12%
11%
6%
5%
5%
TargetedImprovement
11%
6%
IntermodalConnectedness
ComprehensiveImprovement
10%
8%
5%
5%
N/A
8%
N/A
N/A
6%
5%
DevelopmentofPublic
Transport
CostofObtainingaLocalCar
Difficult
Plate
Moderate
RemainstheSame
Source:BainandCompany,Author’sAnalysis
D. KeyFactor:EvolvingPowertrains
Over the past decade, SUVs have become increasingly popular in China and are the fastest
growingpassengervehiclesegment.DatafromJATODynamicsshowthattherapidsalesgrowth
continues,withmorethanfourmillionSUVssoldinChinaduring2014andsixmillionin2015—
accountingfornearly1/3ofallpassengervehiclessoldinthecountry.55
53
GabeCollinsandAndrewErickson,“CountingCars:RisingPrivateAutomobileOwnershipinChineseCitiesPaves
RoadforGasolineDemand,”ChinaSignPost™(洞察中国)71(24June2013),
http://www.chinasignpost.com/2013/06/23/counting-cars-rising-private-automobile-ownership-in-chinese-citiespaves-road-for-gasoline-demand/
54
GabeCollinsandAndrewErickson,“DyingforaSpot:China’sCarOwnershipGrowthisDrivingaNationalParking
SpaceShortage,”ChinaSignPost™(洞察中国)17(10January2011),
http://www.chinasignpost.com/2011/01/10/dying-for-a-spot-chinas-car-ownership-growth-is-driving-a-nationalparking-space-shortage/
55
“FOCUSONCHINA.2015MARKETOVERVIEW:IMPACTSANDOPPORTUNITIESOFCARPURCHASERESTRICTIONS,”
JATO,March2016,http://www.jato.com/wp-content/uploads/2016/04/JATO-Market-Focus-China-Car-Purchaserestrictions-and-Market-overview-2015-Final.pdf
37
SUVs’rapidpenetrationdirectlyimpactsgasolineconsumption,sincethesevehiclestendtohave
higher-displacementenginesthat,allthingsheldequal,generallyconsumemorefuelthanthe
motors used in smaller, lighter sedans. For instance, this author’s model for gasoline use per
millionvehiclesindicatesthatamillionsmalleconomycarsdrivenanaverageof13,000kmper
yearcreateapproximately12kbdofincrementalgasolinedemand.YetamillionmidsizeSUVs
suchastheHondaCR-Vuse21kbdofgasoline--75%more(Exhibit26).AmillionlargeSUV’s—
suchasthePorscheCayenne—raisesthedailygasolinedemandpermillionvehiclesfigureto31
kbd.Toputthegasolinenumbersinperspective,toproduceabarrelofgasoline,Chineserefiners
typicallymustrunapproximately3.8barrelsofcrudeoil.Thismeansthattosupplygasolinefor
each additional million SUVs entering the fleet, the Chinese refining sector will likely need to
process80-to-120kbdofcrudeoil.
Exhibit26:EstimatedIncrementalGasolineDemandPerMillionVehicles,kbd
Source:CompanyReports,Author’sAnalysis
Strong SUV sales may make it difficult for China to meet its ambitious national fuel economy
standards,whichcallforanaverage(i.e.allvehiclesoldandnew)fleetfueluseofnearly48miles
pergallon(“mpg”)by2020.ChinareleasedPhaseIVfuelconsumptionstandardsforpassenger
38
vehicles(GB19578-2014)inDecember2014andtherulestookeffectonJanuary1,2016.56The
PhaseIVstandardsuseacurb-weightindextoestablishmaximumfuelconsumptionlimitsand
determine compliance based on the corporate-average fuel consumption (CAFC) of
manufacturersinagivenyear.Therearetwosetsoflimitvalues:oneforregularcarsandanother
forspecialcars.
To put China’s 2020 fuel economy standards in perspective, U.S. regulators seek a fleet-wide
standardof38.9mpgforlight-dutyvehiclesand44.8mpgforcars.57Thedifficultpartisthatasa
localcarculturetakeshold,Chineseconsumersincreasinglyseekvehiclesthatarefuel-efficient
butalsopowerfulandfuntodrive.Forinstance,in2014,HoneywellTurboTechnologiessurveyed
visitorstoauto.163.comandyangche51.com,twoofChina’smostpopularautomotivenewsand
information portals (combined daily traffic of 340,000 during the survey period) to gauge
consumerexpectationsofnewvehiclescomingtomarket.58
Only 1/3 of respondents were satisfied with their current vehicle’s fuel economy and power,
while80percentwouldseekfueleconomysavingsofatleast10to30percenttobesatisfied
with their next vehicle purchase. Furthermore, 94% of respondents indicated engine
performancewasan“important”or“veryimportant”factorforpurchasingavehicleand68%of
respondentssaidtheywouldprefertobuyasmallercarifitcouldstillperformlikeavehiclewith
alargerengine.59
Results such as these suggest Chinese drivers may follow a path at least somewhat akin to
American drivers, who place a significant premium on powerful vehicles. In the U.S. market,
manufacturers responded strongly to drivers’ dual demand for miserly mileage and generous
horsepower,asfleet-wideaveragefueleconomydoubledbetween1975and2014andaverage
horsepowerpervehiclerosenearly50%(Exhibit27).
56
“China:LightDuty-FuelConsumption,”transportpolicy.net,
http://www.transportpolicy.net/index.php?title=China:_Light-duty:_Fuel_Consumption.
57
“Light-dutyvehicle”meansapassengercarorpassengercarderivativecapableofseating12passengersorless.
42U.S.C.7401https://www.gpo.gov/fdsys/pkg/CFR-2011-title40-vol19/pdf/CFR-2011-title40-vol19-sec86-180301.pdf.
58
“EightyPercentofChineseDriversSeekFuelEconomyImprovementswithNextVehiclePurchase,SurveySays,”
Honeywell,2July2014,https://turbo.honeywell.com/whats-new-in-turbo/press-release/eighty-percent-ofchinese-drivers-seek-fuel-economy-improvements-with-next-vehicle-purchase-survey-says/.
59
Shrinkingvehicledimensionswhilepreservingperformancemayalsobedrivenbythescarcityofparkingspace
inChina.See,forinstance,GabeCollinsandAndrewErickson,“DyingforaSpot:China’scarownershipgrowthis
drivinganationalparkingspaceshortage,”ChinaSignPost™(洞察中国),No.17(10January2011).
39
Exhibit27:FuelEconomyandHorsepowerBothImprovedDramaticallyinUSCarFleetOver
TheLast40Years
Averagehorsepowerpervehicle(leftaxis),averagemilespergallon(rightaxis)
Source:NHTSA,Author’sAnalysis
The U.S. auto sector’s ability to balance fuel economy and power output demands has been
remarkableandcontrastssharplywiththeEUvehiclefleet,wherefueleconomyclearlywonout
overpower.Indeed,comparing2011EUpassengervehiclefleetdata(thelatestavailable)tothe
2010 numbers from the US showed that the typical European passenger car got 57% better
mileagethanitsUSequivalent,buthad61%lesshorsepower.60Thechiefdifferenceappearsto
beinthevehicles’powertrains,asthetypicalEuropeanvehiclewasonly14%lighterthanitsUS
counterpart—likelyinsufficienttoexplainthesubstantialfueleconomyandpowerdisparities.
