Tesla (TSLA): A Long-Term Business Owner's Research Memo

[Fact] The cost structure is heavily manufacturing-like. In 2025 Tesla's total gross profit was 17.094 billion dollars, with a gross margin of 18.0%; within that, total automotive gross margin was 17.8% and energy generation and storage gross margin was 29.8%. By comparison, in 2022 total gross margin was 25.6% and total automotive gross margin was 28.5%. This means that over the past three years, Tesla's profit swings have come mainly from ASP, capacity utilization, product mix, subsidies/credits, raw materials, and depreciation and amortization, not from a high-contribution-margin software revenue pool. Fixed costs and capital spending are both fairly sticky, so this is a company an investor must understand as advanced manufacturing plus a product cycle, not one to which an internet-platform framework can simply be applied.

[Fact] On customer profile, Tesla's auto side is mostly consumer-facing, while the energy side serves more utility/commercial-and-industrial and residential storage customers. In the primary documents retrieved, the company does not provide the share of its top five customers, so I cannot treat "low customer concentration" as fact; I can only confirm that Tesla does not emphasize dependence on a single customer in its 10-K the way a traditional B2B equipment maker with large accounts would. This disclosure gap limits any judgment on "customer stability."

[Reasonable inference] If I had to explain in 200 words how Tesla makes money to a non-specialist, I would put it this way: it sells cars first and collects most of its cash flow up front from a one-time delivery; once a car is sold, it keeps earning afterward through repairs, Supercharging, insurance, software features, used cars, and leasing; and it also sells Megapack, Powerwall, and solar-related products. The market was willing to assign it a high valuation in the past not because software profit is already large, but because it believed Tesla would eventually turn the fleet, the charging network, autonomy, robots, and energy systems into higher-margin, more repeatable revenue. The issue is not whether that story is compelling, but that to this day these incremental profits have not landed in the financials at a large enough scale.

The stock-market test: no. Not because the company would fail, but because if I could not look at the screen for the next five years, my grasp of the return sources as a business owner is still not sufficiently "financials-grounded": medium-to-long-term shareholder returns still depend visibly on option value from autonomy, Robotaxi, Optimus, and the like, which have not yet formed a large-scale, verified economic model, rather than being supported independently by verified owner earnings. Business comprehensibility score: 3/5. I can clearly describe today's car-building, storage, services, and deferred-revenue mechanics; but for the Autonomy/Robotaxi/Optimus economics most critical to how the market prices the stock, there are still one or two big black boxes.

Industry and Competitive Landscape

[Fact] Per public securities-classification sources, Tesla's closest GICS placement is Consumer Discretionary / Automobiles & Components / Automobile Manufacturers; the classification is imperfect, because Tesla also runs storage and energy systems, but it still leads with the capital intensity and competitive structure of an automaker. On the industry anchor, Damodaran's January 2026 Auto & Truck data shows the U.S. industry has an after-tax operating margin of about 2.44% and an ROIC of about 2.62%. This set of anchors tells us the auto industry is not inherently high-return, long-run average capital returns are modest, and in valuation one must beware of mapping a technology dream straight onto manufacturing cash flow.

[Fact] Long-term demand is not fake. IEA data and mainstream-media accounts show that 2025 global EV sales already exceeded 20 million units, about a quarter of new-car sales; 2026 is projected to reach 23 million units, close to 30% of the auto market. In other words, electrification itself is structural demand, not a false premise. The issue is that industry growth does not automatically translate into share gains or margin expansion for Tesla, especially as Chinese makers keep compressing the moat of global competitors on cost, development cycle, and model iteration.

[Fact] The competitive landscape is getting harder. The IEA's 2025 report notes that even after Tesla cut the Model Y price further in 2024, its market share still fell about 10% year over year. A Reuters survey in July 2025 found that the average age of Chinese-brand EVs or plug-in hybrids on sale in the home market is only about 1.6 years, versus about 5.4 years for foreign brands; Chinese automakers can compress the development cycle of an all-new or refreshed model to roughly 18 months. This combination means Tesla faces not just "more competitors," but faster product cycles, lower cost thresholds for trial and error, and more intense price wars.

