Tesla TSLA

Reference table

1. The two segments and where the revenue actually comes from

Tesla reports two operating and reportable segments: automotive (which includes services and other) and energy generation and storage. That is confirmed verbatim in Note 14 of the 10-Q filed 2026-07-23.

The vendor's seg_prod reconciles exactly to reported revenue, which is unusual and worth stating — this is one of the few names where the segment file is clean:

seg_geo is also populated and reconciles: United States $47.627B, China $20.962B, Other countries $26.238B for FY2025, summing to $94.827B. China is 22.1% of revenue and has been flat for three years ($21.745B in 2023, $20.944B in 2024, $20.962B in 2025) while the United States grew.

What the vendor file does NOT give you is the split that matters, and it is the entire analytical question on this name: within "Automotive," how much is vehicle sales, how much is regulatory credits, and how much is leasing. The 10-K and 10-Q disaggregate revenue into five sources — automotive sales, automotive regulatory credits, automotive leasing, energy generation and storage, services and other — but the extracted filing text drops HTML tables. We therefore derive the split from the percentage and dollar movements stated in the Management's Discussion, and label every derived figure as such. The derivation is self-checking: it reconciles to reported totals to within rounding, and to the segment gross margins stated independently in the same document.

Derived June-quarter revenue disaggregation (from 10-Q MD&A deltas; our arithmetic):

Three observations.

First, regulatory credits are collapsing. Down 67% year over year in the quarter, down 49% across the first half, and down 28% across fiscal 2025 (from a derived ~$2.75B to ~$1.98B). The 10-K attributes it to governmental action restricting the programmes; the 10-Q adds that "recent governmental and regulatory actions have restricted certain regulatory credit programs tied to our products." This revenue carries essentially no cost of revenue, so every dollar lost is a dollar of gross profit lost — and its disappearance mechanically depresses the reported automotive margin even when the underlying business improves. That is Section 2.

Second, services and other grew 50% to a derived $4.62 billion, attributed in the 10-Q to "increases in used vehicle sales volume and average selling price, non-warranty maintenance services and collision revenue and paid Supercharging sessions." Note what is not in that list: autonomous ride-hailing. The robotaxi service launched in June 2025 and the 10-Q says it has been expanded, but no revenue from it is separately disclosed anywhere in the filing.

Third, energy is now a real business with a real backlog — $10.05 billion of remaining performance obligations at 2026-06-30, of which $4.56 billion is expected within twelve months, plus 22.3 GWh deployed through the first half of 2026 against 46.7 GWh in all of 2025. It also has a related-party problem: $318 million of Q2'26 energy revenue and $405 million of first-half revenue came from SpaceX buying Megapacks, an entity controlled by the same chief executive. That is roughly 10.5% of the quarter's energy revenue.

2. What the Q2'26 miss actually was — and the automotive margin nobody reports

This is the single most important analytical issue on this name, and the popular version of it is wrong.

The 2026-07-22 print produced actual earnings per share of $0.33 against a $0.50 estimate — a 34% miss. It is being read as a demand failure. It was not.

The entire operating-income decline is operating expense. Gross profit rose $873 million; research and development plus selling, general and administrative rose $1.398 billion, exactly accounting for the $525 million swing. And $472 million of that increase was stock-based compensation: $189 million more in research and development and $283 million more in selling, general and administrative, the latter attributed by the 10-Q to the 2025 CEO Performance Award. That is 34% of the operating-expense increase, and it is non-cash.

Below the operating line, two items produced the profit.

1. A $1.00 billion net gain on the SpaceX equity investment. Tesla invested $2.00 billion in March 2026 (originally a preferred-share investment in a related artificial-intelligence venture, converted to SpaceX common stock), for an ownership interest of less than 1%. It is accounted for under the equity method with the fair value option elected, classified Level 2 with a $238 million discount for lack of marketability, and marked through Other income. In one quarter the stake was written up by half its cost. The counterparty is controlled by Tesla's own chief executive.

2. A $274 million discrete income tax benefit from releasing the valuation allowance on California deferred tax assets following the enactment of California Senate Bill 122. Effective tax rate fell from 23% to 15%.

Stated plainly: those two items total $1.274 billion pre-tax against $1.114 billion of reported net income. Even after tax on the investment gain, they are approximately the whole quarter. Our derivation — labelled as ours — is that ex-mark, ex-tax-benefit, the June quarter was roughly break-even to a small loss. Pre-tax income of $1.329 billion less the $1.00 billion mark is $329 million; the tax provision of $201 million would have been $475 million without the discrete benefit.

Now the part that cuts the other way, and it is the finding of this dive.

The 10-Q states that gross margin for total automotive fell from 17.2% to 16.9%. That looks like continued erosion. It is not. Regulatory credits carry no cost of revenue — they are pure gross profit — so removing them from both numerator and denominator gives the margin on the act of building and selling cars:

Two conclusions. Year over year in the June quarter, automotive gross margin excluding regulatory credits improved by approximately 133 basis points. And across fiscal 2024 to fiscal 2025 it was flat at ~15.4% while the reported figure fell 60 basis points — meaning the entire reported fiscal-2025 automotive margin decline was the credit line, not the car business.

Two independent cross-checks confirm the derivation. The 10-Q separately states that gross margin for the automotive and services and other segment rose from 15.4% to 16.4% in the quarter and from 14.9% to 17.5% across the first half — improvements, not declines. And applying that 16.4% to our derived segment revenue backs out an energy gross margin of 20.4%, exactly matching the 10-Q's independently stated figure.

The offsetting negative is energy. Energy gross margin fell from 30.3% to 20.4% year over year — 990 basis points — because cost of energy revenue rose 29% against revenue up 13%, on "an increase in average cost per MWh primarily driven by sales mix and unfavorable warranty adjustments." Energy gross profit therefore FELL roughly $195 million despite 13% revenue growth. For fiscal 2025 the same margin was 29.8%, up from 26.2%. The 10-Q also warns that "the current tariff regime will have a relatively larger impact on our energy generation and storage business compared to our automotive business."

So the honest summary of the quarter: the car business improved, the energy business gave back a year of margin gains, the highest-margin revenue line shrank 67%, spending on autonomy and robotics grew faster than gross profit, and a related-party mark plus a tax release produced the reported profit. None of that is a demand problem. All of it is a "when does the investment pay" problem.

3. Growth — the reacceleration is real, the operating-margin collapse is also real

*FY2023 net income includes a −$5.001B income tax benefit (a valuation-allowance release). Pre-tax income that year was $9.973B, lower than FY2022's $13.719B.

†The 10-Q reports first-half net income attributable to common stockholders of $1.59B, an increase of $10 million year over year; the vendor's quarterly net-income lines sum to $1.605B before noncontrolling interests. We use the filing's figure.

Read the two columns against each other. Revenue has turned decisively: fiscal 2025 fell 2.9% with automotive sales revenue down 9% on 8% fewer cash deliveries, then the first half of 2026 grew 21.0% and the June quarter grew 25.5% on approximately 25% more cash deliveries. Volumes corroborate: approximately 1.66 million consumer vehicles produced and 1.64 million delivered in fiscal 2025, against approximately 860 thousand produced and 838 thousand delivered through the first half of 2026 — an annualised pace above 1.67 million with the second half historically stronger.

