Summary

Priced past autonomy certainty: initiate at Sell, $333.38 target (−14.6%), Low conviction

InitiationBy Penmark Team
TSLASell$333.38-15%
Page 1 of 30
Equity Research | Initiation of coverage | North America
Automobiles & Autonomous Systems | Coverage: Initiated | Published 14 November 2026, 07:00 ET
Penmark Team | research@penmarkresearch.com
Tesla, Inc. (Nasdaq: TSLA)

Priced past autonomy certainty: initiate at Sell, $333.38 target (−14.6%), Low conviction

Rating
Initiation · Low conviction
Sell
12-month target
$333.38
Price, 13 Nov 26
$390.22
Price return
−14.6%
Total return
−14.6%
Bull case
$490.18
Bear case
$195.86

We initiate at Sell with a $333.38 target (−14.6% return) and Low conviction, as plausible discount-rate or software inputs flip the call. Tesla sells EVs, grid storage, charging and FSD software while funding pre-revenue robotaxi and humanoid programmes. At $390.22, the share price implies 1.20x unrisked robotaxi success ahead of a multi-year cash trough.

  • Venture spend puts FY27E non-GAAP EPS 28.1% below Street ($1.55 vs $2.16) and turns FCF deeply negative. 3Q26 R&D jumped 69% y/y to $2.76bn; capex reaches $25.2bn in FY27E and $30.0bn in FY28E, holding GAAP EBIT margin below 1% and FCF at −$10.6bn and −$12.8bn.
  • $390.22 demands $303.98/sh from robotaxi—7.93m cars by FY35E at certainty—leaving skewed downside. Weighting four analyst-set outcomes (10% fail, 20% niche, 50% success at 6.6m cars, 20% dominant) values robotaxi at $230.18/sh, $73.80/sh below price-implied value.
  • In our SOTP, core cars ($12.76), energy ($7.31) and FSD ($31.65) total $51.71/sh, or $38.86/sh net of $12.85/sh programme spend. Ex $187m net energy tariff refunds ($526m COGS credit less $339m customer pass-through), 3Q26 energy GM is ~24% vs 33.1% reported, and cell supply caps FY30E deliveries at 2.43m vs ~5m ambition.

What the market is missing. Consensus models FY27E non-GAAP EPS rising 24% to $2.16, whereas our model deducts $7.65bn of venture R&D (incl. $2.65bn programme D&A), holding EPS at $1.55 and cutting net cash from $35.9bn (3Q26) to $8.7bn by FY28E.

Catalysts. Dallas Cybercab rides (Nov 2026), 4Q26 results & FY27 capex guide (late Jan 2027), EU FSD ruling (1Q27) and targeted unsupervised customer FSD in 2Q27 (+$25.29/sh unrisked).

Valuation and target. SOTP DCF (11.57% WACC, 11.62% Ke, 3.0% g) gives $316.42/sh base value (peer P/E check $269.02/sh). Weighting Bear ($175.48, 30%), Base ($316.42, 50%) and Bull ($439.16, 20%) yields $298.69/sh today and rolls at Ke to $333.38 (−14.6%; 0.51x Bull/Bear skew).

Risks. On-time 2Q27 customer autonomy (+$25.29/sh) or −100bp WACC ($399.32 target, Hold) are key upside risks to Sell; Congressional AV oversight (−$25.29/sh) and export pricing (−$6.56/sh) lead downside.

Exhibit 1: Revenue compounds 12.2% to FY28E, but venture spend compresses GAAP EPS to $0.45 and turns FCF negative
Dec YE, US$mFY24AFY25AFY26EFY27EFY28E
Revenue97,69094,827107,389120,660133,788
Growth y/y+0.9%−2.9%+13.2%+12.4%+10.9%
Adjusted EBITDA16,05614,59613,77314,32817,357
Margin16.4%15.4%12.8%11.9%13.0%
Net income (GAAP)7,0913,7942,6391,5471,606
EPS, GAAP ($)2.041.080.740.440.45
Growth y/y−52.6%−47.1%−31.1%−41.3%+3.8%
Free cash flow3,5816,220−1,947−10,582−12,848
At $390.22 (13 Nov)FY24AFY25AFY26EFY27EFY28E
P/E (GAAP)191.3x361.3x524.8x893.7x860.9x
EV / Adj. EBITDA93.8x103.2x109.4x105.1x86.8x
P / B18.7x16.8x15.2x14.3x13.3x
FCF yield0.3%0.5%−0.1%−0.8%−0.9%
Dividend yield0.0%0.0%0.0%0.0%0.0%
ROE (GAAP)10.5%4.9%3.1%1.6%1.6%
Net debt / EBITDA−1.8x−2.4x−2.3x−1.5x−0.5x
Target P/E (GAAP)163.4x308.7x448.3x763.6x735.5x

Source: Company filings, Yahoo Finance, Penmark Research model. Non-GAAP EPS: $1.56 FY26E, $1.55 FY27E, $1.66 FY28E (249.6x/251.4x/235.5x at price; 213.3x/214.7x/201.2x at target).

Base SOTP is $316.42/share; our scenario-weighted $333.38 target (−14.6%) has 0.51x Bull/Bear target skew

Exhibit 3: Weighted value rolls to our $333.38 target (-14.6%); only Bull clears $390.22
$100$200$300$400$500Jan-26Jul-26Jan-27Jul-27ActualEstimatesPrice $390Bull $490 (+26%)Base $353 (−10%)Bear $196 (−50%)Target $333 (−15%)

Source: Yahoo Finance; Penmark Research model

Scenarios and price target basis (12m rolled at 11.62% Ke)
Bull$490.18+25.6%20% prob.

Value today $439.16. Deliveries +16% FY27E; hardware GM 18.5–19.5%; energy GM 29%; FSD 21m by FY40E; robotaxi dominant 35%, Optimus 25%.

Base$353.18−9.5%50% prob.

Value today $316.42. Deliveries +10% FY27E; hardware GM 16.5–17.5%; energy GM 26%; FSD 10m by FY34E; robotaxi success 50% (6.6m FY35E), Optimus 35%.

Bear$195.86−49.8%30% prob.

Value today $175.48. Deliveries +2%/yr; hardware GM 14%; energy GM 21%; FSD 7.5m by FY40E; robotaxi fail 25%, Optimus fail 45% on AV oversight.

Price target ($298.69 today)$333.38−14.6% return = Sell
Bull / Bear vs $390.220.51x+$99.96 / −$194.36
Exhibit 4: Venture R&D holds GAAP EBIT under $2.1bn through FY28E
Income stmt, US$mFY24AFY25AFY26EFY27EFY28E
Revenue97,69094,827107,389120,660133,788
Gross profit17,45017,09420,13722,54826,413
Adjusted EBITDA16,05614,59613,77314,32817,357
GAAP EBIT7,0764,3552,0527741,252
Net interest & other1,8529621,3151,210808
Income tax−1,837−1,423−684−436−453
Non-GAAP net inc.n/dn/d5,5195,5025,872
GAAP net income7,0913,7942,6391,5471,606
Exhibit 5: $75.2bn of FY26E–28E capex drives −$25.4bn cumulative FCF
Cash flow, US$mFY24AFY25AFY26EFY27EFY28E
GAAP net income7,0913,7942,6391,5471,606
D&A5,3686,1486,8608,84010,996
SBC & working cap.2,4644,8058,5684,1844,584
Operating cash flow14,92314,74718,06714,57117,186
Core capex−11,342−8,527−8,312−9,653−10,034
Venture capex00−11,702−15,500−20,000
Free cash flow3,5816,220−1,947−10,582−12,848
Credit facility draw00006,813
Exhibit 6: Cash hits our $25.0bn model floor in FY28E with $6.8bn drawn
Balance sheet, US$mFY24AFY25AFY26EFY27EFY28E
Cash & investments36,56344,05941,61731,03525,000
SpaceX stake002,9002,9002,900
Depreciable assetsn/dn/d64,29280,60599,642
of which programme0015,27628,12843,743
Net assets / capital81,12690,513100,209106,472119,999
Debt & leases8,2138,3769,5239,52316,336
Net cash / (debt)28,35035,68332,09421,5128,664
Common equity72,91382,13790,68696,949103,664
Exhibit 7: Non-GAAP EPS stalls near $1.56 despite volume growth
Per share & KPIsFY24AFY25AFY26EFY27EFY28E
Diluted shares (m)3,4983,5283,5443,5443,544
EPS, non-GAAP ($)n/dn/d1.561.551.66
EPS, GAAP ($)2.041.080.740.440.45
FCF per share ($)1.021.76−0.55−2.99−3.63
Book value / sh ($)20.8423.2825.5927.3629.25
Deliveries (k)1,7891,6361,8051,9852,144
Storage (GWh)31.446.752.067.682.5
FSD paid subs (m)n/d1.101.952.803.70

Source: Company filings, Signals KPIs, Penmark Research model. Diluted shares exclude 423.7m unearned 2025 CEO award shares until tranches vest; n/d = annual non-GAAP / combined depreciable line not disclosed in 10-K (9M26A non-GAAP NI $4,303m, EPS $1.22).

GAAP EBIT margin troughs at 0.6% in FY27E as $25.2bn capex and R&D leave our non-GAAP EPS 28.1% below consensus

Exhibit 8: Capex reaches 22.4% of sales in FY28E as FCF margin hits −9.6%
Growth, margins (%)FY24AFY25AFY26EFY27EFY28E
Growth y/y
Sales+0.9%−2.9%+13.2%+12.4%+10.9%
Adjusted EBITDA−3.5%−9.1%−5.6%+4.0%+21.1%
GAAP EBIT−20.4%−38.5%−52.9%−62.3%+61.7%
GAAP EPS−52.6%−47.1%−31.1%−41.3%+3.8%
Non-GAAP EPSn/dn/dn/d−0.7%+6.7%
Margins & intensity
Gross margin17.9%18.0%18.8%18.7%19.7%
Adj. EBITDA margin16.4%15.4%12.8%11.9%13.0%
GAAP EBIT margin7.2%4.6%1.9%0.6%0.9%
FCF margin3.7%6.6%−1.8%−8.8%−9.6%
R&D / sales4.6%6.8%9.5%10.8%11.9%
Capex / sales11.6%9.0%18.6%20.8%22.4%
Exhibit 9: ROE dips to 1.6% while TSLA trades at 251.4x FY27E non-GAAP EPS
Ratios and valuationFY24AFY25AFY26EFY27EFY28E
Returns and leverage
ROE (GAAP)10.5%4.9%3.1%1.6%1.6%
ROA (GAAP)6.2%2.9%1.8%1.0%0.9%
ROCE (NOPAT/cap.)12.6%7.5%3.0%0.9%1.1%
Net debt / equity−38.9%−43.4%−35.4%−22.2%−8.4%
Net debt / EBITDA−1.8x−2.4x−2.3x−1.5x−0.5x
Valuation at $390.22
P/E (non-GAAP)n/dn/d249.6x251.4x235.5x
P/E (GAAP)191.3x361.3x524.8x893.7x860.9x
EV / Adj. EBITDA93.8x103.2x109.4x105.1x86.8x
P / B18.7x16.8x15.2x14.3x13.3x
FCF yield0.3%0.5%−0.1%−0.8%−0.9%
Target P/E (non-GAAP)n/dn/d213.3x214.7x201.2x

Source: Company filings, Signals ratios, Penmark Research model. Multiples at $390.22 (13 Nov 2026), except final row at our $333.38 12-month target.

Exhibit 10: We sit 10.2% below consensus on FY26E EPS and 28.1% below on FY27E
US$m / US$ per sh9M26A4Q26EFY26EFY27E
Revenue (US$m)
Penmark (9M act.)78,66928,720107,389120,660
Consensusn/a29,081107,750122,300
Prior-year period69,92624,90194,827107,389
Diff. vs consensusn/a−1.2%−0.3%−1.3%
Growth y/y (Penmark)+12.5%+15.3%+13.2%+12.4%
Non-GAAP EPS ($)
Penmark (9M act.)1.220.341.561.55
Consensusn/a0.521.742.16
Prior-year periodn/d0.50n/d1.56
Diff. vs consensusn/a−33.9%−10.2%−28.1%
Growth y/y (Penmark)n/d−31.4%n/d−0.7%

Source: Filings, Yahoo Finance 14 Nov snapshot (4Q26E rev & EPS implied as FY26E less 9M26A; standalone 4Q26E EPS $0.50 in live YF, $0.49 prior; live 1Q27E $26,502m/$0.46), Q3 Update, Penmark model.

Exhibit 11: Revenue grows 12% a year to $133.8bn by FY28E, 1.3% below FY27E consensus
Consensus$0bn$50bn$100bn$150bnFY24AFY25AFY26EFY27EFY28EActualEstimates$97.7bn$94.8bn$107.4bn$120.7bn$133.8bn

Source: Company filings (A); Yahoo Finance consensus; Penmark Research estimates (E)

Why our EPS diverges from revenue. Our FY27E revenue ($120.7bn) sits within 1.3% of Street ($122.3bn), so the −28.1% EPS gap comes entirely from cost: we model $13.08bn R&D (incl. $2.65bn programme D&A) and 26.0% energy margin ex tariff refunds.

At 249.6x FY26E P/E versus the 6.1x automaker median, TSLA leaves no cushion if autonomy slips

Exhibit 12: TSLA trades at 40.7x the listed automaker forward P/E median
At $390.22 (13 Nov)TSLAAuto med.Diff.
Valuation
Forward P/E (CY26E)*249.6x6.1x+3,973%
Trailing P/E (13 Nov)*428.8x23.0x+1,761%
EV / EBITDA (TTM)153.9x12.4x+1,139%
Dividend yield0.0%n/aNo payout
Growth, margins and scale
Revenue growth (TTM)12.5%5.9%+6.6pp
Operating margin (TTM)3.3%2.3%+1.0pp
EBITDA margin (TTM)9.5%10.1%−0.6pp
ROE (FY25A GAAP)4.9%n/dDepressed
Net debt / EBITDA−2.4xn/d−$35.9bn cash
Active FSD subs (3Q26)1.74m0.00m+1.74m

Source: Penmark Research model (GM 6.1x, Ford 6.3x, Stellantis 5.1x CY26E P/E; Toyota 9.1x, Ferrari 37.0x trailing), Signals, Yahoo Finance. *TSLA 249.6x on Penmark FY26E EPS (180.7x YF NTM); 428.8x at $390.22 vs 390.6x 3Q26 end.

Exhibit 13: Four dated catalysts over the next 9 months
  • Nov 26Dallas Cybercab rides: commercial autonomous dispatch test beyond Austin.
  • Jan 274Q26 results & FY27 capex: tests ex-refund energy margin (25% est.) & $25.2bn capex.
  • 1Q 27EU FSD decision: wider UNECE/national clearance (+$0.73/sh unrisked; 50% prob.).
  • 2Q 27Customer unsupervised FSD: management target window (+$25.29/sh unrisked; 40% prob.).

Source: Q3 2026 Update & call, Penmark Research model.