[Fact] Tesla's own data also confirms the "trading price for share" pressure. The four quarterly delivery figures in 2025 were 336,700, 384,100, 497,100, and 418,200 vehicles, for a full year of about 1.6361 million units; Q1 2026 deliveries were 358,000, up only 6% year over year. Over the same span, the four quarterly total-revenue figures in 2025 were 19.335 billion, 22.496 billion, 28.095 billion, and 24.901 billion dollars, with Q1 2026 at 22.387 billion dollars; total gross margin recovered from 16.3% in 2025Q1 to 21.1% in 2026Q1, but this recovery is not yet enough to offset the large amount of long-dated expectation already priced into the stock. Tesla remains one of the world's most important pure-EV brands, but it is no longer in a phase of "share rising in one direction, margins expanding in one direction."

[Fact] From a profit-pool view, the most notable thing in 2025 is not autos but energy. Tesla's 2025 energy generation and storage revenue was 12.771 billion dollars, up 27% year over year, with a gross margin of 29.8%; the broad auto-ecosystem revenue was 82.056 billion dollars, down 6% year over year, with a gross margin of 16.2% (auto plus services combined). Put differently, the profit pool is migrating from vehicles toward storage and software/services; if Tesla can eventually prove that energy and software can steadily take a larger share of gross profit, the valuation framework is more likely to shift from "automaker plus options" toward "platform plus industrial software." To date, the direction of this migration is correct, but the scale is still not enough to overwhelm the core auto business.

[Opinion] Between "a good company in a good industry" and "an excellent company in a poor industry," Tesla today still looks more like the latter. Long-term industry demand is good, but the core auto business has a modest industry-average return and fragile price discipline. Historically Tesla earned above-industry-average economic returns through brand, manufacturing efficiency, software, and capital-market credibility, but over the past three years that gap has been narrowing. Industry attractiveness score: 3/5. Industry growth is real, but profit distribution is unstable and competitive intensity has clearly risen.

Moat, Management, and Capital Allocation

Moat Analysis

Source type: company primary disclosure (10-K, Q1 2026 Update), official quarterly updates; industry/regulatory supporting sources are IEA and Reuters.

[Fact] The most certain part of the moat is not the "FSD myth," but the combination of brand plus charging network plus manufacturing vertical integration plus access to capital. In 2026Q1, Tesla's Supercharger stations and connectors reached 8,463 and 79,918 respectively; active FSD subscriptions reached 1.28 million; cumulative deliveries reached 9.2 million. This data shows it has a still-considerable installed base and set of service touchpoints worldwide.

[Reasonable inference] But if the moat is defined as "able to support a stable ROIC at least 5 percentage points above the cost of capital over 10 years," then Tesla's moat has not kept widening over the past three years. On the contrary, automotive gross margin fell from 28.5% in 2022 to 17.8% in 2025, and total operating margin fell back from 16.8% in 2022 to 4.6% in 2025. This is not a monopoly platform collecting rent; it looks more like an advanced manufacturer that still needs price cuts, refreshes, financing stimulus, and capacity adjustment to hold volume.

[Reasonable inference] Its inflation resistance is moderate, not top-tier. During the high-inflation period of 2021–2022, Tesla's gross margin and operating margin rose sharply, but over 2023–2025, as price wars and competition intensified, those margins gave back quickly. This shows that part of the earlier high margin came from supply-demand mismatch and early-stage industry tightness, not from permanent pricing power. Moat trend: stable but narrowing. Difficulty to replicate: a vehicle platform and brand take years, but on per-vehicle cost, product-development speed, and the storage supply chain, Chinese competitors are catching up clearly faster than traditional European and U.S. automakers. Moat-strength score: 3/5.

