In two decades Tesla went from a near-bankrupt startup to the company that forced the entire car industry electric — and then, after a 2023 peak, to two straight years of falling deliveries even as its market value climbed toward $1.5 trillion.

The genuinely open question is not whether Tesla changed the world — it did — but whether today’s valuation belongs to the car business it actually runs or to the autonomy-and-robots business it keeps promising. On the automotive fundamentals the evidence is increasingly hard: deliveries down, BYD ahead, margins squeezed, the profit cushion of regulatory credits fading [51][19][14]. On the AI thesis the evidence is real but unproven: a huge data lead and a live robotaxi pilot, against a Level-2 system, a decade of missed timelines, and a competitor already running driverless at scale [35][41][43]. The evidence cuts both ways on every major question below. This study lays out both cases; the verdict is yours.

The decisive questions

Each links to the section that lays out the evidence on both sides.

The climb, the peak, and the turn

Annual deliveries (units). The rise to a 2023 peak is the bull’s proof of execution; the two declines since are the bear’s proof that the core business has stalled. Hover any point for the figure.

Tesla’s history is a genuine turnaround story — it survived 2018’s “production hell,” posted its first profit in 2020, and peaked at ~1.81M deliveries in 2023 [4][5]. The debate is about the chapter that started after that peak, when growth reversed and the company began describing itself less as a carmaker than as an AI and robotics company [29].

What Tesla is

Tesla, Inc. designs and builds electric vehicles (Models 3, Y, S, X and the Cybertruck), a fast-growing energy generation-and-storage business (Powerwall, Megapack), and a set of software and AI products — the Supercharger network, Full Self-Driving (FSD) software, a robotaxi service, and the Optimus humanoid robot. It is headquartered in Austin, Texas, run by Elon Musk, and is the most valuable automaker in the world by market capitalization. It is also a company whose stated identity is shifting: its September 2025 “Master Plan Part 4” centers on bringing “AI into the physical world” rather than on cars [29].

The dated record

How to read the rest of this study

Global EVs are not a shrinking market: ~20.7M sold in 2025, up ~20%, about a quarter of all cars [7]. The problem for Tesla is where that growth is — China is ~60% of it and Europe’s BEV sales rose ~30%, while US EV sales actually fell ~4% [7][8]. Tesla is most exposed to the slowest region and most pressured in the fastest one.

The EV market is growing — and globalizing away from Tesla’s strengths

The International Energy Agency’s data shows EV adoption accelerating, not stalling: roughly 20.7 million electric cars sold worldwide in 2025, about 20% growth and a quarter of the global car market [7]. But the geography is lopsided. China is about 60% of global EV sales; European BEV sales rose 29.7% to 2.58M; and the US fell ~4% to 1.8M [7][8]. Tesla’s home market is the laggard, and in the two big growth regions it is losing share to local rivals (see Competitive Landscape).

The quieter boom: grid-scale storage

Energy storage is the part of the industry that is unambiguously a tailwind for Tesla. BloombergNEF puts 2025 additions at ~94 GW / 247 GWh and projects a 14.7% CAGR through 2035, reaching 972 GWh of annual additions [10]. Global stationary-storage shipments grew ~85% year-on-year in the first three quarters of 2025, and Tesla was a top-five supplier overall and the leader in residential storage [11]. This is the rare market where Tesla is riding the growth rather than defending share.

2024 and 2025 figures cited ([54]); 2022–2023 are reported historical totals shown for trend.

Two industries, two trajectories

Roughly 73% of 2025 revenue ($69.5B of $94.8B) is automotive, but that line fell 10%, while energy (+27%) and services (+19%) grew [12]. The uncomfortable detail underneath: in some recent quarters Tesla’s profit would have vanished without regulatory-credit sales it no longer controls [15].

Where the money comes from

Tesla’s 2025 revenue split was automotive $69.5B (−10%), energy generation & storage $12.8B (+27%), and services & other $12.5B (+19%) [12]. The mix is shifting in a telling way: the biggest segment is shrinking while the two smaller ones grow. Crucially, energy storage carries a ~29.8% gross margin — nearly double the automotive business— and already drives close to a quarter of Tesla’s gross profit [13]. The economics are quietly tilting away from cars.

