---
mi_id: MIMR-node_7ff0734
type: market_report
canonical_url: "https://www.pheonixresearch.com/automotive-transportation/vehicle-manufacturing-assembly/market-report/global-automotive-market/"
md_url: "https://www.pheonixresearch.com/automotive-transportation/vehicle-manufacturing-assembly/market-report/global-automotive-market.md"

schema:
  "@type": Dataset
  "@id": MIMR-node_7ff0734
  url: "https://www.pheonixresearch.com/automotive-transportation/vehicle-manufacturing-assembly/market-report/global-automotive-market/"
  title: "Global Automotive Market Report, Size, Share and Forecast 2026–2033: 99.8 Million Vehicles, 61% of Them Built in Asia, at a Price That Has Stopped Rising"
  description: "The global automotive market is USD 4,241bn in 2025, reaching USD 5,982bn by 2033 at 4.39%. 99.8m vehicles sold, 96.4m built, 61.4% of output in Asia, 25% of new cars electric, and a US transaction price that rose just 0.8% against incentives of 7.5%."
  datePublished: "2025-09-18T08:26:47+00:00"
  dateModified: "2026-11-17T15:05:15+00:00"
  keywords:
    - automotive market size
    - global vehicle sales 2025
    - automotive market forecast 2033
    - OICA production statistics
    - electric vehicle share
    - automotive value chain
    - Chinese vehicle exports
    - automotive tariffs
    - vehicle affordability
    - automotive aftermarket
  isPartOf:
    - id: MISG-node_fb2d5b0
      type: CollectionPage
      url: "https://www.pheonixresearch.com/automotive-transportation/vehicle-manufacturing-assembly/"
      name: "Vehicle Manufacturing & Assembly"
    - id: MIIN-node_233f7ed
      type: CollectionPage
      url: "https://www.pheonixresearch.com/automotive-transportation/"
      name: "Automotive & Transportation"
    - id: MIWS-root
      type: WebSite
      url: "https://www.pheonixresearch.com/"
      name: Pheonix Research
  mentions:
    - id: MIEN-node_3c2f8253
      type: Organization
      url: "https://global.toyota/en"
      name: Toyota
      sameAs:
        - "https://www.wikidata.org/wiki/Q53268"
    - id: MIEN-node_35f0fca5
      type: Organization
      url: "https://www.volkswagen-group.com/en"
      name: Volkswagen Group
      sameAs:
        - "https://www.wikidata.org/wiki/Q156578"
    - id: MIMR-node_368f895
      type: Dataset
      url: "https://www.pheonixresearch.com/automotive-transportation/passenger-vehicle-tyres/market-report/global-hatchback-sedan-tyres-market/"
      name: "Global Hatchback & Sedan Tyres Market Report, Size, Share and Forecast 2026–2033: 780 Million Tyres, and the Duty That Repriced the Cheapest of Them"
    - id: MIMR-node_f287a2c
      type: Dataset
      url: "https://www.pheonixresearch.com/automotive-transportation/off-highway-specialty-tyres/market-report/global-off-the-road-otr-tyre-market/"
      name: "Global Off-the-Road (OTR) Tyres Market Report, Size, Share and Forecast 2026–2033: 2.03% More Rubber and 4.06% More Money — the Gap Is the Business"
    - id: MIMR-node_66e429f
      type: Dataset
      url: "https://www.pheonixresearch.com/automotive-transportation/commercial-vehicle-chassis-braking/market-report/global-commercial-vehicle-air-disc-brake-market/"
      name: "Global Commercial Vehicle Air Disc Brake Market Report, Size, Share and Forecast 2026–2033: The Conversion Has a Deadline, Not a Demand Curve"
    - id: MIMR-node_b21bbaf
      type: Dataset
      url: "https://www.pheonixresearch.com/automotive-transportation/commercial-vehicle-chassis-braking/market-report/global-commercial-vehicle-brake-friction-products-market/"
      name: "Global Commercial Vehicle Brake Friction Products Market Report, Size, Share and Forecast 2026–2033: Four Dollars in Five Are Spent After the Truck Is Built"
    - id: MIMR-node_690f2c4
      type: Dataset
      url: "https://www.pheonixresearch.com/automotive-transportation/road-safety-enforcement-systems/market-report/global-road-safety-enforcement-systems-market/"
      name: "Global Road Safety & Enforcement Systems Market Report, Size, Share and Forecast 2026–2033: The Camera Was Never the Product"
  about:
    - "@type": Thing
      name: Market Intelligence
      mi_id: "MIMR-node_7ff0734#intelligence"
    - "https://www.wikidata.org/wiki/Q1901028"
  base_year: 2025
  forecast_year: 2033
  value_base_year: 4241
  value_forecast_year: 5982
  value_cagr: 4.39
  value_currency: USD
  value_unit_scale: Trillion