Management and Capital Allocation

[Fact] Elon Musk has served as Tesla CEO since October 2008; Vaibhav Taneja has served as CFO since August 2023, having served as CAO since 2019 and previously worked at SolarCity. Management continuity is fairly strong, especially on the CEO dimension, but a key handover occurred at the CFO level only in 2023.

[Fact] As of March 31, 2024, the 2024 proxy shows Elon Musk's beneficial ownership at 715.0 million shares, or 20.5%, including about 411.1 million shares held directly through a trust and about 304.0 million shares exercisable within 60 days; of these, about 238.4 million shares were pledged as collateral for personal debt. On an "options-excluded" basis, counting only common shares already held, Musk's ownership is closer to 12.9%. This means both alignment of interest and, in governance terms, high concentration, pledging, and key-person risk.

[Fact] The strength of capital allocation lies in continuously plowing cash flow into capacity, storage, supply-chain vertical integration, and AI infrastructure. In 2025 operating cash flow was 14.747 billion dollars and net purchases of property and equipment were 8.527 billion dollars; meanwhile purchases of investments were 37.109 billion dollars and proceeds from maturities of investments were 30.158 billion dollars. The 2025 10-K also states clearly that 2026 capital spending is expected to exceed 20 billion dollars, mainly for AI compute infrastructure, data centers, manufacturing and R&D production lines, and the fleet plus the retail/service/charging network. The company pays no dividend and does no regular buybacks.

[Fact] The most serious problem in capital allocation is dilution. The 2025 10-K shows common shares outstanding rose from 3.216 billion at the end of 2024 to 3.751 billion at the end of 2025; the statement of equity in the same filing shows that in 2025 "common stock issued for equity incentive awards (net)" reached 535 million shares. That amounts to a single-year increase in shares issued of about 16.6%. For a long-term owner, this is not a "non-cash cost" but a real dilution of per-share ownership.

[Fact] Governance red flags must also be written out plainly. The 2024 proxy discusses at length the Delaware Chancery Court's ruling on Musk's 2018 compensation plan and the company's push for shareholder re-approval; in September 2025 the company also rolled out a new CEO Performance Award and filed related 8-K exhibits in November 2025. At the same time, Reuters reported in July 2025 and January 2026 that Musk pushed Tesla to put to a shareholder vote whether to invest in xAI, and that Tesla ultimately invested 2 billion dollars in xAI. This means Tesla, as a public company, has begun to exhibit more complex capital and governance coupling with other entities Musk controls.

[Fact] On the other hand, on audit quality I did not find the classic red flag of "frequent auditor changes." Tesla states in both its 2020 and 2022 annual reports that PricewaterhouseCoopers LLP has served as the company's auditor since 2005; this round of retrieval found no qualified audit opinion. The 2024 proxy further discloses that the company is aware of no late Section 16(a) filings.

[Opinion] My assessment of management and capital allocation is strong innovation-driven capability, but per-share value discipline markedly weaker than the pursuit of revenue/scale/long-dated story. Management and capital-allocation score: 2/5. The main reason for the deduction is not technical execution, but governance, dilution, and the boundaries of related-party capital allocation.

Financial Quality and Owner Earnings

Financial Quality

The table below lists only the core metrics that are fully primary-verified and most useful for long-term judgment. Several turnover ratios in the longer historical series, the complete 10-year medians, and M-Score/F-Score were not primary-verified item by item under this round's public-data chain, so they do not enter the valuation model and are used only as support in the written judgment.

Source type: company primary disclosure; documents: Tesla 2025/2022/2020/2018 Form 10-K; reporting period: FY2021–FY2025; currency and basis: USD, US GAAP, consolidated. 2021–2022 data is from the 2022 10-K; 2023–2025 data is from the 2025 10-K; share count uses period-end shares issued and outstanding.