And operating margin has fallen every single year since 2022, from 16.8% to 2.6%. The causes are identifiable and mostly deliberate: research and development rose from $3.075 billion (3.8% of revenue) in fiscal 2022 to $6.411 billion (6.8%) in fiscal 2025 and is annualising near $9.5 billion on the June quarter's $2.371 billion (8.4% of revenue). Stock-based compensation rose from $1.812 billion (2023) to $1.999 billion (2024) to $2.825 billion (2025) and is running at roughly $3.80 billion trailing. Fiscal 2025 also absorbed $390 million of restructuring charges from "convergence of AI chip design efforts."

Consensus estimates — the forward basis for everything in Section 6:

The 2029 row is the bull case written down by the street. Revenue jumping 51.8% in a single year with earnings per share more than doubling is not something an automotive and energy business does; it is the autonomy and robotics ramp being modelled. It rests on 10 analysts. And the 2028 earnings range of $1.40 to $6.39 — a 4.6x spread across 20 analysts — is the honest measure of how little anyone knows.

4. Balance sheet — $34.4B of net cash, and a share count nobody reconciles

From the 10-Q, at 2026-06-30 (the filing wins over the vendor throughout this section):

• Cash and cash equivalents $15.22B · short-term investments $28.31B · total $43.52B (down $535M from year-end 2025)
• Debt: $9.08B aggregate principal, of which $1.35B is current
• Corrected net position: $43.52B − $9.08B = +$34.44B NET CASH
• $5.00B of unused committed credit; in material compliance with all financial debt covenants
• Foreign-currency balances of $3.80B, primarily euros and Chinese yuan
• Digital assets: 11,509 units of Bitcoin at an acquisition cost of $386 million, marked to market through Other income
• SpaceX equity investment: $2.00 billion cost, marked up $1.00 billion in Q2'26, held in Other non-current assets

The vendor's enterprise value understates net cash by $28.56 billion. km_ttm.enterpriseValueTTM reports $1,287.007B against a market capitalisation of $1,292.884B — an implied net cash position of only $5.877B. That is cash and equivalents minus debt, ignoring the entire $28.31 billion of short-term investments. It is the textbook version of the known failure mode. Corrected enterprise value = $1,292.884B − $34.44B = $1,258.44B. On a company this size the proportional error is small (2.3% of enterprise value) but the absolute error, $28.56 billion, is larger than the market capitalisation of Rivian.

The share count is the bigger problem, and it corrupts every per-share multiple in the file.

Market capitalisation is struck on 3,949,547,394 shares — 3,949,547,394 × $327.35 = $1,292,884,339,426, matching the vendor's figure to the dollar. Every earnings-per-share denominator is struck on 3.28-3.55 billion. The gap is roughly 700 million restricted shares that are legally outstanding but excluded from earnings per share until earned:

• ~423.7 million shares of performance-based restricted stock under the 2025 CEO Performance Award, granted 2025-09-03 and approved by shareholders 2025-11-06. Twelve tranches, each requiring a market-capitalisation milestone and an operational milestone. Vested shares are reduced by an offset of $334.09 per share. Unearned shares are forfeited after ten years.
• ~286.5 million restricted shares from the chief executive's June 2026 exercise of approximately 304.0 million options under the reinstated 2018 CEO Performance Award (17.5 million shares net-settled for the exercise price), subject to a service condition and a five-year holding period.
• Offset by 96 million shares forfeited under the 2025 CEO Interim Award in April 2026, consistent with a "no double dip" principle.

That reconciles: 3,752.4M + 304.0M − 17.5M − 96M ≈ 3,943M, against 3,949.5M reported. Share count rose 5.3% in under six months and roughly 23% against the fiscal-2024 weighted basic count of 3,197 million. There has never been a buyback — commonStockRepurchased is $0 in every year of the cash-flow file — and there is no dividend.

The consequence for valuation, stated precisely. The vendor's trailing price-to-earnings of 303.10x divides $327.35 by diluted earnings per share of $1.08. The honest figure divides the $1.293 trillion market capitalisation by trailing net income of $3.818 billion: 338.6x. Likewise the vendor's price-to-book of 12.20x uses a per-share book value struck on ~3.24 billion shares; market capitalisation divided by trailing equity of approximately $87.5 billion is ~14.8x. Every forward multiple in Section 6 is computed on market capitalisation ÷ consensus net income, not on per-share earnings, for exactly this reason.

One further balance-sheet note. Equity was $82.137B at 2025-12-31 (total stockholders' equity, plus $728M minority interest), against total assets of $137.806B and total liabilities of $54.941B. Interest coverage is 13.1x and debt to equity is 0.108. Inventory was $12.392B at year-end 2025 with 59.7 days of inventory outstanding; the June-quarter inventory table is not in our extract, but the 10-Q discloses $100 million of inventory write-downs in the quarter and $177 million in the first half, both lower than the year-ago $131 million and $210 million.

Deferred revenue, which is the closest thing to a customer-validated autonomy number: deferred revenue related to internet connectivity, driver-assistance features and their maintenance, free Supercharging and over-the-air updates was $4.05 billion at 2026-06-30 against $3.87 billion at 2025-12-31 — 4.7% growth in six months. Of the total deferred revenue balance, only $962 million is expected to be recognised in the next twelve months, and $468 million was recognised in the first half (against $428 million a year earlier). Energy deferred revenue was $2.31 billion against $2.04 billion.

5. Cash flow — the year free cash flow goes negative

The trailing free-cash-flow derivation checks against the vendor independently: evToFreeCashFlowTTM of 223.36 on an enterprise value of $1,287.007B implies $5.762 billion, matching our $5.76 billion.

Now the forward problem, which is the clearest single fact in this dive.

• The 10-K filed 2026-01-29 said: "we currently expect our capital expenditures to exceed $20 billion in 2026."
• The 10-Q filed 2026-07-23 says: "we currently expect our capital expenditures to be in excess of $25 billion in 2026 driven by our AI initiatives, including investments in compute infrastructure and data centers."

That is a $5 billion upward revision in six months. First-half capital expenditure was $8.28 billion, which means the second half must exceed $16.7 billion — a doubling. Against trailing operating cash flow of $18.68 billion, fiscal 2026 free cash flow is very likely to be negative, plausibly in the region of −$5 billion to −$6 billion.

The 10-Q says as much between the lines: "As we make critical high-value investments in our AI initiatives and expand manufacturing capabilities, including for our semiconductor, Optimus and solar operations, we intend to manage the business such that we maintain a strong balance sheet and sufficient liquidity, which may include additional funding."

This is not a crisis — $34.44 billion of net cash absorbs several years of that. But it changes the character of the investment. Tesla has been self-funding since 2020. From 2026 it is not, and the shareholder return equation loses its only non-dilutive leg: there is no dividend, no buyback, and the share count is rising. Shareholder yield is negative.

One more cash-flow item. Tesla invested $2.00 billion in SpaceX common stock during the first half — an investing outflow that is not capital expenditure and not an operating asset, and which subsequently produced the $1.00 billion mark that carried the quarter's earnings. It is disclosed under Related Party Transactions.