acf:
  base_year: 2025
  forecast_year: 2033
  value_base_year: 4241
  value_forecast_year: 5982
  value_cagr: 4.39
  value_currency: USD
  value_unit_scale: Trillion
  coverage_type: Global
  coverage_name: "Asia-Pacific and Middle East, Europe, North America, South America, Africa"
  competitive_intensity_level: high
  market_structure_type: moderately_consolidated
  tier1_player_count: 10
  investment_trend_direction: rising
  capital_intensity_level: high
  recent_mna_activity: yes
  overall_market_risk_level: moderate
  geopolitical_exposure_level: high
  substitution_risk_level: moderate
  regulatory_complexity_level: high
  approval_pathway_structure: multi_jurisdictional
  innovation_intensity_level: high
  technology_maturity_stage: growth
  patent_activity_level: high
  supply_chain_complexity_level: high
  distribution_structure: retail_chain_dominated
  primary_operational_model: hybrid
  executive_summary: "<p><strong>The global automotive market is USD 4,241 billion in 2025, reaching USD 5,982 billion by 2033 at a 4.39 per cent value CAGR, on 99.8 million vehicles sold and 96.4 million built.</strong> The industry returned to growth in 2025 — sales up 4.7 per cent, production up 3.9 per cent — but the growth was almost entirely geographic and almost none of it was financial. Units compound at 2.10 per cent across the forecast window and revenue per vehicle at 2.4 per cent, which is a content story, not an inflation story, and it produces a value CAGR roughly two and a half points below the rate this report previously carried.</p><p><strong>Production has moved east and the move is structural, not cyclical.</strong> Asia-Oceania built 59.2 million vehicles in 2025, 61.4 per cent of world output, and China alone built 34.53 million — up 10 per cent in a year in which Europe's output fell and the Americas' fell faster. Emerging economies out-built the developed world for the first time, 58.2 million against 38.2 million, which is 60.4 per cent of global output. Europe built 17.20 million against sales of 18.63 million, so it imported 1.43 million net; Asia built 4.19 million more than it sold; the Americas sold 6.12 million more than they built. <strong>The geography of the value chain is now the geography of the trade balance, and each region's exposure to tariffs is simply its number.</strong></p><p>The price line is where this report departs from consensus. The US average transaction price hit an all-time high of USD 50,326 in December 2025 — and rose 0.8 per cent year on year, with incentives at 7.5 per cent of transaction price to hold the volume. <strong>The industry is buying its growth at a discount rather than earning it through price, and the average is being held up by mix — full-size pickups at USD 66,386, luxury at close to a fifth of December's shoppers — not by pricing power.</strong> That is why 4.39 per cent, not 6.15, is the defensible rate: a 6.15 per cent value CAGR on 2.10 per cent unit growth requires eight straight years of 4 per cent transaction-price inflation, and the year just closed delivered 0.8.</p><p><strong>Electrification has stopped being an American story and is now a Chinese and European one.</strong> More than 20 million electric cars were sold in 2025, 20 per cent up, for a 25 per cent share of global new car sales — one in four. China sold more than 13 million of them at close to 55 per cent of its new car market, Europe rose 30 per cent to more than 4 million, and the United States held just below 10 per cent with sales falling sharply in the final quarter once tax credits ended. Two million electric cars were sold outside those three markets, more than half of them in countries that have now passed a 10 per cent share. <strong>The technology's growth no longer depends on the market that invented the subsidy.</strong></p>"
  toc: "<ol>
<li><strong>Market Snapshot</strong></li>
<li><strong>Market Definition and the Three Boundary Calls</strong></li>
<li><strong>Market Size and Forecast</strong>
<ol>
<li>3.1 The 2025 Base Was Built From Units and Prices, Not From a Value Estimate</li>
<li>3.2 The Regional Allocation Is the Business Model</li>
<li>3.3 The Decomposition Is the Finding</li>
</ol>
</li>
<li><strong>Four Structural Shifts Defining 2026–2033</strong>
<ol>
<li>4.1 Production Moved East and Is Not Moving Back</li>
<li>4.2 The Transaction Price Stopped Rising</li>
<li>4.3 Electrification Became a Non-American Story</li>
<li>4.4 The 3.4 Million Vehicle Gap Nobody Explains</li>
</ol>
</li>
<li><strong>Market Drivers</strong>
<ol>
<li>5.1 Replacement Demand in a Parc of Roughly 1.6 Billion Vehicles</li>
<li>5.2 Content Per Vehicle Rises Even as Part Count Falls</li>
<li>5.3 Emerging Asia Is Still Buying Its First Car</li>
<li>5.4 The Aftermarket Lags the Sale by Seven Years</li>
<li>5.5 Localisation Subsidies Are Demand Policy in Disguise</li>
</ol>
</li>
<li><strong>Market Restraints</strong>
<ol>
<li>6.1 Affordability Is Now the Binding Constraint</li>
<li>6.2 Trade Friction Is a Cost Line, Not a Risk Line</li>
<li>6.3 Chinese Capacity and the Export Disposal Problem</li>
<li>6.4 Incentive Dependency in the Retail Channel</li>
<li>6.5 Repair and Insurance Costs Became a Purchase Barrier</li>
<li>6.6 The Battery Cost Floor Has Stopped Falling</li>
</ol>
</li>
<li><strong>Segmentation</strong>
<ol>
<li>7.1 By Propulsion</li>
<li>7.2 By Body Type and Segment</li>
<li>7.3 By Region of Assembly</li>
<li>7.4 By Vehicle Class and Duty Cycle</li>
<li>7.5 By Channel</li>
</ol>
</li>
<li><strong>Regional Analysis</strong>
<ol>
<li>8.1 Asia-Pacific and the Middle East</li>
<li>8.2 Europe</li>
<li>8.3 North America</li>
<li>8.4 South America</li>
<li>8.5 Africa</li>
<li>8.6 What the Geography Says</li>
</ol>
</li>
<li><strong>Value Chain</strong></li>
<li><strong>Technology and Innovation</strong></li>
<li><strong>Competitive Landscape</strong></li>
<li><strong>Regulatory Landscape</strong></li>
<li><strong>Investment Activity</strong></li>
<li><strong>Market Risk</strong></li>
<li><strong>Strategic Outlook</strong></li>
<li><strong>Watch List</strong></li>
<li><strong>Methodology</strong></li>
<li><strong>About This Report</strong></li>
</ol>
"
  competitive_landscape_content: "<h2>Competitive landscape: sixty-seven per cent of the market in twelve groups, and none of them with pricing power</h2><p>The twelve largest groups sold approximately <strong>66.7 million of the 99.8 million vehicles sold worldwide in 2025, about 66.8 per cent</strong> of the market. That is a high concentration ratio for a capital-intensive industry, and it would normally imply pricing power. In 2025 it implied the opposite: the same twelve groups competed through 7.5 per cent incentives on a US average transaction price that grew 0.8 per cent. Concentration has bought them cost scale and has bought them no pricing.</p><table border=\"1\" cellspacing=\"0\" cellpadding=\"8\"><tbody><tr><th>Group</th><th>Home</th><th>Units sold</th><th>Basis</th></tr><tr><td>Toyota Group</td><td>Japan</td><td>11,320,000</td><td>With Daihatsu and Hino</td></tr><tr><td>Volkswagen Group</td><td>Germany</td><td>8,983,900</td><td>Deliveries; includes 305,600 trucks and buses</td></tr><tr><td>Hyundai Motor Group</td><td>South Korea</td><td>7,244,408</td><td>Hyundai with Genesis, plus Kia</td></tr><tr><td>General Motors</td><td>United States</td><td>6,180,000</td><td>Including Chinese joint ventures</td></tr><tr><td>Stellantis</td><td>Netherlands</td><td>5,480,000</td><td>Rounding; alternative counts differ</td></tr><tr><td>BYD</td><td>China</td><td>4,602,436</td><td>Battery electric and plug-in hybrid only</td></tr><tr><td>SAIC Motor</td><td>China</td><td>4,507,000</td><td>Including joint ventures</td></tr><tr><td>Ford</td><td>United States</td><td>4,395,000</td><td>Wholesales</td></tr><tr><td>Geely Holding</td><td>China</td><td>4,110,000</td><td>With Volvo Cars, Zeekr, Lynk & Co, Proton</td></tr><tr><td>Honda</td><td>Japan</td><td>3,387,000</td><td>Fiscal year to March 2026</td></tr><tr><td>Suzuki Motor</td><td>Japan</td><td>3,319,975</td><td>Fiscal year to March 2026; cars only</td></tr><tr><td>Nissan</td><td>Japan</td><td>3,150,000</td><td>Fiscal year to March 2026</td></tr></tbody></table><p><em>Group volumes as each group reports them. Four are fiscal-year figures rather than calendar 2025, two are rounded at source, and GM and SAIC include joint ventures that other counts attribute differently. The table is a ranking, not a share calculation to three decimal places.</em></p><h3>What the group table says</h3><p>Four things, and none of them is the ranking. First, <strong>the top of the table is stable and the middle is not</strong>: Toyota and Volkswagen have held the first two positions through a decade in which the industry's technology changed completely, because scale in assembly is transferable across powertrains even when scale in components is not. Second, <strong>three of the twelve are Chinese and two of those are inside the top seven</strong>, which is a different fact from China building 34.53 million vehicles — the country's output was for years assembled by joint ventures whose brand belonged to somebody else. Third, the <strong>spread between first and twelfth is only 3.6 times</strong>, which is unusually flat for a capital-intensive industry and is a direct consequence of the economics in the value chain chapter: a plant is expensive but it is not a barrier, and a brand is transferable within a region. Fourth, and most important, <strong>there is no group on the list whose volume position protects it from the price data</strong>.</p><h3>Two different consolidation stories</h3><p>The industry is consolidating for two unrelated reasons and the two are usually reported as one. The first is <strong>technology consolidation</strong>: platforms, batteries and software have become so expensive that they must be amortised across more units, which drives alliance and platform-sharing behaviour rather than outright merger. The second is <strong>capacity consolidation</strong>: plants that cannot be loaded are closed, and the volume migrates to plants that can. The first increases the number of units per engineering dollar. The second reduces the number of plants. Europe is currently doing the second on a scale it has not attempted since the 1980s, which is why Italy at 474,044 vehicles and the United Kingdom at 764,715 are the chapter's most informative numbers.</p><h3>The Chinese group problem</h3><p>BYD's 4.60 million units are <strong>battery electric and plug-in hybrid only</strong>, which means its position in the electric market is materially different from its position in the total market: for every combustion vehicle sold by a legacy group, BYD has no competing product. That is a vulnerability as much as a strength outside the markets where electrification is policy-driven, and it is why the group's international expansion has concentrated in markets with a regulatory ratchet — Europe above all, where electric sales rose 30 per cent past 4 million on a CO<sub>2</sub> step change.</p><p>The strategic consequence is that <strong>China's export weight is carried by groups whose product range is narrower than their volume suggests</strong>, while Japan's export weight — a 3.84 million unit surplus from a 4.57 million unit home market — is carried by groups with a full range and a distribution network in every deficit region. Those are different kinds of competitive threat and they call for different responses.</p><h3>Share is not margin</h3><p>The most misleading number in this industry is market share, because the segments that carry the volume are not the segments that carry the profit. In the United States, full-size pickups averaged USD 66,386 in December 2025 against an industry average of USD 50,326, and close to a fifth of that month's shoppers bought luxury. A group with 12 per cent of units and 6 per cent of trucks has a worse business than one with 8 per cent of units and 20 per cent of trucks. <strong>The right competitive metric in 2026 is revenue per unit against content per unit, not share of a unit count</strong> — because the industry has proven it can buy share with incentive and has not proven it can buy revenue with anything.</p><h3>Who is actually gaining</h3><p>Gains in 2025 came from three places, and only one of them is a group-level story. <strong>Geography</strong>: Asia-Oceania added 4.2 points of production share in a single year, and emerging economies passed the developed world outright at 58.2 million against 38.2 million. <strong>Powertrain</strong>: battery electric vehicles recovered to 65 per cent of electric sales, reversing two years of hybrid drift, which reordered the competitive field within the electric segment without changing the total. <strong>Price tier</strong>: mix moved up, with luxury and large trucks carrying the average, which benefits the groups with the deepest premium ranges and penalises the ones competing at the entry price point where the affordability constraint binds hardest.</p><h3>The barriers that remain</h3><p>Only four. <strong>Capital at scale</strong>, which is why the top twelve are stable — a new platform and a new plant together are a multi-billion commitment with a five-year payback. <strong>Regulatory certification</strong>, which is expensive, jurisdiction-specific and increasingly about software and emissions conformity rather than crash structure. <strong>Distribution</strong>, which is the moat nobody prices: a group with 3,000 franchised outlets in a deficit region can place volume in a way a new entrant cannot, and that is why Chinese expansion in Europe has been slower than Chinese production growth suggests. And <strong>residual value</strong>, which is the barrier that binds hardest in a market where affordability is the constraint — a brand that holds value can charge more for the same content, and no amount of manufacturing scale buys it.</p><p><strong>The competitive conclusion is uncomfortable: the twelve groups that hold two-thirds of this market are locked in a cost competition they cannot win by raising price, in regions where their capacity is shrinking, against a competitor whose product range is narrower but whose cost base is lower.</strong> That is not a consolidation story. It is a margin story, and it is the reason this report forecasts units and content rather than price.</p>"
  value_chain_content: "<h2>The value chain: where the USD 4.24 trillion actually sits</h2><p>An average vehicle transacts at USD 30,471, and almost nothing about that number is decided by the company whose badge is on the grille. The figure is the output of a ladder that runs from a mine and a steel mill to a dealer forecourt, and the ladder's economics — who holds pricing power, who absorbs a cost shock, who gets paid twice — determine whether a market growing 4.39 per cent a year is a good business to be in. For most of the ladder it is not; for two rungs of it, it is one of the best.</p><h3>The tier structure and the margin ladder</h3><p>The chain has five economically distinct layers. Raw materials — steel, aluminium, resin, glass, copper, and the battery metals — are commodity businesses priced off an index with no pricing power and no differentiation. Tier two and tier three component makers convert those materials into parts on a cost-plus basis, typically with annual productivity concessions written into the contract, which means their revenue rises with volume while their margin is negotiated downward every year. Tier one suppliers integrate parts into systems — a braking module, a seating system, a battery pack, an infotainment head unit — and this is the only supplier layer with genuine engineering pricing power, concentrated in the handful of firms that can carry a system across several vehicle programmes. The assembler integrates systems into a vehicle and sells it, at an operating margin that in a good year is high single digits. Below that sit distribution and retail, and then the aftermarket.</p><p>The structural point that follows is the one the industry keeps rediscovering: <strong>the assembler carries the volume risk and the tier one carries the engineering risk, but the aftermarket carries neither.</strong> Parts and service are 28.3 per cent of this market's value and are sold into an installed base of roughly 1.6 billion vehicles whose owners have already made the purchase decision. That is why the aftermarket block compounds at 4.0 per cent against 2.10 per cent unit growth, and why every assembler that can build a service, subscription or software revenue line is trying to own a piece of it.</p><h3>The powertrain swap re-sorts the tiers</h3><p>Electrification does not simply add a supplier; it deletes a chain and builds another one alongside it. An internal combustion powertrain is a mature, deeply optimised network of perhaps two thousand part numbers spread across engine, transmission, exhaust, fuel and cooling, and it is populated by supplier firms whose entire engineering capital is invested in it. A battery-electric powertrain replaces that with a pack, an e-machine, an inverter and a thermal system: dramatically fewer parts, dramatically fewer suppliers, and a much larger share of the vehicle's cost sitting with a handful of firms that can make cells at scale.</p><p>Two consequences follow, and both are visible in the 2025 data. First, <strong>the part count falls while the value rises</strong>, which is the mechanism behind this report's 2.4 per cent revenue-per-unit line — fewer components, more expensive ones, and less labour content per vehicle. Second, the supplier base is being re-sorted faster than the assembly base is, because an assembler can retool a line in three years while a tier two stamping or castings supplier cannot convert its capital at all. The 6.49 million vehicles India built and the 34.53 million China built were largely assembled by firms that already had the new chain; the 474,044 Italy built were not.</p><h3>Where content per vehicle is actually added</h3><p>Content grows in four places, and they behave differently. The <strong>pack</strong> is the largest single line in a battery-electric bill of materials and is the reason a US electric vehicle transacted at USD 58,034 against an industry average of USD 50,326 — the premium is the pack, amortised through the monthly payment. <strong>Power electronics</strong> add value per vehicle while reducing cost per kilowatt, which is a rare combination and the reason silicon carbide adoption is a cost story rather than an efficiency story. <strong>Driver assistance</strong> adds sensor, compute and wiring content to every vehicle regardless of propulsion, which makes it the most broadly distributed content driver in the industry. <strong>Thermal management</strong> grows in both powertrains, because a pack needs cooling and a fast-charging pack needs more of it, and it is the segment where the internal-combustion and electric chains overlap most profitably.</p><p>Against that, content falls in the places nobody writes about: fewer castings, fewer fasteners, fewer fluid circuits, less machining, less assembly labour. A vehicle built in 2033 will carry more value in fewer parts, and the firms that make their living from the parts being removed are the ones the forecast is least kind to.</p><h3>Raw materials and the asymmetry of pass-through</h3><p>The chain absorbs commodity shocks unevenly, and the asymmetry is structural rather than cyclical. Steel and aluminium pass through to the assembler within one contract cycle, because there are alternate suppliers and a published index. Battery metals do not pass through cleanly, because the cell maker has signed multi-year offtake and the pack price is quoted to the assembler twelve to twenty-four months forward. The result is that <strong>a metal price move lands on the balance sheet of whichever layer happens to be holding the forward position</strong>, which is why pack pricing has been stickier than metal pricing in both directions.</p><p>This matters for the forecast because it means the cost curve that drove the first decade of electrification is not reliably available for the second. The 2025 European step change in electric sales came from a regulatory ratchet, not from a price signal, and the chain's cost structure is the reason.</p><h3>The geography of the chain is the trade balance</h3><p>The chain's location decisions show up in the only fully observed dataset in this report. Asia built 4.19 million more vehicles than it sold in 2025; Europe built 1.43 million fewer than it sold; the Americas built 6.12 million fewer than they sold. Those three numbers are the sum of thousands of individual sourcing decisions, and they produce a market in which <strong>the region with the lowest conversion cost is also the region with the largest surplus, and the deficit regions respond with tariffs that raise the cost of the vehicles their own buyers need.</strong></p><p>Within the surplus there is a second layer worth separating. China built 34.53 million and sold 34.40 million, so its domestic market absorbs almost its entire output — its surplus is small, and its export weight comes from growth rather than from an absolute glut. Japan built 8.41 million and sold 4.57 million, which is a 3.84 million surplus from a market of 4.57 million: the largest export ratio of any major producer. The two are often discussed as one phenomenon and they are not.</p><h3>The aftermarket is the profitable half</h3><p>USD 1,200 billion of parts and service, growing 4.0 per cent, lagging the vehicle sale by roughly seven years, and largely insulated from the new-vehicle price war because the alternative to a repair is not a cheaper repair — it is a different vehicle or no repair at all. This is the block of the chain where a distributor, an independent workshop network or a parts brand can earn a return without carrying model-cycle risk.</p><p>It is also the block most exposed to electrification's second-order effects. A battery-electric vehicle has no engine oil, no filter, no exhaust, no spark plug and materially fewer brake replacements, because regenerative braking does much of the deceleration. In its first several years it therefore generates <strong>less</strong> service revenue than the combustion vehicle it replaced — and more tyre, body and electronics revenue. The aftermarket does not disappear; it changes address, and the firms that own the old address lose.</p><h3>What the chain cannot do</h3><p>It cannot expand capacity quickly: a new assembly plant is a three-to-five year project and a new cell plant is longer. It cannot cut price to create demand, because 7.5 per cent of transaction price is already incentive. And it cannot move a plant to follow a tariff within a model cycle. <strong>Those three inflexibilities, more than any demand assumption, are what turn a 4.7 per cent unit growth year into a 4.39 per cent value forecast for the following eight.</strong></p>"
  investment_activity_content: "<h2>Investment activity: capital is moving faster than demand</h2><p>The industry's investment problem in 2026 is not a shortage of capital. It is that capital is being committed to capacity in one set of geographies while demand sits in another, and the instrument being used to bridge the gap is policy rather than price. The 2025 production data makes the reallocation visible: <strong>emerging economies built 58.2 million vehicles and the developed world 38.2 million</strong>, and within that China added roughly 3.25 million units of output in a single year, India built 6.49 million, Mexico 4.09 million, Morocco 501,965 and Egypt 89,570, up 62 per cent.</p><h3>Where the capital is going, and why</h3><p>Three distinct investment motives are operating at once, and they produce very different returns.</p><p><strong>Capacity for a protected market.</strong> A plant built behind a tariff or a localisation requirement earns a return that is a function of the tariff, not of the product. Mexico's 4.09 million units of output against 1.56 million units of domestic sales is the clearest example in the dataset: a country that builds 2.5 times what it sells is an export platform, and its investment case rests on access rather than on domestic demand. Türkiye, at 1.42 million built and 1.41 million sold, is the opposite configuration and a more balanced one.</p><p><strong>Capacity for a cost base.</strong> Morocco's 501,965 units and Egypt's 89,570 are not large numbers, but they are a cost position — labour, energy and proximity to the European deficit of 1.43 million units. <strong>A plant in a low-cost location serving a high-cost deficit market is the most reliably profitable configuration in this industry right now</strong>, which is precisely why the deficit market is writing localisation rules against it.</p><p><strong>Capacity for a regulatory ratchet.</strong> Cell and pack plants are being built where the standard is tightening, because the standard is the demand forecast. Europe's 30 per cent electric growth to more than 4 million units was created by a CO<sub>2</sub> step change, and the investment that supplies it was justified by the standard rather than by consumer research. That makes it the least elastic and most politically exposed category of the three.</p><h3>The stranded asset problem, in numbers</h3><p>The mirror of new capacity is idle capacity, and the European figures are the clearest reading of what it costs. <strong>Italy built 915,291 vehicles in 2019 and 474,044 in 2025 — a fall of 48.2 per cent.</strong> The United Kingdom built 1,381,405 and 764,715. The Netherlands built zero cars in both 2024 and 2025. These are not plants that were closed; they are plants that were loaded down, which is the more expensive outcome because the fixed cost stays and the contribution disappears.</p><p>The investment consequence is that Europe's remaining capital has to be spread across fewer units. Germany built 4.15 million cars and remains the continent's base, but a continent at 17.20 million units of output against 18.63 million of sales cannot justify the platform investment its regulatory targets require without either importing the vehicles or subsidising the plant. <strong>Both are the same decision: pay for the transition in the exchequer or in the trade balance.</strong></p><h3>M&A and alliances: technology consolidation, not capacity consolidation</h3><p>Deal activity in this industry is concentrated where the fixed cost is, and the fixed cost is now in platforms, software and cells rather than in stamping presses. That produces alliance and platform-sharing behaviour — two groups sharing a platform, three sharing a cell specification — rather than the large horizontal mergers that consolidation ratios usually imply.</p><p>The reason is that a merger buys capacity, and capacity is the thing the industry already has too much of. What a group cannot buy at a sensible price is a battery supply relationship or a validated software stack, so that is where the money goes. <strong>The predictable consequence is that the twelve groups holding 66.8 per cent of the market will still be twelve groups in 2033, but they will share more of their engineering than they do now</strong> — and the suppliers that hold the shared technology will hold the pricing power.</p><h3>Return on capital, and the honest problem</h3><p>The industry's investment case is arithmetic: it must fund a platform transition and a capacity reallocation out of a margin that is being compressed by 7.5 per cent incentives on a transaction price growing 0.8 per cent. <strong>Capital intensity is rising, unit growth is 2.10 per cent, and price is flat.</strong> That combination has one of two resolutions — either the content-per-vehicle line in this forecast holds and revenue per unit rises 2.4 per cent a year, or the industry spends the decade funding a transition out of a flat revenue base.</p><p>The most important investment signal in the 2025 data is therefore not any single plant announcement. It is that <strong>battery-electric vehicles recovered to 65 per cent of electric sales</strong>, reversing two years of drift toward hybrids and range-extenders. A group that had hedged its capital across hybrids and been vindicated by 2023–24 had to re-weight toward pure electric in 2025, which means the capital committed to hybrid lines is now the most exposed position in the industry.</p><h3>What would change the investment case</h3><p>Three things, in order of consequence. <strong>A durable rise in transaction price</strong> would restore the return on content and unlock the higher value path. <strong>A settlement on trade</strong> would let capacity be sited for cost rather than for access, which would materially improve returns on the export-platform investments in Mexico, Morocco and Türkiye. <strong>A step change in pack cost</strong> would move the entry price point down far enough to reach first-time buyers in India, Southeast Asia and Africa, where the demand is not constrained by replacement cycles — and where the 99.8 million unit base has the most room to grow.</p>"