The regulatory-credit crutch

For years, Tesla sold regulatory/emissions credits to other automakers at almost 100% margin — a multibillion-dollar profit cushion that propped up earnings when the core business was thin. That crutch is being kicked away. In Q1 2025, without $595M of credits, Tesla would have posted a loss [15]. And because the 2025 US reconciliation law zeroed out the CAFE penalties that gave the credits value — roughly 75% of Tesla’s credit revenue — analysts project this line falling toward ~$0 by 2027 [14].

The recurring-revenue thesis

Tesla’s answer is to build software-like recurring revenue on top of the installed base. It moved FSD to a subscription-only model (ending one-time purchases in Feb 2026), reporting ~1.1M active FSD users at end-2025 [17]. And as ~15 other automakers adopt its NACS connector, Tesla earns per-kWh fees from rival drivers across 8,182 Supercharger stations / 77,682 stalls[16]. These are real, high-margin, and growing — but still small next to the car business they are meant to offset.

The affordability gap

A mass-market business needs a mass-market price. Tesla’s long-promised sub-$25,000 car never shipped; in October 2025 it instead launched modestly cheaper “Standard” Model 3 ($36,990) and Model Y ($39,990) — stripped of features and still above the $35,000 threshold Tesla promoted back in 2016 [18]. The unit economics of going truly downmarket, against Chinese rivals with a structural cost advantage, remain unproven.

In 2025 BYD sold ~2.26M BEVs to Tesla’s 1.64M— outselling it by over 600,000 units and taking the global crown for the first time — while its revenue passed Tesla’s too [19][21]. Yet Tesla still holds 46% of the US EV market, more than GM and Ford combined[27]. It is simultaneously losing the volume war and holding a premium fortress.

The crown has changed hands

For a decade Tesla was the world’s top EV maker. No longer. BYD ended 2025 with 2,254,714 all-electric vehicles (+27.9%), against Tesla’s 1,636,129 [19]. BYD’s overseas sales topped one million for the first time, and its full-year revenue surpassed Tesla’s — though, tellingly, the price war cut even BYD’s net profit by 19%, showing the pressure is industry-wide [21]. Tesla is also seeing its first-ever annual sales decline in China, squeezed by BYD’s pricing and faster-iterating local rivals like XPeng, NIO and Xiaomi [23].

Tesla dwarfs Western pure-plays Rivian and Lucid, but BYD now dwarfs Tesla. Sources: BYD [19], Tesla [51], VW [24], Xiaomi [25], Rivian [62].

Five Forces: a structurally harder market every year

Click a force for the rated pressure and its basis. Rivalry and buyer power are high and rising; Tesla’s vertical integration is the one force it has partly tamed. The bull case is that Tesla still out-earns most rivals despite this; the bear case is that the whole pond is draining.

Where Tesla sits

A qualitative map (placements are judgments from the cited evidence, not scores). The axes capture the central debate: Tesla is the only player that is both a large-scale automaker andan aggressive autonomy/AI platform — which is exactly why bulls and bears value it so differently. Hover a point for the basis.

Three fronts at once

Tesla’s strongest moats are vertical integration, the Supercharger network (now the North American standard), brand, and the largest real-world driving dataset [32][35]. But its stated strategy has pivoted to AI and robotics — and a moat is only as good as the business it protects. Several of these are quietly commoditizing.

Stated strategy vs. revealed strategy

Tesla’s “Master Plans” chart the official strategy. Part 3 (2023) framed a path to a sustainable global energy economy — an estimated $10 trillion of manufacturing investment, 240 TWh of storage [28]. Part 4 (September 2025) pivoted decisively to AI and robotics — “sustainable abundance” built on autonomy, energy and the Optimus robot [29]. The revealed strategy is messier: even Tesla supporters called Part 4 short on specifics, and Musk conceded the criticism was “fair” [30].

Moat 1 — Vertical integration

Tesla makes more of its own car than almost any rival: in-house 4680 battery cells, lithium refining, software, motors and increasingly its own AI chips. This is a genuine cost-and-speed lever legacy automakers cannot quickly copy. The catch is that BYD is even more vertically integratedin the one place that matters most — its in-house LFP “Blade” battery underpins a structural cost advantage that helped it pass Tesla on volume and revenue [21].

Moat 2 — The Supercharger network (and the NACS double-edge)

Tesla’s ~7,900 stations and 75,000+ connectors were long a closed advantage [31]. Then in 2023–2024, Ford, GM and most automakers adopted Tesla’s connector, and it was standardized as SAE J3400 — the North American Charging Standard [32]. That is a strategic win (Tesla earns fees from every brand’s drivers) and a strategic dilution (a proprietary moat becomes shared infrastructure). Musk then abruptly fired the entire ~500-person Supercharger team in April 2024, a move widely seen as self-inflicted damage to a crown-jewel asset [33].