6. Valuation — priced in or room?

At $327.35 (market capitalisation $1,292.884B on 3,949,547,394 shares; corrected enterprise value $1,258.44B after $34.44B of net cash):

Estimate-basis note. Every forward multiple above uses market capitalisation divided by consensus net income, because per-share earnings are struck on 3.53-3.55 billion shares while market capitalisation is struck on 3.95 billion (Section 4). The italic row shows what screens display; it understates the multiple by roughly 11%. FY2028 consensus net income is computed as EPS $3.123 × the 3,547 million share base implied by the FY2027 row, because the vendor's own FY2028 netIncomeAvg of $12.184B implies 3,901 million shares — inconsistent with every adjacent year (Section 8).

The est EBITDA and EBIT rows are rejected outright. FY2026E ebitAvg is $11.760 billion. Actual first-half 2026 operating income was $1.339 billion. Meeting that estimate would require $10.4 billion of operating income in the second half against a June quarter that produced $398 million. The row is not a forecast anyone holds. All forward valuation here runs on epsAvg and netIncomeAvg, per the data contract.

Peer context, and why no peer multiple is drawn. The vendor peer set contains no financial data beyond market capitalisation, so no peer multiple is computable. It also contains Chijet Motor Company at a market capitalisation of $555,536 — five hundred thousand dollars — alongside Toyota at $224.0 billion, a range of seven orders of magnitude. The set is unusable as a comparison universe without curation. What it does supply is a scale anchor: Toyota $224.0B, General Motors $79.87B, Ferrari $70.86B, Ford $56.78B, Geely $26.33B, Rivian $19.13B, Li Auto $12.99B, NIO $11.82B, ZEEKR $6.85B — a combined $508.6 billion, or 39% of Tesla's market capitalisation.

The vendor's own quality model rates Tesla C+ with an overall score of 2 out of 5, with price-to-earnings and price-to-book each scored 1.

6a. What today's price assumes (the inversion)

At $327.35 — 12.14x trailing sales, 338.6x trailing net income, 159.9x consensus 2027 net income — the price embeds the following falsifiable claims. This is the section the brief demanded: the three businesses priced separately.

• The disclosed automotive business is worth at most a few hundred billion dollars, and the market is not paying for it. (Our derivation, anchored on peer market capitalisations in the vendor file.) Tesla delivered approximately 1.64 million vehicles in 2025 at a derived ~15.4% automotive gross margin excluding credits and a 4.6% consolidated operating margin. Value that entire operation — plus services, plus energy — at Toyota's $224.0 billion market capitalisation, add the corrected $34.44 billion of net cash, and you get $258.4 billion, or ~$65 a share. The remaining $262 of the $327.35 price, 80.0%, is the option. Double the operating-business anchor to $448 billion, twice Toyota, and the option is still 60% of the price.
• Energy storage, the fastest-growing disclosed business, is worth roughly $75-125 billion on any defensible basis — 6-10% of the market capitalisation. (Our derivation.) Fiscal 2025 revenue was $12.771 billion at a 29.8% gross margin, growing 27%, with $10.05 billion of contracted backlog. But June-quarter gross margin fell to 20.4%, so annualising the quarter gives roughly $12.2 billion of revenue at $2.5 billion of gross profit. At a generous 8x sales that is $98 billion. Energy is a good business that cannot carry a trillion-dollar valuation, and the 10-Q warns tariffs hit it harder than automotive.
• Therefore the autonomy and robotics option is being valued at roughly $1.0 trillion — about 80% of enterprise value — with zero separately disclosed revenue. (Our derivation, from the two bullets above against the corrected $1,258.44 billion enterprise value.) The falsifier is specific and public: the revenue disaggregation table in each 10-Q currently has five lines — automotive sales, regulatory credits, automotive leasing, energy, services and other. If a sixth appears, the option starts converting. If it does not appear by 2028, it has not. The closest proxy today is $4.05 billion of deferred revenue on driver-assistance features, growing 4.7% in six months, of which only $962 million is expected to be recognised within twelve months.
• The market assumes no multiple compression despite fiscal 2026 free cash flow going negative. (Consensus-derived: capital expenditure guided "in excess of $25 billion" against trailing operating cash flow of $18.68 billion; first-half capital expenditure $8.28 billion requires >$16.7 billion in the second half.) The company's own filing says liquidity management "may include additional funding."
• The market assumes the earnings path arrives on the street's schedule. (Consensus-derived.) Even taking consensus at face value and applying a rich 30x terminal multiple to fiscal 2030 net income of $31.981 billion gives $959 billion — 26% below today's market capitalisation, four and a half years from now, undiscounted. To earn 10% a year from here through 2030 you would need the market to pay roughly 55x fiscal 2030 consensus net income in 2030, or for 2030 earnings to come in roughly 80% above a consensus built by 10 analysts.
• The most fragile assumption in the price is the third bullet: that a $1.0 trillion option converts into disclosed revenue on a timetable, and that the primary evidence for the timetable — the chief executive's public forecasts, which constitute 22.1% of our entire knowledge-base lane on this name — is reliable.

6b. The return bridge (why the multiple moves)

Expected return decomposes as earnings growth + multiple drift + shareholder yield, and on this name two of the three legs are negative.

Earnings growth is the strong leg. Consensus earnings per share go $1.738 (2026) to $2.280 (2027) to $3.123 (2028) — a 34% annual compound rate off a small base, on 20-25 analysts. We do not dispute it; the revenue reacceleration to +25.5% and the 133-basis-point improvement in automotive margin excluding credits support it.

Multiple drift is a large, structural negative, and our fair value assumes COMPRESSION. The market pays 208.8x fiscal 2026 consensus net income today. Our $250 base assumes it pays 89.1x fiscal 2028 net income and 43.0x fiscal 2029 — still extraordinary multiples in absolute terms, and still crediting the optionality with several hundred billion dollars, but a compression of roughly 60% from where it stands. This is not forward-multiple rolldown; the price would have to fall for those multiples to be reached at those earnings. The driver is not a re-rating of quality — it is that 34% annual earnings growth cannot outrun a 209x starting multiple over any horizon we can underwrite.

Shareholder yield is negative. No dividend (lastDividend 0, no common dividends paid in any year of the cash-flow file), no buyback in any year, and share count up 5.3% in six months to 3,949,547,394 with roughly 710 million further restricted shares already outstanding and awaiting milestones. This is the only leg a shareholder can rely on independent of the narrative, and here it subtracts.

Net: +34% annual earnings growth, roughly −35% annual multiple compression required to reach a defensible valuation, and a negative shareholder yield. The bridge does not close at $327.35. It closes near $250 — and even that credits the option with roughly $700 billion.