  technology_innovation_content: "<h2>Technology and innovation: the content is the market</h2><p>The industry's innovation problem is not a shortage of technology. It is that almost every significant technology in the pipeline adds cost to the vehicle before it adds revenue to the maker, and the 4.39 per cent value forecast is the arithmetic of how much of that cost can be recovered. This chapter examines where the content is being added, what it costs, and which parts of the innovation agenda have stopped paying.</p><h3>The pack is the bill of materials</h3><p>The single most important fact about automotive technology in 2026 is that the battery pack is the largest line item in an electric vehicle's cost, and therefore that <strong>cell economics set the price of the product</strong>. The US transaction data shows the consequence plainly: the average new electric vehicle sold for USD 58,034 in December 2025 against an industry average of USD 50,326, a USD 7,708 premium that is essentially the pack, financed through the monthly payment rather than paid at the till.</p><p>The chemistry roadmap is where the cost fight is being fought. Iron-phosphate chemistry has taken the volume segment because it is cheaper and safer per kilowatt-hour and the range penalty is acceptable in a car used for commuting; sodium-ion is being developed for the entry segment and for stationary applications where energy density matters less than cost and cold-weather behaviour. Neither is a performance story. Both are attempts to move the pack off the part of the cost curve that nickel and cobalt set, and their success is measured in dollars per kilowatt-hour, not in watt-hours per kilogram.</p><p>The 2025 evidence that this is working is indirect but strong: <strong>more than 20 million electric cars sold, 20 per cent up, in a year when the largest single subsidising market withdrew its support mid-year.</strong> A product whose sales survive the removal of a purchase incentive has reached cost parity in at least the segments where it sells — which, given that China sold more than 13 million of them at close to 55 per cent of its new car market, is now a very large segment.</p><h3>Power electronics and the 800-volt question</h3><p>Silicon carbide inverters are the most consequential component technology of the decade that almost nobody outside the supply chain discusses, because the business case for them is not efficiency. It is that a more efficient inverter lets an assembler fit a smaller, cheaper pack to achieve the same range — which means the inverter pays for itself out of the battery budget rather than out of the vehicle's price.</p><p>That is a genuinely different kind of innovation from the ones the industry is used to. It reduces cost while it improves the product, and it therefore diffuses faster than a technology that only improves the product. The 800-volt architecture question follows from the same logic: higher voltage reduces current for a given power, which reduces conductor mass and thermal load, which allows faster charging with less cooling hardware. Whether an assembler adopts it is a decision about its charging network and its pack supplier, not about its ambitions.</p><h3>Software-defined vehicle: the value without the revenue</h3><p>The industry has spent a decade arguing that software will convert a low-margin hardware business into a high-margin services business. The 2025 data does not support the strong version of that claim and does support a weaker one. <strong>Software is unquestionably adding cost per vehicle</strong> — compute, sensors, wiring, and the engineering organisation to write and validate code — and it is adding that cost to a vehicle whose transaction price grew 0.8 per cent. The revenue side has arrived mostly as content bundled into the price rather than as a separate subscription line.</p><p>The weaker claim that survives is this: software's value in this market is currently expressed as <strong>mix and retention</strong> rather than as subscription revenue. It is what allows an assembler to sell a higher trim, to keep a vehicle's residual value up by updating it, and to hold a customer in the franchise for the second purchase. Those are real and they are the reason the investment continues, but they do not change the value forecast, because they do not change the price the market will pay.</p><h3>Driver assistance as content, and as cost</h3><p>Driver assistance is the most broadly distributed content driver in the industry because it applies to every propulsion type and every segment. It adds cameras, radar, ultrasonic sensors, a compute module, wiring harness capacity and a calibration burden at the dealer — and the calibration burden is the part the industry consistently underestimates, because it lands in the aftermarket as a cost the owner pays after a windscreen replacement.</p><p>That is the mechanism connecting this chapter to the restraints in the main body. <strong>Content per vehicle rises, and so does the cost of repairing it.</strong> A vehicle with expensive, calibrated electronics is more expensive to insure, and insurance is part of the monthly payment. The industry has been adding content to the product faster than it has been adding the service capacity to maintain it, and the affordability constraint now sits on both sides of that gap.</p><h3>Manufacturing: the part-count collapse</h3><p>The least discussed innovation in the industry is in the press shop. Large single-piece castings replace dozens of stamped and welded parts, reducing tooling count, cycle time, floor space and headcount at the same time. Combined with the powertrain swap, this is why the forecast assumes revenue per vehicle rising at 2.4 per cent while unit growth is 2.10 — the value is going up by less than the cost of the content that replaces it.</p><p>The labour consequence is straightforward and is visible in the production data. The vehicle built in 2033 needs fewer people per unit than the vehicle built in 2019, which is why a European assembly base that built 17.20 million vehicles in 2025 supports fewer jobs than one that built the same number in 2019, and why the political cost of a plant closure is now higher than the economic one.</p><h3>What did not happen</h3><p>Three expected technologies have not arrived on schedule. <strong>Solid-state batteries</strong> remain a small-volume proposition; the chemistry roadmap that matters for this forecast window is iron-phosphate and sodium-ion, both of which are incremental. <strong>Full autonomy</strong> at scale has not materialised, and the industry has quietly repriced it from a product to a feature. <strong>Hydrogen</strong> has consolidated into heavy duty and industrial applications and is not a passenger-car answer in this forecast.</p><p>Each of the three absences matters, because each was in somebody's model as a content-per-vehicle driver. Their deferral is part of why the value CAGR is 4.39 per cent rather than the higher number a technology-led forecast would have produced. <strong>The industry's realistic innovation agenda for 2026–2033 is a cost agenda wearing a technology costume</strong> — cheaper packs, cheaper inverters, cheaper assembly — and the firms that win it will be the ones that pass the saving to the buyer rather than booking it as margin, because the price data says the buyer has stopped paying.</p>"
  market_risk_content: "<h2>Market risk: the seven things that would break this forecast</h2><p>This forecast is not fragile. Units growing 2.10 per cent and content growing 2.4 per cent produce 4.39 per cent, and both are conservative relative to the industry's own history. What is fragile is the <em>margin</em> inside that number, because the industry has one recovery mechanism — price — and the 2025 data says it has stopped working. The risks below are ordered by how much of the forecast they can actually move, not by how often they are discussed.</p><h3>Risk one: affordability, and the fact that price has stopped</h3><p>This is the only risk that can move the whole forecast. The US average transaction price reached USD 50,326 in December 2025, an all-time high, and grew 0.8 per cent for the year with incentives at 7.5 per cent of transaction price. <strong>A market at a record price, growing under 1 per cent, discounting 7.5 per cent to hold volume, is a market with no remaining capacity to absorb a cost increase.</strong></p><p>The exposure is direct. If transaction prices are flat rather than rising 2.4 per cent, the value forecast falls from 4.39 per cent toward 2.5 per cent over the window — roughly USD 900 billion of 2033 value removed. The mitigation is mix, and mix is a holding pattern rather than an engine: it works while luxury and full-size pickups carry the average, and it fails when affordability pressure reaches the segments that carry the volume.</p><h3>Risk two: trade policy, which is a cost with a date</h3><p>OICA's own assessment of the 2025 data flags growing exposure to trade frictions, tariff risk and localisation pressures. The structural arithmetic makes escalation likely rather than possible: the Americas sell 6.12 million more vehicles than they build, Asia builds 4.19 million more than it sells, and Europe is short 1.43 million units. <strong>Three deficit regions facing one surplus region is not a stable configuration, and the instrument that responds fastest is a tariff.</strong></p><p>The reason this ranks second rather than first is that a tariff redistributes volume and raises cost, but it does not remove demand. The reason it still ranks second is that it can do both at once in a single market if it is large enough, and the American market is where the industry's profit pool sits.</p><h3>Risk three: the export disposal problem</h3><p>China added roughly 3.25 million units of output in 2025 and sold 34.40 million against 34.53 million built, so its own market absorbed nearly all of it. That is the reassuring reading. The unreassuring one is that <strong>a production base of 34.53 million units growing 10 per cent a year cannot be absorbed indefinitely by a domestic market whose sales growth is slowing</strong>, and the surplus has to go somewhere. Every percentage point of Chinese output growth beyond domestic demand is roughly 350,000 vehicles of export pressure, and that number lands on the deficit regions in Risk two.</p><p>Japan's configuration is the contrasting case and it is often conflated with China's: 8.41 million built against 4.57 million sold is a 3.84 million surplus from a home market of 4.57 million, the largest export ratio of any major producer. Japan has run that configuration for forty years without triggering the tariff response China is now meeting, and the difference is distribution and local production rather than volume.</p><h3>Risk four: the incentive-withdrawal shock, and its repeatability</h3><p>2025 proved that removing a consumer purchase incentive removes the marginal buyer — quickly, and within a single quarter. The United States held just below 10 per cent electric share and fell sharply in the fourth quarter once credits ended. <strong>The same mechanism can fire in any market where demand is credit-supported rather than standard-supported</strong>, and the forecast's regional growth rates assume it does not fire again at scale.</p><p>The asymmetry worth noting is that a fleet-average CO<sub>2</sub> standard does not have this failure mode. It produces volume on a schedule because the vehicles must be sold. That is why the European and Chinese trajectories in this forecast are treated as more reliable than the American one.</p><h3>Risk five: the aftermarket transition, which is a revenue timing problem</h3><p>Parts and service are 28.3 per cent of this market, growing 4.0 per cent, and lagging the vehicle sale by roughly seven years. A battery-electric vehicle generates materially less service revenue than the combustion vehicle it replaces — no engine oil, no filters, no exhaust, fewer brake replacements because regenerative braking does the deceleration — and more tyre, body and electronics revenue.</p><p><strong>For a period, the industry will be servicing a parc that is electrifying faster than its service revenue is growing.</strong> The forecast assumes the offset from electronics, tyres and body repair arrives in time; if it does not, the aftermarket block — the most profitable part of the chain — grows slower than 4.0 per cent, and the value forecast loses roughly a percentage point.</p><h3>Risk six: the pack cost floor</h3><p>The first decade of electrification was carried by a falling pack cost. That decline has slowed, and the 2025 European step change came from a regulatory ratchet rather than from a price signal — which is the tell. <strong>If pack cost is flat to 2033, electric vehicles do not reach the entry price points in India, Southeast Asia and Africa where the unit base has room to grow</strong>, and the 2.10 per cent unit CAGR becomes dependent entirely on replacement demand in mature markets.</p><h3>Risk seven: the data itself</h3><p>World vehicle sales exceeded world production by 3.39 million units in the 2025 OICA series, which is physically impossible for new vehicles in one year, and the same series notes that its world totals differ from the sum of its own rows by roughly 25,000 units in both sales and production. <strong>Every regional trade balance in this report is a difference between two series assembled from national registration counts, several of which include imported used vehicles.</strong> The direction of those balances is reliable; the decimal places are not, and a reader using them to size an export opportunity should treat them as an order of magnitude.</p><h3>What would falsify this forecast</h3><p>Three readings, in order of speed. <strong>US average transaction price growth staying under 1 per cent while incentives rise beyond 8 per cent of price</strong> means the value path has no cushion and the forecast is optimistic. <strong>China's output-minus-sales gap widening past 1 million units</strong> means export pressure and tariff escalation arrive faster than the regional growth rates assume. <strong>Battery-electric share of electric sales falling back below 60 per cent</strong> means the pack cost curve has stalled and the entry-price story in Risk six is failing. None of the three has fired yet. All three are visible in published data within one quarter of firing.</p>"
  regulatory_landscape_content: "<h2>Regulatory landscape: three regimes, one market, and a natural experiment</h2><p>2025 was the cleanest natural experiment the automotive industry has ever run, and it was not designed. The same product was sold into three regulatory regimes that moved in opposite directions in the same twelve months, and the outcomes diverged so sharply that the year can be read as a controlled test of whether regulation or product economics sets the pace of electrification.</p><p>Europe tightened. The EU CO<sub>2</sub> standards took a step change and electric car sales rose <strong>30 per cent to more than 4 million units</strong>, after having stagnated in 2024. The United States loosened, with tax credits ending during the year: the electric share held just below 10 per cent and sales <strong>fell sharply in the final quarter</strong>. China continued a policy path it has run for a decade and sold <strong>more than 13 million electric cars at close to 55 per cent of its new car market</strong>. Same technology, same year, three regimes, three results. <strong>The regime is the variable.</strong></p><h3>Europe: the ratchet works, and it is now the demand model</h3><p>The European mechanism is a fleet-average CO<sub>2</sub> standard that tightens on a published schedule, with a penalty per gram of exceedance multiplied by units sold. Its distinguishing feature is not severity but <strong>predictability</strong>: an assembler knows the number four years out, which means the compliance decision is made at the model-cycle stage, which means the product arrives whether or not the consumer is ready. That is why the 2025 step change produced volume rather than a price war — the vehicles had to be sold, and they were.</p><p>The consequence for this report is that Europe's electric growth rate is now a <em>regulatory</em> variable rather than a consumer one, and it will move in steps at the years the standard tightens rather than in a smooth curve. It is also why Europe, which built 17.20 million vehicles and sold 18.63 million, is importing both the vehicles and the compliance: a deficit region that tightens its standard hands the volume to whoever can supply it cheapest.</p><h3>United States: withdrawal, and what it revealed</h3><p>The American mechanism was a consumer purchase credit, and its withdrawal in 2025 revealed something the sector had spent years arguing about. Electric share did not collapse to a niche — it held just below 10 per cent — but it stopped growing, and the fourth quarter fell sharply. <strong>What the credit had been buying was the marginal buyer, not the early adopter.</strong> The early adopter was always going to buy the car; the credit moved the buyer who was indifferent, and when it went, so did that buyer.</p><p>Two things follow. First, a consumer-incentive regime is inherently volatile in a way a fleet-average standard is not, because it can be removed within a budget cycle. Second, the American market will now be the slowest of the three to electrify, which means the largest single profit pool in the industry — full-size pickups at USD 66,386 and a fifth of December's shoppers buying luxury — stays combustion-led for longer. <strong>That is a competitive fact, not a regulatory one, and it advantages the groups with the deepest truck franchises.</strong></p><h3>China: a mandate by other means</h3><p>China's regime is neither a purchase credit nor a fleet standard in the European sense. It is an industrial policy that combines production-side support, local-market access rules, purchase-tax treatment and a state-directed build-out of charging infrastructure, and its result in 2025 was 16.626 million new-energy vehicles built, up 29 per cent, within a total output of 34.53 million. Nearly half of China's production is now new-energy, and the domestic market absorbed almost all of it: 34.40 million sold against 34.53 million built.</p><p>The important regulatory feature is that <strong>China's policy sets the industry's cost curve, not just its own market</strong>. Scale in Chinese cell and pack production is what makes a USD 22,000 electric car possible, and that price point is what makes electrification viable in India, Southeast Asia, Latin America and the Middle East without a subsidy. The two million electric cars sold outside the three major markets in 2025 are, in a real sense, a Chinese policy outcome.</p><h3>Tariffs are a fourth regime, and they are the one that costs most per unit</h3><p>OICA's own reading of the 2025 data warns of growing exposure to trade frictions, tariff risk and localisation pressures, particularly for the United States. The arithmetic in this report's regional chapter explains why: the Americas built 18.74 million and sold 24.86 million, so the region imports roughly 6.1 million units net, while Asia built 4.19 million more than it sold. <strong>A tariff is the only instrument that lets a deficit region respond to a surplus region without waiting for a plant to be built.</strong></p><p>The cost consequences are structural rather than temporary. A tariff raises the landed cost of an imported vehicle, which either raises the transaction price in a market where price growth is 0.8 per cent, or forces the assembler to absorb it out of a high single-digit operating margin, or moves the vehicle to a third market. All three outcomes reduce the industry's return on capital, and the third one redirects volume to whichever region has not yet written a tariff — which is why the watch list in this report tracks the direction of trade policy rather than its level.</p><h3>Compliance is a fixed cost, and fixed costs favour scale</h3><p>Certification, emissions conformity, software validation, battery traceability and cybersecurity approval are all per-platform costs that do not scale down with volume. A platform sold in 200,000 units carries the same regulatory engineering burden as one sold in 2 million, which means <strong>regulatory complexity is itself a consolidation force</strong> — and it is the one force in this industry that operates in the same direction as the cost pressure rather than against it.</p><p>That is also why the twelve groups holding 66.8 per cent of the market have not been dislodged. A new entrant can build a good electric car. Building a car that is <em>certified in forty jurisdictions</em> is a different undertaking, and it is the part of the moat that does not appear in a specification sheet.</p><h3>What regulation does, and does not, do to this forecast</h3><p>It sets the <strong>mix</strong>, reliably and in steps. It does not set the <strong>total</strong>. Europe's standard guarantees that a rising share of the vehicles it registers will be electric; it guarantees nothing about how many vehicles are registered. A market that builds 17.20 million and sells 18.63 million with flat-to-negative growth in both can hit every electrification target it has while its total volume shrinks.</p><p><strong>That distinction is the single most important regulatory judgement in this report: the regulations reliably reallocate demand between powertrains, and they do not create demand.</strong> The 4.39 per cent value forecast follows from that — a mix shift inside a market growing 2.10 per cent in units, with content per vehicle doing the rest.</p>"
  faq_question_1: "How large is the global automotive market in 2025?"
  faq_answer_1: "USD 4,241 billion, on 99.8 million new road vehicles sold at an average transaction value of USD 30,471, plus USD 1,200 billion of parts and service revenue. It reaches USD 5,982 billion by 2033 at a 4.39 per cent value CAGR."
  faq_question_2: "Why is the value forecast lower than the previous edition of this report?"
  faq_answer_2: "Because the price data does not support a higher one. A 6.15 per cent value CAGR on 2.10 per cent unit growth requires transaction prices to rise about 4 per cent a year for eight years. The US average transaction price rose 0.8 per cent in 2025 with incentives at 7.5 per cent of price to hold volume, so the rebuild forecasts units plus content per vehicle, and treats price as flat to slightly positive rather than as a growth engine."
  faq_question_3: "Which region produces the most vehicles?"
  faq_answer_3: "Asia-Oceania, at 59.2 million vehicles in 2025, or 61.4 per cent of world output of 96.4 million. China alone built 34.53 million, Japan 8.41 million, India 6.49 million and South Korea 4.10 million. Europe built 17.20 million, the Americas 18.74 million and Africa 1.23 million."
  faq_question_4: "How large is the electric vehicle share of the market?"
  faq_answer_4: "25 per cent of global new car sales in 2025, on more than 20 million electric cars sold, 20 per cent up on 2024. Battery electric vehicles were 65 per cent of those sales. China was close to 55 per cent electric by share, Europe passed 4 million units after a 30 per cent rise, and the United States stayed just below 10 per cent with a sharp fourth-quarter fall after tax credits ended."