[Fact] The strongest side of financial quality is the balance sheet. At year-end 2025 total assets were 137.806 billion dollars and shareholders' equity was 82.137 billion dollars; interest-bearing debt and finance leases combined were only about 8.376 billion dollars, of which the current portion was 1.640 billion and the non-current portion 6.736 billion. In 2026Q1 cash, cash equivalents, and short-term investments reached 44.743 billion dollars. On this rough calculation, Tesla remains in a net-cash position, with net cash on the order of about 36 billion dollars. For an automaker with heavy-capital characteristics, this significantly lowers bankruptcy-type risk.

[Fact] The weakest side of financial quality is the decline in margins and per-share economics. 2022 was the high point of profitability over the past five years: operating margin 16.8%, net margin 15.4%. By 2025, operating margin had fallen to 4.6% and net margin to 4.0%; yet the share count expanded sharply in 2025. In other words, Tesla's current problem is not "short of money" but the combination of falling per-share profit density and a still-very-expensive valuation.

[Reasonable inference] The degree to which profit converts to cash does not meet the standard of a typical "high-quality consumer/software compounder." Over 2021–2025, FCF/net income to parent was about 0.91x, 0.60x, 0.29x, 0.50x, and 1.64x, with a five-year median of about 0.60x. Only in 2025, because the earnings base was very low while operating cash flow stayed stable, does FCF/net income look unusually high. Overall, Tesla's cash-generation capability is steadier than its income statement, but to sustain growth it still needs continuous heavy capex and working-capital investment. This looks more like an industrial company still expanding capacity and betting on next-generation businesses than a "net income roughly equals free cash flow" asset-light compounder.

[Reasonable inference] The marginal economics of growth are deteriorating. When revenue surged in 2021–2022, gross margin and operating margin expanded in step; but over 2023–2025, with revenue roughly flat or even slightly lower, margins compressed clearly. Treating 2022 as the inflection point, Tesla's incremental revenue, at least in the core auto business, no longer enjoys the early supply-shortage operating leverage, and marginal ROIC has most likely declined. Because several sub-items in the full 10-year invested-capital series were not verified one by one, I do not put a precise ROIC figure into the valuation core, but the trend judgment is clear: marginal returns are no longer rising.

[Fact] On accounting quality, this round of retrieval found no high-risk signals such as auditor changes, qualified audit opinions, or significant late filings; but two items need ongoing tracking. First, in 2025 prepaid and other assets showed a net outflow of 3.181 billion dollars in the operating-cash-flow adjustments, a fairly large amount. Second, the 2025 stock-based compensation and share dilution were very large, and even though they appear as non-cash items under GAAP, they should be treated as an economic cost.

[Stress test] If future revenue falls 20%, starting from the 4.6% operating margin of 2025, with worsening fixed-cost absorption, operating profit could well approach zero; if revenue falls 30%, the core auto business's EBIT would likely come under clear pressure and might even turn negative. But because the company is in a net-cash position, interest expense was only 338 million dollars, and 2025 EBIT was still 4.355 billion dollars, the risk of a financing default or covenant trigger is markedly lower than for a traditional highly leveraged automaker. The direction of this conclusion is high-confidence; the precise covenant test value is stated conservatively because all the terms of the public credit agreements were not broken out item by item in this report.

Owner Earnings

For the definition of Owner Earnings I present two versions at once. The first layer is the "Buffett original": operating cash flow minus maintenance capital expenditure. The second layer is the "economically conservative": on top of the first layer, a further deduction for the economic cost to long-term shareholders, especially heavy SBC dilution. I do this because Tesla's 2025 share increase was so large that treating SBC as costless would systematically overstate per-share owner earnings.