E. CarsSoldinChinaWillIncreasinglyUseTurbochargersandPartialHybridPropulsion
toBalanceFuelEconomyandPowerOutputDemands
China’scarfleetislikelytobeincreasinglypermeatedbypowertraintechnologiesthatenhance
fuel economy without detracting from vehicle performance. There are three primary ways of
accomplishingthis,eachofwhichmustbalanceequipmentcostanduserconvenience.Firstisto
movetowardsafullyelectricvehiclefleet.Thesecondistomakeenginedisplacementsmaller
60
“EUROPEANVEHICLEMARKETSTATISTICS:Pocketbook2014,”TheInternationalCouncilonCleanTransportation,
http://www.theicct.org/sites/default/files/publications/EU_pocketbook_2014.pdf
40
anduseturbochargerstomaintainpowerlevels.Thethirdistoincorporatepartialelectricvehicle
technologies—“mildhybrids”inindustryparlance.
Pureelectricvehiclesareatadisadvantageforseveralreasons.Tobeginwith,fullyelectriccars
likelywillnotreduceChina’ssevereairpollutionburdenandwillsimplytrade“oilsecurity”issues
for“airsecurity”problems.61Despitetherapidgrowthrateinwindandsolargenerationcapacity,
Chinastillreliesoverwhelminglyoncoal-firedplantstogenerateitselectricity.Assuch,electric
carssimplyshiftpollutionproblemsfromasemi-dirtyvehicletailpipetoapotentiallymuchmore
emissions-intensivepowerplantsmokestack.
PureelectriccarsnotsocleaninChinaduetocoal-dependentpowergeneration…
Eachmillionplug-inelectricpassengercarswouldlikelycreateanadditional740,000tonnesper
yearofcoaldemandinChina—equivalentinCO2emissionstermstoapproximatelyonemillion
BYDF0gasoline-poweredpassengersedans.62Inthissense,apureplug-inelectriccarrunningon
grid power in many parts of China (aside from areas where grid supply comes primarily from
hydroornuclearplants)iseffectivelystillascarbonintensiveasafullygasoline-poweredcompact
car,virtuallynegatingtheenvironmentalbenefitsofgoingfullelectric.Moreover,burningcoal
also releases far more sulfur than burning an equivalent amount of gasoline, not to mention
mercuryandothertoxicsnottypicallyemittedinmeaningfulamountsbygasolineengines.
This “road to grid” linkage increases coal demand and air pollution and could disrupt local
electricalgridswheretherearehigherconcentrationsofsuchcars.Additionally,EVinfrastructure
likely cannot be easily or cheaply socialized and hence will represent a barrier to entry for
consumers. Expenses, inconvenience, range limitations, and the costs of chargers and other
supportingsystemsfallontheindividualownerwhereasthecostsofliquid-fuelinfrastructure
likegasstationsaredistributedacrosssocietybecausesomanypeopleusethem.63
As such, China—like other major global markets—is most likely to continue using gasoline
propulsionforpassengervehicles.Butpowertrainswilladoptmuchdifferentconfigurationsthan
driversarecurrentlyaccustomedto.Inshort,theuseofsmallerengineswithturbochargingand
mildhybridswillcontinuegrowingrapidly.
61
AibingGuo,“TeslasMayBeMakingHongKong'sPollutionWorse,”Bloomberg,13April2016,
http://www.bloomberg.com/news/articles/2016-04-14/hong-kong-teslas-linked-to-more-co2-emissions-thangasoline-cars
62
Basedondatafrom“CarbonDioxideEmissionsCoefficients,”EIA,2February2016,
https://www.eia.gov/environment/emissions/co2_vol_mass.cfm.1.854tonnesofCO2emittedpertonneofsubbituminouscoalburned.Burningonegallonofgasolineemits8.89kgofCO2,whichmeans373kgperbarrel,or
3.17tonnesofCO2pertonneofgasoline(8.5barrelspertonneofmotorgasoline).1.9milliontonnesofsubbituminouscoalemits3.52milliontonnesofCO2,thesameas1.1milliontonnesofgasolineor26kbdannualized.
63
GabrielB.CollinsandAndrewS.Erickson,“ChinaCarSales:NewPocketsofOpportunity,”ChinaSignPost™(洞
察中国)79(29December2013.http://www.chinasignpost.com/2013/12/29/china-car-sales-new-pockets-ofopportunity/.
41
Turbochargersenableenginestogeneratesignificantlymorepowerfromsmallerdisplacements,
whichpreservesperformancewhileincreasingfueleconomy.Forinstance,datafromHoneywell
indicatesthata2.0literturbochargedfourcylinderenginecangeneratethesamehorsepower
asa3.0litersix-cylindermotor,whileusing25%lessfuelandemitting20%lessCO2.64Atpresent,
roughlyoneinfourcarsinChinausesturbocharging,butthispenetrationrateisexpectedto
climbtonearly50%by2020.65DozensofcarssoldinChinanowincorporateturbochargers,as
evidencedbytheAutoShanghai2015show,wheremorethan50vehiclesusingturbochargers
wereondisplay.66
“Mildhybrids”areacompromisetechnologyemployinga48-voltelectricsystem(asopposedto
the standard 12-volt auto battery) to boost fuel efficiency by assisting the car’s initial
acceleration.Inamildhybrid,theelectricmotorsprovideaboostatlowerspeedsbutonlyserve
asapowerboosterorstarter-generator,orboth.67Electricmotorsinthesevehiclescan'tpropel
thecarforwardontheirown.68Incontrast,Autotraderdefinesa“fullhybrid”asavehiclethat
has both “an electric motor and a rechargeable battery, which can work independently or in
conjunctionwitheachother.”69Theelectricalmotorsinafullhybridvehiclearepowerfulenough
topropelitwithoutrunningtheinternalcombustionengineduringverylightcruisingandlight
acceleration. Once the vehicle requires additional power is needed, the internal combustion
enginethenactivatesandprovidesfullpower.70
Datasuggestthatmildhybridscandeliverfueleconomygainsofapproximately15%,butatonly
aboutathirdthecostofafullhybridsystem.71LargeChinesecarmakerssuchasGeelyandFAW
plantobeginselling48-voltmildhybridswithinthenexttwoyears.72Furthermore,A123Systems,
aleadingglobalbatterymakerownedbyWanxiangGroupsays“weseeallmajorChineseauto
makersonorsourcing48-voltbatteries”andthattheywillphasesuchsystemsinoverthenext
threeyears.73
64
HoneywellCorporation,InvestorPresentations,2March2016InvestorPresentation,
http://investor.honeywell.com/Events-and-Presentations.
65
Id.
66
“HoneywellPlansTurboFutureinChina;Morethan50VehiclesHighlightedatAutoShanghai2015,”14May
2015,http://www.honeywell.com/newsroom/news/2015/05/honeywell-plans-turbo-future-in-china-more-than50-vehicles-highlighted-at-auto-shanghai-2015
67
“MildorFull?HybridDistinctionsDefined,”Autotrader.com,http://www.autotrader.com/car-news/mild-or-fullhybrid-distinctions-defined-36605(lastaccessedon15September2016)
68
Id.
69
Id.
70
Id.
71
MikeRamsey,“China’sCarMakersPioneer‘MildHybrids’”,TheWallStreetJournal,8March2016,
http://www.wsj.com/articles/china-car-makers-pioneer-mild-hybrids-1457433002;SeealsoContinental
CorporationInvestorPresentation,2February2016,http://www.continentalcorporation.com/www/download/portal_com_en/themes/ir/events/20160202_hsbc_sri_ffm_en.pdf(Slide13)
72
Ibid.
73
Ibid.