global_schema:
  organization:
    "@id": MIWS-root#organization
    "@type": Organization
    url: "https://www.pheonixresearch.com/"
    name: Pheonix Research
  website:
    "@id": MIWS-root
    "@type": WebSite
    url: "https://www.pheonixresearch.com/"
    name: Pheonix Research
  api:
    "@id": MIWS-root#api
    "@type": WebAPI
    url: "https://www.pheonixresearch.com/"

graph:
  node_id: MIMR-node_7ff0734
  graph_node_endpoint: "https://graph.pheonixresearch.com/api/v1/query/live/node/MIMR-node_7ff0734"
  graph_snapshot: "https://graph.pheonixresearch.com/api/v1/read/live/graph"
  graph_exclusions: "https://graph.pheonixresearch.com/api/v1/read/live/exclusions"
  graph_meta: "https://graph.pheonixresearch.com/api/v1/read/live/meta"
  graph_bundle: "https://graph.pheonixresearch.com/api/v1/read/live/bundle"

discovery:
  discovery_json: "https://www.pheonixresearch.com/.well-known/pheonix-discovery.json"
  llms_txt: "https://www.pheonixresearch.com/llms.txt"
  sitemap: "https://www.pheonixresearch.com/sitemap.xml"
---
# Global Automotive Market Report, Size, Share and Forecast 2026–2033: 99.8 Million Vehicles, 61% of Them Built in Asia, at a Price That Has Stopped Rising

- **Market Size (Base Year):** USD 4241 Trillion
- **Forecast Value:** USD 5982 Trillion
- **CAGR:** 4.39%
- **Forecast Period:** 2026 - 2033
- **Coverage:** Global - Asia-Pacific and Middle East, Europe, North America, South America, Africa

## Executive Summary

The global automotive market is USD 4,241 billion in 2025, reaching USD 5,982 billion by 2033 at a 4.39 per cent value CAGR, on 99.8 million vehicles sold and 96.4 million built. The industry returned to growth in 2025 — sales up 4.7 per cent, production up 3.9 per cent — but the growth was almost entirely geographic and almost none of it was financial. Units compound at 2.10 per cent across the forecast window and revenue per vehicle at 2.4 per cent, which is a content story, not an inflation story, and it produces a value CAGR roughly two and a half points below the rate this report previously carried.Production has moved east and the move is structural, not cyclical. Asia-Oceania built 59.2 million vehicles in 2025, 61.4 per cent of world output, and China alone built 34.53 million — up 10 per cent in a year in which Europe's output fell and the Americas' fell faster. Emerging economies out-built the developed world for the first time, 58.2 million against 38.2 million, which is 60.4 per cent of global output. Europe built 17.20 million against sales of 18.63 million, so it imported 1.43 million net; Asia built 4.19 million more than it sold; the Americas sold 6.12 million more than they built. The geography of the value chain is now the geography of the trade balance, and each region's exposure to tariffs is simply its number.The price line is where this report departs from consensus. The US average transaction price hit an all-time high of USD 50,326 in December 2025 — and rose 0.8 per cent year on year, with incentives at 7.5 per cent of transaction price to hold the volume. The industry is buying its growth at a discount rather than earning it through price, and the average is being held up by mix — full-size pickups at USD 66,386, luxury at close to a fifth of December's shoppers — not by pricing power. That is why 4.39 per cent, not 6.15, is the defensible rate: a 6.15 per cent value CAGR on 2.10 per cent unit growth requires eight straight years of 4 per cent transaction-price inflation, and the year just closed delivered 0.8.Electrification has stopped being an American story and is now a Chinese and European one. More than 20 million electric cars were sold in 2025, 20 per cent up, for a 25 per cent share of global new car sales — one in four. China sold more than 13 million of them at close to 55 per cent of its new car market, Europe rose 30 per cent to more than 4 million, and the United States held just below 10 per cent with sales falling sharply in the final quarter once tax credits ended. Two million electric cars were sold outside those three markets, more than half of them in countries that have now passed a 10 per cent share. The technology's growth no longer depends on the market that invented the subsidy.