Exhibit 14: Key risks to our Sell rating and target
  • Unsupervised customer FSD on schedule (upside risk to Sell). Paid driverless customer cars in 2Q27 lift success odds. Med (40%); +$25.29/sh.
  • Congressional AV oversight. Federal approval framework slows driverless rollout. Med (30%); −$25.29/sh.
  • Chinese EV export pricing. Aggressive export ASPs cut Tesla hardware ASP by 3%. Med (40%); −$6.56/sh.
  • Post-refund energy margin reset. 3Q26’s 33.1% margin normalises as tariff refunds end. Med (35%); −$1.67/sh.
Monitoring & rating triggers: Hold requires −10% to +10% return: a $351.20–$429.24 target at $390.22 (or a $303.07–$370.42 share price at our $333.38 target), or broad 2Q27 customer FSD; net expected catalyst skew is +$1.44/sh.
Exhibit 15: 3Q26 quarter-end P/E of 390.6x (428.8x at $390.22) sits 1.7 SD above 168.5x mean
0x100x200x300x400x500x4Q213Q222Q231Q244Q243Q252Q26Mean 168.5x+1 SD 301.6x−1 SD 35.4x390.6x, +1.7 SD

Source: Signals quarter-end key metrics (4Q21-3Q26); Yahoo Finance

Exhibit 16: At 153.9x EV/EBITDA on 9.5% margin, TSLA sits far above auto and platform peers
0x50x100x150x200x0%10%20%30%40%EV / EBITDATrailing EBITDA marginTSLAGMFordStellantisToyotaBYDFerrariUberAlphabetR² 0.23, n=8

Source: Yahoo Finance; Signals peer statistics as of 13 Nov 2026

PART 1 · SUMMARY · 3 THE THESIS IN CHARTS

FSD attach drives sentiment, but 1.98x opex puts FY27E EPS 28% below Street while $390 embeds 7.9m robotaxis

Exhibit 17: Active FSD subscriptions rose 83% to 1.74m as TSLA traded around autonomy milestones
Active FSD subscriptionsQuarter-end TSLA price (right)0.0m0.5m1.0m1.5m2.0m$300$350$400$4502Q253Q254Q251Q262Q263Q26

Source: Signals KPI series (2Q25-3Q26); Yahoo Finance

Exhibit 18: Annual opex rises 1.98x from FY25A to $25.2bn by FY28E, squeezing GAAP EBIT margin below 1%
Operating expensesGAAP operating margin (right)$0bn$10bn$20bn$30bn0%2%4%6%8%FY24AFY25AFY26EFY27EFY28EActualEstimates

Source: Company filings (A); Penmark Research estimates (E)

Exhibit 19: Our FY27E non-GAAP EPS of $1.55 sits 28% below the $2.16 Street mean
Street rangeConsensus meanPenmark$1$2$3$4$5FY26EFY27E$1.56, −10%$1.55, −28%

Source: Yahoo Finance consensus (35/33 analysts); Penmark Research estimates

Exhibit 20: The $390.22 share price requires $303.98/sh from robotaxi (1.20x success at certainty)
$0$100$200$300$400Core, FSD &cashOptimus(weighted)Ex-robotaxivalueImpliedrobotaxiShare price+$49.70+$36.54$86.24+$303.98$390.22

Source: Penmark Research model

Exhibit 21: 3Q26 quarter-end P/B of 14.1x (15.5x at $390.22) sits below its 17.5x 5-year mean
0x10x20x30x40x4Q213Q222Q231Q244Q243Q252Q26Mean 17.5x+1 SD 25.3x−1 SD 9.8x14.1x, −0.4 SD

Source: Signals quarter-end key metrics (4Q21-3Q26); Yahoo Finance

Exhibit 22: TSLA trades at 249.6x FY26E P/E despite 12.5% 9M26 growth, detached from peer multiples
0x100x200x300x0%5%10%15%20%25%Forward P/ERevenue growthTSLAGMFordToyotaBYDUberAlphabetR² 0.78, n=6

Source: Penmark Research model; Yahoo Finance as of 13 Nov 2026

PART 1 · SUMMARY · 4 KEY DEBATES

We sit above Morgan Stanley on robotaxi (+$118/sh) but $163/sh lower across FSD, cars and energy

Exhibit 23: Four debates separate our $316.42 base SOTP ($333.38 target) from the $390.22 share price and $416.26 Street mean
DebateMarket / StreetPenmark viewKey evidenceChanges viewResolves
1. Robotaxi scale & economicsPrice embeds $303.98/sh (7.93m cars); MS $112/sh today.$230.18/sh across 4 analyst-set outcomes (6.6m FY35E success).Waymo runs ~4k AVs & 500k rides/wk; Tesla has hundreds & $0.30/mi unreached.>5k paid driverless customer cars in 2027.2Q 2027
2. FSD ARPU & fleet attachMS puts FSD/network at $130/sh today ($145 12m target).$31.65/sh ($110.7bn EV): 10m subs FY34E at $3,323/yr FY36E ARPU.Subs hit 1.74m in 3Q26 (75% NA attach), but EU/China lag & R&D is $3.5bn/yr.EU unsupervised approval & >3.5m subs end-27.1Q–2Q 27
3. Optimus external salesMS credits $35/sh today ($39 12m) on factory rollout.$36.54/sh across 4 outcomes (20% fail; 8m/yr FY40E success).Fremont S/X lines converting; external sales start late 2027, 1m/yr >1yr later.>50k binding external orders at >25% GM.End 2027
4. Core cars & energy marginStreet $2.16 FY27E EPS; MS puts core & energy at $85/sh.$20.07/sh gross, or $7.22/sh net of −$12.85/sh programme spend.3Q26 EBIT ($718m) held $447m net tariff refunds ($187m net energy: ~24% GM vs 33.1%).Auto ex-credit GM >19% & energy >29% clean.Late Jan 27

Source: Company filings & Q3 call, Waymo, Penmark Research model; Morgan Stanley public SOTP discounted 12m at 11.62% Ke (not like-for-like: MS network services incl. insurance, charging & licensing carried in our cars/services).

Exhibit 24: We sit $98/sh below Morgan Stanley's broader network line and $65/sh below on cars and energy
Morgan StanleyPenmark SOTPCars & services$49.28$12.76Energy storage$35.84$7.31FSD / network*$129.91$31.65Robotaxi$111.99$230.18Optimus$34.94$36.54

Source: Penmark Research model; Morgan Stanley public SOTP (*not like-for-like: incl. insurance/charging)

How the debates map to the thesis

Three thesis chapters test every dollar of our gap to the share price. Debate 4 anchors Chapter 10 (Venture spend & cash trough, p. 16 & 20–21), showing why FY27E EPS sits 28.1% below consensus. Debate 1 anchors Chapter 11 (Robotaxi risk-reward, p. 17). Debates 2–3 anchor Chapter 12 (FSD & core hardware/energy, p. 18).

Where consensus and price diverge from our build

Extrapolating 3Q26 margins while pricing unrisked autonomy omits the investment bridge. 3Q26 included $447m net tariff refunds, while scaling autonomy and robotics requires an estimated $62.2bn of FY26E–30E venture capex in our model.

Exhibit 25: All four key debates hit measurable signposts between November 2026 and December 2027
25 Nov 2026Dallas Cybercab launchTests unsupervised riderdispatchLate Jan 20274Q26 results & FY27capexResolves Debate 4: cleanex-refund margins1Q 2027EU FSD decision windowResolves Debate 2a:European attach ramp2Q 2027Customer unsupervisedFSDResolves Debates 1 & 2b:fleet scale & pricingEnd 2027Optimus externalshipmentsResolves Debate 3:commercial humanoiddemand

Source: Q3 2026 call; company calendar; Penmark Research

PART 2 · THE BUSINESS · 5 THE STORY

A $94.8bn car and battery maker priced at $1.54tn for a robotaxi fleet still measured in hundreds

Tesla builds electric vehicles, grid-scale battery packs and fast-charging stations that brought in $94.8bn of revenue in FY25, yet 87% of its $390.22 share price (78% from robotaxi alone) rests on driverless fleets and humanoid robots that earn almost nothing today. Between its factories and that autonomous future lies an estimated $75.2bn three-year capital-spending bill that turns free cash flow negative.

23 yrsFounded Jul 2003
1.64mFY25 car deliveries
$94.8bnFY25A revenue
84,465Superchargers, 3Q26
$1,542bnMarket cap, 13 Nov 26

The scene

On an autumn afternoon in Austin, a Tesla operating without a human at the wheel threads city traffic while engineers watch telemetry from a control room. Miles away at Gigafactory Texas, stamping presses turn out Model Y crossovers and Cybertrucks for retail buyers alongside utility-scale Megapacks bound for AI data centres. Those two scenes capture Tesla in November 2026: an industrial manufacturer using car and battery cash flows to fund a high-wire bet on physical artificial intelligence.

What it actually does

Today Tesla sells four things. Passenger electric cars—led by Model 3 and Model Y—are sold online and through 1,553 store and service sites (end-FY25) without franchised dealers. Stationary battery systems—Powerwall for homes and Megapack for grids—reached 46.7 GWh in FY25 and 36.0 GWh in 9M26. Ecosystem services span Supercharging, parts, repair and insurance. And Full Self-Driving (Supervised) software counted 1.74m active subscribers by 3Q26.

Why it works, and why it might not

Direct retail, seamless fast charging and 75% North American FSD attach in 3Q26 keep drivers loyal. What could break the loop (Debates 1 & 4) is timing: if driverless customer cars slip past 2Q27 while $62.2bn of FY26E–30E venture capex drains net cash from $35.9bn to $8.7bn, shareholders are paying upfront for certainty they do not yet own.

How the money comes in

Buyers pay upfront on delivery for cars and Megapacks, per kilowatt-hour at Superchargers, and monthly or upfront for FSD software. In FY25, automotive generated $69.5bn (73.3% of sales, incl. $2.0bn of zero-cost regulatory credits), energy storage $12.8bn (13.5%) and services $12.5bn (13.2%). Out of each revenue dollar, 82.0 cents built and serviced the product, leaving 18.0 cents of gross profit; R&D (6.8 cents) and SG&A plus restructuring (6.7 cents) left 4.6 cents ($4.36bn) of GAAP operating profit and 6.6 cents ($6.22bn) of free cash flow.

Where the story turns

From 2022 to 2026, price cuts, Chinese EV competition and the July 2025 repeal of US consumer EV tax credits compressed GAAP operating margin from 16.8% in FY22 to 2.6% in 9M26. Rather than harvest cash, Tesla doubled down: 3Q26 R&D jumped 69% y/y to $2.76bn, Fremont Model S/X bays began converting to Optimus robot assembly, and FY26 capex was set near $20bn ahead of a Terafab and AI-compute wave peaking at $30.0bn in FY28E.

78%Robotaxi share of $390.22
At $390.22, $303.98 of every share ($1.06tn) rests on a robotaxi network that management does not expect to post noticeable revenue until around 2Q 2027.Source: Q3 2026 call; Penmark model. Equals 1.20x our unrisked 6.6m-car FY35E success case, driving our $333.38 target (−14.6%).
Exhibit 26: One buyer, start to finish: a $41.7k FY27E vehicle leaves $2.42k core operating profit before venture R&D
1 · DIRECT ORDER
0 dealersOnline order & 1,553 company sites (FY25A); core SG&A is 6.3% of sales.
2 · VEHICLE SALE
$41.7kFY27E hardware ASP ex credits & FSD across 1,985k deliveries (+10.0% y/y).
3 · ECOSYSTEM
75% attach3Q26 North America FSD attach (1.74m subs) & 84,465 chargers.
4 · CORE PROFIT
$2.42k / car$6.88k hardware GP + $0.18k credit − $4.64k core opex, before venture R&D.

Source: FY2025 10-K, Signals KPIs, Penmark Research model (FY27E hardware & allocated core opex).

Why the story matters for the stock: At $390.22, the share price implies 1.20x our robotaxi success path at certainty (Chapter 11, p. 17), while consensus EPS sits 28.1% above our FY27E build through the FY26E–28E capital cycle (Chapter 10, p. 16).
PART 2 · THE BUSINESS · 5 HOW THE COMPANY MAKES MONEY

Hardware and services generated all $94.8bn of FY25A revenue, leaving $4.4bn (4.6%) of GAAP operating profit

Exhibit 27: Direct retail and fleet buyers generated $94.83bn in FY25A across vehicles, storage and ecosystem services
Retail & fleet EV buyers$68.73bnUtilities & C&I grid$12.77bnInstalled fleet (>7m cars)$13.33bnVehicles, FSD & credits$69.53bnMegapack & Powerwall$12.77bnSupercharging & service$12.53bnCost of revenue$77.73bnR&D, SG&A & other$12.74bnGAAP operating profit$4.36bn

Source: FY2025 Form 10-K; Penmark Research model (opex allocated by revenue)

1,636kFY25A deliveries · −8.6% y/y
46.7 GWhFY25A storage · +48.7% y/y
1.74m3Q26 FSD subs · +67.3% y/y
18.1%3Q26 auto GM ex cred. · +180bp q/q
33.1%3Q26 energy GM · ~24% ex $187m net refunds

What it sells, to whom, and why customers buy

Tesla monetises a shared battery, power-electronics and AI software stack across retail drivers, utilities and fleet owners. Retail and commercial EV buyers purchase Model 3/Y, Cybertruck and Model S/X ($69.5bn in FY25A; $57.3bn in 9M26A) for total cost of ownership, integrated NACS Supercharging (84,465 connectors in 3Q26) and over-the-air FSD updates. Utilities, IPPs and data-centre operators buy turnkey Megapack systems ($12.8bn in FY25A; $8.4bn in 9M26A) to buffer renewable intermittency and AI-load spikes. Services & other ($12.5bn in FY25A; $12.9bn in 9M26A) monetises the >7m-vehicle installed base via paid charging, parts, collision repair and insurance.

Pricing, payment model and the revenue equation

Consolidated revenue follows a hybrid unit-times-price plus recurring-software equation. Hardware revenue equals deliveries × ASP ex credits/FSD plus zero-cost regulatory credits ($1,993m FY25A, $639m 9M26A); energy equals GWh deployed × price/GWh ($273m/GWh FY25A, $234m/GWh 9M26A); FSD (carved out from FY27E; ~$800m est. in FY26E) equals average paid subscribers × annual ARPU. Vehicles and Megapacks settle in cash or third-party financing on delivery, whereas FSD bills monthly ($99/mo in North America) alongside upfront purchases.

Cash versus profit: working capital, deferred revenue & capex

Negative net operating working capital cushions cash flow, but venture capex overwhelms operating cash generation in FY26E–28E. At 30 Sep 2026, accounts payable ($17.2bn) and deferred revenue ($4.66bn, incl. unrecognised FSD and service obligations) exceeded receivables ($4.64bn) and inventory ($14.09bn), leaving −$8.63bn of other net operating assets. Even so, capex jumping from $8.53bn in FY25A to $20.01bn–$30.03bn in FY26E–28E turns FCF from +$6.22bn to −$12.85bn.

Concentration, bottlenecks & reporting

Geographic sales are anchored by the US and China, while battery supply is the binding physical constraint. In 3Q26 ($28,046m revenue), the US contributed $12,495m (44.6%), China $5,209m (18.6%) and rest-of-world $10,342m (36.9%). No single buyer exceeds 10% of revenue, but LFP cell imports exposed 2025–26 margins to tariff swings until the Feb 2026 Supreme Court IEEPA ruling triggered $447m net 3Q26 tariff refunds ($526m energy & $260m auto COGS credits less $339m owed to storage customers).