6c. Variant perception (where we differ, what would surprise)

• We differ from the bears on the operating business, and we are more constructive than the print suggests. The consensus read of the 2026-07-22 miss is demand deterioration. Revenue beat by 6.9% and grew 25.5%; the miss was entirely operating expense, and $472 million of the $1.398 billion increase was non-cash stock compensation. More importantly, automotive gross margin excluding regulatory credits improved approximately 133 basis points year over year (our derivation, ~14.98% to ~16.31%) and was flat across fiscal 2024-2025 while the reported figure fell — meaning the reported margin decline is a regulatory-credit story, not a pricing story. Watchable number: automotive gross margin excluding credits in the 2026-10-28 print, derivable from the same MD&A disclosures.
• We differ from the bulls on what the price is actually paying for. The knowledge base is emphatic — a manager with a roughly $5 billion position expecting "four to five times their money over the next 10-15 years," another calling Tesla "100% in pole position to manufacture humanoid robots at scale." We do not dispute the assets; we dispute that they are worth $1.0 trillion before a single dollar of them appears on a revenue line. The one customer-validated autonomy number in the filings — $4.05 billion of deferred revenue on driver-assistance features, growing 4.7% in six months, with $962 million expected within twelve months — is not consistent with $1.0 trillion of value being months away. Watchable event: a sixth line in the revenue disaggregation table.
• We differ on earnings quality, and this is where we are most sharply at odds with screens. The June quarter's $1.114 billion of net income is exceeded by the sum of a $1.00 billion unrealised mark on a sub-1% stake in a private, related-party company and a $274 million discrete tax benefit. A Level 2 fair-value mark that wrote a $2.00 billion investment up 50% in one quarter, on an entity controlled by the same chief executive, is not earnings. It will also reverse if the mark reverses. Watchable number: Other income, net, in each subsequent quarter, and the carrying value of the SpaceX stake.
• We differ on the effective valuation because of the share count. Screens show 303x trailing earnings; on market capitalisation over trailing net income the figure is 338.6x, and every forward multiple is understated by roughly 11% for the same reason. That is not a rounding difference — it is roughly 700 million shares.
• Positive surprise that would force us to re-rate: autonomous ride-hailing or humanoid-robot revenue appearing as a separately disclosed line in the revenue disaggregation, at any material size; energy gross margin recovering toward 30% with the $10.05 billion backlog converting; automotive gross margin excluding credits pushing above 18%; or capital expenditure guidance being met while free cash flow stays positive. Any of the first two would move our base materially.
• Negative surprise that would break even our reduced base: a capital raise on unfavourable terms; the SpaceX mark reversing and taking reported earnings negative; the chief executive selling shares to meet the tax obligation the Implementation Agreement leaves explicitly unresolved; a further step down in regulatory credits (already −67% year over year) with no offsetting margin gain; or an adverse driver-assistance liability outcome scaling from the $129 million compensatory and $200 million punitive award already assessed in a single case.

Synthos fair values

All three anchors are expressed as market capitalisation divided by consensus net income, then divided by 3,949,547,394 shares outstanding, with corrected enterprise value and enterprise-value-to-sales as cross-checks. Stated arithmetic, not a discounted cash flow.

• Bear ~$145 — market capitalisation $572.7B, corrected enterprise value $538.2B = 5.19x trailing sales; 51.7x fiscal 2028 consensus net income; 24.9x fiscal 2029. The scenario: autonomy and robotics revenue does not appear in the disaggregation by 2028, energy gross margin stays near 20%, free cash flow is negative through 2027 and capital is raised, regulatory credits go to zero, and the market marks the option down toward the operating business. Note that even here the option retains roughly $280 billion of value above the Toyota-anchored fundamentals. −55.7%.
• Base ~$250 — market capitalisation $987.4B, corrected enterprise value $952.9B = 9.20x trailing sales; 89.1x fiscal 2028 consensus net income; 43.0x fiscal 2029; 30.9x fiscal 2030. The scenario: consensus is roughly right, the revenue reacceleration holds, automotive margin excluding credits keeps grinding up, energy stabilises, autonomy contributes on the street's own schedule rather than management's — and the multiple compresses toward 31x on 2030 earnings, undiscounted and four and a half years out. This base still credits the autonomy and robotics option with roughly $700 billion. −23.6%.
• Bull ~$420 — market capitalisation $1,658.8B, corrected enterprise value $1,624.4B = 15.68x trailing sales; 72.2x fiscal 2029 consensus net income; 51.9x fiscal 2030. The scenario: the 2029 step-change arrives early — a disclosed autonomy or robotics revenue line by 2027-28, energy margin recovering, and the market paying for re-acceleration rather than waiting for it. This is approximately the street's own central case: consensus price target $436, median $435, high $491. +28.3%.

Base is 23.6% BELOW spot. Asymmetry from here is roughly 1:0.5 against — 55.7% of downside to the bear against 28.3% of upside to the bull, with the base itself negative. That, plus a negative shareholder yield, negative 2026 free cash flow, and a knowledge-base lane whose largest single voice is the chief executive, is an Avoid.

7. Knowledge base — 489 claims, and the worst attribution problem in the batch

Tracked knowledge-base claims naming Tesla: 489. Raw hits: 489. Discarded: 17. The search covered TSLA, Tesla and Tesla, Inc. as exact entity matches across all 51,928 records, spanning 2015-03-18 to 2026-07-31.

Before any claim is quoted, the composition must be stated, because it is the finding.

108 of 489 claims — 22.1% of the entire lane — are Tesla's own chief executive. They are 92% bullish, they are forward-looking promises about autonomy timelines, humanoid robots, energy and market capitalisation, and they are the primary evidence base for exactly the $1.0 trillion of enterprise value that no filing supports. A conviction lane in which the single largest voice, by a factor of more than two, is the company's own founder is not independent conviction. It is amplified guidance. Every management claim is half-weighted here and none of them is used as evidence for anything in Sections 6a or 6c.

The 17 quarantined claims are excluded and named. They are channel-attributed rather than speaker-attributed and cannot be trusted to a voice; several are bullish Tesla calls that would otherwise strengthen the lane, which is precisely why the exclusion matters. One further hygiene note: one claim carries the thinker slug "Chamath Palihapitiya" — a raw name rather than a normalised slug, indicating the source registry has not fully normalised this lane.

Management lane — half weight, quoted so readers can judge the register for themselves:

> 2026-02-27 · elon_musk_mgmt · bullish · conviction 80

> "Over 5 to 10 years Tesla has an extremely bright future; hold on to your Tesla stock, it's going to be worth a lot — that's my bet."

> 2026-07-25 · elon_musk_mgmt · bullish · conviction 70

> "Tesla should focus on 5-to-10-year bets like the robotaxi network that depress short-term earnings; long-term retail investors see the long arc and justify keeping voting control."

> 2026-05-18 · elon_musk_mgmt · bullish · conviction 63

> "Tesla already runs vehicles with no people or safety monitors in three Texas cities; driverless will probably be widespread across the US by end of this year."

> 2026-03-12 · elon_musk_mgmt · bullish · conviction 80

> "Optimus 3 is by far the most advanced robot in the world; production starts this summer, reaching high-volume production around summer next year."

These are dated, specific and gradeable, which is to their credit. They are also the company selling its own stock, and they are 22% of the lane.

Independent bullish lane — full weight, and genuinely substantial:

> 2026-07-24 · compound_and_friends · bullish · conviction 80

> "Bought Tesla 2014-16 (~$14 split-adjusted); now ~$5B position; expects to make four to five times their money over the next 10 years."

> 2026-07-24 · compound_and_friends · bullish · conviction 70

> "Tesla is on the cusp of transforming from a hardware company selling a box on four wheels for one-time profit into a software company 'selling the mile,' via vertical integration and low costs."

> 2026-06-29 · compound_and_friends · bullish · conviction 90

> "Tesla is '100%' in pole position to manufacture humanoid robots at scale, working on a 10-million-robot-per-year facility versus peers making thousands to tens of thousands."

> 2026-03-20 · gavin_baker · bullish · conviction 70

> "Tesla is today the most advanced AI in the world — doing the most complex task (driving) at the edge on an inferior inference chip with no cloud, at roughly human-brain power consumption (~20-50W)."

> 2026-02-12 · no_priors · bullish · conviction 70

> "There is no internet of driving data, so a large sensor-rich car park is the moat; Tesla has by far the largest fleet."