Maintenance-capex estimation method: [Assumption] Anchoring on 2025 depreciation and amortization of 6.148 billion dollars, I set maintenance capex at 0.9x–1.1x of D&A, i.e., about 5.5–6.8 billion dollars. The reasoning: Tesla's current total capex is already heavily mixed with expansionary projects such as AI compute, data centers, and Semi/Cybercab/Megapack 3, so not all capex can be treated as maintenance; but it is, after all, a high-intensity manufacturer, so maintenance capex cannot be set too low either. The 2025 10-K's guidance for 2026 capex exceeding 20 billion dollars also indirectly proves that "total capex no longer represents maintenance spending."

On this basis, the 2025 estimate is as follows:

Units: billions of dollars. Source type: company primary disclosure plus research assumptions; document: Tesla 2025 Form 10-K; reporting period: FY2025; currency and basis: USD, US GAAP, consolidated. The SBC deduction is an [Assumption], used to protect the per-share valuation from dilution distortion.

[Opinion] In the valuation I use a neutral Owner Earnings of 6.3 billion dollars, conservative 4.9, optimistic 7.7. Using the 2026-05-19 share price and about 3.752 billion shares, Tesla's current market cap is about 1.52 trillion dollars. On that basis, the implied P/Owner Earnings is roughly 197x–310x. For a company whose 2025 revenue fell year over year, whose automotive gross margin is far below the 2022 peak, and which still needs large-scale reinvestment, this multiple is extremely high.

Valuation, Margin of Safety, and Cross-Opportunity Comparison

Intrinsic-Value Estimate

Method A: Owner Earnings discounting

I separate Tesla's primary-verifiable financials from the "high-uncertainty AI option." The neutral DCF does not treat the future large-scale commercialization of Robotaxi/Optimus as verified fact; it discounts only the already-existing automotive, storage, services, and software base. The discount rate uses: a risk-free rate of 4.67% (the U.S. Treasury 10-year on 2026-05-19); plus a higher uncertainty compensation, forming conservative 12% / neutral 10.5% / optimistic 9.5% equity discount rates. The extra risk premium here is an [Assumption], not a judgment on future market rates.

The key DCF assumptions are as follows: Conservative scenario: starting OE of 4.9 billion dollars, 4% compound growth over the next 5 years, 3% over the following 5, terminal growth 2.5%, discount rate 12%. Neutral scenario: starting OE of 6.3 billion dollars, 8% compound growth over the next 5 years, 4% over the following 5, terminal growth 3%, discount rate 10.5%. Optimistic scenario: starting OE of 7.7 billion dollars, 15% compound growth over the next 5 years, 5% over the following 5, terminal growth 3.5%, discount rate 9.5%. Without explicitly capitalizing long-dated Robotaxi/Optimus earnings, the per-share intrinsic value I get is roughly: conservative 40–60 dollars, neutral 80–130 dollars, optimistic 180–260 dollars. This is a valuation of the "verified-financials economy," excluding a large amount of unproven AI-monetization premium.

Method B: relative valuation

Working back from the verified financials, Tesla's 2025 net income to parent EPS was 1.08 dollars, so the current price implies a trailing P/E of about 374x; year-end 2025 shareholders' equity was 82.137 billion dollars and period-end share count 3.751 billion, implying a P/B of about 18.4x; on 2025 revenue of 94.827 billion dollars and net cash of about 36 billion dollars, EV/Sales is about 15–16x; on free cash flow of 6.220 billion dollars, P/FCF is about 244x. These multiples are closer to a high-growth software platform than to a capital-intensive automaker.

Secondary external models give an entirely different conclusion: Morningstar in April 2026 raised Tesla's long-term fair value to 400 dollars, explicitly citing a higher valuation for Robotaxi, the humanoid robot, and autonomy-software adoption; DBS in May 2026 assigned a rating of HOLD with a 12-month price target also of 400 dollars. The publicly visible analyst-target mean is about 411.89 dollars, with a high of about 600 dollars and a low of about 123 dollars. This shows that the current price is not mainly valuing the auto and storage base, but valuing the Autonomy/AI option.