Management states passenger-car backlog is at a four-year high, with cell supply capping output. That bottleneck explains why Megapack growth (+30% in FY27E to 67.6 GWh as Megafactory Texas ramps) competes directly with vehicle packs until in-house cell and Terafab capacity scales later in the decade.

EBITDA definition note: Company Adjusted EBITDA ($14,596m FY25A; $10,841m 9M26A) adds back SBC ($2,825m FY25A; $3,346m 9M26A), whereas standardised Signals EBITDA ex SBC add-back is $9,385m for FY25A. Model EV/Adj. EBITDA multiples use the company definition.
PART 2 · THE BUSINESS · 5 REVENUE ARCHITECTURE, SEGMENTS & GEOGRAPHY

FY27E revenue rises 12.4% to $120.7bn across four engines, led by 30% storage and 140% FSD growth

Exhibit 28: Of $94.8bn FY25A revenue, $17.1bn reached gross profit and $4.4bn (4.6%) became operating income
Automotive$69.53bnEnergy storage$12.77bnServices & other$12.53bnRevenue$94.83bnCost of revenue$77.74bnGross profit$17.09bnR&D$6.41bnSG&A$5.83bnRestructuring$0.49bnOperating profit$4.36bn

Source: FY2025 Form 10-K; Penmark Research model

Exhibit 29: FY27E revenue of $120.7bn (+12.4%) builds from deliveries, carved-out FSD, storage GWh and services
Deliveries1,985k (+10.0%)ASP ex FSD$41.7k (-1.0%)Hardware &credits$83.1bn (+8.3%)Avg paid subs2.38mARPU/yr$809FSD software$1.92bn (+140%)Deployed67.6 GWh (+30%)Price/GWh$234m (0%)Energy storage$15.8bn (+30.0%)Services & other$19.7bn (+12.0%)Total revenueFY27E$120.7bn (+12.4%)

Source: Penmark Research model

Exhibit 30: FSD and Energy lead on FY27E margin and growth while core hardware anchors revenue scale
22%100%0%20%0%150%FY27E gross marginFY27E revenue growthCircle area: FY27E revenue ($bn)High-margin growthScale &ecosystemEnergy storageFSD softwareHardware & creditsServices & other

Source: Penmark Research model

Exhibit 31: The US generated 44.6% ($12.50bn) and China 18.6% ($5.21bn) of Q3 FY26 revenue
$28.05bn3Q26 revUnited States45%$12.50bnChina19%$5.21bnOtherinternational37%$10.34bn

Source: Q3 FY2026 Form 10-Q; Signals geographic KPIs

What the four diagrams prove together: Core hardware supplies 68.9% of FY27E revenue ($83.1bn incl. $350m credits) at 16.5% ex-credit gross margin, while higher-margin Energy (26.0% GM) and FSD (80.9% GM after crash liability, before $3.5bn autonomy R&D) drive long-duration margin expansion. Geographic exposure (18.6% China, 36.9% ROW) keeps non-US regulatory approvals central to Debate 2.
PART 2 · THE BUSINESS · 5 FIXED VS VARIABLE COSTS, CAPEX & UNIT ECONOMICS

Core cars earn $2.42k operating profit per unit, but $62.2bn of FY26E–30E venture capex drives negative FCF

Exhibit 32: Estimated venture capex adds $62.2bn in FY26E-FY30E, pushing total capex to a $30.0bn FY28E peak
Core automotive & energy capexEstimated venture capex (AI compute, fab, Cybercab, Optimus)$0bn$10bn$20bn$30bn$40bnFY25AFY26EFY27EFY28EFY29EFY30EActualEstimates$8.5bn$20.0bn$25.1bn$30.0bn$20.3bn$16.2bn

Source: Company filings (A); Penmark Research estimates (E)

Exhibit 33: Each FY27E vehicle averages $41.7k hardware ASP and leaves $2.4k core operating profit before venture R&D
$0k$10k$20k$30k$40k$50kHardware ASPManufacturing COGSHardware gross profitReg. credit / carCore R&D & SG&ACore EBIT / car$41.71k−$34.83k$6.88k+$0.18k−$4.64k$2.42k

Source: Penmark Research model (FY27E hardware & allocated core opex per delivery)

Fixed versus variable cost structure

Separating variable factory costs from fixed AI/programme spend reveals a profitable core masked by pre-revenue ventures. In FY27E, variable vehicle manufacturing COGS absorbs 83.5% of hardware revenue ($34.83k per car on $41.71k ASP), leaving $6.88k hardware gross profit plus $0.18k of regulatory credits. Allocating core R&D (4.5% of revenue) and SG&A ex CEO award (6.3%) deducts $4.64k per delivery, leaving $2.42k ($5,762m across cars & services) of clean core EBIT.

Consolidated EBIT collapses to $774m (0.6% margin) only because fixed venture R&D ($7,648m) and CEO award SBC ($1,095m) sit on the income statement today. Within FY27E venture R&D, $3,500m is cash autonomy R&D, $1,500m cash programme R&D and $2,648m non-cash depreciation on programme compute and fab tooling.

Why estimated venture capex peaks at $30.0bn in FY28E

Management guided FY26 capex to ~$20bn (10-Q: “in excess of $20bn”) and flagged expanding multi-year spend for AI5/AI6 compute, Cybercab, Optimus and Terafab. We hold core automotive & storage capex at 8.0% of hardware/energy revenue ($8.3bn–$11.2bn in FY26E–30E) and layer an estimated $62.2bn of venture capex ($11.7bn FY26E, $15.5bn FY27E, $20.0bn FY28E, $10.0bn FY29E, $5.0bn FY30E).

Capex runs at 2.85x D&A in FY27E and 2.73x in FY28E, pushing programme assets to $43.7bn by end-FY28E. Under our modelled $25.0bn minimum liquidity rule (~2.2 months of FY28E revenue), cumulative FY26E–28E FCF of −$25.4bn requires drawing $6.81bn on Tesla’s $30.0bn credit facilities in FY28E before repaying in FY29E–30E.

PART 2 · THE BUSINESS · 6 THE MARKET

Global EV sales topped 20m (+20%) and BESS hit 275 GWh (+61%) in 2025, while BYD overtook Tesla in BEV units

Exhibit 34: Tesla integrates cells, vehicles, charging, storage and AI software, while ride-hailing is still nascent
Cells & refiningCATL, Panasonic, LGES,TeslaLFP Nevada & lithiumrefiningVehicles & BESSBYD, Toyota, GM, Tesla,Fluence$82.3bn FY25A auto &energyDirect retail & charging84,465 Superchargerconnectors$12.5bn FY25A servicesAutonomy softwareTesla FSD, Mobileye,Huawei1.74m active subs in 3Q26Autonomous fleetsWaymo (500k rides/wk),BaiduFirst meaningful TSLA revFY27EConsolidated revenue streams today ($94.8bn FY25A)

Source: Company filings; Signals KPIs; IEA; Waymo disclosures; Penmark Research

Exhibit 35: Global EV sales topped 20m in 2025 (+20%) while utility & grid BESS surged 61% to 275 GWh
EV sales, m2025: 20.4m20232025BESS, GWh2026E: 35320242026E

Source: IEA Global EV Outlook 2026; InfoLink Consulting BESS estimates

Exhibit 36: BYD overtook Tesla in annual pure-BEV volume in 2025 (+27.9% to 2.26m vs Tesla -8.6% to 1.64m)
20242025BYD (BEV) 1.76m2.26m BYD (BEV)Tesla 1.79m1.64m Tesla

Source: Company production & delivery filings; BYD 2025 sales disclosures

Bottom-up market size, cycle & demand drivers

Global electric car volume exceeded 20m units in 2025 (+20% y/y, 25% of global light-vehicle sales), but regional growth is diverging sharply. China accounted for over 13m EVs (~55% new-car share) and Europe 4.2m (28% share), whereas North American EV affordability tightened after Congress repealed the US consumer EV tax credit in July 2025 and 10-year Treasury yields held at 5.15%. Against ~90m annual global car sales cited on calls, EV penetration is mid-S-curve.

Stationary battery storage is in structural hypergrowth on AI data-centre and grid demand. InfoLink estimates global BESS deployments rose 61.3% to 275.3 GWh in 2025 and reach 353.4 GWh in 2026E (US 2025 additions: 57.6 GWh per SEIA/Benchmark vs 52.1 GWh InfoLink; IEA tracks 108 GW of power capacity added, which is not interchangeable with GWh energy capacity).

Profit pools, share shifts & regulatory threats

Hardware margins are migrating toward lowest-cost Chinese cell-to-pack integrators, pushing Tesla toward software and grid profit pools. BYD grew pure-BEV passenger sales 27.9% in 2025 to 2,256,714 units, overtaking Tesla’s 1,636,129 deliveries (−8.6% y/y), and grew overseas shipments 67.8% in 1H26. In grid storage, cell price deflation pulled Tesla’s realised price from $321m/GWh (FY24A) to $234m/GWh (9M26A), though turnkey software (Autobidder) preserves ~26% clean gross margins vs loss-making integrators like Fluence (5% GM).

Autonomous ride-hailing remains city-by-city regulated. Waymo leads US commercial operations at ~500,000 paid rides/week across 15 cities with ~4,000 vehicles as of Sep 2026, proving technical viability while underscoring regulatory and fleet-support hurdles ahead of a Democratic Congress in Jan 2027.

PART 2 · THE BUSINESS · 7 COMPETITIVE POSITION

Charging scale (84,465 connectors) and 75% NA FSD attach anchor Tesla’s moat, but BYD leads on cost and Waymo on paid rides

Exhibit 37: Tesla grows faster than legacy US/EU automakers but trails BYD on growth and Toyota on TTM margin
2%20%−80%10%0%20%TTM operating marginRevenue growthProfitable scaleEx-growth / restructuringTSLABYDToyotaGMFordRivian

Source: Penmark Research model; Yahoo Finance as of 13 Nov 2026

Exhibit 38: Penmark assessment: Tesla leads in charging and FSD fleet data; BYD leads EV cost and Waymo paid rides
TSLABYDWaymoGMCharging networkOwner-car autonomyManufacturing costPaid robotaxi opsGrid storage scaleNoneWeakFairStrongLeading

Source: Penmark Research assessment from company filings & KPIs

Exhibit 39: Penmark competitive assessment across four arenas: strong ecosystem lock-in, contested hardware cost and nascent robotaxi
ArenaKey players / cohortCustomer choice criteriaTesla rank & factual evidenceDurability / vulnerability
1. Passenger EVs ($69.5bn FY25A)BYD (2.26m BEVs), Toyota, GM, Ford; comparable to Rivian & Lucid.Price/range, fast-charger access, residual value, ADAS software.#2 global BEV volume (1.64m FY25A); #1 NACS network (84,465 connectors). 18.1% 3Q26 auto GM ex credits beats Western EV cohorts.Chinese export ASPs (−$6.56/sh Bear 2 risk) & aging core lineup ex Model Y refresh.
2. Grid & home BESS ($12.8bn FY25A)CATL, Sungrow, BYD, Fluence Energy (1.42x EV/sales, 5% GM, Jun 26 10-Q).Levelised cost/kWh, bankability, dispatch software, lead time.Top-tier turnkey integrator (46.7 GWh FY25A; 36.0 GWh 9M26A). Megapack nearly sold out for 2027 as Megafactory Texas scales toward 50 GWh.LFP tariff sensitivity; 3Q26 33.1% GM held $187m net refunds (~24% ex refunds, incl. warranty true-downs).
3. Consumer ADAS / FSD (~$0.8bn FY26E)Huawei ADS, XPeng/Xiaomi NOA, Mobileye, GM Super Cruise.Door-to-door urban capability, monthly price ($99), regulatory legality.#1 Western paid consumer base (1.74m subs; 75% NA attach in 3Q26). End-to-end neural net trained on multi-billion fleet miles & ~225 MW AI compute.Supervised-only liability ceiling ($900/yr); EU/China regulatory delays & free Chinese ADAS.
4. Autonomous ride-hailing (pre-rev)Waymo (~4k AVs, 500k rides/wk; $126bn Feb 26 private round), Baidu, Uber.Safety record, regulatory permits, wait time, cost per paid mile.Challenger: hundreds of Austin/Bay Area cars (~2m unsupervised miles by 3Q26). Camera-only architecture targets lower vehicle capex ($30k) if safety holds.$0.30/mi target not yet reached ($0.75/mi FY27E model); federal/state permit risk.

Source: Company filings, Signals profile & KPIs, BYD, Waymo, Fluence FQ3 26 10-Q, Penmark Research model & assessment.

Exhibit 40: North America FSD attach jumped 20pp QoQ to 75% in 3Q26, underpinning 1.95m year-end paid subs
2Q26 NA FSD attach3Q26 NA FSD attach55%75%

Source: Q3 FY2026 call & Signals KPI series

Why ecosystem switching costs last—and their limits

North America FSD attach jumping from 55% in 2Q26 to 75% in 3Q26 proves software pull inside Tesla’s hardware base. Standardisation on Tesla’s NACS plug, direct service integration and recurring software create high owner retention without dealer leakage.

Yet Musk noted in Nov 2025 that no other automaker wants to license FSD, capping software to Tesla’s own fleet. Unlike Waymo’s sensor-redundant city fleets or Uber’s multi-OEM demand aggregator, Tesla must self-fund every Cybercab ($57bn FY35E success-case capex) to capture mobility economics.

PART 2 · THE BUSINESS · 8 MANAGEMENT, GOVERNANCE & CAPITAL ALLOCATION

Musk holds 19.8% beneficial ownership with a 423.7m-share $2tn–$8.5tn award, while $62.2bn est. venture capex needs $6.8bn debt

Exhibit 41: Key-person control, 2025 CEO pay award & dilution mechanics
Governance itemReported terms & Penmark model treatment
CEO & insider stakeElon Musk beneficial ownership: 699.58m shares (19.8% at 30 Sep 2026 proxy); all directors & officers 20.0%. CFO: Vaibhav Taneja.
2025 CEO stock award423.74m restricted shares across 12 tranches (35.31m each) vesting at $2.0tn–$8.5tn market-cap & operating milestones (8-K 7 Nov 2025).
Dilution & P&L impact0 tranches met at $1.54tn market cap; excluded from 3,544.0m diluted shares until met, while GAAP SG&A books $275m in 3Q26 ($1,095m/yr FY27E–34E).
Credit lines (Sep 26)$30.0bn unsecured facilities signed 29 Sep 2026 ($20bn 3Y DDTL, $8bn 5Y rev., $2bn 364d rev.); $0 drawn in 2026, $6.81bn in FY28E.

Source: DEF 14A proxy, Form 8-K (29 Sep 2026; 7 Nov 2025), Q3 10-Q, Penmark model.