42
Asvehicleswithturbochargersand48-voltmildhybridtechnologycometocomprisealarger
portionofChina’svehiclefleet,theincrementalgasolinedemandofeachmillionvehiclessold
coulddeclineby15-to-20%,assumingmileageanddrivinghabitsremainconstant.Theeffectsin
2016and2017willbemoremuted,butpotentiallystillmaterial.ThisauthorbelievesChina’s
gasolinedemandwillfeelthefullimpactofmoreefficientpropulsiontechnologiesinthe201920timeframe.Accordingly,akeyquestionwillbe:“doesnumericalgrowthincarsalesremain
strongenoughtopoweroverallannualgasolinedemandgrowthinthe200kbdrangedespite
carsbecomingmorefuel-efficient?”Thenextanalysiswillassessrecentpassengervehiclesales
trendsinChina,aswellaspotentialgrowthtrajectoriesmovingforward.
F. ChinaPassengerCarFleetGrowthOutlookOvertheNextDecade
China’soverallpassengerfleetcargrowthwilllikelybethesinglemostconsequentialelement
influencing China’s gasoline demand growth profile over the next two to five years. On a 12monthrollingbasis,newcarsalesarestillonanupwardtrend,buthaveslowedfromthegrowth
rates seen a few years ago (Exhibit 28). The China Association of Automobile Manufacturers
forecaststhatnewpassengervehiclesaleswillriseby6%YoYin2016.74Lastyearsaw21.2million
newpassengervehiclessoldinChina,soCAAM’sforecastsuggestsnewvehiclesalescouldrise
toapproximately22.5millionvehiclesin2016.
74
NatashaLi,“CAAMPredicteda6%Growthinthe2016ChineseAutoMarket,”Gasgoo.com,13January2016,
http://autonews.gasgoo.com/china-news/caam-predicted-a-6-growth-in-the-2016-chinese-aut-160113.shtml.
43
Exhibit28:ChinaNewandUsedPassengerCarSales
Source:CAAM,CADA
Usedvehiclesalesarealsoincreasinglyimportantfromanoiluseperspective,astheyhelpdrive
furthermarketpenetration.December2015markedthefirsttimemonthlyusedcartransactions
inChinabroketheonemillionmark.China’snascent—butincreasinglyrobust—usedcarmarket
isapositivesignforthenext24monthsinthecountry’scarmarketoverall.Detailedusedcar
salesdatafrom2015suggestthatChinesenewcarownersinmanycasesturntheirvehiclesover
quiterapidly:vehiclesone-to-threeyearsoldaccountedfor26%ofusedcarssoldin2015,while
three-to-fiveyearoldvehiclesmadeup37%oftotalsales.75Vehicleseightyearsandoldermade
up19%oftotalusedcarsalesin2015.76Lightlyusedvehiclesofferaspiringcarownersthechance
toacquirepremiumbrandssuchasVolkswagenatsignificantdiscountsrelativetowhatanew
vehiclewouldcost.
75
“2015ChinaUsedCarMarketReport”(2015年中国⼆⼿车交易数据报告),QQ.com,22January2016,
http://auto.qq.com/a/20160122/052375_all.htm.
76
Ibid.
44
Considerthefollowingexample:Sina.com’susedcarportallistsa2009VWPassatwithamanual
transmissionand86,000kmontheodometerfor86,800RMB.77The2016modelyearofthe
vehicle—whichislistedforsaleinJining,Shandong—wouldcostnearly230,000RMBincluding
taxes.78Basedontheseeconomicfacts,itbecomesclearthattheheavydiscounts(62%from
currentnewvehiclestickerpriceinthiscase)stronglyattractbuyerswhomaynotbeflushwith
money,butseekqualityvehiclescapableofcomfortablycarryingseveralpassengersplusgoods
andluggage.
Nearly70%ofChina’susedcartransactionsin2015weredonefor100,000RMBorless,which
suggeststhatthisisanimportantpricingpointforpotentialbuyers.79Inturn,asthecarssold
duringthesalesspikein2008-11begintoageintothe5-year-pluscategory,usedcartransactions
couldrisesharplyoverthenext2-3yearsaswealthierownersseeknewrideswhilelesswellheeleddriverscapitalizeonthegrowingavailabilityofqualitymid-sizedvehiclescomingontothe
usedmarket.
Inthe2011timeframe,manyindependentusedcardealersinChinacomplainedaboutbeing
unabletosecuresufficientinventory.80Yetascurrentvehicleownersupgradetheircarsandmore
quality vehicles hit the used market, the growing availability of attractive bargains may spur
further purchasing and ownership penetration in both core markets and areas on their
periphery—forinstancetheconstellationsofcitiessurroundingShanghaiandGuangzhou,twoof
China’slargestmarketsfornewandusedvehicles.
On the policy side, several factors contribute to growing month-to-month volatility in sales
numbers.TheChinesegovernmentcuttaxesonsmalldisplacementvehicles(enginesizelessthan
1.6L)inOctober2015,whichhelpedpropupsales.81Inaddition,overthepasttwoyears,periodic
discussionsofvariousownershiprestrictionsbykeymunicipalitieshavefueledpanicbuyingas
prospectivecarownersrushtobeatownershipcaps,higherlicenseplatefees,andthelike.82
Amidst rising monthly sales variance between months, one trend is clear: China’s wealthier
Easternprovinceshavebeentheforcebehindthebulkofcivilianvehiclefleetadditionsoverthe
past several years. China’s seven largest civilian vehicle owning provinces—Shandong,
77
“VWPassatSedan,2009Model,1.8T,High-EndEditionWithManualTransmission”(⼤众帕萨特领驭09款1.8T
⼿动尊品型),Sina.com,http://usedcar.auto.sina.com.cn/buycar/d/1543632.Accessed22June2016
78
Ibid.
79
“2015ChinaUsedCarMarketReport”(2015年中国⼆⼿车交易数据报告),QQ.com,22January2016,
http://auto.qq.com/a/20160122/052375_all.htm.
80
BobTita,“DeereProfitTumblesAmidGlutofUsedFarmEquipmentatDealers,”TheWallStreetJournal,21
August2015,http://www.wsj.com/articles/deere-reports-decline-in-profit-as-sales-tumble-1440158935;Bill
Russo,ArnoldSun,EdwardTse,KenZhong,“ANewModelforSellingUsedCarsinChina,”Booz&Co.,March2011.
81
ChristianShepherdandTimMitchell,“Chinacarsalesaccelerateaftertaxcuts,”FT,12January2016,
http://www.ft.com/cms/s/0/c44622a0-b914-11e5-b151-8e15c9a029fb.html#axzz4CK3Y1gfo.
82
“China'scarsalesup13%asbuyersrushtobeatownershiplimits,”AutomotiveNews,9May2014,
http://www.autonews.com/article/20140509/GLOBAL03/140509801/chinas-car-sales-up-13-as-buyers-rush-tobeat-ownership-limits.
45
Guangdong, Jiangsu, Zhejiang, Hebei, and Henan—held approximately 46% of China’s total
civilianvehiclefleetatyear-end2014.Morenotably,thesesevenprovincesaccountedfornearly
50%ofincrementalfleetgrowthbetween2010and2014(Exhibit29).Itappearsthatthebig,
predominantlycoastal,vehicleowningregionscontinuetodrivethenationalmarket.
Exhibit29:ChinaCivilianVehicleFleetbyProvinceandProvincialContributionstoIncremental
FleetGrowthBetween2010and2014(民用汽车保有量)
Source:NBSChina,Author’sAnalysis
46
G. Diving Deeper Into China Vehicle Sales Picture—Detailed Provincial-/City-Level
Analysis
Togainprovince-andcity-levelinsightsintothebehaviorofChinesenewcarbuyers,thisanalysis
drawsonandupdatesananalysisoriginallydoneinDecember2014forChinaSignPost™that
usedVolkswagenasananalyticallenstogaininsightsintotheevolvinggeographyofpassenger
carsales.83VWwas—andremains—thelargestpassengercarsellerinChina’smassivemarket,
accountingforroughlyoneineveryfivepassengercarssoldinthecountry.Assuch,theChina
SignPoststudyviewedVW’ssalesandserviceinfrastructuredecisionsasausefulbarometerfor
assessing how the “smart money” saw passenger car demand unfolding in various regions of
China.84 The dataset was updated in June 2015 and May 2016 by obtaining additional VW
disclosuresofprovincial-leveldealershipdata.