## Market Report

# Global Automotive Market Forecast Snapshot, 2026–2033

 | Metric |  | Value |

 | Market size, 2025 |  | USD 4,241 billion |

 | Market size, 2033 |  | USD 5,982 billion |

 | Value CAGR, 2026–2033 |  | 4.39 per cent |

 | Units sold, 2025 |  | 99.8 million |

 | Units sold, 2033 |  | 117.9 million |

 | Unit CAGR, 2026–2033 |  | 2.10 per cent |

 | Average transaction value per vehicle, 2025 |  | USD 30,471 |

 | Largest producing region |  | Asia-Oceania, 59.2 million units, 61.4 per cent of world output |

 | Largest single market |  | China, 34.40 million units sold |

 | Electric share of new car sales, 2025 |  | 25 per cent, on more than 20 million units |

 | Largest net vehicle exporter |  | Asia, 4.19 million units more built than sold |

 | Regulatory drivers |  | EU CO2 standards, US credit withdrawal, tariff and localisation regimes |

The global automotive market in 2025. Pheonix Research.

## Market Definition and the Three Boundary Calls

This report measures the value of new road vehicles at the point of sale, plus the parts and service revenue that vehicles generate once they are in use. In 2025 that is **USD 3,041 billion of new-vehicle transaction value and USD 1,200 billion of aftermarket parts and service**, giving USD 4,241 billion on 99.8 million units and an average of USD 30,471 a vehicle. It reaches USD 5,982 billion in 2033 at 4.39 per cent.

Three boundaries decide what is and is not in that number, and a reader who disagrees with one of them should be able to see immediately what it costs.

**The first is exclusions.** Fuel, insurance, road tax, used-vehicle transfers and captive finance are outside the boundary. The automotive industry is routinely described as a five or six trillion dollar industry by including them, and it is not wrong to do so — it is simply measuring the cost of owning a vehicle rather than the revenue of making and servicing one. This report measures the second.

**The second is used vehicles.** They are excluded, which matters more than it looks in 2025, because affordability pressure has been pushing buyers toward the used parc in several markets and that substitution shows up here as a loss of new-vehicle revenue even where total household spending on cars has not fallen.

**The third is the border.** Value is counted where the vehicle is sold, not where it is built. A Mexican-built pickup sold in Texas is North American sales and North American production; a Chinese-built electric car sold in Rotterdam is European sales and Asian production. Getting this consistent is what makes the regional chapter of this report a trade-balance chapter as much as a demand chapter, and it is the only way the OICA production and sales series can be used together without double counting.

## Market Size and Forecast

Global automotive market value, 2025–2033 (USD billion). Pheonix Research.

### 3.1 The 2025 Base Was Built From Units and Prices, Not From a Value Estimate

The base year rests on one fully observed figure and two disclosed builds. The observed figure is 99,798,406 new road vehicles sold worldwide in 2025 against 96,383,650 built, up 4.7 and 3.9 per cent respectively on 2024 (OICA correspondents survey, 23 April 2026). The first build is a transaction-value line per region rather than a single global average price, because a global average is the one number guaranteed to be wrong everywhere: a vehicle transacts at roughly USD 49,000 in the United States, USD 38,000 in Europe, USD 22,000 in China and USD 13,000 in India, and a blended figure built on any one of those would misprice the other three. The second build is the aftermarket, at 0.395 times new-vehicle value.

The US line is the only one anchored on a published transaction series: the average transaction price reached **USD 50,326 in December 2025**, an all-time high and up 0.8 per cent year on year, against an average MSRP of USD 52,627 and incentives at 7.5 per cent of transaction price. The average new electric vehicle transacted at USD 58,034 (Kelley Blue Book, Cox Automotive, 12 January 2026). Every other regional line is a Pheonix estimate and is labelled as one in the methodology.

### 3.2 The Regional Allocation Is the Business Model

The allocation is not a courtesy to geography. It is the whole commercial story, because in 2025 production and sales moved in opposite directions in three of the four regions and the differences are large enough to be a business model.

Asia-Oceania and the Middle East **built 59.2 million vehicles and sold 55.02 million** — a net surplus of 4.19 million units. Europe **built 17.20 million and sold 18.63 million** — a net deficit of 1.43 million. The Americas **built 18.74 million and sold 24.86 million** — a net deficit of 6.12 million, of which the United States alone accounts for the bulk: 10.24 million built against 16.68 million sold. Africa built 1.23 million and sold 1.29 million, which at that scale is a rounding error in units and a strategic question in every other respect.

**Read together, those four numbers describe an industry in which the exporting region also holds the cost advantage, and the deficit regions are the ones writing tariff policy.** That is not a coincidence and it is not stable.

### 3.3 The Decomposition Is the Finding

Value grows 4.39 per cent a year. Units grow 2.10 per cent. Revenue per vehicle grows 2.4 per cent. The gap between the headline and the unit count is therefore *content*, not price inflation, and the distinction decides how the decade should be read.

Content per vehicle rises because a battery-electric vehicle carries a large and expensive pack, power electronics and a thermal system where a combustion vehicle carried an engine, a transmission and an exhaust. It falls at the same time because that powertrain swap removes several hundred to a couple of thousand part numbers and the labour and logistics attached to them. The net is a vehicle with fewer parts and more value, and an industry whose supplier base is being re-sorted faster than its assembly base is.

What the forecast deliberately does *not* assume is price. A 6.15 per cent value CAGR on 2.10 per cent unit growth requires transaction prices to rise roughly 4 per cent a year for eight consecutive years. The year just closed delivered 0.8 per cent, with incentives raised to 7.5 per cent of transaction price to hold the volume. **The industry is buying its growth at a discount, and the rebuild forecasts what the price data supports rather than what the growth rate would prefer.**

## Four Structural Shifts Defining 2026–2033

### 4.1 Production Moved East and Is Not Moving Back

Asia-Oceania built **59.2 million vehicles in 2025, 61.4 per cent of world output**, up 7.6 per cent in a year in which Europe's output slipped 0.8 per cent and the Americas' fell 2.1 per cent. China alone built 34.53 million, adding roughly 3.25 million units in a single year; India built 6.49 million; Japan 8.41 million; South Korea 4.10 million. **Emerging economies out-built the developed world for the first time: 58.2 million against 38.2 million, or 60.4 per cent of global output.**

The mirror image is Europe, where the adjustment has stopped being gentle. Italy built 915,291 vehicles in 2019 and 474,044 in 2025 — a fall of 48.2 per cent, and a volume that is now below Morocco's 501,965. The United Kingdom built 1,381,405 in 2019 and 764,715 in 2025. Germany remains the continent's largest producer at 4.15 million cars, but a continent that builds 17.20 million and sells 18.63 million is running a structural deficit it did not have a decade ago.

### 4.2 The Transaction Price Stopped Rising

The US average transaction price set an all-time record in December 2025 at USD 50,326 and grew 0.8 per cent for the year. Incentives reached 7.5 per cent of transaction price, the highest of the year. That combination — record headline, flat growth, rising discounts — is what a market looks like when mix is doing the work that pricing used to do: full-size pickups averaged USD 66,386 and generated more than USD 15 billion of combined retail and fleet revenue in December alone, and close to a fifth of December's shoppers bought luxury.

**Mix is not a growth engine, it is a holding pattern.** It works while the top of the market is deep enough to absorb it, and it stops working the moment affordability pressure reaches the segments that carry the volume. Both the drivers and the restraints in this report come back to that single asymmetry.

### 4.3 Electrification Became a Non-American Story

More than **20 million electric cars were sold in 2025, 20 per cent up, for a 25 per cent share of global new car sales**. Battery electric vehicles were 65 per cent of those sales, reversing two years of drift toward hybrids and range-extenders. China sold more than 13 million at close to 55 per cent of its new car market and produced 16.626 million new-energy vehicles, up 29 per cent. Europe rose 30 per cent to more than 4 million units after a step change in EU CO2 standards. The United States held just below 10 per cent, with sales falling sharply in the final quarter once tax credits ended.

The number that matters is not any of those. It is the **two million electric cars sold outside China, Europe and the United States**, more than half of them in countries that have now passed a 10 per cent sales share. The technology's growth has decoupled from the market that invented the subsidy, and the forecast's regional growth rates follow that decoupling rather than assuming American policy sets the pace.

### 4.4 The 3.4 Million Vehicle Gap Nobody Explains

World sales exceed world production by 3.39 million units in the OICA series. That is physically impossible for new vehicles in a single year, and it is worth being explicit about: the gap is a definitional artefact of assembling national registration counts, several of which include imported used vehicles and some of which count vehicles that another country has already counted. The same series notes that its own world totals differ from the sum of its rows by roughly 25,000 units in both sales and production.

**The correct handling is to trust the production series for physical output and the sales series for demand direction, and not to difference them as if the result were inventory.** Every regional trade balance in this report is stated as the gap between the two series and is described as a directional measure, not an observed flow.

## Market Drivers

**Replacement demand in a parc of roughly 1.6 billion vehicles.** The installed base is many times annual sales, so the replacement cycle — not first purchase — sets the floor under demand in every mature market. When buyers delay replacement, as they have under affordability pressure, the deferred volume accumulates rather than disappearing, and it is the single largest source of upside in the forecast.

**Content per vehicle rises even as part count falls.** This is the mechanical driver of the 2.4 per cent revenue-per-unit line, and it is the most reliable part of the model because it is a design fact rather than a demand assumption.

**Emerging Asia is still buying its first car.** India sold 5.52 million vehicles and built 6.49 million, with record domestic passenger vehicle sales above 4.6 million and more than a million trucks and buses in fiscal 2025–26. Türkiye rose 10.0 per cent, Uzbekistan 34.5 per cent, Pakistan 39.8 per cent and Egypt 69.1 per cent. These are not replacement markets and their growth does not depend on interest rates in Frankfurt.

**The aftermarket lags the sale by roughly seven years.** Parts and service revenue in 2033 is largely determined by the vehicles sold between 2021 and 2029, which is why the aftermarket block grows at 4.0 per cent against 2.10 per cent unit growth even as its ratio to new-vehicle value falls.

**Localisation subsidies are demand policy in disguise.** Where a government pays for a plant, it is also paying for the domestic volume that plant needs in order to load, which shows up as demand support that no consumer incentive programme accounts for.

## Market Restraints

**Affordability is now the binding constraint.** A record average transaction price growing 0.8 per cent a year, held in place by 7.5 per cent incentives, is the definition of a market that has run out of price room.

**Trade friction is a cost line, not a risk line.** OICA's own reading of the 2025 data warns of growing exposure to trade frictions, tariff risk and localisation pressures, particularly for the United States. A tariff is not a forecast risk that may or may not materialise; it is a landed cost with a date.

**Chinese capacity and the export disposal problem.** A market that adds 3.25 million units of output in a year while its domestic growth slows has to place that volume somewhere, and the surplus arithmetic in section 3.2 says where it goes. That is the mechanism behind the tariff policy in the deficit regions.

**Incentive dependency in the retail channel.** When 7.5 per cent of transaction price is incentive, the effective price is the discounted one, and every quarter of volume becomes a decision about how much discount the channel needs rather than about how much the product is worth.

**Repair and insurance costs became a purchase barrier.** Content per vehicle has a second-order effect that the industry consistently underweights: a vehicle with expensive electronics and a large battery is more expensive to insure and to repair, and for a buyer financing on a monthly payment, insurance is part of the affordability calculation.

**The battery cost floor has stopped falling.** The decade-long decline in pack cost did much of the work of electrification. It cannot do the next stage's work on its own, which is why the European step change in 2025 came from a regulatory ratchet rather than from a price signal.

## Segmentation

**By propulsion.** Combustion, hybrid, plug-in hybrid, battery electric and fuel cell. The 2025 data supports only one strong statement about this split: battery electric vehicles took 65 per cent of electric sales, reversing two years of drift toward hybrids and range-extenders, and the reversal happened in the same year that EU standards stepped and US credits ended. Propulsion mix is now set by regulation and by China's domestic market, in that order.

**By body type and segment.** The mix has become the pricing system. In the United States, full-size pickups and luxury together carried December's record average, which means body-type mix is doing the work that list-price increases used to do.

**By region of assembly.** Asia-Oceania 59.2 million, the Americas 18.74 million, Europe 17.20 million, Africa 1.23 million. This is the only segmentation in the report where the shares are observed rather than estimated, and it is the one a supply-chain reader should start from.

**By vehicle class and duty cycle.** Passenger cars, light commercial, medium and heavy commercial, and buses. Duty cycle is what determines whether electrification is economic today: a depot-based bus or van electrifies on total cost of ownership, while a long-haul tractor electrifies on policy.

**By channel.** Franchised retail, independent used, fleet and corporate, rental, and government. The channel mix is where affordability pressure shows up first, because the fleet channel can absorb a price increase that the retail buyer cannot.

## Regional Analysis

**Asia-Pacific and the Middle East.** 59.2 million built, 55.02 million sold, up 7.1 per cent in sales and 7.6 per cent in output. China is 34.53 million of the production and 34.40 million of the sales, with 16.63 million new-energy vehicles built. India is the second story: 6.49 million built, 5.52 million sold, and a first-purchase cycle that has barely started. Japan built 8.41 million but sold only 4.57 million, which makes it the region's quiet exporter.