Exhibit 42: Insider activity & Musk-ecosystem related-party dealings
Date / entityDisclosed transaction & governance assessment
16 Jun 2026 (CEO)Musk exercised 303.96m options at $23.34; 17.53m shares settled at $404.66 for strike/tax withholding (net +286.43m shares).
8 Sep 2026 (CFO)CFO Taneja sold 2,605.75 shares at $360.13 ($0.94m); no 10b5-1 flag in filing feed, which does not prove the sale was unplanned.
SpaceX ($2.0bn cost)Bought ~$2.00bn SpaceX stock on 12 Mar 2026; $2.90bn fair value at 3Q26 (lock-up ends Dec 2026). FY25 rev $143.3m, exp $11.4m.
xAI & X Corp.xAI revenue $430.1m in FY25 & $78.1m Jan–Feb 26 (Megapack sales), exp $4.0m FY25; X advertising $3.3m in FY25.

Source: Signals Form 4 insider feed, DEF 14A related-party disclosures, Q3 2026 10-Q.

Exhibit 43: Across 4Q26E-FY28E, $62.2bn of est. capex draws cash from $45.4bn to our $25.0bn floor plus $6.8bn debt
Cash, Sep 2026$45.45bnOperating cash flow$34.93bnFacility draw$6.81bnTotal cash pool$87.19bnCore capex$21.98bnVenture capex$40.21bnCash, Dec 2028E$25.00bn

Source: Penmark Research model (4Q26E-FY28E cumulative)

Exhibit 44: Guidance record: numeric FY26 capex and credit-line guides are on track, while multi-year volume timelines have slipped
Management statement / guidePeriodReported / PenmarkAssessment & implication for capital allocation
FY26 capex: >$20bn (Jan), >$25bn (Apr/Jul), ~$20bn (Oct)FY26E$13.01bn 9M26A; $20.01bn FY26EInside revised ~$20bn range: spend shifted into 4Q26E ($7.00bn) & FY27E–28E ($25.2bn / $30.0bn) as AI compute & Terafab ramp.
2.6m annualised production run rate by end-2026 (Nov 25 AGM)3Q26A1.86m annualised (464.4k × 4)Currently below 2.6m end-2026 target: reinforces caution on long-dated volume goals (~5m cars by 2030 vs our cell-constrained 2.43m FY30E).
No draws on $30bn credit facilities in 2026; zero dividends/buybacksFY26E–28E$0 draw FY26E–27E; $6.81bn FY28EConfirmed for FY26E: zero shareholder payouts as 100% of OCF ($49.8bn FY26E–28E) plus cash/debt funds $75.2bn capex.
SpaceX combination press reports (12 Nov 2026; questions declined)12MNot valued ($374.25 Hold check)Governance watchpoint: at market value ($1.855tn SpaceX), a share swap alters paper ownership rather than Tesla operating value.

Source: Q4 2025–Q3 2026 earnings calls, Form 8-K, Penmark Research model.

PART 3 · THE THESIS · 9 WHAT MOVES THE STOCK

Idiosyncratic residual is 51% of weekly variance (after S&P/XLY); TSLA fell on 5 of 8 earnings days

Tesla trades on physical-AI optionality and multi-year investment shocks rather than current vehicle EPS. Across the last eight calls, autonomy and Optimus took 55.0% of analyst Q&A turns while margins and storage took 10.1%; yet on results days, opex and capex step-ups drove sell-offs (−14.5% in 2Q26, −3.1% in 3Q26), leaving a $91.53/sh gap between the $390.22 price and our $298.69 probability-weighted value.

Exhibit 45: Confirmed stock drivers: $390.22 implies $73.80/sh more robotaxi value than our model while pricing lower venture spend
DriverWhat moves the stockTests passedMarket reaction vs Penmark model valueDebateNext number to watch
1. Robotaxi & AV milestonesUnsupervised miles, city launches & federal AV policy shifts.●●●●
All 4 tests
+8.9% 5-day rally (San Antonio launch, Nov 26); −10.8% on 5 Nov (Nasdaq −5.5%). Price implies $303.98/sh vs our $230.18/sh (+$73.80/sh, +32.1%).Debate 1:
Robotaxi scale
Dallas Cybercab & customer AVs (Nov 26–2Q27): 15k avg FY27E fleet (±$25.29/sh).
2. Venture R&D & AI capexQuarterly R&D/SG&A surprises & multi-year fab/AI capex guides.●●●○
Events, Ret., Var.
−14.5% on 2Q26 & −3.1% on 3Q26 as R&D jumped +69% y/y to $2.76bn. Model deducts −$12.85/sh PV and sits −28.1% vs FY27E EPS ($1.55 vs $2.16).Debate 4:
Core & capex
4Q26 capex & FY27 guide (late Jan 27): we $7.00bn / $25.15bn vs ~$20bn FY26 guide.
3. FSD attach, ARPU & EUPaid FSD subscriber growth, attach rates & EU/China approvals.●●●○
Events, Ret., Q&A
Morgan Stanley credits $129.91/sh (not like-for-like: incl. insurance/charging) vs our $31.65/sh pure FSD net of $3.5bn/yr R&D (−$98.26/sh, −75.6%).Debate 2:
FSD software
FY26E paid FSD subs & EU (Jan–1Q27): we 1.95m subs after 75% NA attach.

Source: Filings, Signals transcripts & KPIs, Yahoo Finance, Penmark model. Confirmed drivers pass ≥2 of 4 tests (events, returns, variance, Q&A).

Exhibit 46: TSLA fell on 5 of the last 8 earnings days, led by -14.5% in 2Q26 when capex and opex stepped up
−15%−10%−5%0%+5%+10%4Q241Q252Q253Q254Q251Q262Q263Q26+2.9%+5.4%−8.2%+2.3%−3.5%−3.6%−14.5%−3.1%

Source: Yahoo Finance next-session closes after earnings releases

Exhibit 47: TSLA lagged the S&P 500 by 15.2pp over 1Y and 46.0pp over 5Y as GAAP EPS fell while multiples expanded
1M rel. S&P 5003M rel. S&P 50012M rel. S&P 5003Y rel. S&P 5005Y rel. S&P 500+7.1pp+17.8pp−15.2pp+3.0pp−46.0pp

Source: Yahoo Finance closes to 13 Nov 2026

Exhibit 48: We credit robotaxi $118/sh above Morgan Stanley, but sit $98/sh below its broader network line
Morgan Stanley (price-date eq.)Penmark base SOTPRobotaxi & mobilityFSD & network*Cars & servicesEnergy storageOptimus humanoid$112$230$32$130$13$49$7$36$35$37

Source: Penmark Research model; Morgan Stanley public SOTP (*not like-for-like: incl. insurance/charging)

PART 3 · THE THESIS · 9 WHAT MOVES THE STOCK (CONTINUED)

Autonomy took 44.9% of call Q&A turns (n=89), while delivery beats failed to hold results-day gains

Exhibit 49: Residual variance is 51% of 3Y weekly TSLA returns after S&P 500 and XLY (which holds TSLA)
Each square is 1%Idiosyncratic residual(unexplained by S&P/XLY)51%51S&P 500 market factor25%25XLY sector factor (incl.TSLA weight)24%24

Source: Yahoo Finance 3-year weekly OLS regression (n=156, R2=0.489)

Exhibit 50: Autonomy & robotaxi took 44.9% of Q&A turns across the last 8 calls; margins & energy took 10.1%
Autonomy, FSD & robotaxiGovernance & otherOptimus humanoidVehicle margins & demandCapex & AI computeEnergy & Megapack44.9%32.6%10.1%7.9%2.2%2.2%

Source: Signals transcripts (4Q24-3Q26, n=89 Q&A turns)

DRIVER 1 · CONFIRMED BY 4 OF 4 TESTSRobotaxi milestones drive $303.98/sh of price—$73.80/sh (+32%) above our $230.18/sh base value

What the market does

AV headlines coincide with sharp multi-day swings. Shares fell 10.8% on 5 Nov (vs Nasdaq −5.5% amid AV oversight concerns), then rose 8.9% on 9–13 Nov (San Antonio launch); 44.9% of Q&A (40/89) focused on autonomy with hedged dates (“~2Q27”).

What our model says

Our four-outcome DCF credits robotaxi $230.18/sh ($805bn)—$118.19/sh above Morgan Stanley yet $73.80/sh below price-implied value. Clearing $390.22 needs $1,063bn (1.20x our 6.6m-car FY35E success path at certainty).

Debate and next number

Debate 1 (Robotaxi scale). Watch Dallas Cybercab (Nov–Dec 26) and unsupervised customer cars (~2Q27): we model 30k year-end FY27E fleet ($900m success rev); hitting or missing Q2 swings value ±$25.29/sh.

DRIVER 2 · CONFIRMED BY 3 OF 4 TESTSVenture R&D and AI capex step-ups drive earnings sell-offs and our 28.1% FY27E EPS gap

What the market does

Spend shocks punished 5 of 8 earnings days. TSLA fell 14.5% on 2Q26 and 3.1% on 3Q26 despite $0.48 non-GAAP EPS (aided by $447m net tariff refunds) as R&D surged 69% y/y to $2.76bn and 5Y GAAP EPS fell.

What our model says

Programme spend costs −$12.85/sh (−$44.9bn PV) and pulls FY27E non-GAAP EPS to $1.55 (−28.1% vs $2.16 consensus). Venture R&D reaches $9.89bn and total capex $30.03bn in FY28E, turning FY26E–28E FCF to −$25.4bn.

Debate and next number

Debate 4 (Core & capex). Watch 4Q26 capex and FY27E capex guide (late Jan 27): we forecast $7.00bn 4Q26E ($20.01bn FY26E) and $25.15bn FY27E; each +$2bn/yr R&D step-up trims $0.88/sh.

DRIVER 3 · CONFIRMED BY 3 OF 4 TESTSPaid FSD attach (75% NA) lifts software growth, but shared R&D caps FSD at $31.65/sh vs $129.91/sh broader network mark

What the market does

Market SOTP estimates extrapolate 75% 3Q26 NA attach without deducting shared autonomy R&D. Over 3Y, TSLA rose +74.4% (+3.0pp vs S&P) on multiple expansion to 180.7x consensus forward P/E (249.6x Penmark FY26E) as FSD subs hit 1.74m.

What our model says

FSD is worth $31.65/sh ($110.7bn)—$98.26/sh (−75.6%) below Morgan Stanley’s $129.91/sh network mark (not like-for-like: incl. insurance, charging & licensing). Charging $3.5bn/yr of shared autonomy R&D keeps FSD FCF negative through FY28E (−$814m).

Debate and next number

Debate 2 (FSD software). Watch 4Q26 active subs (Jan 27; Penmark 1.95m) and EU FSD ruling (1Q27; +0.4m subs = +$0.73/sh); MS’s $130/sh FSD mark would flip our target to $443.06 (Buy).

Exhibit 51: Candidates tested and dropped: 3 widely cited metrics failed to pass two empirical tests
Candidate driverTests passedEmpirical & model reason it does not make the confirmed list
Quarterly EV delivery beats●○○○ Return (1/4)3Q26 beat (486.5k cars) still fell −3.1% next day; ±4pt annual delivery growth moves SOTP by just ±$3.6/sh.
Reported energy GM spikes●○○○ Partial (1/4)3Q26 hit 33.1% GM (+12.7pp q/q) on $187m net energy tariff refunds ($526m COGS less $339m rev.), yet stock fell −3.1%.
Isolated 10Y Treasury moves●○○○ Sector (1/4)Rates hit auto affordability (−$1.31/sh) and 11.6% Ke, but add negligible weekly R² beyond S&P 500 + XLY (R²=0.489).

Source: Signals KPIs, Q3 10-Q, Yahoo Finance, Penmark Research model.

Limits note: Event reactions span 4Q24–3Q26 plus Nov 2026 policy/launch sessions; weekly OLS spans 156 weeks to 13 Nov 2026 (note XLY holds TSLA). Results-day moves bundle multiple disclosures and measure association rather than standalone causality.
PART 3 · THE THESIS · 10 THESIS CHAPTER ONE · DRIVER 2 & DEBATE 4
THE CLAIM · RESTS ON SECTION 9 DRIVER 2 (VENTURE R&D & AI CAPEX)Venture R&D ($9.9bn by FY28E) and $75.2bn of FY26E–28E capex drive a 28.1% FY27E EPS gap and −$25.4bn FCF

What the market believes

Consensus models a smooth FY27E rebound ($2.16 non-GAAP EPS, +24.1% y/y), extrapolating 3Q26 margins without full AI capex/R&D P&L drag. Street FY26E EPS ($1.74 snapshot; live $1.72) implies ~$0.52 vs 9M26A $1.22 (live 4Q26E $0.50)—near 3Q26’s refund-boosted $0.48—while stale Jan–Apr 2026 EBITDA snapshots still show $19.91bn for FY27E.

The evidence (three independent sources)

  • 1. Q3 10-Q & 8-K filings: 3Q26 GAAP EBIT ($718m, 2.6% margin) held $447m net IEEPA refunds ($526m energy & $260m auto COGS less $339m owed to storage customers) as R&D rose 69% y/y to $2,761m and SBC hit $1,165m.
  • 2. Capex & credit filings: FY26 capex was guided to ~$20bn ($13.01bn in 9M26A) alongside $30.0bn unsecured credit lines signed 29 Sep 2026 for AI compute and Terafab.
  • 3. Call guidance vs revisions: CFO Taneja guided rising multi-year opex and capex, yet FY27E consensus EPS drifted down just 3% over 90 days ($2.22 to $2.16).
Exhibit 52: R&D nearly quadruples from $4.5bn to $17.6bn while capex peaks at $30.0bn in FY28E
R&D, $bnFY30E: $17.6bnFY24AFY30ECapex, $bnFY30E: $16.2bnFY24AFY30E

Source: Company filings (A); Penmark Research estimates (E)

Exhibit 53: Penmark vs consensus: 28.1% below FY27E non-GAAP EPS as venture R&D & D&A surge
$m, except EPSFY25AFY26EFY27EFY28E
Revenue, Penmark94,827107,389120,660133,788
Revenue, consensus (YF Nov 26)94,827107,750122,300n/a
Revenue gap vs consensus—−0.3%−1.3%—
Total R&D, Penmark (incl. venture)6,41110,17013,07815,905
Adj. EBITDA, Penmark14,59613,77314,32817,357
Adj. EBITDA, Jan–Apr 26 snapshot*14,59615,08419,907n/a
Non-GAAP EPS ($), Penmarkn/d1.561.551.66
Non-GAAP EPS ($), consensusn/d1.742.16n/a
EPS gap vs consensus—−10.2%−28.1%—
Free cash flow, Penmark6,220−1,947−10,582−12,848

Source: Filings, Yahoo Finance (14 Nov 26; live $1.72/$2.16), *MarketScreener (Jan–Apr 26, stale), Penmark model.

What it is worth

Isolating estimated programme spend (−$12.85/sh PV) and valuing cars/services ($12.76/sh) and storage ($7.31/sh) on clean cash flows puts core ex-FSD at $7.22/sh—$65.05/sh below Morgan Stanley’s broader core + energy lines. Across FY26E–28E, $75.2bn of capex ($47.2bn est. venture) lifts D&A from $6.15bn in FY25A ($6.86bn FY26E) to $11.00bn in FY28E, compressing GAAP EBIT margin from 1.9% (FY26E) to 0.6%–0.9% (FY27E–28E) and cutting net cash from $35.9bn to $8.7bn.