VWdiscloseddatafordealershipsbyprovinceatapproximatelyquarterlyintervalsbetweenthe
fourthquarterof2012andthesecondquarterof2016.Ouranalysisofthedatashowedthatthe
10 provinces with the largest increase in number of dealerships during this time were (in
descendingorder):(1)Jiangsu,(2)Zhejiang,(3)Hebei,(4)Shandong,(5)Guangdong,(6)Sichuan,
(7)Henan,(8)Yunnan,(9)Anhui,and(10)atiebetweenAnhuiandHunan.85These10provinces
reflect a significant East Coast bias in the China sales strategy of one of the world’s premier
automakers.Indeed,bytheauthor’scountusingVW’sowndata,morethan45%ofitsnewChina
dealerships added between 4Q2012 and 2Q2016 were opened in East Coast provinces that
accountforlittlemorethan1/3ofChina’spopulation(Exhibit30).
Provinces in Central and Southwestern China saw significant growth in the number of VW
dealerships,butataproportionlowerthanonewouldexpectbasedontheprovinces’population
andlatentdemographicandeconomicpotentialforincreasingautosales.Dealershipadditions
inNortheast,Northwest,andWesternChina(Xinjiang,Tibet,Ningxia,Qinghai)werelowrelative
topopulation,butthisislesssurprisinggiventheseregions’relativepovertycomparedtoplaces
likeJiangsuandSichuan.
83
NotethatthisanalysiscountsdealershipsforallofthethreecoreVWbrands:VWitself,Audi,andSkoda.
GabrielB.CollinsandAndrewS.Erickson,“Volkswagen’sChinaDealershipExpansionsSuggestSubstantial
UpsideforContinuedGrowthinCarSales—andGasolineConsumption,”ChinaSignPost™(洞察中国)87(17
December2014).
85
Ibid.
84
47
Exhibit30:VWDealershipAdditionsbyRegioninChina,4Q2012to2Q2016
Source:CompanyReports,NBSChina,Author’sAnalysis
Examiningtrajectoriesofdifferentcitiesinthesamehighgrowthprovince—inthiscase,Jiangsu
and Zhejiang—further reinforces the theme of just how locally concentrated much of China’s
passengercarfleetgrowthhasbeen.Thisisanimportantwrinklebecauseitsuggeststhatthe
marketmayneedtopriceinamorenuancedautosalesandgasolinedemandgrowthpicture
that is more complex than simply envisioning a wave of cars washing over a vast a populous
country.
Consider,forinstance,thecitiesofSuzhouinJiangsuProvinceandTaizhouinZhejiangProvince,
twoadjacentareaswhichcompriseoneofChina’slargestpassengercar(andgasoline)markets.
TheSuzhoumetroarea,whichliesapproximately60milesnorthwestofShanghai,ishometo
roughly11millionpeople.Inaseven-yearspanfrom2008to2014,Suzhou’spassengervehicle
fleetgrewby191%to2.4millionvehicles(Exhibit31).Incontrast,theTaizhoumetroarea,which
sits190milessouthofShanghaiandcountsapproximately6millionresidents,sawitspassenger
vehiclefleetgrowby240%duringthesametimeframe,risingto470thousandvehicles.While
Taizhou’sfleetdidgrowsomewhatfasterthanSuzhou’s,itstartedfromamuchlowerabsolute
numerical baseline and was not nearly rapid enough to compensate for the significant
arithmeticaldisparitybetweenthecities’respectivevehiclefleets.
48
Exhibit31:CivilianVehicleOwnershipinSuzhou,JiangsuandTaizhou,Zhejiang
Source:NBSChina,Author’sAnalysis
H. SalesofNewCarsAreSlowing
VWdatasuggeststhatthecompanysoldapproximately1,655vehiclesperdealershipinChina
during2012.86By2015,thisnumberdeclinedtoanestimated1,265vehiclesperdealer.Thetrend
isimportantherebecauseittranscendsacommonargument—thatthereare“toomany”car
dealersinChina.87Fromtheperspectiveofdealershipeconomicandprofitability,thisthesismay
well hold. But even consolidating dealerships will not reverse the slowdown in new car sales
growthrates,whicharethecorefocusofthisanalysisduetotheirsignificanteffectsonChina’s
gasolinedemandoutlook.
ConsiderthesignificantrecentslowdowninVW’sChinadealershipbuildoutrate,from425net
dealershipsaddedbetween2Q2013and2Q2014,306netdealershipsaddedbetween2Q2014
and2Q2015,tojust110netdealershipsaddedfrom2Q2015to2Q2016(Exhibit32).Whenone
ofthelargestcardealersinamarketrapidlyscalesbacktheirdealershipadditionrate,theaction
strongly suggests the operator sees much less sales upside than they perceived even a few
quartersprior.
86
Theauthorusestheword“suggests”becausewhilecarsalesarecalculatedonanannualizedbasis,VWhas
continuallyaddeddealershipsduringeachoftheyearsinquestion.Becausewedonotknowwhenthese
dealershipsopenedduringagivenquarterorwhattheiroperational“rampup”periodis,wetreattheaverage
salesperdealerasanapproximationthatreflectsanoveralltrend,asopposedtoanexact,time-adjustednumber.
87
“AutoDealersinChina,UbiquitousasKFCs,FaceShakeout,”Bloomberg,24April2016,
http://www.bloomberg.com/news/articles/2016-04-24/foreign-car-dealers-in-china-ubiquitous-as-kfcs-faceshakeout.
49
Exhibit32:NetVWDealershipAdditionsinChinabyRegionHaveSlowedDramaticallyinPast
TwoYears
Source:CompanyReports,Author’sAnalysis
PerhapsthemostnotabletrendemergingfromthisgranularVWdataisthatbetween2015and
2016,thecompanyclearlybegantotestthelimitsofdealershippenetrationinCoastalChina.88
DealershipexpansionalsoslowedsignificantlyinSouthwestChina,whichbasedonitsstilllow
ratesofpersonalcarownership,wouldbeconsideredafertilegrowthmarket.VWisnotalonein
realizingthatpassengercardemandinWesternChinahasgrownslowerthanexpected.Several
other multinational and domestic automakers saw their sales underperform expectations in
WesternChinabeginningasearlyasJune2015.89
CentralChinadealershipadditionsshranktheleast,whichsuggeststhatconsumerstherestill
haveahealthyappetiteforVWvehicles.Inturn,ifconsumerscanaffordpremiumforeignmarks
such as VW, this strongly implies that there is also sufficient local purchasing power to drive
continuedsalesgrowthamongothermakes,particularlylowercostdomesticbrandsandused
88
ItispossiblethatsomeofthesalesanddealershipexpansionslowdownstemsfromVW’semissionsscandal.Yet
thecompany’ssalesinChinaaredominatedbygasoline-poweredcars,asopposedtotheTDIdieselvehiclesVW
waspushingintheUnitedStatesmarket.