Method C: asset/liquidation value

This has the weakest explanatory power for Tesla. Its year-end 2025 book shareholders' equity was about 82.1 billion dollars, or roughly 21.9 dollars per share; the rough 2026Q1 net cash of about 36 billion dollars is roughly 9–10 dollars per share. Book assets matter somewhat for the floor, but if Tesla's real value exists, it comes mainly from brand, software, network, and future optionality, not liquidation recovery. The asset method here is more of a valuation floor, not enough to support the current price.

Combining the three methods, I give the following ranges:

Note: the high end of the relative valuation depends heavily on secondary research's discounting of long-dated Robotaxi/Optimus profit; I place it under "optimistic reference" rather than in "neutral verifiable value." Current price versus neutral value: taking the neutral combined intrinsic-value midpoint at about 125 dollars, the current price carries a premium of about 223%. Even at the optimistic combined ceiling of 350 dollars, the current price is still above that ceiling.

Margin of Safety and Cross-Opportunity Comparison

[Fact] The current price is not at a discount to conservative intrinsic value but at a huge premium. Even using a valuation framework friendly to Tesla, 404.11 dollars is already near or above the top of the optimistic range. For a value-investing framework, this means the question now is not "how thick is the margin of safety" but "the margin of safety basically does not exist."

The three most fragile assumptions in the valuation, ranked by sensitivity, are in my view: First, the monetization rate of autonomy/Robotaxi. If this assumption fails to deliver, Tesla looks more like a high-quality but heavy-capital automaker plus storage company than an AI asset that can be valued as a software platform. Second, whether automotive operating margin can return to double digits. The 16.8% operating margin of 2022 is an important reference for the valuation's imagination, but 2025 was only 4.6%. Third, whether share dilution can slow markedly. If the dilution pace of 2025 repeats, per-share value will be continuously diluted.

In a zero-growth scenario, if Tesla is viewed as a long-term Owner Earnings business of about 6 billion dollars with no long-dated AI premium, then the yield against the current 1.52 trillion dollar market cap is only about 0.4%; even using 2025 free cash flow of 6.2 billion dollars, the FCF yield is only about 0.4%, clearly below the 10-year U.S. Treasury of about 4.67% on 2026-05-19. In other words, waiting on price is itself not expensive.

The margin-reverting-to-industry-median scenario is also worth spelling out. Damodaran's current-2026 Auto & Truck industry after-tax operating margin is about 2.44%, while Tesla's 2025 operating margin was 4.6%. If Tesla's auto-plus-services base can only sustain a normalized margin close to the industry's, and energy and software fail to scale quickly, then much of the "technology premium" implied in the current valuation needs to compress. For a long-term holder, this is not short-term volatility risk but the risk of permanently overpaying.

On cross-opportunity comparison, I set Tesla against the "risk-free rate" and "the index" at a minimum. For a long-term investor who can hold only five assets, at the current price Tesla is not clearly superior to buying a broad index or the 10-year Treasury. The index at least does not require you to be continuously right about a string of high-uncertainty variables such as Robotaxi, Optimus, FSD subscription penetration, and xAI-related capital allocation; and the 10-year Treasury's current nominal yield of about 4.67%, while lacking upside imagination, has the virtue of not requiring you to prepay an EV/Sales above 15x for a long-dated story. Conclusion: in a "hold only five assets" portfolio, I would not let Tesla at the current price crowd out a broad index or a high-grade long-term government bond.

Risk List, Bear Case, Final Conclusion, and Self-Check

Risks and the Bear Case

I rank the risks with the largest impact on permanent capital loss by importance rather than probability, as follows.

Source type: company primary disclosure plus Reuters plus IEA plus U.S. Treasury.