Dated signposts

  • Late Jan 2027 (4Q26 results): Confirms if 4Q26 capex hits ~$7.0bn, non-GAAP EPS dips toward $0.34 ex refunds, and FY27E capex is guided ≥$24bn.
  • Apr & Jul 2027 (1H27 10-Qs): Confirms if quarterly R&D tops $3.1bn ($13.1bn FY27E pace) ahead of FY28E facility draws.
Exhibit 54: Venture R&D rises from $0.1bn in FY24A to $9.9bn in FY28E, driving 62% of total R&D
Core R&D (4.5% of rev)Venture R&D (autonomy & programmes)$0bn$5bn$10bn$15bn$20bnFY24AFY25AFY26EFY27EFY28EFY29EFY30EActualEstimates$4.5bn$6.4bn$10.2bn$13.1bn$15.9bn$17.2bn$17.6bn

Source: Company filings (A); Penmark Research estimates (E)

Exhibit 55: Free cash flow turns negative (-$10.6bn FY27E, -$12.8bn FY28E) as GAAP EBIT margin troughs at 0.6%
Free cash flowGAAP operating margin (right)−$15bn−$10bn−$5bn$0bn$5bn$10bn−15%−10%−5%0%5%10%FY24AFY25AFY26EFY27EFY28EFY29EFY30EActualEstimates

Source: Company filings (A); Penmark Research estimates (E)

What would prove us wrong: Holding FY27E capex below $16bn and quarterly R&D near $2.4bn while lifting hardware GM above 19% would restore >$2.10 FY27E non-GAAP EPS and preserve Tesla’s $35.9bn net cash without debt draws.
PART 3 · THE THESIS · 11 THESIS CHAPTER TWO · DRIVER 1 & DEBATE 1
THE CLAIM · RESTS ON SECTION 9 DRIVER 1 (ROBOTAXI & AV MILESTONES)At $390.22, the share price implies $303.98/sh ($1.06tn) for robotaxi—1.20x our 6.6m-car FY35E success path at certainty

What the market believes

Holding non-robotaxi pillars at base ($86.24/sh, or $301.6bn equity at Sep 26), $390.22 embeds $303.98/sh ($1,063.1bn) for robotaxi. At $1.00/mi mature price, $0.28/mi opex and $30k capex/car, the price requires 7.93m robotaxis by FY35E at 100% certainty—8.4x Waymo’s reported $126bn Feb 2026 private round.

The evidence (three independent sources)

  • 1. Q2–Q3 call disclosures: Austin/Bay Area cars remain in the hundreds (~2m unsupervised miles); noticeable revenue starts ~2Q27 and cost/mile exceeds the $0.30/mi target.
  • 2. Industry & policy reality: Waymo runs ~4k AVs across 15 cities (~500k rides/wk) as of Sep 2026, highlighting city permitting ahead of a Democratic Congress in Jan 2027.
  • 3. Fleet capex physics: Our analyst-set 6.6m-car FY35E success path burns $13.47bn FCF in FY27E–30E and needs $57.0bn/yr fleet capex by FY35E (62.9% terminal value share).
Exhibit 56: At $390.22, price implies 7.93m FY35E robotaxis at certainty vs ~4k Waymo AVs (Sep 2026)
Tesla unsupervised (Sep26)Waymo commercial (Sep26)Morgan Stanley 2030EPenmark FY35E successPrice-implied FY35E(100%)0.2k4.0k30.0k6,600.0k7,931.5k

Source: Waymo disclosures (Sep 2026); Q3 call; Penmark Research model (vehicles in thousands)

Exhibit 57: Four-outcome venture matrix: $230.18/sh robotaxi credit trails price by $73.80/sh
Venture outcome (analyst-set)Mult.Prob.Sep 26 ($bn)Base ($/sh)
Robotaxi (success = 6.6m cars, $300bn rev FY35E)
Failure (supervised only)0.00x10%0.00.00
Niche (regional approved roads)0.05x20%44.212.65
Success (unrisked path)1.00x50%884.7252.95
Dominant (2x success scale)2.00x20%1,769.3505.90
Robotaxi wt. / Price-implied0.91x100%805.0230.18 / 303.98
Optimus (success = 8.0m units FY40E, net of $2.3bn dev PV)
Fail (20%) & Niche 0.15x (30%)—50%7.7 wt2.20 wt
Succ. 1x (35%) & Dom. 2x (15%)—50%120.1 wt34.34 wt
Optimus probability-weighted0.69x100%127.836.54

Source: Penmark Research model. 6.6m robotaxi & 8.0m Optimus paths are analyst-set assumptions.

What it is worth

We credit robotaxi at $230.18/sh ($805bn) and Optimus at $36.54/sh ($128bn)—$118.19/sh above Morgan Stanley ($111.99/sh, 30k cars in 2030)—rolling our base case to a $353.18 12m target (−9.5%, Hold) and weighted call to $333.38 (−14.6%, Sell). Using Morgan Stanley’s mobility mark cuts our target to $201.46 (−48.4%, Sell).

Dated signposts

  • Nov–Dec 2026 (Dallas Cybercab): Paid driverless rides in Texas metro #3 after Austin and San Antonio.
  • 2Q 2027 (Customer unsupervised FSD): Retail cars entering network (±$25.29/sh) vs NHTSA/Congressional permit delays (−$25.29/sh).
Exhibit 58: Even in the 6.6m-car success case, robotaxi burns $13.5bn cumulative cash through FY30E before inflecting
Success-case FCFSuccess-case revenue (right)−$25bn$0bn$25bn$50bn$75bn$100bn−$100bn$0bn$100bn$200bn$300bn$400bnFY27EFY29EFY31EFY33EFY35E

Source: Penmark Research model (unrisked success path)

Exhibit 59: At 60% success and 2.5x dominant scale, 12m base target reaches $413.6 (+6.0%, Hold), below $429 Buy
$292.7$308.8$325.0$341.0$357.1$306.9$323.0$339.1$355.2$371.3$321.0$337.1$353.2$369.3$385.4$335.1$351.2$367.3$383.4$399.5$349.2$365.3$381.4$397.5$413.6Robotaxi success probability40%45%50%55%60%Dominant multiple1.50x1.75x2.00x2.25x2.50x

Source: Penmark Research model (base SOTP rolled 12m at 11.62% Ke)

What would prove us wrong: Nationwide unsupervised customer-car approval in mid-2027 with sub-$0.35/mi verified cost and >50k active robotaxis exiting FY27E would shift weight to Dominant (Bull target $490.18, +25.6%).
PART 3 · THE THESIS · 12 THESIS CHAPTER THREE · DRIVER 3 & DEBATES 2 & 4
THE CLAIM · RESTS ON SECTION 9 DRIVER 3 (FSD ATTACH, ARPU & CLEAN ENERGY MARGINS)Shared autonomy R&D ($3.5bn/yr) caps FSD at $31.65/sh and ex-refund margins (26%) keep Energy at $7.31/sh

What the market believes

Third-party SOTP frameworks value FSD/network services up to $129.91/sh ($145 target; broader scope incl. insurance & charging) and Energy up to $35.84/sh ($40 target) without deducting shared AI R&D or 3Q26’s one-off tariff refunds.

The evidence (three independent sources)

  • 1. Signals FSD KPIs & OEM reality: Active subs hit 1.74m in 3Q26 (+67% y/y; 75% NA attach), yet FY26E FSD revenue is ~$800m and Musk confirmed no OEM wants to license FSD ($0 third-party line).
  • 2. 3Q26 Energy margin bridge: Reported 33.1% energy GM held $187m net IEEPA refunds ($526m COGS less $339m owed to storage customers; ~24% ex refunds, still incl. unquantified warranty true-downs) as $/GWh fell 27% from $321m (FY24A) to $234m (9M26A).
  • 3. R&D & peer discipline: Fluence trades at 1.42x EV/sales (5% GM, Jun 2026 10-Q); charging FSD $3.5bn/yr of shared autonomy R&D (−$25.9bn PV) cuts FSD by $7.41/sh and keeps FCF negative to FY28E.
Exhibit 60: FSD FY36E revenue spans $7.0bn if supervised-only to $46.8bn if personal autonomy dominates
Dominant ($4,200/yr)Success ($3,600/yr)Niche ($2,200/yr)Failure ($900/yr)$0bn$10bn$20bn$30bn$40bn$50bnFY27EFY29EFY31EFY33EFY35E$47bn$44bn$25bn$7bn

Source: Penmark Research model (FSD outcomes)

Exhibit 61: FSD outcomes & clean Energy build: $38.95/sh combined vs $165.75/sh broader Street lines
FSD & Energy buildProb/25AFY27EFY30E$/sh
FSD outcomes (10m base subs FY34E; net of $3.5bn/yr R&D)
Failure ($900/yr cap)10%−$2,056m−$907m1.29
Niche ($2,200/yr roads)20%−$1,727m+$1,295m19.66
Success ($3,600/yr AV)50%−$1,602m+$3,558m38.62
Dominant ($4,200/yr)20%−$1,640m+$3,844m41.38
FSD wt. (pre-R&D $39.06)100%$1,921m rev$9,920m rev31.65
Energy storage (normalised ex 3Q26 refunds)
GWh / Price ($m/GWh)46.7/$27367.6/$234107.1/$221—
Energy GM / Pillar DCF29.8%26.0%26.0%7.31

Source: Filings, Signals KPIs, Penmark model (FCF in $m for FSD rows).

What it is worth

On a 12-month target basis, doubling FSD subs (20m) or ARPU lifts our base target to $394.57–$394.69 (+1.1%, Hold), while +4pt Energy GM reaches $355.57 (−8.9%, Hold). Cutting WACC by 100bp lifts our weighted target to $399.32 (+2.3%, Hold); reaching Buy (>$429.24) requires Morgan Stanley’s $129.91/sh network mark ($443.06 target, +13.5%), setting Low conviction.

Dated signposts

  • 1Q 2027 (EU FSD ruling): Tests non-US FSD approval (+0.4m subs = +$0.73/sh) vs regulatory ceiling.
  • 1Q–2Q 2027 (Energy GM ex refunds): Tests whether Megapack GM settles at 24–26% once 3Q26’s $187m net energy refunds and unquantified warranty true-downs roll off.
Exhibit 62: Storage deployments reach 82.5 GWh by FY28E, while gross margin normalises to 26.0% ex refunds
Storage deployedEnergy gross margin (right)0GWh20GWh40GWh60GWh80GWh100GWh26%27%28%29%30%FY24AFY25AFY26EFY27EFY28EActualEstimates

Source: Company filings (A); Penmark Research estimates (E)

Exhibit 63: FSD 2x or -1pt WACC lifts 12m target to $395-$399 (Hold); MS FSD ($443) clears the $429 Buy line
Base target (as modelled)Energy GM +4ptStorage growth +5pt/yrHardware GM +2pt/yrTerminal growth 4.0%FSD subs doubledFSD ARPU doubledWACC -1.0pt (wt. target)MS FSD mark (wt. target)$353$356$357$358$383$395$395$399$443

Source: Penmark Research model (12m target $/sh; MS public SOTP)

What would prove us wrong: Multi-OEM third-party FSD licensing deals or >25m paid autonomous cars clearing >$4,500/yr ARPU (well beyond our 10m-sub, $3,600/yr success case) without extra R&D would lift FSD toward $130/sh and flip our Low-conviction Sell to Buy.
PART 4 · THE NUMBERS · 13 TRACK RECORD AND QUALITY OF REPORTED NUMBERS

GAAP operating margin fell from 16.8% (FY22A) to 2.6% (9M26A), while 3Q26 non-GAAP NI was 2.16x GAAP

Exhibit 64: Five-year track record (FY21A–FY25A & 9M26A): storage offset vehicle contraction, but opex halved GAAP profit
$m, except EPS, k, GWhFY21AFY22AFY23AFY24AFY25A9M26A
Total revenues53,82381,46296,77397,69094,82778,669
Revenue growth y/y+70.7%+51.4%+18.8%+0.9%−2.9%+12.5%
Deliveries (k) / BESS (GWh)936 / 4.01,314 / 6.51,809 / 14.71,789 / 31.41,636 / 46.71,325 / 36.0
Gross / GAAP EBIT margin25.3% / 12.1%25.6% / 16.8%18.2% / 9.2%17.9% / 7.2%18.0% / 4.6%19.0% / 2.6%
GAAP NI / Diluted EPS ($)5,519 / 1.6312,556 / 3.6214,997 / 4.307,091 / 2.043,794 / 1.082,378 / 0.67
OCF / Free cash flow11,497 / n/d14,724 / n/d13,256 / 4,35714,923 / 3,58114,747 / 6,22014,897 / 1,883
FCF/NI conv. / ROE (avg eq.)n/d / 21.1%n/d / 33.5%29.1% / 27.9%50.5% / 10.5%163.9% / 4.9%79.2% / 3.7%

Source: FY2021–FY2025 Form 10-K, Q3 2026 Form 10-Q, Penmark Research model (9M26A ROE annualised). FY23A GAAP NI included a $5.9bn tax valuation-allowance release.

Exhibit 65: Services (+41%) and Auto (+12%) offset Energy (-4.6%) in FY26E before all three grow in FY27E-28E
AutomotiveEnergy storageServices & otherTotal growth−10pp−5pp0pp5pp10pp15ppFY25AFY26EFY27EFY28EActualEstimates−2.9%+13.2%+12.4%+10.9%

Source: Company filings (A); Penmark Research estimates (E)

Exhibit 66: 3Q26 non-GAAP net income ($1,697m) was 2.16x GAAP ($787m) as SBC reached $1,165m
GAAP net incomeNon-GAAP net income$0m$500m$1,000m$1,500m$2,000m2Q253Q254Q251Q262Q263Q26

Source: Company quarterly updates & Form 10-Q (2Q25-3Q26)

Exhibit 67: Earnings-quality audit: 2 flags, 1 watch, 1 clean
ItemReported accounting & cash impactStatus
SBC & 2025 CEO award9M26A SBC hit $3,346m (30.9% of $10,841m Adj. EBITDA), incl. $275m 3Q26 CEO award. Model treats 3% employee SBC as cash cost.Flag
3Q26 tariff & warranty$447m net IEEPA refunds ($526m energy & $260m auto COGS less $339m energy rev. = $187m net energy) plus unquantified warranty true-downs lifted 3Q26 EBIT ($718m) & energy GM (33.1% vs ~24% ex refunds).Flag
Zero-cost reg. credits100%-margin credits fell from $2,763m (FY24A) to $1,993m (FY25A) and $639m (9M26A); modelled at $350m FY27E & $150m FY29E+.Watch
Deferred rev. & leases$4,660m 3Q26 deferred rev. is cash-collected upfront; operating lease cost sits in opex, avoiding EV double count.Clean

Source: Q3 2026 10-Q, Signals, Penmark model.

Our read on reported profit quality

Reported non-GAAP EPS ($1.22 in 9M26A; $0.48 in 3Q26) overstates recurring economic profit because it adds back $3.35bn of ongoing SBC while retaining $447m of net 3Q26 tariff refunds. In our DCF, employee SBC (3.0% of revenue) is expensed as a real economic cost and the 2025 CEO award ($1,095m/yr) is added back to NOPAT only because tranche dilution is tested directly in share count.