Moat 3 — Brand and direct sales

Tesla built the most valuable car brand of the 2010s with no advertising and no dealers — only word of mouth and Musk’s reach (it began “a little advertising” only in 2023) [34]. That brand is now a contested asset: the same founder-led intensity that built it is, per multiple surveys, now damaging it (see Leadership & Brand).

Moat 4 — Fleet data

Tesla’s deepest potential moat is data: 10 billion+ cumulative FSD miles, accumulating at ~29 million miles a day — a real-world dataset no rival can easily match [35]. Whether that translates into deployable autonomy is the trillion-dollar question (see Autonomy, AI & Robotics). Skeptics note that “not all data is equal”: Waymo already runs a true driverless service and accepts liability, while Tesla keeps the human responsible [36].

Almost everything above Tesla’s auto business in the valuation is here. The bull case is concrete — a robotaxi service launched in Austin in June 2025 and 10B+ FSD miles of data [37][35]. The bear case is equally concrete — FSD is still Level 2 (supervised), the robotaxi ran ~42 cars at 19% availability versus a promised 500, Waymo already does ~500,000 driverless rides a week, and Musk has promised full autonomy “next year” every year since 2016 [41][40][43][42].

Robotaxi: launched, but small and supervised

Tesla launched a robotaxi service in Austin on June 22, 2025 — but with about ten Model Ys, a Tesla “safety monitor” in the front passenger seat, and a narrow geofence [37]. Eight months later it was running ~42 vehicles at just 19% availability, far short of Musk’s promise of 500 cars by year-end 2025 [40]. Regulators took notice: within weeks NHTSA opened an inquiry after videos showed robotaxis driving on the wrong side of the road and braking erratically [38], and the service logged 17 reported automated-driving incidentsfrom July 2025 to March 2026 — with Tesla the only operator to fully redact its crash narratives, citing “financial harm” [39].

FSD: real progress, but still Level 2

Tesla’s “Full Self-Driving” remains, by its own labeling, FSD (Supervised)— a Level 2 system that requires the driver to stay responsible [41]. As of April 2026 Musk again pushed unsupervised consumer FSD to “probably” the fourth quarter of 2026 [41]. The pattern is the heart of the skeptics’ case:

Similar promises — “feature complete this year,” “a million robotaxis next year” — recur from 2016 through 2024, none delivered on schedule [42].

Optimus and the AI stack

Musk calls the Optimus humanoid robotpotentially “the biggest product ever,” targeting 1,000+ units in 2025 scaling toward a million a year at ~$30,000 each [46]. The credibility gap showed at the October 2024 “We, Robot” event, where the robots that chatted and poured drinks were teleoperated by humans— autonomous only in walking — a fact not disclosed on stage[45]. On compute, Tesla shut down its in-house Dojo supercomputerin August 2025, with Musk calling the design “an evolutionary dead end” and refocusing on AI5/AI6 chips [47].

The valuation that rides on all of this

The bull thesis is explicit that cars are almost beside the point. ARK Invest projects the robotaxi business could drive ~90% of Tesla’s enterprise value by 2029, implying a market cap up to $8.3 trillion [48]; Wedbush’s Dan Ives values the autonomy/robotics opportunity at “at least $1 trillion” [49]. The bear thesis is that this prices a future that keeps slipping — Tesla’s stock hit record highs in late 2025 even as EV sales fell, with robotaxis driving a valuation well above $1 trillion despite generating under 1% of revenue[50].

2025 was Tesla’s first-ever annual revenue decline (−3% to $94.8B), with net income down 46% to $3.79B and operating margin slipping to 5.7% [52][53]. Deliveries fell ~9% — a second straight drop [51]. Against that, the market cap sits near $1.56 trillion, or roughly 400× trailing earnings [57].

The core business contracted

After peaking in 2023, Tesla’s fundamentals turned down. Deliveries fell to 1,636,129 in 2025, a ~9% decline and the second consecutive annual drop, “and the decline is accelerating” [51]. Revenue posted its first annual decline on record, and quarterly results showed the squeeze: Q2 2025 revenue fell 12% with operating income down 42% [55], and Q4 GAAP profit fell 61% year-on-year[53]. With Musk leading DOGE in early 2025, Q1 profit had cratered 71% to $409M [58].