**Europe.** 17.20 million built against 18.63 million sold, both marginally down. Germany built 4.15 million cars; Italy fell to 474,044 and the United Kingdom to 764,715. Electric sales rose 30 per cent past 4 million on the CO2 step change, which means Europe is now importing both the vehicles and the compliance.

**North America.** The United States built 10.24 million and sold 16.68 million; Mexico built 4.09 million and sold 1.56 million; Canada built 1.24 million and sold 1.93 million. The region builds 15.57 million and sells 20.17 million, so it imports roughly 4.6 million net — which is precisely why it is the region writing tariffs.

**South America.** Brazil built 2.64 million and sold 2.69 million, up 2.1 per cent, and Argentina rose 48.2 per cent in sales off a low base. The region is the clearest example of a market where a locally built vehicle and an imported one can differ in price by a third.

**Africa.** 1.23 million built, 1.29 million sold, with sales up 22 per cent to 1.29 million from 1.05 million. Morocco built 501,965 and Egypt 89,570, up 62 per cent. The continent's importance is not its volume; it is that it is the only region where both production and sales are growing from a base low enough that the growth rate is not constrained by replacement demand.

**What the geography says.** A report that treats the world as one market would have concluded in 2025 that the industry grew 4.7 per cent. A report that separates exporter from importer concludes that the growth was placed by the region with the cost advantage into the regions with the demand and the tariffs, and that the second half of the decade is a negotiation over that flow.

## Strategic Outlook

The next eight years are not a demand problem. Units grow 2.10 per cent, which is unremarkable and close to the long-run trend. The problem is that **the industry has three ways to grow and only one of them is currently working**: volume in emerging Asia, content per vehicle everywhere, and price, which has stopped. Volume and content together produce 4.39 per cent. Price would have produced 6.15. The distance between those two numbers is the value at stake in the affordability debate.

The second strategic fact is that the industry's capacity now sits in a different hemisphere from its demand, and that gap is being closed with policy on both sides — incentives to build where the demand is, tariffs to slow the flow from where the capacity is. **Both are capital-allocation decisions masquerading as trade policy, and both raise the cost per unit of the vehicles they are meant to make cheaper.**

The third is that electrification's centre of gravity has moved. A technology that grows 20 per cent to 25 per cent share in a year in which its largest subsidising market withdrew support, and in which two million of its units were sold outside the three big markets, is no longer a policy artefact. It is a product cycle, and it is now being set by Chinese cost and European regulation rather than by American incentives.

## Watch List

- US transaction price against incentive level. If average transaction price growth stays under 1 per cent while incentives rise beyond 8 per cent of price, the 4.39 per cent value path has no cushion.

- China's output-minus-sales gap. A widening surplus is the leading indicator for export volume and therefore for tariff escalation in Europe and the Americas.

- Europe's build-sell deficit. At 1.43 million units net it is manageable; at 2.5 million it stops being a demand story and becomes an industrial-policy one.

- India's production curve. 6.49 million built is the single largest untapped demand pool outside China, and it is growing faster than any market of comparable size.

- Battery-electric share inside the electric mix. The 2025 move back to 65 per cent is the hinge; if it reverses again, the pack-cost curve is not doing what the forecast assumes.

- The two million. Electric sales outside China, Europe and the US are the cleanest read on whether the technology has become self-sustaining.

- Italy and the UK. 474,044 and 764,715 vehicles built are the two clearest readings of what a European assembly base looks like when it loses its volume.

## Table of Contents

Market Snapshot
Market Definition and the Three Boundary Calls
Market Size and Forecast

3.1 The 2025 Base Was Built From Units and Prices, Not From a Value Estimate
3.2 The Regional Allocation Is the Business Model
3.3 The Decomposition Is the Finding


Four Structural Shifts Defining 2026–2033

4.1 Production Moved East and Is Not Moving Back
4.2 The Transaction Price Stopped Rising
4.3 Electrification Became a Non-American Story
4.4 The 3.4 Million Vehicle Gap Nobody Explains


Market Drivers

5.1 Replacement Demand in a Parc of Roughly 1.6 Billion Vehicles
5.2 Content Per Vehicle Rises Even as Part Count Falls
5.3 Emerging Asia Is Still Buying Its First Car
5.4 The Aftermarket Lags the Sale by Seven Years
5.5 Localisation Subsidies Are Demand Policy in Disguise


Market Restraints

6.1 Affordability Is Now the Binding Constraint
6.2 Trade Friction Is a Cost Line, Not a Risk Line
6.3 Chinese Capacity and the Export Disposal Problem
6.4 Incentive Dependency in the Retail Channel
6.5 Repair and Insurance Costs Became a Purchase Barrier
6.6 The Battery Cost Floor Has Stopped Falling


Segmentation

7.1 By Propulsion
7.2 By Body Type and Segment
7.3 By Region of Assembly
7.4 By Vehicle Class and Duty Cycle
7.5 By Channel


Regional Analysis

8.1 Asia-Pacific and the Middle East
8.2 Europe
8.3 North America
8.4 South America
8.5 Africa
8.6 What the Geography Says


Value Chain
Technology and Innovation
Competitive Landscape
Regulatory Landscape
Investment Activity
Market Risk
Strategic Outlook
Watch List
Methodology
About This Report

## Competitive Landscape

Competitive landscape: sixty-seven per cent of the market in twelve groups, and none of them with pricing power
The twelve largest groups sold approximately 66.7 million of the 99.8 million vehicles sold worldwide in 2025, about 66.8 per cent of the market. That is a high concentration ratio for a capital-intensive industry, and it would normally imply pricing power. In 2025 it implied the opposite: the same twelve groups competed through 7.5 per cent incentives on a US average transaction price that grew 0.8 per cent. Concentration has bought them cost scale and has bought them no pricing.



Group
Home
Units sold
Basis


Toyota Group
Japan
11,320,000
With Daihatsu and Hino


Volkswagen Group
Germany
8,983,900
Deliveries; includes 305,600 trucks and buses


Hyundai Motor Group
South Korea
7,244,408
Hyundai with Genesis, plus Kia


General Motors
United States
6,180,000
Including Chinese joint ventures


Stellantis
Netherlands
5,480,000
Rounding; alternative counts differ


BYD
China
4,602,436
Battery electric and plug-in hybrid only


SAIC Motor
China
4,507,000
Including joint ventures


Ford
United States
4,395,000
Wholesales


Geely Holding
China
4,110,000
With Volvo Cars, Zeekr, Lynk & Co, Proton


Honda
Japan
3,387,000
Fiscal year to March 2026


Suzuki Motor
Japan
3,319,975
Fiscal year to March 2026; cars only


Nissan
Japan
3,150,000
Fiscal year to March 2026



Group volumes as each group reports them. Four are fiscal-year figures rather than calendar 2025, two are rounded at source, and GM and SAIC include joint ventures that other counts attribute differently. The table is a ranking, not a share calculation to three decimal places.
What the group table says
Four things, and none of them is the ranking. First, the top of the table is stable and the middle is not: Toyota and Volkswagen have held the first two positions through a decade in which the industry’s technology changed completely, because scale in assembly is transferable across powertrains even when scale in components is not. Second, three of the twelve are Chinese and two of those are inside the top seven, which is a different fact from China building 34.53 million vehicles — the country’s output was for years assembled by joint ventures whose brand belonged to somebody else. Third, the spread between first and twelfth is only 3.6 times, which is unusually flat for a capital-intensive industry and is a direct consequence of the economics in the value chain chapter: a plant is expensive but it is not a barrier, and a brand is transferable within a region. Fourth, and most important, there is no group on the list whose volume position protects it from the price data.
Two different consolidation stories
The industry is consolidating for two unrelated reasons and the two are usually reported as one. The first is technology consolidation: platforms, batteries and software have become so expensive that they must be amortised across more units, which drives alliance and platform-sharing behaviour rather than outright merger. The second is capacity consolidation: plants that cannot be loaded are closed, and the volume migrates to plants that can. The first increases the number of units per engineering dollar. The second reduces the number of plants. Europe is currently doing the second on a scale it has not attempted since the 1980s, which is why Italy at 474,044 vehicles and the United Kingdom at 764,715 are the chapter’s most informative numbers.
The Chinese group problem
BYD’s 4.60 million units are battery electric and plug-in hybrid only, which means its position in the electric market is materially different from its position in the total market: for every combustion vehicle sold by a legacy group, BYD has no competing product. That is a vulnerability as much as a strength outside the markets where electrification is policy-driven, and it is why the group’s international expansion has concentrated in markets with a regulatory ratchet — Europe above all, where electric sales rose 30 per cent past 4 million on a CO2 step change.
The strategic consequence is that China’s export weight is carried by groups whose product range is narrower than their volume suggests, while Japan’s export weight — a 3.84 million unit surplus from a 4.57 million unit home market — is carried by groups with a full range and a distribution network in every deficit region. Those are different kinds of competitive threat and they call for different responses.
Share is not margin
The most misleading number in this industry is market share, because the segments that carry the volume are not the segments that carry the profit. In the United States, full-size pickups averaged USD 66,386 in December 2025 against an industry average of USD 50,326, and close to a fifth of that month’s shoppers bought luxury. A group with 12 per cent of units and 6 per cent of trucks has a worse business than one with 8 per cent of units and 20 per cent of trucks. The right competitive metric in 2026 is revenue per unit against content per unit, not share of a unit count — because the industry has proven it can buy share with incentive and has not proven it can buy revenue with anything.
Who is actually gaining
Gains in 2025 came from three places, and only one of them is a group-level story. Geography: Asia-Oceania added 4.2 points of production share in a single year, and emerging economies passed the developed world outright at 58.2 million against 38.2 million. Powertrain: battery electric vehicles recovered to 65 per cent of electric sales, reversing two years of hybrid drift, which reordered the competitive field within the electric segment without changing the total. Price tier: mix moved up, with luxury and large trucks carrying the average, which benefits the groups with the deepest premium ranges and penalises the ones competing at the entry price point where the affordability constraint binds hardest.
The barriers that remain
Only four. Capital at scale, which is why the top twelve are stable — a new platform and a new plant together are a multi-billion commitment with a five-year payback. Regulatory certification, which is expensive, jurisdiction-specific and increasingly about software and emissions conformity rather than crash structure. Distribution, which is the moat nobody prices: a group with 3,000 franchised outlets in a deficit region can place volume in a way a new entrant cannot, and that is why Chinese expansion in Europe has been slower than Chinese production growth suggests. And residual value, which is the barrier that binds hardest in a market where affordability is the constraint — a brand that holds value can charge more for the same content, and no amount of manufacturing scale buys it.
The competitive conclusion is uncomfortable: the twelve groups that hold two-thirds of this market are locked in a cost competition they cannot win by raising price, in regions where their capacity is shrinking, against a competitor whose product range is narrower but whose cost base is lower. That is not a consolidation story. It is a margin story, and it is the reason this report forecasts units and content rather than price.

- **Structure:** moderately_consolidated
- **Tier 1 Players:** 10
- **Intensity:** high