PART 4 · THE NUMBERS · 14 FORECASTS

Revenue compounds at 11.1% (FY25A–30E) to $160.7bn, but venture spend puts FY27E EPS 28.1% below consensus

Exhibit 68: Reused revenue equation (FY27E) & multi-year operating driver path (FY24A–FY30E)
Deliveries1,985k (+10.0%)ASP ex FSD$41.7k (-1.0%)Hardware &credits$83.1bn (+8.3%)Avg paid subs2.38mARPU/yr$809FSD software$1.92bn (+140%)Deployed67.6 GWh (+30%)Price/GWh$234m (0%)Energy storage$15.8bn (+30.0%)Services & other$19.7bn (+12.0%)Total revenueFY27E$120.7bn (+12.4%)
Operating & cost driversFY24AFY25AFY26EFY27EFY28EFY29EFY30E
Vehicle deliveries (k)1,789.21,636.11,804.71,985.12,144.02,294.02,431.7
Delivery growth y/y−1.1%−8.6%+10.3%+10.0%+8.0%+7.0%+6.0%
Hardware ASP ex FSD ($k)41.5*41.3*42.141.741.741.741.7
Paid FSD subs, end (m)n/d1.101.952.803.704.705.80
Wt. FSD software rev. ($m)n/dn/d8001,9213,5026,0529,920
Storage deployed (GWh)31.446.752.067.682.595.7107.1
Energy rev. ($m) / GM (%)10,086/26%12,771/30%12,188/29%15,844/26%18,944/26%21,535/26%23,637/26%
Services & other rev. ($m)10,53412,53017,63119,74721,72123,67625,570
Core / Venture R&D ($bn)4.4 / 0.14.3 / 2.14.8 / 5.35.4 / 7.66.0 / 9.96.6 / 10.67.2 / 10.3
Core / Venture capex ($bn)11.3 tot8.5 tot8.3 / 11.79.7 / 15.510.0 / 20.010.3 / 10.011.2 / 5.0

Source: Filings (A), Signals KPIs, Penmark estimates (E; *FY24A–25A ASP incl. FSD; 9M26A: 1,325k cars, 36.0 GWh, 1.74m subs).

Exhibit 69: Penmark vs consensus & guidance line by line: the FY27E gap sits in opex, D&A and clean margins, not revenue
Metric ($m / $)PeriodPenmarkCons.GuideGapAnalytical reason for the gap
Revenue26E/27E107.4/120.7b107.8/122.3bNone−0.3/−1.3%+10% FY27E deliveries (cell-capped) & +30% storage match Street top line (live YF $107.9b/$122.4b).
Gross margin26E/27E18.8%/18.7%~19.5% impBESS up−0.8ppStrip $447m net 3Q26 tariff refunds ($187m net energy; 26% FY27E GM) & model 16.5% hardware GM on Chinese exports.
Adj. EBITDA26E/27E13,773/14,32815,084/19,907*Opex up−8.7/−28.0%*MarketScreener snapshot (Jan–Apr 26) predates 2Q/3Q R&D surge (+69% y/y); $5.0bn cash venture R&D.
Non-GAAP EPS26E/27E1.56 / 1.551.74 / 2.16None−10.2/−28.1%4Q26E drops to $0.34 ex 3Q refunds; FY27E D&A rises +$1.98bn to $8.84bn on $25.2bn capex (live YF $1.72/$2.16).
Capex26E/27E20,014/25,153~20,000/n/d~$20bnIn guideMatches ~$20bn FY26 call guide (>$20bn 10-Q), then steps up for AI5/AI6, Cybercab & Terafab.
Free cash flow26E/27E−1,947/−10,582Pos. impNo 26 drawDeficit$14.57bn FY27E OCF falls $10.58bn short of $25.15bn capex (−$14.20bn FCF after employee SBC).

Source: Yahoo Finance (14 Nov 26; live), *MarketScreener (Jan–Apr 26, stale), Q3 10-Q & call, Penmark model.

PART 4 · THE NUMBERS · 14 FORECASTS (CONTINUED)

Core margin sensitivity is muted (−$5.6 to +$9.8/sh), while negative FY26E–28E FCF draws $6.81bn of credit in FY28E

Exhibit 70: A +/-4pt shift in annual delivery growth and +/-2pt in hardware margin moves SOTP by just $15.4/sh
$310.8$312.5$314.1$315.8$317.4$311.3$313.2$315.2$317.1$319.0$311.9$314.2$316.4$318.7$320.9$312.7$315.4$318.0$320.7$323.3$313.7$316.9$320.0$323.1$326.3Hardware gross margin vs base-2pt-1ptBase+1pt+2ptDeliveries growth vs base-4pt-2ptBase+2pt+4pt

Source: Penmark Research model

Exhibit 71: Deliveries compound to 2.43m by FY30E (+8.2% CAGR from FY25A), below the 5m management ambition
0.0m0.5m1.0m1.5m2.0m2.5mFY24AFY25AFY26EFY27EFY28EFY29EFY30EActualEstimates1.79m1.64m1.80m1.99m2.14m2.29m2.43m

Source: Company filings (A); Penmark Research estimates (E)

$774m (0.6%)FY27E GAAP EBIT (trough)
$8,064m (5.0%)FY30E GAAP EBIT (inflection)
−$25,376mCum. FY26E–28E FCF
$6,813mFY28E credit-line draw

Reasoning for each revenue & margin path

Vehicle deliveries (+10.3% FY26E, +10.0% FY27E, tapering to +6.0% in FY30E = 2.43m units): Passenger-car backlog sits at a four-year high, but cell supply caps output and installed car capacity is ~2.35m units. Unlike management’s ~5m 2030 ambition, we cap core deliveries near existing factory expansion and hold hardware ASP flat at $41.7k as affordable trims offset mix.

Energy storage (+30.0% FY27E to 67.6 GWh; 107.1 GWh FY30E at 26.0% GM): Megafactory Texas (50 GWh planned) and near-sold-out 2027 Megapack demand drive volume, while LFP competition nudges $/GWh down 2%/yr from FY28E and strips 3Q26’s refund spike.

FSD software (1.95m subs FY26E to 5.80m FY30E; $809 to $1,890 wt. ARPU): Probability-weighted across our four autonomy outcomes, FSD revenue scales from ~$800m (FY26E) to $1,921m (FY27E) and $9,920m (FY30E) at ~81% gross margin after crash-liability accruals.

Fixed vs variable costs, working capital & funding

Core vs venture cost split: Core R&D (4.5% of sales) and SG&A ex CEO award (6.3% in FY27E, scaling to 5.7% in FY30E) grow with revenue, whereas fixed cash venture R&D ($5.0bn–$5.5bn/yr) plus programme D&A ($2.65bn FY27E to $5.82bn FY30E) depresses near-term EBIT before tapering as ventures commercialise.

Cash, working capital & balance-sheet bridge: Working capital investment absorbs a modest ~$530m/yr ($531m FY27E, $525m FY28E) thanks to direct-sale payables and FSD deferred revenue. However, total capex of $20.01bn (FY26E), $25.15bn (FY27E) and $30.03bn (FY28E) pulls cash from $45.45bn (3Q26A) to our $25.00bn liquidity floor in FY28E, triggering a $6.81bn draw on the $30.0bn Sep 2026 facilities (pushing debt from $9.52bn to $16.34bn and net cash down to $8.66bn) before positive FY29E–30E FCF ($0.53bn / $8.88bn) repays the draw in full.

Why sensitivity is asymmetric (Section 9 confirmation): Across the entire ±4pt delivery growth and ±2pt hardware gross-margin grid in Exhibit 70, base SOTP moves by just $15.4/sh ($310.8 to $326.3/sh), whereas FY27E non-GAAP EPS swings ±$0.38/sh (±24%) on ±2pt hardware GM—explaining why equity holders focus on autonomy probabilities and WACC rather than quarterly car beats.
PART 4 · THE NUMBERS · 15 VALUATION

Primary SOTP values TSLA at $316.42/sh (base) and peer P/E cross-check at $269.02/sh (−15.0%)

Exhibit 72: Robotaxi ($230.18/sh; 78% of price implied) and Optimus ($36.54/sh) drive 84% of base SOTP (87% of price)
$0$100$200$300$400Auto & energyFSD softwareProgrammespendCore EV / shRobotaxiOptimusNet cash &SpaceXBase SOTP+$20.07+$31.65−$12.85$38.87+$230.18+$36.54+$10.83$316.42

Source: Penmark Research model (at 13 Nov 2026 price date)

Exhibit 73: Every central valuation mark sits below $390.22; only Bull and upper Street targets clear the price
SOTP value (bear-bull)12m target (bear-bull)12m target, WACC +/-1ptPeer P/E check & thesisStreet 12m targets$0$200$400$600$175$439$196$490$282$399$267$326$25$600Price $390Target $333

Source: Penmark Research model; FactSet / Yahoo Finance Street targets

Exhibit 74: Base SOTP spans $254.3 to $412.8 across beta (1.31-1.74) and terminal growth (2.0%-4.0%)
$347.0$360.4$375.6$392.9$412.8$319.8$331.3$344.2$358.8$375.4$295.6$305.4$316.4$328.8$342.8$273.9$282.4$291.8$302.3$314.2$254.3$261.7$269.8$278.9$289.0Terminal growth2.0%2.5%3.0%3.5%4.0%Equity beta1.311.421.5251.631.74

Source: Penmark Research model

Exhibit 75: Primary SOTP build (11.57% WACC) & peer forward P/E cross-check reconciliation
Valuation pillar / bridge lineMethod & key assumptionSep 26 ($m)$/shShare
Auto hardware, credits & services14Y DCF (FQ4 26E–FY40E), 17.5% mature GM, 3% g44,63112.764.0%
Energy generation & storage14Y DCF, 26.0% clean GM, 107 GWh FY30E, 3% g25,5497.312.3%
FSD software (net of $3.5bn R&D)4-outcome DCF (10m base subs; $39.06 pre-R&D)110,68131.6510.0%
Estimated programme spendPV of programme R&D & $62.2bn est. capex (run-off)−44,939−12.85−4.1%
Core enterprise valueMatches consolidated Core DCF (62.3% TV share)135,92238.8612.3%
Robotaxi (4 outcomes, analyst)0.91x of $884.7bn 6.6m-car FY35E success DCF805,035230.1872.7%
Optimus (4 outcomes, analyst)0.69x of $184.2bn 8.0m-unit FY40E success DCF127,80036.5411.5%
Net cash, SpaceX, NCI & litig.$35.92bn net cash + $2.90bn SpaceX − $0.93bn NCI/lit.37,89510.843.4%
Primary base SOTP equity3,544.0m sh; rolled 44d at 11.62% Ke (1.0133x)1,106,652316.42100.0%
Cross-check: 6.13x peer P/E6.13x FY28E core NI ($5.4bn) − $58.7bn venture PV + vent.940,859269.02−15.0%

Source: Penmark Research model; Yahoo Finance. CAPM: 5.15% Rf + 1.525 beta × 4.24% Damodaran ERP = 11.62% Ke (11.57% WACC).

Method fit & 15.0% cross-check gap: Upfront venture spend distorts consolidated FY27E P/E (251.4x at $390.22). Our 6.13x GM/Ford/Stellantis P/E cross-check ($269.02/sh) sits 15.0% below base SOTP ($316.42/sh) because FY28E core profit ($28.8bn Sep 26 core equity) misses post-FY28E FSD scaling ($110.7bn DCF). Weighting Bear (30%), Base (50%) and Bull (20%) yields $298.69/sh today and rolls at 11.62% Ke to our $333.38 12-month target.
PART 4 · THE NUMBERS · 15 VALUATION (CONTINUED)

Scenario weighting yields $298.69/sh and a $333.38 12m target (−14.6%, Sell), with Low conviction on six flips

Exhibit 76: 80% of scenario probability weight sits below the $390.22 price, giving $298.69 weighted value
$0$100$200$300$400$500$175$316$439Bear30%Base50%Bull20%Weighted $299Price $390

Source: Penmark Research model (price-date scenario values)

Exhibit 77: Six robustness tests move the $333.38 target into Hold or Buy territory, setting Low conviction
MS robotaxi vs MS FSDDiscount rate +1pt / -1ptWeighted vs all MS linesBeta 1.74 vs 1.31Weighted vs SpaceXpaperWeighted vs Base alone12m target $333−$132+$110−$51+$66+$62−$47+$59+$41+$20

Source: Penmark Research model; Morgan Stanley public SOTP

Exhibit 78: Bull, Base & Bear scenarios and exact reconciliation to the Page 1 call box ($333.38 target, −14.6% return)
Scenario / lineProb.Key operating & venture assumptionsVal ($)Tgt ($)Ret.Call
Bear case30%+2% FY27E cars, 14% hw GM, 21% BESS GM; 7.5m FSD; robotaxi 25% fail/35% succ; Opt. 45% fail.175.48195.86−49.8%Sell
Base case50%+10% FY27E cars, 16.5–17.5% hw GM, 26% BESS GM; 10m FSD; robotaxi 10% fail/70% succ+dom.316.42353.18−9.5%Hold
Bull case20%+16% FY27E cars, 18.5–19.5% hw GM, 29% BESS GM; 21m FSD; robotaxi 5% fail/35% dom; Opt. 25% dom.439.16490.18+25.6%Buy
Weighted / Target100%$298.69 × (1 + 11.616% Ke), $0 div.; 214.7x FY27E non-GAAP P/E at target (251.4x at $390.22)298.69333.38−14.6%Sell

Source: Penmark Research model. Unrounded return: ($333.3839 / $390.22) − 1 = −14.565% (Sell: <−10.0%). Subjective scenario weights carry wide bounds.

Exhibit 79: Reverse DCF (what $390.22 implies) & robustness checks that drive our Low-conviction label
Reverse-DCF / robustnessImplied or alternative inputVal ($)Tgt ($)Ret.Rating
Implied robotaxi ($390.22)Needs $1,063.1bn ($303.98/sh) = 1.20x success path (7.93m FY35E AVs at certainty)390.22—0.0%Price
Implied Core EV / Term. gNeeds 2.90x model Core EV ($394bn); max g at WACC−1pt (10.57%) gives $312.80312.80——Below
WACC −1.0pt / Beta 1.31100bp lower WACC or 1.31 relevered peer beta lifts 12m target into Hold band361 / 355399 / 393+2.3/+0.6%Hold
MS FSD / All MS lines*MS network services ($129.91/sh) flips to Buy; all MS lines ($415 tgt) give Hold397 / 354443 / 395+13.5/+1.3%Buy/Hold
SpaceX market-ratio dealPaper swap at $1.855tn SpaceX mark (not operating value) shifts target to Hold335.30374.25−4.1%Hold

Source: Penmark Research model; *Morgan Stanley public SOTP (not like-for-like: network services incl. insurance/charging). Base case alone ($353.18 target, −9.5%) also lands in Hold.