89
“CarSalesSlowinChina’sWest,DentingHopesofGlobalAutoIndustry:Salesshiftdownagearinregionwhere
automakerssawpotentialforfastergrowth,”TheWallStreetJournal,10June2015,
http://www.wsj.com/articles/car-sales-slow-in-chinas-west-denting-hopes-of-global-auto-industry-1433936986
50
vehicles.DealersarealsolikelytargetingCentralChinafornewcarsalesbecausethepastfew
years’ intensive focus on East Coast markets is driving saturation and increases in the traffic
congestionissuesthatdissuadedriversfromusingtheirvehicles—orinsomecases,evenbuying
acartobeginwith.90
I. China’sNewCarSalesOutlook
China’snewcarmarketistransitioningintoaperiodwheresaleswilllikelyexpandmoreslowly
thantheboomtimesofthepasteightyears.IndustrydataproviderIHSexpectsacompensated
annualsalesgrowthrateof5%betweennowand2020.91Thebaselinesalesvolumeisalready
highenoughthatevena5%CAGRmeansthatsalescangrowbymorethan1millioncarsper
year.Moreover,China’scarfleetremainsyoungenoughthatsignificantscrappingactivitymost
likelywillnotbeginforatleast3-4moreyears.Assuch,netfleetgrowthwillhavecommensurate
impactsonoildemand.EachmillionnewcarssoldinChinacouldconsumeavolumeofgasoline
equivalenttoapproximately15-to-40kbdofcrudeoil.92
Evenifcarsalesslowmoresignificantlythantheconsensusforecast,thevolumeofvehicleson
theroadtoday,coupledwiththefactthateachcar’sservicelifeislikelytobeatleast10years,
meanssignificantgasolinedemandisalready“bakedin.”Thisholdstrueevenasmoreefficient
vehicleslikethosedescribedearlierinthissectioncometocompriseagreatershareofChina’s
passengercarfleetincomingyears.
DemandSector2:FuelOilandJetFuel
A. FuelOil
Unlikegasoline,whichhasenjoyedayears’longstreakofsteadilyincreasingconsumption,fuel
oildemandinChinahassteadilydeclinedoverthepastdecade.In2004,fueloilaccountedfor
nearlyoneineveryfourbarrelsofpetroleum-derivedliquidfuelconsumedinChina.Byearly
2015, this proportion had steadily declined to approximately 8%—roughly a third of its level
slightlyoveradecadebefore(Exhibit33).
90
GabrielCollins,“ExploringPotentialLimitstoGasolineDemandGrowthinChina,”ChinaSignPost™(洞察中国)
96(13June2016),http://www.chinasignpost.com/2016/06/13/exploring-potential-limits-to-gasoline-demandgrowth-in-china/.
91
VolkswagenGroupChina–Balancingthe“NewNormal,”13June2016,
http://www.volkswagenag.com/content/vwcorp/info_center/en/talks_and_presentations/talks_and_presentation
s.html.
92
Theimpliedvolumeofcrudeoilnecessarytoproduceagivenamountofgasolinegiventhepredominantproduct
outputslatesatChineserefineries.See,forinstance,GabrielCollins,“RisingGasolineDemandMeansChina’s
CrudeOilUseisDecouplingfromtheIndustrialEconomy,”ChinaSignPost™(洞察中国)95(3June2016),
http://www.chinasignpost.com/2016/06/03/rising-gasoline-demand-means-chinas-crude-oil-use-is-decouplingfrom-the-industrial-economy/.
51
Exhibit33:FuelOilLosingImportanceasaDriverofOilProductsDemandinChina
Source:JODI,Author’sAnalysis
TheprincipalfactorbehinddecliningfueldemandistheriseofChina’sincreasinglysophisticated
independentoilrefiners.Broadlydescribedas“teapots”(茶壶),theseplantsformerlycouldnot
importcrudeoilandwerethusforcedtouseheavyfueloilasafeedstock.Regulatorychanges
dramaticallyshiftedthelandscapein2015,astheNDRCbegangrantingsignificantimportquotas
toarangeofindependentrefiners.Theactivitypeakedinthesecondhalfof2015,andbylate
December, the NDRC had issued quotas for these plants to import approximately 69 million
tonnesperyearofcrude—roughly1.4millionbpd,avolumeequivalenttonearly21%ofChina’s
totalcrudeoilimportsin2015.93
Many of the teapots run significantly-sized facilities. For instance, Dongming Petrochemical’s
Shandong plant can handle 240,000 bpd of crude.94 Overall, the teapots account for
93
“FACTBOX-China'snewcrudeoilimporters,”Reuters,29December2015,http://in.reuters.com/article/chinacrude-imports-idINL3N14I2JN20151229
94
“EnterpriseProduction”(集团产业),ShandongDongmingPetrochemicalGroup,
http://www.dmsh.com.cn/about/?29.html(accessed16June2016)
52
approximately20%ofChina’snameplaterefiningcapacity.95Assuch,theirswitchtocrudeoil
feedstock definitively shows up in the monthly provincial-level crude oil processing data. The
author determines this by taking national monthly crude oil processing volume data and
comparing it to data from Guangdong—one of China’s largest oil refining provinces—and
Shandong—wherethebulkofteapotcapacityislocated.
WhentherawcrudeprocessingvolumedataareconvertedintoanindexwithJune2013asthe
start point, it becomes apparent that when the NDRC began issuing crude import quotas to
independent refiners in late summer 2015, Shandong’s processing volumes grew rapidly and
decisivelypulledawayfromthepaceofcruderefininginGuangdongandnationwide.Hereit
bearsnotingthat“crudeprocessing”reflectsactualrefiningactivity,andgivesacleareranalytical
picturethanotherpotentialmetrics,suchasimportvolumes.
Exhibit34:RiseoftheTeapotsDrivesGreaterCrudeOilDemand
Monthlycrudeoilprocessingvolumeindex,June2013=100
Source:NBSChina,NDRC,Author’sAnalysis
95
LucyHornby,“China’s‘teapot’oilrefineriesposechallengetomajors,”FT,7April2016,
http://www.ft.com/intl/cms/s/0/7fc95106-fc71-11e5-b5f5-070dca6d0a0d.html#axzz4BlyBRIje
53
To date, Shandong’s processing volumes remain elevated and local refiners have imported
cargoesfromAngola,Brazil,Libya,Oman,Russia,andVietnam.96ShandongDongming,oneofthe
largestteapotoperatorsbycapacity,signedalong-termcrudeoilsupplyagreementwithBritish
Petroleum (BP) in November 2015.97 In short, the evidence strongly reflects that China’s
independentrefinershavebecomeaforceintheglobalcrudemarketandthattheirdemandfor
fueloilasabaselinerefineryfeedstockhaslikelybeenpermanentlyerased.
Changesintheintensityofwaterbornefreighttransport—aswellaspolicymeasuresaimedat
cutting pollution from ship exhausts—are also chipping away at fuel in oil use in China.
Historically,railcongestionledcoal-firedpowerplantsinChinatosourcecoalbysea.Coalwould
be mined in north-central China, transported by rail to ports on the Bohai Gulf, and then be
carriedbycoastwiseshippingtopowerplantsinSouthernChina.TheDaqinRailway,whichlinks
therichcoalfieldsatDatongwiththemassiveQinhuangdaocoalportontheBohaiGulf,isChina’s
single-largestdedicatedcoaltransportline.Assuch,thetonnageofcoalitcarries—mostofwhich
isshippedonwardtopowerplantsinSouthandEastChina—givesasenseofthehealthofChina’s
coastalshippingindustry,forwhichcoalisthemaincargoandfueloilisthemainpowersource
forshipengines.Volumesstagnatedoverthepastfouryears,declinedslightlyin2015,andare
likelytocontinuedecliningin2016andbeyondascoastalChinesecitiesstrivetoreducelocal
coalconsumption(Exhibit35).