The scoreboard

*2023 net income included a one-time ~$5.9B deferred-tax benefit, so it overstates underlying profitability and isn’t comparable to 2024–2025. 2023/2024 segment lines are reported historical figures shown for trend.

The valuation gap

Here is the crux. On 2025’s $3.79B of profit, Tesla’s ~$1.56T market cap implies roughly 400× trailing earnings— an order of magnitude above any traditional automaker[57]. The stock hit record highs in late 2025 even as deliveries fell, because investors are pricing robotaxis, Optimus and AI rather than cars — robotaxis drive a >$1T valuation while still under 1% of revenue [50]. Whether that is foresight or froth is the single most contested question about Tesla — and the market reminds you it is also a Musk-sentiment stock: his late-April 2025 retreat from DOGE added ~$158B of value in days [59].

On volume Tesla is now second to BYDand shrinking; on profitability it still towers over Western pure-plays like Rivian (42,247 deliveries, −18%) [60][62]. But its ~$1.56T market cap on $3.79B of profitmeans its real “comparables” aren’t carmakers at all — even longtime bull Morgan Stanley values the auto business at only ~$55/share[61].

Side by side (2025)

BYD’s figure is pure-EV (BEV); it sold ~4.55M vehicles in total including plug-in hybrids. Profitability descriptions mix disclosed figures (Tesla, BYD) with reported results (Rivian) and estimates (Xiaomi, VW EV unit).

The two ways to read it

Read as a carmaker, Tesla is mid-story: it lost the volume crown to a lower-cost rival but remains the most profitable Western EV maker by a wide margin, while Rivian and Lucid burn billions to deliver a tiny fraction of its cars [60][62]. Read as a stock, it is in a different universe: Morgan Stanley — historically a bull — downgraded Tesla to Equal-Weight and pegged the auto business at just ~$55 a share, meaning the overwhelming majority of the price is autonomy, robots and energy optionality, not cars [61].

Tesla’s board calls Musk “critical to the brand,” and in November 2025 shareholders approved a performance award worth up to ~$1 trillion [65]. In the same period, Musk’s politics were tied to a brand-value slide and a ~28% European sales collapse, and his attention is split across four companies [69][22][68]. Both the “indispensable visionary” and the “central risk” readings are well-evidenced.

The pay saga: rescinded, then restored

Musk’s 2018 pay package — the largest in corporate history — was rescinded by the Delaware Court of Chancery in January 2024, with Chancellor McCormick reaffirming in December 2024 and awarding $345M in plaintiff fees [63]. Then, in December 2025, the Delaware Supreme Court reversed and ordered the ~$55B package restored, awarding Tesla nominal damages of $1 [64]. In between, Tesla reincorporated in Texas and, in November 2025, shareholders approved an entirely new award.

Key-man risk, in the board’s own words

The bull and bear cases here use the same fact — Tesla isMusk — and reach opposite conclusions. Chair Robyn Denholm framed the 2025 vote as existential, warning that rejecting the package “could risk Musk’s departure” and that his leadership “is critical to the brand’s success” [65]. Critics read the same dependence as a liability: a single point of failure whose attention is divided across Tesla, SpaceX, xAI and X.

Governance critics also flag self-dealing risk — Musk negotiating both sides of inter-company deals (SolarCity, the X→xAI merger), owning ~20% of Tesla but controlling ~85% of SpaceX’s votes [70].

The brand became political

Tesla’s brand, once an unambiguous asset, is now contested. It fell from #8 in 2021 to #95 of 100 in the 2025 Axios/Harris reputation poll — last in “character” — with the decline tied to Musk’s political activism [69]. YouGov tracked US net favorability hitting a nine-year low[80]. A grassroots “Tesla Takedown” movement staged protests in 250+ cities, and one poll found 31% of Tesla drivers had sold or considered selling over Musk’s actions [71].

The risks are not hypothetical — most are already visible in the 2025 numbers: fading credits, a demand cliff after the EV tax credit expired, Cybertruck’s struggles, a regulatory probe covering 3.2M vehicles, and a ~400× valuation that leaves no room for disappointment [14][73][44][57]. The contested part is how much each one ultimately matters.

The concrete risks

Demand and incentives

The US federal $7,500 EV tax credit expired Sept 30, 2025, pulling demand forward and forcing Tesla into its most aggressive incentives yet — 0% APR financing and $0-down leases [73]. Demand support is being withdrawn just as competition intensifies.