Why conviction is Low: Six plausible single-input shifts (base case alone, −1pt WACC, 1.31 beta, MS FSD, all MS lines, or a market-ratio SpaceX swap) move our Sell into Hold or Buy. Our Sell reflects negatively skewed risk-reward (30% Bear −49.8% vs 20% Bull +25.6%), not operational distress.
PART 4 · THE NUMBERS · 16 CATALYSTS

Eight dated catalysts over the next 14 months test our EPS gap and ±$25.29/sh autonomy probabilities

Exhibit 80: Catalyst calendar (Nov 2026–Dec 2027): we sit below consensus on 4Q26/FY27E earnings and unrisked AV timing
DateEventPenmark expectationConsensus / market viewUnrisked $/sh & mechanismRating impact
Nov–Dec 26Dallas Cybercab launchPilot fleet in hundreds; supervised backup in edge cases.Priced as smooth multi-city scale-up after San Antonio (+8.9%).Part of Bull 1 (+$25.29/sh if paired with Q2 customer AVs).Neutral
Dec 2026+SpaceX lock-up & deal talk$2.9bn stake lock-up expires; no operating value creation.Speculative premium after 12 Nov Bloomberg merger report.Market-ratio swap tests $374.25 Hold on paper mix only.Could flip Hold
Late Jan 274Q26 results & FY27 capexRev $28.72bn, non-GAAP EPS $0.34, capex $7.00bn; FY27E capex ~$25bn.Implied rev ~$29.1bn (YF $29.3bn), EPS $0.50–$0.52.Exposes ~$0.10/sh 3Q refund roll-off & Bear 1 (−$0.88/sh).Supports Sell
Jan–Jun 27120th Congress AV oversightTighter NHTSA reporting & scrutiny of 2,500-car exemption cap.5 Nov −10.8% drop (vs Nasdaq −5.5%) partly retraced 9–13 Nov.Bear 3: −10pt robotaxi success prob. = −$25.29/sh (30% prob).Supports Sell
1Q 2027EU-wide FSD rulingConditional supervised approval; gradual attach outside NA.Musk expects approval by 1Q27; high SaaS multiple.Bull 3: +0.4m FY27E FSD subs = +$0.73/sh (50% prob).Neutral
1Q–2Q 27Energy GM ex IEEPA refundsEnergy GM normalises to 25–26% as $187m net 3Q26 energy refunds end.33.1% 3Q26 energy GM widely extrapolated.Bear 4 (−4pt GM = −$1.67/sh) vs Bull 4 (+1pt car GM = +$2.19/sh).Supports Sell
2Q 2027Unsupervised customer AVsLimited geo-fenced rollout; noticeable revenue slips past 2Q27.Management guides noticeable robotaxi P&L impact ~2Q27.Bull 1: +10pt robotaxi success = +$25.29/sh (40% prob).Could flip Hold
Apr–Dec 27Megafactory TX & Optimus67.6 GWh FY27E storage; pilot external Optimus deliveries late 27.50 GWh TX ramp & AI5/Optimus commercial scale priced in.Bull 2 (+12 GWh = +$1.93/sh); Bull 5 (+5pt Opt. = +$2.63/sh).Neutral

Source: Q3 2026 10-Q & call, Yahoo Finance, Penmark Research model.

Exhibit 81: Dallas Cybercab, Megafactory Texas, EU FSD and customer autonomy cluster in the next nine months
Nov-26Jan-27Mar-27May-27Jul-27Sep-27Nov-27Jan-28AutonomyDallas Cybercab ridesEU-wide FSD approvalCustomer unsupervised FSDEnergy & FabMegafactory Texas 50 GWh rampMarginsIEEPA tariff refund wind-downVenturesAI5 chip & Optimus external14 Nov 2026+$0.73/sh+$25.29/sh+$2.63/sh

Source: Q3 2026 Update & call; Penmark Research model

Exhibit 82: Section 9 ‘next numbers to watch’ & exact rating-band crossing thresholds
Section 9 driverNext number & datePenmark vs consensusExact threshold that changes our Sell rating
1. Robotaxi & AV scaleActive unsupervised AVs & $/mi (2Q27 call, Jul 27)15k avg / 30k end-FY27E success fleet; $0.75/mi FY27E vs $0.30/mi goal.Hold band (−10% to +10%): Target $351.20–$429.24 at $390.22 (or share price $303.07–$370.42 at our $333.38 target).
2. Venture R&D & capex4Q26 capex & R&D (late Jan 27 release)$7.00bn capex, $3.09bn R&D, $0.34 non-GAAP EPS vs $0.50–$0.52 Street.Buy boundary (>+10.0%): Target >$429.24 at $390.22 (or share price <$303.07 at our $333.38 target).
3. FSD subs & ARPUYear-end paid FSD subs (late Jan 27) & EU ruling1.95m end-FY26E subs (vs 1.74m 3Q26); ~$800m FY26E FSD rev.Adopting MS’s $129.91/sh network mark lifts target to $443.06 (+13.5%, Buy).

Source: Signals KPIs, Yahoo Finance, Penmark Research model (rating rule: Buy >+10%, Hold −10% to +10%, Sell <−10%).

PART 4 · THE NUMBERS · 17 RISKS

For our Sell call, upside autonomy execution (+$25.29/sh) and WACC shifts pose the main rating risks

Exhibit 83: Quantified risks to rating and target: upside risks to our Sell call and operational/policy downside risks
Risk (direction vs Sell)Mechanism & scenario impactProb.Unrisked $/shObservable signpost to watch
Bull 1: Q2 customer AVs (Upside risk to Sell)Dallas Cybercab scales cleanly & retail owners deploy unsupervised FSD in 2Q27 (+10pt robotaxi success from failure).40%+$25.29NHTSA/state driverless permits & removal of remote safety monitors by Jun 2027.
Bull 5: Optimus & AI5 (Upside risk to Sell)AI5 chip enters Optimus ~Apr 2027 & paid external factory shipments start late 2027 (+5pt Optimus success).30%+$2.63Third-party industrial purchase orders & Fremont humanoid yield disclosures.
Bull 4 & 2: Tariffs & TX BESS (Upside to Sell)Lower inbound duties lift FY27E car GM +1pt (+$2.19/sh, 50% prob); Megafactory TX adds +12 GWh (+$1.93/sh, 40% prob).40–50%+$4.12 comb.1H27 automotive ex-credit GM >18.5% & quarterly BESS deployments >18 GWh.
WACC / Street FSD re-rate (Model risk to Sell)−100bp WACC lifts target to $399.32 (Hold); MS FSD valuation ($129.91/sh) lifts target to $443.06 (Buy).Low conv.+$65.94 target10Y Treasury falling below 4.15% or OEM FSD licensing deals (currently $0).
Bear 3: Congressional AV (Downside to target)Democratic Congress (from Jan 2027) tightens federal AV rules & Musk-company oversight (−10pt robotaxi success).30%−$25.29House/Senate AV hearing subpoenas or NHTSA recall restrictions in 1H27.
Bear 2, 4, 5 & 1: Core (Downside to target)Chinese export ASP −3% (−$6.56/sh), energy GM −4pt ex refunds (−$1.67), 5% rates (−$1.31), R&D +$2bn/yr (−$0.88).35–50%−$10.42 comb.BYD EU/ROW share gains, 1Q27 energy GM <24%, & quarterly R&D >$3.3bn.

Source: Q3 2026 Form 10-Q & call, Penmark Research model (net probability-weighted catalyst/risk skew: +$1.44/sh).

Exhibit 84: Autonomy milestones (+/-$25.29/sh) and Chinese export pricing (-$6.56/sh) dominate 12-month value swings
Bull 1: Dallas Cybercab & Q2 customer carsBull 5: Optimus external & AI5 chipBull 4: IEEPA tariff relief on car GMBull 2: Megafactory Texas (+12 GWh)Bull 3: EU FSD approval (+0.4m subs)Bear 1: Venture R&D +$2bn/yrBear 5: 5% rates & no US EV creditBear 4: Energy GM ex refunds (-4pt)Bear 2: Chinese export ASP (-3%)Bear 3: Congressional AV oversight+$25.29+$2.63+$2.19+$1.93+$0.73−$0.88−$1.31−$1.67−$6.56−$25.29

Source: Penmark Research model (unrisked $/sh)

What would end our Sell thesis

Our Sell call ends if probability-weighted 12-month target upside crosses −10.0% ($351.20 target at $390.22, or a $303.07–$370.42 share price at our $333.38 target). Operationally, commercial unsupervised FSD in customer cars by 2Q27 without safety incidents (+$25.29/sh unrisked) combined with disciplined FY27E capex (<$20bn) would lift our weighted target into Hold ($351–$429) or Bull ($490.18).

Regulatory & product-liability risks

Autonomous vehicles face federal FMVSS limits (2,500-car exemption cap for pedal-free Cybercabs) and tort liability. We deduct $329m in our EV bridge for the appealed Benavides Autopilot verdict ($129m compensatory + $200m punitive, FY25 10-K) and expense $80m–$661m/yr of FSD crash liability in FY27E–31E.

Key-person, governance & related-party risks

Tesla’s AI valuation is uniquely tied to CEO Elon Musk (19.8% holder), who simultaneously leads SpaceX, xAI and X Corp. Related-party links deepened in 2025–26 via $430.1m of FY25 xAI Megapack sales and Tesla’s $2.0bn SpaceX investment.

Two governance overhangs matter for per-share value: (1) the 423.7m-share 2025 CEO award (~10.7% potential dilution if $2.0tn–$8.5tn milestones are met, while costing $1,095m/yr in GAAP SG&A today); and (2) potential SpaceX merger terms, where any off-market exchange ratio could dilute TSLA holders (at market prices, it alters paper mix to $374.25 Hold).

PART 4 · THE NUMBERS · 18 OWNERSHIP AND TRADING

Insiders hold 20.0%, short interest sits at 1.75 days to cover (66.0m shares), and 44% of Street ratings are Buy

Exhibit 85: Top 10 institutional 13F holders at Q2 2026 (partial 42.5% Signals coverage)
Institution (Q2 2026 13F)Shares (m)% S/O*Δ (m)Δ %
Vanguard Group263.376.67%+2.22+0.85%
BlackRock214.125.42%+5.99+2.88%
State Street117.462.97%+2.76+2.41%
Geode Capital Mgmt68.911.74%+0.91+1.34%
Invesco55.461.40%+17.52+46.17%
Capital World Investors45.741.16%+2.83+6.60%
JPMorgan Chase41.971.06%−4.10−8.90%
FMR (Fidelity)39.771.01%+3.74+10.38%
Morgan Stanley38.870.98%−0.31−0.78%
Norges Bank Inv. Mgmt38.810.98%n/dn/d
Reported 13F (4,262)*1,677.1442.46%——

Source: Signals 13F feed (30 Jun 2026). *42.46% of 3,949.55m shares; incomplete non-13F coverage (Sep 26 proxy lists Vanguard at 5.3%).

Exhibit 86: Short interest fell 16.6% from its June peak to 65.96m shares (1.75 days to cover) at Oct-end
65m70m75m80mJun-26Jul-26Aug-26Sep-26Oct-2666.0m

Source: Signals biweekly short-interest filings to 30 Oct 2026

Exhibit 87: Public Street targets span $24.86 to $600.00 (mean $416.26), with Penmark at $333.38
GLJ low $25Price $390BofA $460Penmark $333Street mean $416Wedbush high $600

Source: FactSet / Yahoo Finance public consensus & broker targets (13-14 Nov 2026); Penmark model

Exhibit 88: Liquidity, index membership, lock-ups & Street ratings
3M avg. daily volume / value37.72m sh / $13,905.9m
52-week intraday / close$297.38–$498.83 / $298.32–$489.88
Shares out (15 Oct 2026)3,952.76m (incl. 423.74m unearned)
Diluted shares modelled3,544.01m (ex unearned CEO award)
Insider beneficial stake20.0% (Musk 699.58m / 19.8%)
Short interest (30 Oct 26)65.96m sh (1.75x 3M vol; 1.57 feed)
Index membershipS&P 500, Nasdaq-100, Russell 1000
Street ratings (43)19 Buy (44%), 20 Hold (47%), 4 Sell (9%)
Street 12m target range$416.26 mean ($24.86 to $600.00)
Lock-ups & overhangsSpaceX $2.9bn lock-up Dec 26; 423.7m award

Source: Signals, Q3 10-Q, Yahoo Finance, Penmark model.

Who sets the marginal share price

At $13.9bn average daily turnover and 1.75 days to cover on 3M volume (1.57 in FINRA’s settlement feed), TSLA is priced by AI/options flow rather than short-squeeze risk. Passive index funds anchor core ownership while growth funds trade autonomy milestones.

Insider activity & dilution overhangs

Musk’s June 2026 exercise of 303.96m options at $23.34 (17.53m withheld at $404.66) lifted his beneficial stake to 19.8%. CFO Taneja’s $0.94m Sep 2026 sale carried no 10b5-1 flag (not proof it was unplanned); 423.7m 2025-award shares stay unearned below $2.0tn.

PART 5 · APPENDIX · 19 FINANCIAL MODEL

Annual consolidated statements (FY24A–FY30E): GAAP EBIT troughs at $774m in FY27E before FSD inflects

Exhibit 89: Consolidated income statement, balance sheet & cash flow statement (FY24A–FY30E, US$m, Dec year-end)
$m, Dec year-endFY24AFY25A9M26AFY26EFY27EFY28EFY29EFY30E
Income statement
Automotive revenues (incl. credits & FSD)77,07069,52657,31077,57085,06993,123101,883111,492
Energy generation & storage revenue10,08612,7718,42812,18815,84418,94421,53523,637
Services & other revenue10,53412,53012,93117,63119,74721,72123,67625,570
Total revenues97,69094,82778,669107,389120,660133,788147,095160,699
Revenue growth y/y+0.9%−2.9%+12.5%+13.2%+12.4%+10.9%+9.9%+9.2%
Gross profit17,45017,09414,96520,13722,54826,41330,92635,874
Gross margin17.9%18.0%19.0%18.8%18.7%19.7%21.0%22.3%
Research & development (core + venture)4,5406,4117,07810,17013,07815,90517,22517,555
SG&A (incl. $1,095m/yr CEO award FY27E+) & restruc.5,8346,3285,8307,9148,6979,2569,77410,255
Income from operations (GAAP EBIT)7,0764,3552,0572,0527741,2523,9288,064
GAAP operating margin7.2%4.6%2.6%1.9%0.6%0.9%2.7%5.0%
Net interest, other income & income taxes15−561321586774355−700−1,587
GAAP net income to common7,0913,7942,3782,6391,5471,6063,2286,477
Non-GAAP net income (company definition)n/dn/d4,3035,5195,5025,8727,80911,381
Adjusted EBITDA (company definition)16,05614,59610,84113,77314,32817,35722,06127,256
Balance sheet (model net-asset basis from 30 Sep 2026)
Cash, cash equivalents & short-term investments36,56344,05945,44741,61731,03525,00025,00027,596
Depreciable operating assets (excl. $2.9bn SpaceX stake)n/dn/d59,09164,29280,60599,642107,313110,287
of which programme assets (SI, fab, Cybercab, Optimus)n/dn/d10,95615,27628,12843,74348,13847,314
Debt & finance leases8,2138,3769,5239,5239,52316,33615,8039,523
Total stockholders’ equity72,91382,13789,28990,68696,949103,664112,399124,792
Net cash (cash & investments less debt)28,35035,68335,92432,09421,5128,6649,19718,073
Cash flow statement
Net cash from operating activities (OCF)14,92314,74714,89718,06714,57117,18620,82925,124
Capital expenditure (core + estimated venture)11,3428,52713,01420,01425,15330,03420,29716,249
Free cash flow (company definition)3,5816,2201,883−1,947−10,582−12,8485338,875
FCF after employee SBC* / Facility draw (repay)n/dn/dn/d−4,692 (4Q26E) / 0−14,201 / 0−16,861/+6,813−3,880/−533+4,054/−6,280

Source: FY2024–25 10-K, Q3 2026 10-Q, Penmark Research model. Shaded columns are Penmark estimates; *4Q26E only (not FY26E annual); n/d = not disclosed on model basis.