## Value Chain

The value chain: where the USD 4.24 trillion actually sits
An average vehicle transacts at USD 30,471, and almost nothing about that number is decided by the company whose badge is on the grille. The figure is the output of a ladder that runs from a mine and a steel mill to a dealer forecourt, and the ladder’s economics — who holds pricing power, who absorbs a cost shock, who gets paid twice — determine whether a market growing 4.39 per cent a year is a good business to be in. For most of the ladder it is not; for two rungs of it, it is one of the best.
The tier structure and the margin ladder
The chain has five economically distinct layers. Raw materials — steel, aluminium, resin, glass, copper, and the battery metals — are commodity businesses priced off an index with no pricing power and no differentiation. Tier two and tier three component makers convert those materials into parts on a cost-plus basis, typically with annual productivity concessions written into the contract, which means their revenue rises with volume while their margin is negotiated downward every year. Tier one suppliers integrate parts into systems — a braking module, a seating system, a battery pack, an infotainment head unit — and this is the only supplier layer with genuine engineering pricing power, concentrated in the handful of firms that can carry a system across several vehicle programmes. The assembler integrates systems into a vehicle and sells it, at an operating margin that in a good year is high single digits. Below that sit distribution and retail, and then the aftermarket.
The structural point that follows is the one the industry keeps rediscovering: the assembler carries the volume risk and the tier one carries the engineering risk, but the aftermarket carries neither. Parts and service are 28.3 per cent of this market’s value and are sold into an installed base of roughly 1.6 billion vehicles whose owners have already made the purchase decision. That is why the aftermarket block compounds at 4.0 per cent against 2.10 per cent unit growth, and why every assembler that can build a service, subscription or software revenue line is trying to own a piece of it.
The powertrain swap re-sorts the tiers
Electrification does not simply add a supplier; it deletes a chain and builds another one alongside it. An internal combustion powertrain is a mature, deeply optimised network of perhaps two thousand part numbers spread across engine, transmission, exhaust, fuel and cooling, and it is populated by supplier firms whose entire engineering capital is invested in it. A battery-electric powertrain replaces that with a pack, an e-machine, an inverter and a thermal system: dramatically fewer parts, dramatically fewer suppliers, and a much larger share of the vehicle’s cost sitting with a handful of firms that can make cells at scale.
Two consequences follow, and both are visible in the 2025 data. First, the part count falls while the value rises, which is the mechanism behind this report’s 2.4 per cent revenue-per-unit line — fewer components, more expensive ones, and less labour content per vehicle. Second, the supplier base is being re-sorted faster than the assembly base is, because an assembler can retool a line in three years while a tier two stamping or castings supplier cannot convert its capital at all. The 6.49 million vehicles India built and the 34.53 million China built were largely assembled by firms that already had the new chain; the 474,044 Italy built were not.
Where content per vehicle is actually added
Content grows in four places, and they behave differently. The pack is the largest single line in a battery-electric bill of materials and is the reason a US electric vehicle transacted at USD 58,034 against an industry average of USD 50,326 — the premium is the pack, amortised through the monthly payment. Power electronics add value per vehicle while reducing cost per kilowatt, which is a rare combination and the reason silicon carbide adoption is a cost story rather than an efficiency story. Driver assistance adds sensor, compute and wiring content to every vehicle regardless of propulsion, which makes it the most broadly distributed content driver in the industry. Thermal management grows in both powertrains, because a pack needs cooling and a fast-charging pack needs more of it, and it is the segment where the internal-combustion and electric chains overlap most profitably.
Against that, content falls in the places nobody writes about: fewer castings, fewer fasteners, fewer fluid circuits, less machining, less assembly labour. A vehicle built in 2033 will carry more value in fewer parts, and the firms that make their living from the parts being removed are the ones the forecast is least kind to.
Raw materials and the asymmetry of pass-through
The chain absorbs commodity shocks unevenly, and the asymmetry is structural rather than cyclical. Steel and aluminium pass through to the assembler within one contract cycle, because there are alternate suppliers and a published index. Battery metals do not pass through cleanly, because the cell maker has signed multi-year offtake and the pack price is quoted to the assembler twelve to twenty-four months forward. The result is that a metal price move lands on the balance sheet of whichever layer happens to be holding the forward position, which is why pack pricing has been stickier than metal pricing in both directions.
This matters for the forecast because it means the cost curve that drove the first decade of electrification is not reliably available for the second. The 2025 European step change in electric sales came from a regulatory ratchet, not from a price signal, and the chain’s cost structure is the reason.
The geography of the chain is the trade balance
The chain’s location decisions show up in the only fully observed dataset in this report. Asia built 4.19 million more vehicles than it sold in 2025; Europe built 1.43 million fewer than it sold; the Americas built 6.12 million fewer than they sold. Those three numbers are the sum of thousands of individual sourcing decisions, and they produce a market in which the region with the lowest conversion cost is also the region with the largest surplus, and the deficit regions respond with tariffs that raise the cost of the vehicles their own buyers need.
Within the surplus there is a second layer worth separating. China built 34.53 million and sold 34.40 million, so its domestic market absorbs almost its entire output — its surplus is small, and its export weight comes from growth rather than from an absolute glut. Japan built 8.41 million and sold 4.57 million, which is a 3.84 million surplus from a market of 4.57 million: the largest export ratio of any major producer. The two are often discussed as one phenomenon and they are not.
The aftermarket is the profitable half
USD 1,200 billion of parts and service, growing 4.0 per cent, lagging the vehicle sale by roughly seven years, and largely insulated from the new-vehicle price war because the alternative to a repair is not a cheaper repair — it is a different vehicle or no repair at all. This is the block of the chain where a distributor, an independent workshop network or a parts brand can earn a return without carrying model-cycle risk.
It is also the block most exposed to electrification’s second-order effects. A battery-electric vehicle has no engine oil, no filter, no exhaust, no spark plug and materially fewer brake replacements, because regenerative braking does much of the deceleration. In its first several years it therefore generates less service revenue than the combustion vehicle it replaced — and more tyre, body and electronics revenue. The aftermarket does not disappear; it changes address, and the firms that own the old address lose.
What the chain cannot do
It cannot expand capacity quickly: a new assembly plant is a three-to-five year project and a new cell plant is longer. It cannot cut price to create demand, because 7.5 per cent of transaction price is already incentive. And it cannot move a plant to follow a tariff within a model cycle. Those three inflexibilities, more than any demand assumption, are what turn a 4.7 per cent unit growth year into a 4.39 per cent value forecast for the following eight.

- **Model:** hybrid
- **Distribution:** retail_chain_dominated
- **Supply Complexity:** high

## Investment Activity

Investment activity: capital is moving faster than demand
The industry’s investment problem in 2026 is not a shortage of capital. It is that capital is being committed to capacity in one set of geographies while demand sits in another, and the instrument being used to bridge the gap is policy rather than price. The 2025 production data makes the reallocation visible: emerging economies built 58.2 million vehicles and the developed world 38.2 million, and within that China added roughly 3.25 million units of output in a single year, India built 6.49 million, Mexico 4.09 million, Morocco 501,965 and Egypt 89,570, up 62 per cent.
Where the capital is going, and why
Three distinct investment motives are operating at once, and they produce very different returns.
Capacity for a protected market. A plant built behind a tariff or a localisation requirement earns a return that is a function of the tariff, not of the product. Mexico’s 4.09 million units of output against 1.56 million units of domestic sales is the clearest example in the dataset: a country that builds 2.5 times what it sells is an export platform, and its investment case rests on access rather than on domestic demand. Türkiye, at 1.42 million built and 1.41 million sold, is the opposite configuration and a more balanced one.
Capacity for a cost base. Morocco’s 501,965 units and Egypt’s 89,570 are not large numbers, but they are a cost position — labour, energy and proximity to the European deficit of 1.43 million units. A plant in a low-cost location serving a high-cost deficit market is the most reliably profitable configuration in this industry right now, which is precisely why the deficit market is writing localisation rules against it.
Capacity for a regulatory ratchet. Cell and pack plants are being built where the standard is tightening, because the standard is the demand forecast. Europe’s 30 per cent electric growth to more than 4 million units was created by a CO2 step change, and the investment that supplies it was justified by the standard rather than by consumer research. That makes it the least elastic and most politically exposed category of the three.
The stranded asset problem, in numbers
The mirror of new capacity is idle capacity, and the European figures are the clearest reading of what it costs. Italy built 915,291 vehicles in 2019 and 474,044 in 2025 — a fall of 48.2 per cent. The United Kingdom built 1,381,405 and 764,715. The Netherlands built zero cars in both 2024 and 2025. These are not plants that were closed; they are plants that were loaded down, which is the more expensive outcome because the fixed cost stays and the contribution disappears.
The investment consequence is that Europe’s remaining capital has to be spread across fewer units. Germany built 4.15 million cars and remains the continent’s base, but a continent at 17.20 million units of output against 18.63 million of sales cannot justify the platform investment its regulatory targets require without either importing the vehicles or subsidising the plant. Both are the same decision: pay for the transition in the exchequer or in the trade balance.
M&A and alliances: technology consolidation, not capacity consolidation
Deal activity in this industry is concentrated where the fixed cost is, and the fixed cost is now in platforms, software and cells rather than in stamping presses. That produces alliance and platform-sharing behaviour — two groups sharing a platform, three sharing a cell specification — rather than the large horizontal mergers that consolidation ratios usually imply.
The reason is that a merger buys capacity, and capacity is the thing the industry already has too much of. What a group cannot buy at a sensible price is a battery supply relationship or a validated software stack, so that is where the money goes. The predictable consequence is that the twelve groups holding 66.8 per cent of the market will still be twelve groups in 2033, but they will share more of their engineering than they do now — and the suppliers that hold the shared technology will hold the pricing power.
Return on capital, and the honest problem
The industry’s investment case is arithmetic: it must fund a platform transition and a capacity reallocation out of a margin that is being compressed by 7.5 per cent incentives on a transaction price growing 0.8 per cent. Capital intensity is rising, unit growth is 2.10 per cent, and price is flat. That combination has one of two resolutions — either the content-per-vehicle line in this forecast holds and revenue per unit rises 2.4 per cent a year, or the industry spends the decade funding a transition out of a flat revenue base.
The most important investment signal in the 2025 data is therefore not any single plant announcement. It is that battery-electric vehicles recovered to 65 per cent of electric sales, reversing two years of drift toward hybrids and range-extenders. A group that had hedged its capital across hybrids and been vindicated by 2023–24 had to re-weight toward pure electric in 2025, which means the capital committed to hybrid lines is now the most exposed position in the industry.
What would change the investment case
Three things, in order of consequence. A durable rise in transaction price would restore the return on content and unlock the higher value path. A settlement on trade would let capacity be sited for cost rather than for access, which would materially improve returns on the export-platform investments in Mexico, Morocco and Türkiye. A step change in pack cost would move the entry price point down far enough to reach first-time buyers in India, Southeast Asia and Africa, where the demand is not constrained by replacement cycles — and where the 99.8 million unit base has the most room to grow.

- **Trend:** rising
- **Capital Intensity:** high
- **Recent M&A:** yes

## Technology & Innovation

Technology and innovation: the content is the market
The industry’s innovation problem is not a shortage of technology. It is that almost every significant technology in the pipeline adds cost to the vehicle before it adds revenue to the maker, and the 4.39 per cent value forecast is the arithmetic of how much of that cost can be recovered. This chapter examines where the content is being added, what it costs, and which parts of the innovation agenda have stopped paying.
The pack is the bill of materials
The single most important fact about automotive technology in 2026 is that the battery pack is the largest line item in an electric vehicle’s cost, and therefore that cell economics set the price of the product. The US transaction data shows the consequence plainly: the average new electric vehicle sold for USD 58,034 in December 2025 against an industry average of USD 50,326, a USD 7,708 premium that is essentially the pack, financed through the monthly payment rather than paid at the till.
The chemistry roadmap is where the cost fight is being fought. Iron-phosphate chemistry has taken the volume segment because it is cheaper and safer per kilowatt-hour and the range penalty is acceptable in a car used for commuting; sodium-ion is being developed for the entry segment and for stationary applications where energy density matters less than cost and cold-weather behaviour. Neither is a performance story. Both are attempts to move the pack off the part of the cost curve that nickel and cobalt set, and their success is measured in dollars per kilowatt-hour, not in watt-hours per kilogram.
The 2025 evidence that this is working is indirect but strong: more than 20 million electric cars sold, 20 per cent up, in a year when the largest single subsidising market withdrew its support mid-year. A product whose sales survive the removal of a purchase incentive has reached cost parity in at least the segments where it sells — which, given that China sold more than 13 million of them at close to 55 per cent of its new car market, is now a very large segment.
Power electronics and the 800-volt question
Silicon carbide inverters are the most consequential component technology of the decade that almost nobody outside the supply chain discusses, because the business case for them is not efficiency. It is that a more efficient inverter lets an assembler fit a smaller, cheaper pack to achieve the same range — which means the inverter pays for itself out of the battery budget rather than out of the vehicle’s price.
That is a genuinely different kind of innovation from the ones the industry is used to. It reduces cost while it improves the product, and it therefore diffuses faster than a technology that only improves the product. The 800-volt architecture question follows from the same logic: higher voltage reduces current for a given power, which reduces conductor mass and thermal load, which allows faster charging with less cooling hardware. Whether an assembler adopts it is a decision about its charging network and its pack supplier, not about its ambitions.
Software-defined vehicle: the value without the revenue
The industry has spent a decade arguing that software will convert a low-margin hardware business into a high-margin services business. The 2025 data does not support the strong version of that claim and does support a weaker one. Software is unquestionably adding cost per vehicle — compute, sensors, wiring, and the engineering organisation to write and validate code — and it is adding that cost to a vehicle whose transaction price grew 0.8 per cent. The revenue side has arrived mostly as content bundled into the price rather than as a separate subscription line.
The weaker claim that survives is this: software’s value in this market is currently expressed as mix and retention rather than as subscription revenue. It is what allows an assembler to sell a higher trim, to keep a vehicle’s residual value up by updating it, and to hold a customer in the franchise for the second purchase. Those are real and they are the reason the investment continues, but they do not change the value forecast, because they do not change the price the market will pay.
Driver assistance as content, and as cost
Driver assistance is the most broadly distributed content driver in the industry because it applies to every propulsion type and every segment. It adds cameras, radar, ultrasonic sensors, a compute module, wiring harness capacity and a calibration burden at the dealer — and the calibration burden is the part the industry consistently underestimates, because it lands in the aftermarket as a cost the owner pays after a windscreen replacement.
That is the mechanism connecting this chapter to the restraints in the main body. Content per vehicle rises, and so does the cost of repairing it. A vehicle with expensive, calibrated electronics is more expensive to insure, and insurance is part of the monthly payment. The industry has been adding content to the product faster than it has been adding the service capacity to maintain it, and the affordability constraint now sits on both sides of that gap.
Manufacturing: the part-count collapse
The least discussed innovation in the industry is in the press shop. Large single-piece castings replace dozens of stamped and welded parts, reducing tooling count, cycle time, floor space and headcount at the same time. Combined with the powertrain swap, this is why the forecast assumes revenue per vehicle rising at 2.4 per cent while unit growth is 2.10 — the value is going up by less than the cost of the content that replaces it.
The labour consequence is straightforward and is visible in the production data. The vehicle built in 2033 needs fewer people per unit than the vehicle built in 2019, which is why a European assembly base that built 17.20 million vehicles in 2025 supports fewer jobs than one that built the same number in 2019, and why the political cost of a plant closure is now higher than the economic one.
What did not happen
Three expected technologies have not arrived on schedule. Solid-state batteries remain a small-volume proposition; the chemistry roadmap that matters for this forecast window is iron-phosphate and sodium-ion, both of which are incremental. Full autonomy at scale has not materialised, and the industry has quietly repriced it from a product to a feature. Hydrogen has consolidated into heavy duty and industrial applications and is not a passenger-car answer in this forecast.
Each of the three absences matters, because each was in somebody’s model as a content-per-vehicle driver. Their deferral is part of why the value CAGR is 4.39 per cent rather than the higher number a technology-led forecast would have produced. The industry’s realistic innovation agenda for 2026–2033 is a cost agenda wearing a technology costume — cheaper packs, cheaper inverters, cheaper assembly — and the firms that win it will be the ones that pass the saving to the buyer rather than booking it as margin, because the price data says the buyer has stopped paying.