Cybertruck

Tesla’s halo product has badly missed expectations: it sold just ~6,406 units in Q1 2025 (a 50% sequential drop), drew a ~46,000-unit recall over a trim panel that can detach in motion, and depreciates ~35% in a single year [75][74][77]. In Q4 2025, SpaceX bought 1,279 Cybertrucks — ~18% of US registrations — without which registrations would have fallen ~51% [76].

Regulatory and safety

NHTSA escalated its FSD low-visibility investigation to an Engineering Analysis covering ~3.2 million vehicles(nine incidents, one fatality) — the final step before a possible recall — alongside a separate robotaxi inquiry [44].

Valuation

At ~400× trailing earnings, the stock prices in autonomy and robotics success that has repeatedly slipped; any delay or safety setback to the robotaxi/FSD narrative could compress the multiple sharply[57][50].

What reasonable people disagree about

This study does not resolve these for you — that is the point. The evidence on each is laid out in the sections above; the weighting is yours. See Methodology & Limits for what is disclosed, what is estimated, and where this analysis could be wrong.

Method

Research proceeded by fanning out across many web searches and then directly fetching the primary and reputable secondary sources they surfaced — Tesla’s own disclosures (as reported), regulatory and court documents (NHTSA, the Delaware courts), the financial and trade press (Reuters via secondary reports, Fortune, CNBC, just-auto, Euronews, CnEVPost, PBS, Axios, Al Jazeera), specialist EV outlets (Electrek, TechCrunch, CleanTechnica, EVXL) and compiled references (Wikipedia, Waymo’s own blog). Every URL cited here was actually opened and read; the claims it supported were transcribed into a structured source manifest that tags each entry with a tier (1 = primary/official, 2 = reputable secondary, 3 = forums/soft or compiled reference), a confidence level and a stance (supporting / critical / neutral). The load-bearing figures for Tesla are its delivery and revenue trajectory, automotive gross margin ex-credits, regulatory-credit revenue, net income, and the forward-looking robotaxi, Optimus and autonomy claims on which the bull/bear gap mostly turns.

Frameworks used

The analysis applies the Pyramid Principle for the answer-first executive summary, Porter’s Five Forces for the industry structure, a peer-comparables table, a 2×2 positioning map placing manufacturing scale against autonomy/AI ambition, a SWOT, and a bull/base/bear scenario read — each chosen because Tesla competes on both a manufacturing axis and a software/AI axis and the contested questions live in how those interact. They are applied even-handedly, with weaknesses and high-pressure forces given the same weight as strengths; frameworks organize the evidence rather than render a verdict. A formal discounted-cash-flow valuation was deliberately skipped because the outcomes hinge on robotaxi and Optimus assumptions wide enough that any single DCF would imply false precision.

Disclosed vs. estimated

Tesla is a public company, so most headline figures — deliveries, revenue, margins, net income — are disclosedin its filings and quarterly updates and are reported here on that basis. Where a number is presented on a comparable or directional basis (for example, automotive margins quoted ex–regulatory credits, or region-level trend lines) it is flagged as such rather than read as an official line item. Distinct from both are the third-party estimates and forward claims— regulatory-credit run-off projections, robotaxi fleet counts, Optimus production targets, analyst price targets and brand-value surveys — which are labeled as estimates or claims, not facts. One mechanical caveat: Tesla’s SEC filings and IR decks repeatedly blocked the automated fetcher, so several disclosed figures are cited to the reputable secondary outlet that reported them rather than to the filing itself — the numbers are Tesla’s own, but the link is to the reporting.

Neutrality & independence

This is a compilation, not an argument. Every section pairs the case for and the case against, and critical and positive claims alike are attributed to their sources so a reader can weigh them directly. The author is not affiliated with, sponsored by, or endorsed by Tesla, Inc. or any other company named here. Everything is point-in-time as of 2 June 2026; in a fast-moving story the figures may be stale soon after that date, and corrections are welcome.

Cross-checked at build time by an automated link checker. Several primary sources — Tesla’s SEC filings and IR quarterly decks, plus some CNBC/Reuters pages — block automated fetchers (HTTP 403); where a figure originated in one of those, it is cited to the fetched, reputable secondary outlet (Electrek, just-auto, TechCrunch, Fortune, Euronews, CnEVPost) that reported it. See Methodology & Limits.