96
http://www.platts.com/latest-news/oil/singapore/china-teapot-refineries-crude-oil-imports-continue-26337972
“DongmingPetrochemicalSignsaLong-TermCrudeOilSupplyAgreementWithBP”(东明⽯化与BP签订原油长
期供应协议),CNPC,11November2015,http://wap.cnpc.com.cn/system/2015/11/12/001566793.shtml
97
54
Exhibit35:SlowingDaqinRailwayCoalShipmentsSuggestLessCoalBeingBargedtoSouthern
China
Coaltransported(煤炭发送量),milliontonnes
Source:DaqinRailway
Second,Chineseauthoritieshavebeguncrackingdownontheuseofhigh-sulfur,highpollution
bunker fuels by ships in key Chinese ports. In December 2015, the Ministry of Transport
designatedtheBohaiGulf,PearlRiverDelta,andYangtzeRiverDeltaasEmissionControlAreas
(ECAs)whereshipsmustnowusebunkerfuelwithsulfurcontentof0.50%orless.98Thesulfur
restrictionsrepresentasignificantdeparturefromthetraditionalpracticeofusinghigh-sulfur
bunkerfuelsthatcancontainasmuchas3.5%sulfurbymass.PortsintheECAsarescheduledto
imposerestrictionsasfollows:by1January2017,shipsmustusecompliantfuelwhileberthedin
keyportsintheECAs;by1January2018shipsberthedinallportsintheECAsmustusecompliant
fuel;andby1January2019shipsoperatinganywhereintheECAsmustusefuelwithasulfur
contentof0.50%orless.99
98
“GardAlert:Chinaintroduces0.50percentsulphurcapinmajorportregions,”Gard,
http://www.gard.no/web/updates/content/20904877/gard-alert-china-introduces-050-per-cent-sulphur-cap-inmajor-port-regions
99
Id.
55
Thenewrules’bunkerfueldemandeffectswilllikelybesubstantial,asthe11”keyports”that
enforce the rules earliest are among the world’s busiest: Guangzhou, Huanghua, Nantong,
Ningbo-Zhoushan,Qinhuangdao,Shanghai,Shenzhen,Suzhou,Tangshan,Tianjin,andZhuhai.100
ItremainstobeseenhowstringentlyChineseauthoritieswillenforcetherulesandhowcoastal
shipperswillcomplywiththenewemissionsrules.Whatisclearisthatvirtuallyallseabornecoal
shipmentsfromNorthtoSouthChinaemanatefromtheBohaiGulf,whichisasulfur-restricted
ECA.Lowersulfurfuelssuchaslowsulfurdieselandliquefiednaturalgas(LNG)arecostly,asare
otherpotentialsubstitutessuchasbiodieselanddimethylether.Ifshipownershavetoexitthe
tradebecausethecostsofcomplyingwithnewfuelstandardsaretoohigh,morecoalmaymove
southbyrailasnewrailcorridorsopenbetweenthecoalfieldsandmarketsinEastandSoutheast
China. Any of these scenarios stands to reduce marine bunker oil demand, and by extension,
overallfueloilconsumptionintheChinamarket.
B. JetFuel
JetfuelisakeydriverofincrementaloilproductsdemandgrowthinChina,withaverageannual
demandgrowthof11.1%between2005and2012,andhasbeengrowingatapproximately10%
peryearsincethen.101China’skerosenedemandisoverwhelminglydrivenbyjetfueldemand,as
evidencedintheclosecorrelationbetweendomesticpassengerairmilestravelledandkerosene
consumption(Exhibit36).
100
Ibid.
“SinopecandChinaAviationOilWorkHand-in-HandtoDeepenCooperation”(中国⽯油与中国航油携⼿深化
合作),Sinopec,24February2014,
http://www.sinopecgroup.com/group/xwzx/hgzc/20140226/news_20140226_372874125527.shtml;
“CORRECTED-China'sjetfueldemandgrowthtoholdsteadydespiteslowingeconomy-CAOexec,”Reuters,6
August2014,http://www.reuters.com/article/china-aviatn-oil-jetfuel-demand-idUSL4N0QA31V20140806
101
56
Exhibit36:China’sRisingKeroseneDemandCloselyTracksDomesticAirTravelActivity
Source:JODI,NBSChina,Author’sAnalysis
Fromacrudeoilmarketperspective,increasingjetfueldemandplaysanimportantrolebecause
likedieselfuel,itisamiddledistillateandhelpsoffsetfallingconsumptionofdieselfuel.Asof
August2016,theIEAprojectsthatin2017,jetfueldemandgrowthinChinawillbasicallyoffset
the loss in diesel fuel consumption.102 To put the “road” versus “airways” middle distillate
demand interaction in perspective, consider the following: a Boeing 737 jetliner that flies an
averageof7hoursperdayfor350daysoftheyearusesasmuchfuelasapproximately325heavy
dieseltruckstravelling52,000kmperyearapiece.103
102
IEA,OilMarketReport:11August2016,
https://www.iea.org/oilmarketreport/omrpublic/currentreport/#Demand(lastaccessed15September2016)
103
Boeing737fuelusedataarederivedfromJohnCox,“AsktheCaptain:ReaderquestionsabouttheBoeing737,”
USA Today, 12 October 2014, http://www.usatoday.com/story/travel/columnist/cox/2014/10/12/boeing737/17036931/. Boeing 737 average annual flight times are derived from the following source, but adjusted
downwards:JohnCox,“AsktheCaptain:Howfardoesajetflyduringitslifetime?,”USAToday,19November2012,
http://www.usatoday.com/story/travel/columnist/cox/2012/11/19/ask-the-captain-how-far-does-a-jet-fly-duringits-lifetime/1712269/.ChinaheavytruckannualfuelusedataarederivedfromThisestimateusesthefollowing
assumptions:(1)Fueluse—EachheavytruckinChinaconsumes44litersofdieselfuelper100km(BenSharpeand
RachelMuncrief,“LiteratureReview:Real-WorldFuelConsumptionOfHeavy-DutyVehiclesInTheUnitedStates,
China,
And
The
European
Union,”
The
ICCT,
White
Paper,
January
2015,
http://www.theicct.org/sites/default/files/publications/ICCT_HDV_FC_lit-review_20150209.pdf) while China’s
57
DemandSector3:CrudeOil-DerivedChemicalFeedstocks
A. NaphthaandLPG
ChemicalfeedstocksaccountforameaningfulproportionofChina’soilproductsdemand.The
twocoreproductscomprisingthechemicalfeedstockstreamarenaphtha,whichaccountedfor
about60%ofethyleneproducedinthecountryduring2014andliquefiedpetroleumgas(LPG),
whichhasbecomeanimportantfeedforpolymerproduction(Exhibit37).104
Exhibit37:NaphthaandLiquefiedPetroleumGasDemandinChina,‘000bpd
Source:JODI,Author’sAnalysis
Naphthademandhasbeenbasicallyrange-boundforthepastsixyears,whileLPGdemandhas
risen sharply. Rising LPG use comes primarily from a rapid expansion in propane dehydrogenationcapacity,whichusesLPG(primarilypropane)asafeedstocktoproducepropylene
and ultimately, polypropylene plastics. China had an estimated 9 million tonnes per year of
passenger car fleet had an average fuel economy of 7.12 liters/100km in 2015 (“China Passenger Vehicle Fuel
Consumption,”2015DevelopmentAnnualReport,TheInnovationCenterforEnergyandTransportation,
December2015,http://www.icet.org.cn/english/admin/upload/2015120241280305.pdf.)(2)Annualvehiclekm
travelled—52,000kmperyearforheavytrucks(H.Huoetal./EnergyPolicy43(2012)6–16(10),
http://www.tsinghua.edu.cn/publish/ess/7835/20120719144509070242457/[8]%20Huo_EP_2012.pdf)
104
http://www.platts.com/latest-news/petrochemicals/singapore/analysis-chinas-naphtha-imports-surge-amidattractive-26103910
58
propyleneproductioncapacityin2015andby2020,couldhaveasmuchas20milliontonnesper
year,accordingtoIHSandPlatts.105
To give a sense of how propane dehydrogenation-based propylene production drives LPG
demand, consider the following example: Haiyue New Material Company’s Ningbo plant can
process 720 thousand tpy of propane into 600 thousand tpy of propylene.106 The last several
years have featured a major slug of PDH capacity additions in China. Every additional million
tonnes per year of PDH capacity that are built (and actually run) in China would likely add
approximately34kbdofpropanedemand,whichtranslatesintoapproximately36kbdofoverall
LPGdemand,assumingthePDHplantsrunat90%oftheirnameplatecapacity.107
Ultimately, Chinese LPG demand will be substantially influenced by how much the new PDH
plantsactuallyrun.Thefacilities’utilizationrateswill,atafundamentallevel,dependoncrude
oilprices.Inanutshell,PDHunitsneedpropanetobecheaprelativetocrudeoil,fromwhich
napththaisderived.Thus,thelargerthepricespreadbetweencrudeoilandpropane,thebetter
theeconomicsofrunningPDHunitsandthegreaterthatpetrochemical-drivenLPGdemandwill
beinChina.108
IV.