> 2026-02-07 · jordi_visser · bullish · conviction 85

> "Tesla will be the best-performing large-cap stock this year over any Mag 7 name — it already built its scaling and is moving to robotics/humanoids/edge AI; valuation goes 'Bitcoinish' on narrative, with path risk."

The bearish lane — 66 independent claims, given full airtime because the brief demanded it and because it is the more instructive half:

> 2025-10-16 · we_study_billionaires · bearish · conviction 58

> "Tesla's $8T market-cap target implies ~$150B operating profit at a 50x PE versus ~$100B revenue today — an extraordinarily tall task for Elon."

> 2025-10-16 · we_study_billionaires · bearish · conviction 55 · tagged FACT

> "Tesla's gross margins are near their lowest in 10 years; competition is playing a role and they are not yet seeing substantial benefits from their massive scale."

> 2025-10-16 · we_study_billionaires · bearish · conviction 55

> "BYD, vertically integrated with far cheaper vehicles (~$10k+ in China vs $50-70k Tesla), is Tesla's biggest global rival; Tesla's US EV share has fallen from over 70% in 2020 to ~40%."

> 2026-07-24 · mike_green · bearish · conviction 50

> "Like Tesla post-S&P inclusion, these names hold very high valuations even as fundamentals deteriorate; Tesla has gone sideways-to-down since."

> 2026-06-18 · odd_lots · bearish · conviction 75

> "The magic seven were seven easy monopolies dominating separate niches; now they're all fighting in one AI market — capital-heavy this time, not capital-light — a bloodcurdling dogfight where only one can be first mover."

> 2026-01-02 · compound_and_friends · bearish · conviction 55

> "Take the under on regular people turning their cars into robotaxis; like Airbnb, few want to make their car a taxi."

> 2025-06-10 · chinatalk · bearish · conviction 60

> "Tesla's heavy dependence on China for production and profit margins is an underappreciated vulnerability."

> 2022-10-31 · marko_papic · bearish · conviction 55

> "Tesla is just a car company with all a car company's competition problems; no reason to be worth more than the next four or five carmakers combined — sign of a top."

> 2021-05-31 · lyn_alden · bearish · conviction 70 · tagged FACT

> "Tesla is heavily reliant on regulatory/carbon tax credits for its recent profitability."

What the lane establishes, and what it does not.

First, the bear case is old and the bull case is new — and that asymmetry is itself a signal to be careful with. Of 343 independent claims, 84 are dated 2026 or later, and only 8 of those 84 are bearish. The independent lane ran 19.2% bearish across its full history but 9.5% bearish in 2026. A lane that stopped arguing with itself while the stock fell 33% is a lane with reduced information content, and it is why net conviction is scored positive-moderate rather than positive-high despite 358 bullish claims.

Second, two bearish claims have already graded correct and are worth reading against our own Section 2. The 2021 claim about reliance on regulatory credits is now precisely the mechanism compressing reported automotive margin — credits down 67% year over year. And the October 2025 observation that gross margins were near ten-year lows is confirmed by the annual series (25.6% in 2022, 18.0% in 2025). When the knowledge base names a mechanism years ahead of the filings, it earns weight. Conversely, the 2025 claim that "Tesla's US EV share has fallen from over 70% in 2020 to ~40%" is not verifiable from any source in this file and is reported as a claim only.

Third, the bull case in the lane is a thesis about identity, not about numbers. "Selling the mile" rather than selling a box on wheels; the largest sensor-rich fleet as a data moat; pole position in humanoid manufacturing. None of those claims is falsified by anything in the 10-Q. None is evidenced by it either. That is the honest state of play, and it is why exponential potential is scored 8 while the verdict is Avoid.

Why Moderate rather than Moderate-High or High. 489 claims across 46 sources over eleven years is the broadest lane in this batch by a wide margin. But 22.1% is the chief executive; 92% of the management lane is bullish; 17 claims are quarantined; the bearish half is concentrated in 2020-2025 while 2026 runs 73% bullish; and not one claim in the entire lane engages with the valuation at the current price. Breadth without recent disagreement is not the same as conviction.

8. Data integrity — what we rejected from the vendor file and why

Fourteen items. Where the filing and the vendor disagree, the filing wins, and it is named.

1. km_ttm.enterpriseValueTTM ($1,287.007B) implies net cash of only $5.877B; the 10-Q shows $34.44B — a $28.56B error. The vendor nets debt against cash and cash equivalents only, ignoring $28.31 billion of short-term investments. The 10-Q states at 2026-06-30: cash and cash equivalents $15.22B, short-term investments $28.31B, total $43.52B; debt principal $9.08B of which $1.35B is current. We use $34.44B of net cash and a corrected enterprise value of $1,258.44B throughout. The same defect appears in bal_a, where FY2025 netDebt of −$8.137B equals total debt $8.376B less cash $16.513B, ignoring $27.546B of short-term investments.

2. Every per-share multiple in the file is struck on a different share count from market capitalisation — the largest analytical distortion in this file. Market capitalisation of $1,292,884,339,426 = 3,949,547,394 shares (10-Q cover, 2026-07-16) × $327.35, confirmed to the dollar. Q2'26 weighted-average diluted shares are 3,540 million; basic approximately 3,276 million. The gap is roughly 700 million restricted shares — ~423.7 million unearned under the 2025 CEO Performance Award and ~286.5 million from the June 2026 exercise of the 2018 award — legally outstanding but excluded from earnings-per-share denominators. The vendor's priceToEarningsDilutedRatioTTM of 303.10x is therefore rejected in favour of market capitalisation ÷ trailing net income = 338.6x, and priceToBookRatioTTM of 12.20x in favour of ~14.8x. All forward multiples in Section 6 use net income, not per-share earnings.

3. est.ebitAvg and est.ebitdaAvg are corrupt — REJECTED entirely. FY2026E ebitAvg is $11.760B against actual first-half 2026 operating income of $1.339B, which would require $10.4B in the second half from a June quarter that produced $398M. FY2026E ebitdaAvg of $17.440B is a 48% step-up on FY2025 actual EBITDA of $11.764B on 11.7% revenue growth, in a year of exploding operating expense. No EBITDA or EBIT multiple appears anywhere in this dive; forward valuation runs on epsAvg and netIncomeAvg per the data contract.

4. est FY2028 row is internally inconsistent on share count. netIncomeAvg $12.184B ÷ epsAvg $3.12316 implies 3,901 million shares, against 3,563 million implied by the FY2026 row, 3,547 million by FY2027, and 3,528 million by FY2029 and FY2030. We use EPS $3.123 × 3,547 million = ~$11.08B for FY2028 and say so. Note also that FY2028 has 32 revenue analysts but only 20 EPS analysts, the widest gap in the set.

5. bal_a debt is not comparable between FY2024 and FY2025 — a presentation change, not a deleveraging. FY2024 totalDebt of $13.623B includes capitalLeaseObligations of $5.745B; FY2025 totalDebt of $8.376B has capitalLeaseObligations of $0. The 10-K reports $8.18 billion of debt principal at 2025-12-31 and, separately, $7.96 billion of total minimum lease payments. The apparent $5.2 billion debt reduction is an artefact. We use the 10-Q's $9.08 billion principal figure at 2026-06-30 for the enterprise-value rebuild and note leases separately.