If I put the bear case in the sharpest professional-short voice, the core would be this:

First, Tesla is now priced as an "AI company not yet verified," but its financials still come mainly from the low-return, price-volatile auto and manufacturing business. 2025 revenue fell 3% year over year and operating margin was only 4.6%, yet the market assigned it more than 1.5 trillion dollars of market cap. A short would say this is not "cheaply buying the future" but "prepaying in one shot for multiple long-dated tracks all at once."

Second, governance is eroding per-share ownership. In 2025 shares issued and outstanding rose from 3.216 billion to 3.751 billion, a single-year increase of about 16.6%; and the 2024–2025 proxy materials again pushed CEO incentives and a shareholder re-vote to center stage. A short would read this as: the grander Tesla's capital story, the thinner the real claim common shareholders actually get on a per-share basis.

Third, the regulatory and safety risks of autonomy remain very real. Reuters reported in October 2024 that NHTSA had opened a probe into FSD-related accidents across 2.4 million Tesla vehicles; in March 2026, Reuters again reported that NHTSA would escalate a probe into FSD visibility-detection issues across 3.2 million Tesla vehicles, involving multiple crashes and at least one fatality. A short would say: if the business that truly determines the valuation is still subject to regulatory scrutiny to this day, then you cannot pay up front for large-scale commercialization.

Fourth, the boundaries of related-party capital allocation are blurring. Reuters reporting shows Tesla invested 2 billion dollars in xAI, while xAI is also another core entity in Musk's control structure. A short would question on this basis whether the public company's cash flow will increasingly serve the Musk ecosystem rather than strictly serve the maximization of Tesla's per-share intrinsic value.

If the short is right, then three years out Tesla would likely become a company that is financially safe and still has technical bright spots, but is repriced by the market as a "low-growth automaker plus storage-equipment maker plus an unrealized AI option." It would not be a bad company, but it would be one that delivers poor returns to anyone who bought at the 2026 highs.

Conversely, if the following three hard facts appear, I would be willing to admit my conservative judgment is wrong: First, FSD/Robotaxi obtains continuous commercialization licenses in major jurisdictions and discloses verifiable safety and unit-economics data. Second, without relying on regulatory credits, automotive gross margin stabilizes back above 20% and operating margin returns to double digits. Third, dilution slows markedly to no more than 2% per year, while capital investment in related entities such as xAI is subject to clear governance constraints.

Checklist, Final Conclusion, and Self-Check

[Final rating] Watch [Confidence level] Moderate. Confidence is not high, because the AI/Autonomy economic model that sets the stock's ceiling has not yet been verified in the financials; but on the point that "the current price lacks a margin of safety," my grasp is in fact fairly high.

[One-sentence investment thesis] Tesla is still a globally influential advanced-manufacturing and energy-platform company, but as of May 2026, the long-dated Autonomy/AI value reflected in the share price has clearly run ahead of verified owner earnings.

[Core bull case] First, the balance sheet is strong: the company is in a net-cash position, with resistance to downturns and to financing contraction markedly stronger than traditional highly leveraged automakers. Second, the second curve is forming: 2025 energy generation and storage revenue grew 27% year over year with a gross margin of 29.8%, clearly better than autos. Third, the infrastructure and installed base are large: by 2026Q1 there were 8,463 Supercharger stations, 79,918 connectors, 9.2 million cumulative deliveries, and 1.28 million active FSD subscriptions. Fourth, if Robotaxi/FSD/Optimus commercialize successfully, the upside option is enormous; Morningstar's and DBS's 400-dollar-class fair value/target reflect this part of the secondary-market consensus.

[Core bear case] First, the current valuation requires the future delivery of a large amount of profit not yet in the financials. Second, 2025 shares rose about 16.6% in a single year, clearly diluting per-share value. Third, the core auto business's margin has fallen sharply from the 2022 highs, and evidence of pricing power is insufficient. Fourth, FSD/Autonomy is still under regulatory scrutiny, and an escalating NHTSA probe constrains the most core long-dated narrative for the valuation. Fifth, the xAI investment and Musk's cross-entity capital allocation raise the governance discount.