ConclusionsandPolicyInsights
China’soildemandgrowthisnowfundamentallydrivenbygasoline(leveragedtopassengercars)
andtoalesserextent,middledistillates(basicallyjetfuelusedrivenbyrisingdomesticairtravel).
Asthecountry’soildemandbecomesincreasinglydependentonconsumerdemand,itwillalso
likelybecomemorevolatileintheshort-term.Previously,oildemandgrowthinChinawasmost
closelyleveredtolarge,fixedassetinvestmentprojectsandotherdiesel-intensiveactivity.Yet
now,ratherthanbeingtiedtothe5-YearPlans,agreatershareofChina’soildemandwilllikely
dependonday-to-dayandweek-to-weekchangesinconsumersentiments.
105
JesseTijerina,“TrendsinthePolymersMarket,”IHS,2015,
http://www.fullplast.cl/Descargas/Presentaciones_FULL_PLAST/IHS_CHEMICAL_Jesse_Tijerina.pdf;“Thecyclical
natureofspreadsandeffectsonPDHpropylene,”Platts,18December2015,
http://blogs.platts.com/2015/12/18/cyclical-nature-spreads-pdh-propylene/
106
“China'sYantaiWanhuatestrunsShandongPDHplant,eyesMarchstartup,”Platts,8January2015,
http://www.platts.com/latest-news/petrochemicals/singapore/chinas-yantai-wanhua-test-runs-shandong-pdhplant-27998848;“ChinaNingboHaiyue'sPDHplanttostartupinJuly;receivesfirstpropanecargo,”Platts,1July
2014,http://www.platts.com/latest-news/petrochemicals/singapore/china-ningbo-haiyues-pdh-plant-to-start-upin-21836529
107
Producing1tonneofpropyleneviaPDHrequiresapproximately1.2tonnesofpropanefeedstock(Dr.Charles
nd
Fryer,“TheImpactsofFallingCrudeOilPricesonChemicalBuildingBlocks,”TecnonOrbiChem,2 International
ChemicalDownstreamConference,Mumbai,29-30January2015.).Onetonneofpropanehas11.6barrels
(https://www.extension.iastate.edu/agdm/wholefarm/html/c6-87.html.)PlattsAsiaLPGspecusesSaudiAramco’s
LPGexportstandards,whichcallforLPGsoldintotheAsianmarkettoconsistofaminimumof95%propaneby
volume.(https://www.platts.com/IM.Platts.Content/methodologyreferences/methodologyspecs/asia_lpg.pdf)
108
“CharlesFryer,“AftertheFall—TheImpactofOilPriceVolatilityonPetrochemicalMarkets,”TecnonOrbiChem
MarketingSeminaratAPIC2015,Seoul,7May2015.
59
Diesel demand is likely to continue gradually declining as the country works to re-balance its
economyandreducerelianceonpollutingheavyindustriesandfixedassetinvestment.Thereis
someironyhere,asincreasingeconomicprosperityisstimulatingarisingpushbyChina’smiddle
class to reduce pollution, which is also likely to trim the rate of oil demand growth. Fuel oil
demandhasalsotakenapermanenthitduetoBeijing’sdecisiontograntindependentrefineries
licensetoimportcrudeoilfeedstock.Atthesametime,thecountryhasaddedsignificantnew
netoilrefiningcapacity.Refinerycapacityadditionsspringfromtwocoresources.Ononehand,
the“Big3”stateoilcompanieshavebuiltlargenewplants.Ontheother,byadoptingnewrules
thatenabletheindependent“teapot”refinerstoimportcrudeoil,amorecompetitivelypriced
feedstock,thegovernmenthaseffectivelyinducedthemtoruntheirplantsathigherratesand
bringcapacitythathadbeenmothballedforyearsbackonline.
Both dynamics are unfolding concurrent to a slowing industrial sector and lower demand for
diesel fuel. Since Chinese refiners are geared to favor diesel production, running plants to
produce more gasoline while diesel demand weakens is a recipe for surplus diesel, which is
preciselywhyChinesenetexportsofdieselfuelhaverisensosharplyinthepast12months.
Withthosefactorsinmind,China’s“newnormal”foroildemandgrowthfeaturesthreeespecially
notableelements.First,thenextfiveyearsarelikelytobeaperiodofmuchloweroildemand
growth rates than the previous five years. A broad range of structural factors, including the
cumulativeeffectsofpollution,chronicdisease,anddemographicdistortionsthatmeanChina
will grow gray before it grows rich have begun to materially restrain the country’s economic
growth.109Thisstudy’sbasecasepositsoilproductsdemandslowingtoanannualexpansionrate
ofapproximately2.8%peryearoverthenextfouryears(Exhibit38).
Tobesure,suchfiguresstillpointtosignificantannualincreasesinoildemand.Demandvolume
growthof2.8%inacountrythatinsomemonthsnowputsmorethan11.5millionbpdofcrude
oilintoitsrefineriescouldmeanannualcrudeoilthroughputincreasesontheorderof325kbd.
Thisisapproximatelythe2015companywidedailyliquidsproductionofEOGResources,oneof
thelargestU.S.independentoilproducersbyvolume.110
Againstthatcontextualbackdrop,itisclearthatevenasChinamovesontoaflatteroildemand
trajectory,itsabsoluteimpactsonglobalcrudeoildemandwillremainmaterial.Butforaworld
andanoilindustrythathadcometoexpectsteady5-to-8%annualdemandgrowthfromChina,
109
See,forinstance,MarkMagnier,“China’sEconomicGrowthin2015IsSlowestin25Years,”TheWallStreet
Journal,19January2016,http://www.wsj.com/articles/china-economic-growth-slows-to-6-9-on-year-in-20151453169398.ForadetailedandprescientexaminationofthestructuralfactorsbehindChina’sslowdown,see
GabeCollinsandAndrewErickson,“China’sS-CurveTrajectory:StructuralFactorsWillLikelySlowtheGrowthof
China’sEconomyandComprehensiveNationalPower,”ChinaSignPost™(洞察中国)44(15August2011),
http://www.chinasignpost.com/2011/08/15/chinas-s-curve-trajectory-structural-factors-will-likely-slow-thegrowth-of-chinas-economy-and-comprehensive-national-power/.