- **Innovation:** high
- **Patent Activity:** high
- **Maturity:** growth

## Market Risk

Market risk: the seven things that would break this forecast
This forecast is not fragile. Units growing 2.10 per cent and content growing 2.4 per cent produce 4.39 per cent, and both are conservative relative to the industry’s own history. What is fragile is the margin inside that number, because the industry has one recovery mechanism — price — and the 2025 data says it has stopped working. The risks below are ordered by how much of the forecast they can actually move, not by how often they are discussed.
Risk one: affordability, and the fact that price has stopped
This is the only risk that can move the whole forecast. The US average transaction price reached USD 50,326 in December 2025, an all-time high, and grew 0.8 per cent for the year with incentives at 7.5 per cent of transaction price. A market at a record price, growing under 1 per cent, discounting 7.5 per cent to hold volume, is a market with no remaining capacity to absorb a cost increase.
The exposure is direct. If transaction prices are flat rather than rising 2.4 per cent, the value forecast falls from 4.39 per cent toward 2.5 per cent over the window — roughly USD 900 billion of 2033 value removed. The mitigation is mix, and mix is a holding pattern rather than an engine: it works while luxury and full-size pickups carry the average, and it fails when affordability pressure reaches the segments that carry the volume.
Risk two: trade policy, which is a cost with a date
OICA’s own assessment of the 2025 data flags growing exposure to trade frictions, tariff risk and localisation pressures. The structural arithmetic makes escalation likely rather than possible: the Americas sell 6.12 million more vehicles than they build, Asia builds 4.19 million more than it sells, and Europe is short 1.43 million units. Three deficit regions facing one surplus region is not a stable configuration, and the instrument that responds fastest is a tariff.
The reason this ranks second rather than first is that a tariff redistributes volume and raises cost, but it does not remove demand. The reason it still ranks second is that it can do both at once in a single market if it is large enough, and the American market is where the industry’s profit pool sits.
Risk three: the export disposal problem
China added roughly 3.25 million units of output in 2025 and sold 34.40 million against 34.53 million built, so its own market absorbed nearly all of it. That is the reassuring reading. The unreassuring one is that a production base of 34.53 million units growing 10 per cent a year cannot be absorbed indefinitely by a domestic market whose sales growth is slowing, and the surplus has to go somewhere. Every percentage point of Chinese output growth beyond domestic demand is roughly 350,000 vehicles of export pressure, and that number lands on the deficit regions in Risk two.
Japan’s configuration is the contrasting case and it is often conflated with China’s: 8.41 million built against 4.57 million sold is a 3.84 million surplus from a home market of 4.57 million, the largest export ratio of any major producer. Japan has run that configuration for forty years without triggering the tariff response China is now meeting, and the difference is distribution and local production rather than volume.
Risk four: the incentive-withdrawal shock, and its repeatability
2025 proved that removing a consumer purchase incentive removes the marginal buyer — quickly, and within a single quarter. The United States held just below 10 per cent electric share and fell sharply in the fourth quarter once credits ended. The same mechanism can fire in any market where demand is credit-supported rather than standard-supported, and the forecast’s regional growth rates assume it does not fire again at scale.
The asymmetry worth noting is that a fleet-average CO2 standard does not have this failure mode. It produces volume on a schedule because the vehicles must be sold. That is why the European and Chinese trajectories in this forecast are treated as more reliable than the American one.
Risk five: the aftermarket transition, which is a revenue timing problem
Parts and service are 28.3 per cent of this market, growing 4.0 per cent, and lagging the vehicle sale by roughly seven years. A battery-electric vehicle generates materially less service revenue than the combustion vehicle it replaces — no engine oil, no filters, no exhaust, fewer brake replacements because regenerative braking does the deceleration — and more tyre, body and electronics revenue.
For a period, the industry will be servicing a parc that is electrifying faster than its service revenue is growing. The forecast assumes the offset from electronics, tyres and body repair arrives in time; if it does not, the aftermarket block — the most profitable part of the chain — grows slower than 4.0 per cent, and the value forecast loses roughly a percentage point.
Risk six: the pack cost floor
The first decade of electrification was carried by a falling pack cost. That decline has slowed, and the 2025 European step change came from a regulatory ratchet rather than from a price signal — which is the tell. If pack cost is flat to 2033, electric vehicles do not reach the entry price points in India, Southeast Asia and Africa where the unit base has room to grow, and the 2.10 per cent unit CAGR becomes dependent entirely on replacement demand in mature markets.
Risk seven: the data itself
World vehicle sales exceeded world production by 3.39 million units in the 2025 OICA series, which is physically impossible for new vehicles in one year, and the same series notes that its world totals differ from the sum of its own rows by roughly 25,000 units in both sales and production. Every regional trade balance in this report is a difference between two series assembled from national registration counts, several of which include imported used vehicles. The direction of those balances is reliable; the decimal places are not, and a reader using them to size an export opportunity should treat them as an order of magnitude.
What would falsify this forecast
Three readings, in order of speed. US average transaction price growth staying under 1 per cent while incentives rise beyond 8 per cent of price means the value path has no cushion and the forecast is optimistic. China’s output-minus-sales gap widening past 1 million units means export pressure and tariff escalation arrive faster than the regional growth rates assume. Battery-electric share of electric sales falling back below 60 per cent means the pack cost curve has stalled and the entry-price story in Risk six is failing. None of the three has fired yet. All three are visible in published data within one quarter of firing.

- **Overall Risk:** moderate
- **Geopolitical Exposure:** high
- **Substitution Risk:** moderate

## Regulatory Landscape

Regulatory landscape: three regimes, one market, and a natural experiment
2025 was the cleanest natural experiment the automotive industry has ever run, and it was not designed. The same product was sold into three regulatory regimes that moved in opposite directions in the same twelve months, and the outcomes diverged so sharply that the year can be read as a controlled test of whether regulation or product economics sets the pace of electrification.
Europe tightened. The EU CO2 standards took a step change and electric car sales rose 30 per cent to more than 4 million units, after having stagnated in 2024. The United States loosened, with tax credits ending during the year: the electric share held just below 10 per cent and sales fell sharply in the final quarter. China continued a policy path it has run for a decade and sold more than 13 million electric cars at close to 55 per cent of its new car market. Same technology, same year, three regimes, three results. The regime is the variable.
Europe: the ratchet works, and it is now the demand model
The European mechanism is a fleet-average CO2 standard that tightens on a published schedule, with a penalty per gram of exceedance multiplied by units sold. Its distinguishing feature is not severity but predictability: an assembler knows the number four years out, which means the compliance decision is made at the model-cycle stage, which means the product arrives whether or not the consumer is ready. That is why the 2025 step change produced volume rather than a price war — the vehicles had to be sold, and they were.
The consequence for this report is that Europe’s electric growth rate is now a regulatory variable rather than a consumer one, and it will move in steps at the years the standard tightens rather than in a smooth curve. It is also why Europe, which built 17.20 million vehicles and sold 18.63 million, is importing both the vehicles and the compliance: a deficit region that tightens its standard hands the volume to whoever can supply it cheapest.
United States: withdrawal, and what it revealed
The American mechanism was a consumer purchase credit, and its withdrawal in 2025 revealed something the sector had spent years arguing about. Electric share did not collapse to a niche — it held just below 10 per cent — but it stopped growing, and the fourth quarter fell sharply. What the credit had been buying was the marginal buyer, not the early adopter. The early adopter was always going to buy the car; the credit moved the buyer who was indifferent, and when it went, so did that buyer.
Two things follow. First, a consumer-incentive regime is inherently volatile in a way a fleet-average standard is not, because it can be removed within a budget cycle. Second, the American market will now be the slowest of the three to electrify, which means the largest single profit pool in the industry — full-size pickups at USD 66,386 and a fifth of December’s shoppers buying luxury — stays combustion-led for longer. That is a competitive fact, not a regulatory one, and it advantages the groups with the deepest truck franchises.
China: a mandate by other means
China’s regime is neither a purchase credit nor a fleet standard in the European sense. It is an industrial policy that combines production-side support, local-market access rules, purchase-tax treatment and a state-directed build-out of charging infrastructure, and its result in 2025 was 16.626 million new-energy vehicles built, up 29 per cent, within a total output of 34.53 million. Nearly half of China’s production is now new-energy, and the domestic market absorbed almost all of it: 34.40 million sold against 34.53 million built.
The important regulatory feature is that China’s policy sets the industry’s cost curve, not just its own market. Scale in Chinese cell and pack production is what makes a USD 22,000 electric car possible, and that price point is what makes electrification viable in India, Southeast Asia, Latin America and the Middle East without a subsidy. The two million electric cars sold outside the three major markets in 2025 are, in a real sense, a Chinese policy outcome.
Tariffs are a fourth regime, and they are the one that costs most per unit
OICA’s own reading of the 2025 data warns of growing exposure to trade frictions, tariff risk and localisation pressures, particularly for the United States. The arithmetic in this report’s regional chapter explains why: the Americas built 18.74 million and sold 24.86 million, so the region imports roughly 6.1 million units net, while Asia built 4.19 million more than it sold. A tariff is the only instrument that lets a deficit region respond to a surplus region without waiting for a plant to be built.
The cost consequences are structural rather than temporary. A tariff raises the landed cost of an imported vehicle, which either raises the transaction price in a market where price growth is 0.8 per cent, or forces the assembler to absorb it out of a high single-digit operating margin, or moves the vehicle to a third market. All three outcomes reduce the industry’s return on capital, and the third one redirects volume to whichever region has not yet written a tariff — which is why the watch list in this report tracks the direction of trade policy rather than its level.
Compliance is a fixed cost, and fixed costs favour scale
Certification, emissions conformity, software validation, battery traceability and cybersecurity approval are all per-platform costs that do not scale down with volume. A platform sold in 200,000 units carries the same regulatory engineering burden as one sold in 2 million, which means regulatory complexity is itself a consolidation force — and it is the one force in this industry that operates in the same direction as the cost pressure rather than against it.
That is also why the twelve groups holding 66.8 per cent of the market have not been dislodged. A new entrant can build a good electric car. Building a car that is certified in forty jurisdictions is a different undertaking, and it is the part of the moat that does not appear in a specification sheet.
What regulation does, and does not, do to this forecast
It sets the mix, reliably and in steps. It does not set the total. Europe’s standard guarantees that a rising share of the vehicles it registers will be electric; it guarantees nothing about how many vehicles are registered. A market that builds 17.20 million and sells 18.63 million with flat-to-negative growth in both can hit every electrification target it has while its total volume shrinks.
That distinction is the single most important regulatory judgement in this report: the regulations reliably reallocate demand between powertrains, and they do not create demand. The 4.39 per cent value forecast follows from that — a mix shift inside a market growing 2.10 per cent in units, with content per vehicle doing the rest.

- **Complexity:** high
- **Approval Pathway:** multi_jurisdictional

## Frequently Asked Questions

**Q: How large is the global automotive market in 2025?**

USD 4,241 billion, on 99.8 million new road vehicles sold at an average transaction value of USD 30,471, plus USD 1,200 billion of parts and service revenue. It reaches USD 5,982 billion by 2033 at a 4.39 per cent value CAGR.

**Q: Why is the value forecast lower than the previous edition of this report?**

Because the price data does not support a higher one. A 6.15 per cent value CAGR on 2.10 per cent unit growth requires transaction prices to rise about 4 per cent a year for eight years. The US average transaction price rose 0.8 per cent in 2025 with incentives at 7.5 per cent of price to hold volume, so the rebuild forecasts units plus content per vehicle, and treats price as flat to slightly positive rather than as a growth engine.

**Q: Which region produces the most vehicles?**

Asia-Oceania, at 59.2 million vehicles in 2025, or 61.4 per cent of world output of 96.4 million. China alone built 34.53 million, Japan 8.41 million, India 6.49 million and South Korea 4.10 million. Europe built 17.20 million, the Americas 18.74 million and Africa 1.23 million.

**Q: How large is the electric vehicle share of the market?**

25 per cent of global new car sales in 2025, on more than 20 million electric cars sold, 20 per cent up on 2024. Battery electric vehicles were 65 per cent of those sales. China was close to 55 per cent electric by share, Europe passed 4 million units after a 30 per cent rise, and the United States stayed just below 10 per cent with a sharp fourth-quarter fall after tax credits ended.