[Key assumptions the investment requires to hold] One, FSD/Robotaxi needs to move from "technology demo" toward "regulatable, insurable, replicable unit economics." Two, the energy business must, over the next few years, lift its revenue and profit share to a level that materially changes the company's overall profit pool. Three, share dilution must slow markedly. Four, related-party capital allocation must not keep eroding per-share value. Five, the core auto business's operating margin must at least stabilize at a level sufficient to cover heavy expansionary capex.

[Reasonable buy-price range] Under the strict framework of "valuing only verified owner earnings," the range is about 90–160 dollars; if one accepts a higher proportion of long-dated AI option value and is willing to bear execution and regulatory uncertainty, the watch ceiling can be widened to 220–350 dollars. 404 dollars is not within a range I consider to have a margin of safety.

[Target holding horizon] If future price and facts align, the holding period should be measured in 5–10 years, not a bet on the short-term stock price. The current conclusion is not a guess at next week's price but a judgment on "whether buying today pays too much."

[Expected annualized return] On the basis of the current price, my long-term annualized-return expectations under the conservative / neutral / optimistic scenarios are about -15% / -7% / +2%, respectively. The optimistic case is still not high because the high valuation has already taken away a large amount of future return up front. This estimate excludes dividends, because Tesla pays none.

[Maximum loss risk] The worst case is not bankruptcy but multiple compression: Autonomy commercialization is delayed, the core auto business keeps fighting a price war, energy grows but lacks scale, the share count keeps diluting, and the market reprices Tesla from "AI-narrative asset" to "high-quality automaker." Along that path, the permanent capital loss in the stock could reach 50%–75%.

[Tracking metrics for the next 12–24 months] I recommend focusing on: quarterly deliveries, auto ASP and auto gross margin, energy deployment in GWh, active FSD subscriptions, Supercharger stations/connectors, operating cash flow, total capex, share-count growth, NHTSA/DOJ probe progress, and xAI-related capital moves. The main data sources are, respectively, Tesla's official quarterly updates, the 10-Q/10-K, SEC proxy materials, and Reuters regulatory news.

[Hard signals that trigger a reassessment] If any of the following appears, I would immediately rebuild the model: One, FSD/Robotaxi obtains scalable commercial licenses in major markets and continuously publishes operating data; Two, automotive gross margin restabilizes above 20% without relying on regulatory credits; Three, full-year share dilution falls below 2%; Four, related-party transactions such as xAI expand significantly; Five, NHTSA/DOJ reaches a harsher conclusion on the autonomy system.

[Final recommendation] My recommendation is: stay disciplined, keep watching, and do not prepay for the story. If key future assumptions are falsified—say autonomy commercialization fails to grow the financials' profit for a long time, or the share count keeps diluting rapidly—I will acknowledge that the original judgment "do not pay a high price for long-dated options" was right, and keep avoiding it; if key future assumptions are verified by hard facts—say Robotaxi unit economics and regulatory licenses land clearly—I will acknowledge that I underestimated its commercialization speed and revalue on the new factual basis, rather than clinging to the old model.

Report self-check I did not assign a "buy/sell" rating early in the first half of this report; only here is the final conclusion given. The data entering the valuation model comes essentially from Tesla's 10-Ks, Tesla's official quarterly updates, and the U.S. Treasury; the industry 10-year medians, complete sell-side deep-dive reports, and verbatim call content that could not be rigorously primary-verified were not put into the neutral valuation. The body avoids unanchored adjectives such as "solid, excellent, leading," using margins, share-count growth, cash flow, and regulatory facts instead. The bear case is genuinely damaging, because it focuses on the four points—valuation, dilution, regulation, and related-party capital allocation—that could cause permanent capital loss. If Buffett/Munger reviewed it, I think they would first ask: "Buying now, what verified per-share cash flow earns the money back, rather than the story?"—which is exactly the core question this report answers.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.