110
EOGResources,2015AnnualReport,http://investors.eogresources.com/Annual-Reports-and-Proxy-Materials
60
theshiftontothenewnormalpathisagamechanger,astheindustryalreadyseesfromthe
currentglobalsupplydemandimbalance.Moreover,thereisamaterialriskthatChinainfact
movesmoretowardthisstudy’slow-growthcase,underwhichoilproductsdemandslowstoan
average annual growth rate of 1.9% between now and 2020. Coupled with declining oil
consumptionintheEurozoneandstagnantconsumptionintheAmericasanddevelopedAsia,
suchaChineseslowdownwouldexertpowerfuldownwardpressureonglobaloildemandand
could potentially push prices below $45/bbl on a sustained basis for the next several years.
Conversely,underthestudy’shigh-growthscenario,China’soilproductdemandwouldgrowby
anaverageof5.4%betweennowand2020,withdemandreaching15.5millionbarrelsperday
bytheendof2020,witharesultingupwardpressureonoilprices.
Exhibit38:ChinaOilProductsDemandForecastto2020,kbd
61
KeyDemandModelAssumptions
BaseCase
GDP,Billion
RMB
2010
2011
2012
2013
2014
2015
2016E
2017F
2018F
2019F
2020F
40,890
48,412
53,412
58,802
63,591
67,671
71,054
74,252
76,851
79,540
82,324
BaseCase
GDP,YoY
change
LowCase
GDP,YoY
change
HighCase
GDP,YoY
change
18.4%
10.3%
10.1%
8.1%
6.4%
5.0%
4.5%
3.5%
3.5%
3.5%
4.0%
2.0%
1.5%
1.5%
1.5%
6.0%
6.0%
5.0%
5.0%
5.0%
BaseCaseOil
Intensity,kb
perbillion
RMBofGDP
LowCaseOil
Intensity,kb
perbillion
RMBofGDP
HighCaseOil
Intensity,kb
perbillion
RMBofGDP
84
-
-
74
-
-
70
-
-
67
-
-
64
-
-
65
-
-
62
58
65
59
58
65
59
55
63
59
55
63
59
55
63
Sources:BPStatisticalReview,NBSChina,Author’sAnalysis
Second, the combination of slowing domestic middle distillate demand and substantial new
refiningcapacitycomingonlineismakingChinaintooneoftheworld’slargestrefinedproducts
exporters.InApril2016,China’snetexportsofgasolineanddieselfuelflirtedforthefirsttime
withthe500kbdmark—onparwiththenetproductexportnumbersfromSaudiArabia(Exhibit
39).
62
Exhibit39:ChinaNetExportsofGasolineandDieselFuel,kbd
Source:JODI,Author’sAnalysis
ThedatasuggestChina’sshifttobecomingamajornetexporterofoilproductsisonethatwill
endureforyearstocome.Considerthefollowingdatapoints:betweenJanuary2003andJuly
2013,only10monthsoutof127sawChinahavepositivenetexportsof“core”refinedproducts:
diesel,gasoline,andkerosene(i.e.jetfuel).Butofthenext36months—July2013toJune2016—
only three months were not positive (Exhibit 40). To boot, the overall volume of net refined
productexportsgrewsignificantlyinabsolutetermsandasapercentageoftotalcrudeoilintake
atrefineriesinChina.
Thedurableshifttohighernetrefinedproductexportvolumesandthefactthatthesevolumes
have come to occupy a larger share of crude feedstock put into Chinese refineries suggest a
watershedmomenthasoccurred.Inessence,Chinahasalmostcertainlystructurallyoverbuiltits
refiningsystemandisnoweffectivelyexportingthatovercapacitytotheworldmarketinthe
formofgasoline,diesel,andotherrefinedproducts.111
111
LucyHornby,“Chinaplaguedbyoilrefiningovercapacity,”FT,7June2016,
http://www.ft.com/cms/s/0/b1c06f2a-2965-11e6-8b18-91555f2f4fde.html#axzz4Co9HpGDa
63
Exhibit40:NetExportsofCoreRefinedProducts(gasoline/diesel/kero/fueloil)asPercentage
ofRefineryIntakeofCrudeOil
Source:JODI,Author’sAnalysis
Third,China’spoliticalleadershiphasturnedontoapathofnationalismandcultofleadership
personalitythatthreatenstoblunttheformidabletechnicalcompetenceresidentinthetopranks
of China’s leadership The government’s bungled handling of the summer 2015 stock market
declinessuggestitisoverconfidentinitsabilitytoshapeandinfluencemarketactivity.112The
stockmarketfiascoalsosuggeststhatChina’seconomicpolicyisincreasinglyvulnerabledueto
over-centralizationofpresidentialauthoritythatslowsand/orskewseconomicpolicymaking.113
Of nearly equal importance, China’s thousands of local officials’ retain significant ability to
passivelyresistreformsthatmaythreatenenergy-intensivelocalindustriesthatareimportant
employers while large state-owned enterprises’ massive economic heft and commensurate
behind-the-scenespoliticalinfluencealsoslowreforms.114Thesesomewhatcountervailingforces
112
GabrielB.CollinsandAndrewS.Erickson,“ThreeSignPoststoGuideYouasChina’sStockMarketCrisis
Unfolds,”ChinaSignPost™(洞察中国)90(5July2015),http://www.chinasignpost.com/2015/07/06/threesignposts-to-guide-you-as-chinas-stock-market-crisis-unfolds/.
113
Id.
114
AndrewS.EricksonandGabrielB.Collins,“Physician,HealThyself:ModestExpectationsinOrderforChina’s
ReformsasThirdPlenumAnniversaryApproaches,”ChinaSignPost™(洞察中国)81(23October2014),
64
could,atfirstglance,suggestcontinuedindustrialoildemandgrowth.Butthisauthorincreasingly
fearsthatlocalandcorporatevestedinterests’resistancetocentraledictsaimedatre-balancing
theChineseeconomycouldactuallyperpetuatefurtherpoliticaldysfunction,reduceclarityand
certainty,andultimatelydeepenChina’sstructuraleconomicchallengesinwaysthatdepressoil
demandgrowth.
Ultimately, these realities—combined with ongoing issues of data transparency—can magnify
pricevolatility.Forinstance,announcingastimuluspackagecandriveupglobalcrudeoilprices,
thenamplifyaselloffashorttimelaterifthemarketrealizesthatthepotentialdemandeffects
wereoverpriceddueinparttolackofsufficientdatatoproperlycontextualizethemeasure.
ThestakesarehighbecauseChinahassimultaneouslybecomeacoredriverofglobaloildemand
growth while also fueling this growth through the construction of a debt-based, increasingly
unsustainable economic structure. Some of that risk has been priced into the global crude
market,butlikelytoaninsufficientdegree.Considerforinstancethatinthe1997AsianFinancial
Crisis,SouthKorea,amajorindustrializedeconomy,sawoilproductsdemandcrashbymorethan
14%year-on-yearin1998.Toboot,SouthKoreanoildemandtooknearlyfiveyearstorecover
afterthatmassiveinternalshock.
China’ssituationisdifferentinmanyimportantways,butinanenvironmentofsloweroverall
global growth, rising pressure from monetary policy distortions, and a turn to populist and
nationalist politics in many key economies (arguably including China), the potential for a
significantnear-termdemandshockshouldnotbeunderestimated.ThebottomlineisthatChina
was a primarily upside risk factor in the oil markets for much of the past 15 years, but now
presentsacomplexcombinationofcontinueddemandgrowthpotential,thatrisksbeingoffset
byincreasinglysignificantlatentdownsiderisks.
http://www.chinasignpost.com/2014/10/23/physician-heal-thyself-modest-expectations-in-order-for-chinasreforms-as-third-plenum-anniversary-approaches/.
65