6. bal_a property, plant and equipment is a line re-mapping — REJECTED as a comparable series. FY2024 $51.507B falls to FY2025 $40.643B, a $10.864 billion decline in a year with $8.53 billion of capital expenditure and $5.03 billion of depreciation (both from the 10-K) and no material disposals in the cash-flow statement. Simultaneously otherNonCurrentAssets rises from $4.215B to $21.204B, +$16.99B. These cannot both be economic. Only totals are used; fixedAssetTurnoverTTM (1.791) and capexToDepreciationTTM (1.995) are rejected.

7. inc_q Q2'26 depreciationAndAmortization is 0, so the quarter's EBITDA of $1.410B is understated by roughly $1.4 billion. The 10-Q reports $1.37 billion of depreciation expense in the June quarter ($2.71 billion in the first half). The Q2'26 EBITDA figure is rejected.

8. The revenue disaggregation the analysis needs is NOT in the vendor file and the extracted filing text drops HTML tables. seg_prod gives Automotive / Energy / Services and other, which reconcile exactly to reported revenue for every year — genuinely clean. But the five-line disaggregation (automotive sales, regulatory credits, automotive leasing, energy, services and other) exists only in filing tables that our extraction does not preserve. All sub-line figures in Sections 1 and 2 are derived from percentage and dollar deltas stated in the Management's Discussion and are labelled as derived. They reconcile to reported totals within rounding and independently reproduce the 20.4% energy gross margin the 10-Q states separately.

9. tech and quote disagree on the 52-week range. quote reports yearHigh $498.83 / yearLow $297.38; tech reports hi52 $489.88 / lo52 $298.32. We use quote for the range (−33.2% from high, +9.7% from low computed on the tech basis, which is materially identical) and tech for RSI, MACD, moving averages and relative returns, and say so. The 50-day moving average also differs ($390.76 in quote, $387.05 in tech); we use tech.

10. ratios_ttm.dividendPayoutRatioTTM (0.0183) contradicts dividendYieldTTM (0) and lastDividend (0) — REJECTED. The cash-flow statement shows $0 of common dividends paid in every year.

11. km_ttm.intangiblesToTotalAssetsTTM is 0 and tangibleBookValuePerShareTTM equals bookValuePerShareTTM exactly ($27.020) — both wrong but immaterial. The FY2025 balance sheet carries $257M of goodwill and $135M of intangibles.

12. The peers set is unusable as a comparison universe and contains no financial data. It ranges from Chijet Motor Company at a $555,536 market capitalisation to Toyota at $224.0 billion — seven orders of magnitude. No peer multiple is drawn anywhere in this dive; peer market capitalisations are used only as scale anchors.

13. insider misrepresents the chief executive's June 2026 transactions if read as open-market activity. The 2026-06-16 rows show an "acquisition" of 303,960,630 shares at $23.34 and a disposition of 17,531,857 at $404.66. These are the exercise and net settlement of the 2018 CEO Performance Award options at their 2018 strike price, not a purchase. Any screen reporting "insider bought 304 million shares" is wrong. Section 10 treats them correctly.

14. earn_cal reports Q2'26 epsActual of $0.33 while inc_q reports GAAP diluted EPS of $0.32. The calendar is on a marginally different basis. Immaterial; both are stated.

Not defects, correctly reported: seg_geo is fully populated and reconciles exactly for every year — a welcome exception; the price-target set shows genuine dispersion (high $491, low $370, median $435, consensus $436.08) and is not a single-source target; the analyst distribution (0 strong buy / 33 buy / 33 hold / 15 sell across 81 analysts) is granular and plausible; stockBasedCompensationToRevenueTTM of 3.665% is consistent with the cash-flow statement.

Two gaps that are not vendor faults and must be stated. Earnings-call transcripts are not available on this plan — management question-and-answer from the 2026-07-22 call is absent, and no quote from it appears here. And the 8-K exhibits are not in our extract: the 2026-07-02 8-K is an Item 2.02 press release, almost certainly the second-quarter production and deliveries report, but only the cover page was captured. The substitute is the 10-Q's own statement that approximately 860 thousand consumer vehicles were produced and 838 thousand delivered through the second quarter of 2026, with 22.3 GWh of energy storage deployed. The 2026-04-22 and 2026-07-22 8-Ks are the Q1 and Q2 earnings releases respectively. The 10-Q contains no Subsequent Events note — Notes 1 through 14 run from accounting policies to segment reporting — so there is no post-quarter disclosure to report.

9. Technicals — washed out, and still in a downtrend

• Price $327.35, up 1.64% on the day. −33.2% from the 52-week high of $498.83, only +9.7% above the 52-week low of $297.38. Maximum drawdown from peak over the trailing period is −33.2% — the stock is essentially at its worst level of the year.
• Below both moving averages, and both are declining: 15.4% below a 50-day of $387.05 and 20.3% below a 200-day of $410.52. The 50-day sits below the 200-day.
• RSI 26.7 — genuinely oversold, one of the lowest readings in this batch. MACD −23.22 — deeply negative, no crossover.
• Relative performance: 3-month −16.6% against SPY +7.6% and QQQ +7.7% (a 24-point deficit); 6-month −22.4% against SPY +11.1% and QQQ +15.7% (a 34-point deficit); 12-month +8.2% against SPY +24.3% and QQQ +30.8% (16 to 23 points behind).
• Volume of 32.3 million against a 47.1 million average — the down move is not being made on capitulation volume.
• On today's move and what it means for entry. Semiconductors moved hard on 2026-08-04 and Tesla participated for 1.64%. That is beta, not a signal. The stock is 15% below a falling 50-day average with an RSI of 26.7 and no scheduled catalyst for 85 days. Buying today's candle is chasing a bounce inside a downtrend; the technical structure offers no support until the $297.38 low, 9.1% below. An oversold RSI in a persistent downtrend is a reason to wait for a base, not a reason to buy.
• Sentiment corroboration is mixed, which is itself informative. The covering street shows 33 buy, 33 hold, 15 sell across 81 analysts with a consensus target of $436 — 33.2% above spot. The street has not capitulated even as the stock lost a third of its value. Neither has the knowledge base (73% bullish across 2026). Maximum pessimism this is not, which is a material argument against treating a 33% drawdown as a value entry.

10. Insiders — one option exercise that is not a purchase, and one small sale

The insider file is dominated by a single event that must not be misread.

The reading, stated carefully. The 303,960,630-share "acquisition" at $23.34 is the exercise of the 2018 CEO Performance Award options at their 2018 strike price, following the Delaware Supreme Court's reversal of the rescission order and the Court of Chancery's final implementing order of 2026-03-18. It is not an open-market purchase and carries no valuation signal. The 17,531,857 shares disposed at $404.66 were net-settled to cover the exercise price. The resulting 710,172,677 shares — 18.0% of shares outstanding — are Restricted Shares subject to a service condition requiring continuous service as chief executive or as an executive officer responsible for product development or operations, plus a five-year holding period after vesting.

What matters more is what the 10-K says about the tax. The Implementation Agreement "provides for cooperation between the Company and our CEO to create a mutually-agreeable plan to address our CEO's satisfaction of applicable tax obligations" — i.e. the tax bill on a 304-million-share exercise has no disclosed funding plan. The 10-K carries a dedicated risk factor: "Mr. Musk may also sell a portion of his shares in order to satisfy tax obligations relating to the 2018 CEO Performance Award. Any such sales could cause the price of our common stock to decline." That is an overhang the company has told us about in advance.