Balance-sheet & cash-floor mechanics: From 30 Sep 2026, net assets equal cash + SpaceX ($2,900m) + depreciable operating assets + other net operating assets (−$8,626m at 3Q26A, moving to −$6,467m by FY30E as ~$530m/yr working-capital investment accumulates). Under our model rule, cash is held at or above $25,000m (~2.2 months of FY28E revenue) via FY28E facility draws.
PART 5 · APPENDIX · 19 FINANCIAL MODEL (CONTINUED)

Quarterly path (2Q25A–4Q26E), pillar profit build & valuation multiples at $390.22 and $333.38 target

Exhibit 90: Quarterly trajectory (2Q25A–3Q26A reported & 4Q26E Penmark estimate; post-FY26E modelled annually)
$m, except per share, k, GWh, m2Q25A3Q25A4Q25A1Q26A2Q26A3Q26A4Q26E
Total revenues22,49628,09524,90122,38728,23628,04628,720
Automotive (of which reg. credits)16,661 (439)21,205 (417)17,693 (542)16,234 (380)20,516 (146)20,560 (113)20,260 (100)
Energy storage / Services & other2,789 / 3,0463,415 / 3,4753,837 / 3,3712,408 / 3,7453,139 / 4,5812,881 / 4,6053,760 / 4,700
Gross profit (Gross margin %)3,878 (17.2%)5,054 (18.0%)5,009 (20.1%)4,720 (21.1%)4,751 (16.8%)5,494 (19.6%)5,172 (18.0%)
R&D / SG&A1,589 / 1,3661,630 / 1,5621,783 / 1,6551,946 / 1,8332,371 / 1,9822,761 / 2,0153,092 / 2,084
GAAP operating income (EBIT)9231,6241,409941398718−5
GAAP NI / Non-GAAP NI1,172 / 1,3931,373 / 1,770840 / 1,761477 / 1,4531,114 / 1,153787 / 1,697261 / 1,216
GAAP EPS ($) / Non-GAAP EPS ($)0.33 / 0.400.39 / 0.500.24 / 0.500.13 / 0.410.32 / 0.330.22 / 0.480.07 / 0.34
Adj. EBITDA / Free cash flow3,401 / 1464,227 / 3,9904,154 / 1,4203,668 / 1,4443,273 / −1,0923,900 / 1,5312,932 / −3,830
Deliveries (k) / BESS (GWh) / FSD (m)384 / 9.6 / 0.95497 / 12.5 / 1.04418 / 14.2 / 1.10358 / 8.8 / 1.28480 / 13.5 / 1.48487 / 13.7 / 1.74480 / 16.0 / 1.95

Source: Q2 & Q3 2026 Form 8-K Updates, Q3 10-Q, Signals KPIs, Penmark model (4Q26E).

Exhibit 91: Pillar operating profit/FCF build, per-share metrics & valuation ratios at $390.22 and $333.38 target
Pillar & ratio detailFY24AFY25AFY26EFY27EFY28EFY29EFY30E
Core pillar EBIT ($m, ex $1,095m/yr CEO award)
Automotive & services EBIT / Energy EBITn/dn/dn/d5,762 / 2,4086,858 / 2,9178,017 / 3,3598,766 / 3,735
FSD EBIT (after $3.5bn R&D) / Programme EBITn/dn/dn/d−2,154 / −4,148−1,044 / −6,385752 / −7,1053,482 / −6,823
Per-share data, returns, leverage & multiples ($390.22 price, 13 Nov 2026)
Diluted shares ex unearned award (m)3,498.03,528.03,544.03,544.03,544.03,544.03,544.0
GAAP EPS ($) / Non-GAAP EPS ($)2.04 / n/d1.08 / n/d0.74 / 1.560.44 / 1.550.45 / 1.660.91 / 2.201.83 / 3.21
ROE (GAAP NI / avg. common equity)10.5%4.9%3.1%1.6%1.6%3.0%5.5%
ROCE (NOPAT / avg. invested capital)12.6%7.5%3.0%0.9%1.1%3.1%6.0%
Net debt / Adj. EBITDA (negative = net cash)−1.8x−2.4x−2.3x−1.5x−0.5x−0.4x−0.7x
Non-GAAP P/E (x) at $390.22 (GAAP P/E)n/d (191.3x)n/d (361.3x)249.6x (524.8x)251.4x (893.7x)235.5x (860.9x)177.1x (428.4x)121.5x (213.5x)
EV / Adj. EBITDA (x) at $390.22 ($1,506.5bn EV)93.8x103.2x109.4x105.1x86.8x68.3x55.3x
FCF yield (%) / Dividend yield (%) at $390.220.3% / 0.0%0.5% / 0.0%−0.1% / 0.0%−0.8% / 0.0%−0.9% / 0.0%0.0% / 0.0%0.6% / 0.0%
Non-GAAP P/E (x) at our $333.38 targetn/dn/d213.3x214.7x201.2x151.3x103.8x

Source: Filings, Penmark Research model. Per-share multiples & FCF yield use diluted shares (3,544.01m FY26E+; $1,382.9bn diluted cap), while EV ($1,506.5bn) uses 3,952.76m total cover shares ($1,542.4bn cap incl. 423.74m unearned CEO award).

Forecast cadence & multiple interpretation: Beyond 4Q26E, our model forecasts annually (FY27E–FY40E) because robotaxi, Optimus and Terafab milestones lack quarterly guidance. Non-GAAP P/E stays above 200x through FY28E even at our $333.38 target because venture R&D and D&A depress near-term EPS while 84% of base SOTP (and 87% of the $390.22 share price, incl. 78% from robotaxi alone) sits in post-2030 autonomy and robotics cash flows.
PART 5 · APPENDIX · 20 REFERENCE

Reference: GAAP-to-non-GAAP reconciliation, glossary, definitions, peer set & methodology

Exhibit 92: Reconciliation from GAAP net income to non-GAAP net income & company Adjusted EBITDA ($m)
$m, Dec year-endFY24AFY25A9M26A4Q26EFY26EFY27EFY28E
GAAP net income to common stockholders7,0913,7942,3782612,6391,5471,606
Add: Stock-based compensation (employee + CEO award)1,9992,8253,3461,1374,4834,7155,109
Less: Tax / SpaceX mark & other company non-GAAP adj.n/dn/d−1,421−181−1,602−760−843
Non-GAAP net income (company definition)n/dn/d4,3031,2165,5195,5025,872
Memo: GAAP EBIT + D&A + SBC = Company Adj. EBITDA*16,05614,59610,8412,93213,77314,32817,357

Source: Q3 2026 Update, Penmark model. *FY24A–25A incl. restructuring/NCI lines; Signals standardised FY25A EBITDA ex SBC add-back is $9,385m.

Glossary of company & sector terms

  • FSD (Supervised / Unsupervised): Tesla’s camera-based end-to-end neural-network driver-assistance ($99/mo in NA; 1.74m active subs in 3Q26) and target autonomous software stack.
  • Cybercab: Dedicated two-seat autonomous robotaxi without steering wheel or pedals; Dallas pilot from Nov 2026.
  • Megapack & GWh: Utility-scale lithium-ion BESS measured in gigawatt-hours of energy capacity (distinct from GW power rating).
  • Optimus: Humanoid bi-pedal robot in pilot Fremont assembly; external sales targeted late 2027.
  • Terafab & AI5/AI6: Planned high-volume semiconductor/compute and cell manufacturing facilities driving FY27E–28E venture capex.
  • IEEPA refunds: 3Q26 customs duty credits of $526m in energy COGS and $260m in auto COGS less $339m owed to storage customers ($447m net company-wide; $187m net energy).

Methodology note & departures

Hybrid archetype: Primary Products & units (vehicles, BESS) + Recurring revenue (FSD carved out FY27E+) + Pipeline (4-outcome robotaxi & Optimus DCFs), valued via SOTP across one 14-year DCF engine (FQ4 26E–FY40E). Situation flags: controlled/key-person (Musk 19.8%), heavy AI capex cycle, Low conviction.

Departures & limits: Annual historical non-GAAP NI/EPS not disclosed on a comparable basis are shown as n/d; post-FY26E forecasts are annual; SpaceX merger is unvalued (paper-swap check shown). Reporting currency: US$ (domestic US issuer, no FX/ADR adjustment).

Primary data sources (cut-off 13–14 Nov 2026)

  • SEC filings & Signals data: FY21–25 10-K, 3Q26 10-Q (22 Oct 26), 8-Ks, DEF 14A, calls (4Q24–3Q26), KPIs/ratios, 13F, Form 4, short interest.
  • Market & industry: Yahoo Finance (13–14 Nov 26), MarketScreener (Jan–Apr 26 stale EBITDA), US Treasury (5.15% 10Y), Damodaran ERP (May 26), IEA, InfoLink/SEIA, BYD & Waymo.
Exhibit 93: Key ratio and valuation metric definitions
MeasureExact definition in Penmark modelBasis / Scope
Implied P/E (actual & target)Share price ($390.22) or 12m target ($333.38) ÷ Non-GAAP diluted EPS ($1.56 FY26E, $1.55 FY27E).3,544.0m diluted sh.
Enterprise value (EV)Cover market cap ($1,542.45bn on 3,952.8m sh.) − 3Q26 cash & ST inv. ($45.45bn) + debt ($9.52bn) = $1,506.52bn.Consolidated EV
ROE / ROCE / FCF yieldROE = GAAP NI ÷ avg. common equity; ROCE = NOPAT (22% tax) ÷ avg. invested capital; FCF yield on diluted sh.Standardised annual
Rating & 12m target ruleProbability-weighted value ($298.69) × (1 + 11.616% Ke) = $333.38; Buy >+10%, Hold −10% to +10%, Sell <−10%.Price return ex div.
Exhibit 94: Comparable cohorts & outside valuation marks used in Exhibit 12, Section 7 & Section 15
Company / markRole & why comparable to Tesla’s business linesScale / markFwd P/E / mult.
GM, Ford, StellantisMass-market automaker cohort setting our 6.13x CY26E forward P/E core cross-check (repriced 13 Nov 26).$86 / $12 / $4 sh6.1x / 6.3x / 5.1x
Toyota, BYD, Rivian, FerrariGlobal EV/auto cohorts comparable on scale (Toyota/BYD), pure-play EV (Rivian) & luxury margin (Ferrari).2.26m BYD BEV9.1x–37.0x trl.
Waymo & Fluence EnergyOutside SOTP benchmarks: Waymo autonomous ride-hailing (~4k AVs; Feb 26 private round) & Fluence BESS (FQ3 26 10-Q).$126bn Waymo1.42x EV/sales

Source: Penmark Research model, Signals, Yahoo Finance, Fluence FQ3 26 Form 10-Q (30 Jun 2026), reported Feb 2026 Waymo round.

PART 5 · APPENDIX · 21 DISCLOSURES

Analyst certification and important disclosures

Ratings

Ratings are based on the upside from the share price to our 12-month price target, excluding dividends. The price target is weighted across bull, base and bear scenarios.

BuyUpside to price target above +10%
HoldUpside to price target from −10% to +10%, inclusive
SellUpside to price target below −10%

We review ratings at each set of results and whenever the facts change materially. A share price can move outside its band before the rating changes, and targets may not be reached.

Analyst certification

The Penmark Team certifies that (1) the views in this report accurately reflect its views about the subject company and its securities, and (2) no part of its compensation was, is or will be directly or indirectly related to the specific ratings, price targets or views in this report.

Contact

Penmark Research

Questions about our research:
research@penmarkresearch.com

All our research and coverage:
penmarkresearch.com

Basis of preparation

This report uses information publicly available up to its publication date. Share prices are closing prices on the trading day before the publication date, unless stated otherwise. Forecasts, valuations and price targets are Penmark Research’s own, and forecast periods are marked E. Sources are shown beneath each exhibit.

Valuation and price targets

Price targets are based on the valuation methods set out in each report and are weighted across bull, base and bear scenarios. The main risks to each rating and price target are described in the report. Penmark Research updates its coverage after each set of results but is not obliged to update a report at any other time.

Independence and conflicts of interest

Penmark Research is independent. It does not provide investment banking, brokerage, market-making or asset management services, does not trade securities, and does not accept payment from any company for research coverage. Companies do not see or review Penmark Research reports before publication.

No member of the Penmark Team holds a long or short position, directly or through derivatives, in the securities of any company Penmark Research covers.

Regulatory status

Penmark Research publishes general, impersonal investment research on a regular basis. It is not registered as an investment adviser with the US Securities and Exchange Commission or any state securities authority, is not a registered broker-dealer and is not a member of FINRA. Penmark Research does not give advice tailored to any person’s circumstances.

Not investment advice or an offer

This report is for information only. It is not personalised investment advice and is not an offer or solicitation to buy or sell any security or other financial instrument. It does not take into account any reader’s objectives, financial situation or needs. Readers should make their own investment decisions and seek independent professional advice where appropriate.

Forward-looking statements and risk

Forecasts, valuations and price targets are estimates based on assumptions that may prove wrong, and they may not be achieved. Opinions are current as of the publication date and may change without notice. Investing in equities involves risk, including the loss of capital. Share prices can be volatile, and past performance is not a reliable indicator of future results.

Data and sources

Company financials, operating data and filings come from company filings, earnings releases and earnings calls. Market data and any consensus estimates come from third-party sources named beneath the relevant exhibit; consensus figures from aggregators are unofficial. Penmark Research may reference publicly available third-party research, broker estimates, consensus data and market targets for comparison, while all Penmark Research forecasts, ratings and valuations remain proprietary. Third-party data belong to their providers and may not be extracted from this report.

Accuracy and liability

Information is taken from sources believed to be reliable, but Penmark Research makes no representation or warranty as to its accuracy, completeness or timeliness. To the fullest extent permitted by law, Penmark Research accepts no liability for any loss arising from use of this report. Descriptions of litigation and regulatory matters reflect allegations and public records, not findings, unless stated. If a published report is corrected, the corrected version carries an “Updated” date at penmarkresearch.com.

Distribution and access

Penmark Research reports are available to registered users of penmarkresearch.com. Downloads and programmatic access are governed by the Penmark Research subscription terms. This report is not directed at, or intended for, any person in Singapore or India, and access from those countries is blocked. It is not intended for distribution to, or use by, any person in any jurisdiction where that would be contrary to law or regulation or would subject Penmark Research to registration or licensing requirements there.

Copyright and permitted use

© Penmark Research. All rights reserved. Penmark Research is a trade name of Brightriver Labs. Readers may share a link to this report but may not copy, reproduce, redistribute, sell or use its content, including for text and data mining, without written consent. Contact: research@penmarkresearch.com.

Keep reading

Sign in to Penmark