There is no open-market insider buying in this file at all. The only genuine open-market transaction is the chief financial officer selling 2,605.5 shares at $402.197 on 2026-06-08 — trivially small, and above today's price. Against a 33% drawdown, not one insider bought a share on the open market. For a name where the bull case rests on management's forward view, the absence of insider accumulation at $327 is worth noting plainly.

11. Verdict, kill-criteria and flip conditions

Avoid.

This is a de-rating in progress, not an opportunity — and the reason is not the one most people are giving.

The operating business is doing better than the print. Revenue grew 25.5% in the June quarter and beat the estimate by 6.9%. Automotive gross margin excluding regulatory credits improved approximately 133 basis points year over year and was flat across fiscal 2024-2025 while the reported figure fell, meaning the headline margin erosion is a regulatory-credit story that will exhaust itself. Energy has $10.05 billion of contracted backlog. The balance sheet holds $34.44 billion of net cash. The car company is fine.

The price is not about the car company. Corrected enterprise value is $1,258.44 billion against $103.6 billion of trailing revenue and $3.818 billion of trailing net income — 12.1x sales and 338.6x earnings. Value the entire automotive and energy operation at Toyota's $224.0 billion market capitalisation, add the net cash, and the fundamentals-only figure is $65 a share. Everything above that — 80% of the price — is an option on autonomy and humanoid robots that generates no separately disclosed revenue in the 10-K or either 10-Q, whose closest customer-validated proxy is $4.05 billion of driver-assistance deferred revenue growing 4.7% in six months, and whose loudest advocate in our own knowledge base is the chief executive, who accounts for 22.1% of the lane.

Meanwhile the funding profile deteriorated inside six months. Capital expenditure guidance went from "exceed $20 billion" to "in excess of $25 billion" between the January 10-K and the July 10-Q, requiring second-half spending above $16.7 billion against $8.28 billion in the first half. Trailing operating cash flow is $18.68 billion. Fiscal 2026 free cash flow is very likely negative, and the 10-Q says liquidity management "may include additional funding." There is no dividend, no buyback in any year on file, and share count rose 5.3% in under six months to 3,949,547,394 with roughly 700 million further restricted shares outstanding awaiting milestones. Shareholder yield is negative.

Our base fair value of $250 is 23.6% below spot and still credits the option with roughly $700 billion. We could not construct a defensible number above the price without adopting management's timeline as an input, which house rules forbid and the attribution data makes unwise.

Pre-registered KILL criteria — what would confirm the de-rating and take us toward the $145 bear:

• A capital raise on unfavourable terms, or capital expenditure guidance rising again above $25 billion without a corresponding operating cash flow response.
• Two consecutive quarters of negative free cash flow with the net cash balance falling below $25 billion.
• The SpaceX mark reversing. A $1.00 billion write-up on a $2.00 billion Level 2 related-party stake in a single quarter is as reversible as it was creatable, and it exceeded the quarter's entire net income.
• Automotive gross margin excluding regulatory credits falling below 15% on our derivation, which would mean the underlying car business is deteriorating after all.
• Energy gross margin staying below 22% for two more quarters, which would confirm the 30.3%-to-20.4% collapse is structural rather than mix.
• A decisive break below the 52-week low of $297.38 on company-specific news.
• Material share sales by the chief executive to fund the disclosed and unresolved tax obligation on the June 2026 exercise.

Pre-registered FLIP conditions — what would move this to Watch or Buy — Tactical:

• A sixth line in the revenue disaggregation table. Autonomous ride-hailing or robotics revenue disclosed separately, at any material size, in any 10-Q. This is the single cleanest falsifier of our entire thesis and it is publicly checkable every quarter.
• Driver-assistance deferred revenue accelerating materially from the $4.05 billion at 2026-06-30, with the twelve-month recognition figure rising well above $962 million.
• Energy gross margin recovering toward 29-30% with the $10.05 billion backlog converting on schedule.
• A price below approximately $230 (roughly 8x trailing sales on corrected enterprise value, 67x fiscal 2028 consensus net income), at which the option premium falls to a size we could underwrite. Below $180 this becomes actively interesting on the same arithmetic.
• Free cash flow returning positive while capital expenditure guidance is met.
• Any evidence of open-market insider buying, of which there is currently none.

Where TSLA fits in the Synthos Framework Portfolio. No sleeve, no position, 0% weight, at this price. It is not a short — the knowledge-base lane is 73% bullish in 2026, the street holds a $436 consensus target, the balance sheet is unassailable and the option genuinely might convert; borrowing against a narrative this well-funded is a poor risk. It is simply not ownable at 12x sales with negative free cash flow ahead. Logged as a tracked Synthos call (Avoid) as of 2026-08-04 at $327.35, with pre-registered flip levels at $230 and $180 and a pre-registered falsifier — the revenue disaggregation table — that resolves quarterly.

On the exponential score of 8 sitting alongside an Avoid. These are not in tension and the distinction is the whole discipline. Exponential potential measures the slope the business could achieve; the verdict measures whether the price leaves anything for the buyer. Tesla scores 8 because the assets — the largest camera-equipped over-the-air fleet in the world, a running autonomous ride-hailing service, custom inference silicon with booked capacity, a humanoid programme, and a scaling storage business — are exactly the ingredients a step-change requires. The verdict is Avoid because 80% of the enterprise value already assumes the step-change happened.

Single biggest risk: that the autonomy and robotics option, roughly $1.0 trillion of enterprise value, is being priced as a certainty and delivered as a schedule. Nothing in the 10-K or either 10-Q discloses revenue from it. The primary public evidence for its timing is the chief executive's own forecasts, which constitute 22.1% of our knowledge-base lane on this name and 92% of the management lane is bullish. If the option converts, $327.35 will look cheap. If it converts five years late, the intervening period is a car and battery company with a 4.6% operating margin, negative free cash flow and a rising share count, carrying a trillion-dollar premium.

Most fragile assumption in the price: that the market keeps paying above 150x forward net income while free cash flow is negative and the share count is rising. Consensus earnings per share genuinely compound 34% a year through 2028. That is not enough. Applying a generous 30x multiple to fiscal 2030 consensus net income of $31.981 billion — four and a half years out, undiscounted, on estimates from 10 analysts — gives $959 billion, 26% below today's market capitalisation. The price does not merely require the plan to work; it requires the plan to work and the market to still be paying a growth multiple at the end of it.

Provenance & disclosures

• Traceability: 489 tagged knowledge-base claims name Tesla (breadth 41 independent sources; 46 sources including management; net conviction positive-moderate). Composition: 129 management claims (elon_musk_mgmt 108 — 22.1% of the entire lane — plus jensen_huang 18, lisa_su_mgmt, sundar_pichai_mgmt and eli_lilly_ceo 1 each), half-weighted per house rules and used as evidence for nothing in Sections 6a or 6c; 17 quarantine_misattributed claims EXCLUDED and named; 343 independent claims across 41 sources, running 222 bullish / 66 bearish / 55 neutral. Promoter-class weighting applied: anthony_pompliano in his own voice (8 claims) carries low weight; anthony_pompliano_show guests (9 claims) are first-class. Searches were exact entity matches on TSLA, Tesla and Tesla, Inc. across 51,928 records, spanning 2015-03-18 to 2026-07-31, sorted newest-first, with a deliberate full sweep of all 67 bearish claims so the lane is not cherry-picked bullish. All quotes are verbatim from the stored claim text. One hygiene defect noted: one claim carries the raw name "Chamath Palihapitiya" as its thinker slug rather than a normalised identifier.
• Data as-of: fundamentals — income statement and cash flow through the quarter ended 2026-06-30 (Q2'26, reported 2026-07-22, 10-Q filed 2026-07-23, thirteen days before this dive); balance-sheet detail from the 10-Q at 2026-06-30 and the 10-K at 2025-12-31 (the vendor's bal_a is annual only, so no quarterly balance sheet is in the payload) · estimates 2026-08-04 · prices 2026-08-04 ($327.35 close, +1.64%; 50-DMA $387.05; 200-DMA $410.52; RSI 26.7; MACD −23.22) · knowledge-base claims 2026-08-04. Every figure comes from the Synthos vendor data file for TSLA or from the SEC filings at scripts/deepdive/filings/TSLA/; no figure comes from memory, recall or external retrieval.
• Filings used: 10-K filed 2026-01-29 (FY2025); 10-Q filed 2026-04-23 (Q1'26); 10-Q filed 2026-07-23 (Q2'26); 8-Ks filed 2026-04-22 (Q1 earnings release), 2026-07-02 (Item 2.02 press release, almost certainly the Q2 production and deliveries report) and 2026-07-22 (Q2 earnings release). 8-K exhibit content is not in our extract — only cover pages — so the 2026-07-02 delivery figures are unavailable; the 10-Q's "approximately 860 thousand consumer vehicles produced and 838 thousand delivered through the second quarter" is used instead. The 10-Q contains no Subsequent Events note (Notes 1-14). The 10-Q states there have been no material changes to risk factors since the 10-K; the 10-K itself carries a new risk factor addressing the 2025 CEO Performance Award and a standing one warning that the chief executive may be forced to sell shares to satisfy loan or tax obligations.
• Data rejections (detailed in Section 8): vendor enterprise value corrected from $1,287.007B to $1,258.44B (implied net cash of $5.877B corrected to $34.44B using the 10-Q's $15.22B cash + $28.31B short-term investments − $9.08B debt principal); priceToEarningsDilutedRatioTTM 303.10x rejected in favour of 338.6x on market capitalisation ÷ trailing net income, and priceToBookRatioTTM 12.20x in favour of ~14.8x, because market capitalisation is struck on 3,949,547,394 shares while every earnings-per-share denominator is struck on 3.28-3.55 billion; est.ebitAvg and est.ebitdaAvg rejected entirely (FY2026E EBIT of $11.760B against actual first-half operating income of $1.339B); est FY2028 netIncomeAvg rejected as implying 3,901M shares against 3,528-3,563M in every adjacent year; bal_a FY2024-vs-FY2025 debt rejected as non-comparable (FY2024 totalDebt $13.623B includes $5.745B of lease obligations, FY2025 $8.376B includes none); bal_a property, plant and equipment rejected as a re-mapping ($51.507B to $40.643B in a year with $8.53B of capital expenditure and $5.03B of depreciation, while otherNonCurrentAssets rose $16.99B), with fixedAssetTurnoverTTM and capexToDepreciationTTM discarded; inc_q Q2'26 EBITDA rejected because depreciationAndAmortization is 0 against $1.37B reported in the 10-Q; dividendPayoutRatioTTM (0.0183) rejected as contradicting a zero dividend; intangiblesToTotalAssetsTTM (0) and tangibleBookValuePerShareTTM (identical to book) flagged as wrong but immaterial; tech 52-week range ($489.88/$298.32) superseded by quote ($498.83/$297.38); peers rejected as a comparison universe (Chijet Motor at a $555,536 market capitalisation alongside Toyota at $224.0B, and no peer financials supplied); insider June 2026 rows reinterpreted as option exercise and net settlement, not open-market purchases.
• Basis note — the most important one in this dive: the revenue disaggregation into automotive sales, regulatory credits, automotive leasing, energy and services exists only in filing tables that our text extraction does not preserve. All sub-line figures in Sections 1, 2 and 6a are derived from the percentage and dollar deltas stated in the Management's Discussion of the 10-K and 10-Q, and are labelled as derived throughout. The derivation reconciles to reported total revenue within rounding for both Q2'26 and Q2'25, and independently reproduces the 20.4% energy gross margin the 10-Q states separately — two independent checks. Automotive gross margin excluding regulatory credits assumes credits carry no cost of revenue, which is how the company describes and reports them.
• Cash-flow note: trailing operating cash flow of $18.68B and capital expenditure of $12.92B are derived (fiscal 2025 full-year less first-half 2025, plus first-half 2026, all from the filings); the resulting $5.76B of trailing free cash flow independently matches the vendor's evToFreeCashFlowTTM of 223.36x. Fiscal 2026 free cash flow is projected negative on the 10-Q's "in excess of $25 billion" capital expenditure guidance against $8.28 billion spent in the first half. There is no dividend and no buyback in any year of the file; shareholder yield is negative because share count is rising.
• Estimate coverage: 35 revenue and 25 EPS analysts on fiscal 2026 and 2027 — deep. Only 10 EPS analysts on fiscal 2029 and 2030, which is where the entire step-change lives (revenue +51.8% in 2029, EPS more than doubling). The fiscal 2028 EPS range of $1.40 to $6.39 across 20 analysts is a 4.6x spread. No conclusion in this dive rests on the 2029 or 2030 rows other than as a stated upper bound in Section 6a.
• Peer note: the vendor peer set supplies market capitalisations only. Toyota $224.0B, General Motors $79.87B, Ferrari $70.86B, Ford $56.78B, Geely $26.33B, Rivian $19.13B, Li Auto $12.99B, NIO $11.82B and ZEEKR $6.85B total $508.6 billion — 39% of Tesla's market capitalisation. Chijet Motor Company, at a market capitalisation of $555,536, is excluded from that total as unusable. No peer multiple comparison is drawn because no peer financials are supplied.
• Fair-value caveat: the $145 / $250 / $420 anchors are market capitalisation divided by consensus net income, converted at 3,949,547,394 shares outstanding, cross-checked against corrected enterprise value to trailing sales (5.19x / 9.20x / 15.68x against 12.14x today). Stated arithmetic, not a discounted cash flow. The base case still credits the autonomy and robotics option with roughly $700 billion of value; the bear case still credits it with roughly $280 billion. No fair value here assumes the option is worthless.
• Non-equity tripwire: not triggered. TSLA is common equity on NASDAQ Global Select (CUSIP 88160R101, ISIN US88160R1014), beta 1.827, no dividend, a 52-week range of $297.38-$498.83 and 32.3 million shares traded on the day — nothing preferred-like or par-anchored.
• Timing: Q2'26 was reported 2026-07-22, thirteen days before this dive, with the 10-Q filed the following day, so the most recent quarter is fully reflected. The next print is 2026-10-28, 85 days away — no earnings-within-five-days banner applies. Earnings-call transcripts are not available on this plan, so management's question-and-answer from the July call is absent and no quote from it appears here; the 8-K exhibits carry the earnings release, not the call.
• Not investment advice. Independent research, educational and informational only, never personalised. No recommendation to buy, sell or hold any security is made to any person.
• Version: 2026-08-04-full.