TSLA
At $370, Tesla is 5 independent bets. Here's what the market implies for each.
Tesla trades at $370/share, a $1.3 trillion market capitalization. This report decomposes the current price into 5 independent bets, each with its own bull/base/bear scenarios. It does not tell you whether Tesla is cheap or expensive. It answers: at $370, what probability is the market assigning to each bet, and do you have an edge on any of them? If not, the index is the better bet.
What The Price Implies
At $370, Tesla breaks down into: Net Cash $10/share (fact), Auto $11/share, Energy $35/share, FSD Software $10/share, Robotaxi $282/share, Optimus $22/share. The existing businesses (auto + energy + FSD SW) account for $56/share. Robotaxi and Optimus — businesses that do not exist at scale — account for $304/share (82% of equity value).
The Beliefs Embedded in the Price
Musk's dual role (Tesla CEO + SpaceX + xAI + DOGE) creates execution risk across all 5 bets. Tesla invested $2B in Musk's xAI. Terafab chip factory is 80% for SpaceX/xAI, 20% ground-based. $6.5B SBC in FY2025 (+32% YoY). Brand damage from DOGE controversy contributed to Q1 2025's 13% delivery decline.
At $370, you are paying $304/share (82% of equity) for robotaxi and Optimus — businesses that do not exist at scale today. The existing businesses justify ~$66/share. The gap is the price of belief. Do you have an edge on whether these bets will pay off? If not, the index is the better bet.
What Does $370 Buy You?
Price-Implied Expectations decomposition. Every dollar accounted for.
The Key Debates
The 16 questions that determine whether this stock is worth owning.
Bull: Humans drive with vision only. Tesla's neural network is improving 700x per year. FSD v14 goes 9,200 miles between interventions. More data + better AI will solve remaining edge cases. Bear: NHTSA found FSD cameras 'lost track of vehicles directly ahead' in glare/fog. Tesla removed radar 'against the advice of its own engineers.' Camera recognition drops 40% in heavy rain. No camera-only L4 exists anywhere. China mandates LiDAR for L3+.
This contributes ~4% first-order Sobol variance but interacts with HW4 and data advantage for ~30-40% of total uncertainty. If you have no view on camera-only viability, you are guessing on the largest driver of the stock price.
This is perhaps the most important technical question in the entire valuation. Tesla's approach -- cameras only, no LiDAR -- is unique among L4 aspirants. Every commercially deployed L4 system (Waymo, Baidu Apollo) uses LiDAR + camera fusion. If cameras alone CAN achieve L4, every Tesla on the road is a potential robotaxi. If they can't, Tesla needs a fundamental hardware redesign.
- +Humans drive with vision only. Tesla's neural network is improving 700x per year. FSD v14 goes 9,200 miles between interventions. More data + better AI will solve remaining edge cases.
- −NHTSA found FSD cameras 'lost track of vehicles directly ahead' in glare/fog. Tesla removed radar 'against the advice of its own engineers.' Camera recognition drops 40% in heavy rain. No camera-only L4 exists anywhere. China mandates LiDAR for L3+.
What Would Change the Price
The highest-impact events, ranked by potential price impact.
The Beliefs Behind the Price
Each assumption embedded in the current price. Do you have an edge on any of them?
MARKET-IMPLIED EV/Revenue for Tesla's auto business. v5.1 CHANGE: reduced from 2.0x to 1.5x. UPDATED COMPS (March 2026): Toyota 1.5x, GM 0.95x, BYD 1.0x, Stellantis 0.23x. Median legacy OEM: ~0.95x. Tesla's 1.5x = Toyota parity, a 58% premium to median OEM. RATIONALE FOR REDUCTION: Two consecutive years of declining deliveries (1.81M -> 1.79M -> 1.636M). BYD outsold Tesla globally. Auto margins converging (17.9% vs historical 25-28%). Revenue declined 10% in 2025 -- worse than most legacy OEMs. A premium TO Toyota is no longer justified; parity is generous given the delivery trajectory.
FY2025 actual automotive revenue: $69.5B, down 10% YoY. Includes vehicle sales and regulatory credits but excludes services. Two consecutive years of auto revenue decline. Source: Tesla 10-K FY2025.
Can cameras alone (no LiDAR) achieve L4? MARKET-IMPLIED: 50%. NHTSA upgraded probe EA26002 covering 3.2M vehicles for camera visibility failures (glare, fog, dust). One fatality. China mandates LiDAR for L3+. Zero camera-only L4 systems deployed anywhere globally. Waymo (LiDAR+cameras) has 92% fewer serious crashes than humans over 170.7M miles. FSD v14 achieved 9,200 miles between critical interventions -- impressive but still not L4.
Does 8.3B+ miles of fleet data give Tesla a real advantage over Waymo's 22M miles? MARKET-IMPLIED: 60%. Tesla has 375x more miles but from human drivers with cameras. Waymo has richer 3D LiDAR annotations per mile. Tesla shut down Dojo ($500M+) in Aug 2025, suggesting compute architecture matters more than raw data volume.
MARKET-IMPLIED EV/Revenue for Tesla Energy. v5.1 UPDATE: Fluence has re-rated to 1.41x EV/Rev (from 0.78x in v5.0). Enphase has declined to ~3.7x (from ~5.7x). The 5.0x is now split for transparency: COMP-ANCHORED (2.0x): 42% premium to Fluence 1.41x, justified by Tesla's 30% GM vs Fluence 12% and profitability. GROWTH PREMIUM (3.0x): belief in 27%+ growth sustaining, 80% US market share, IEA 14x storage expansion, Autobidder software. The growth premium ($38.4B or $10.9/share) is as much a BELIEF as any robotaxi assumption. Range: 1.4x (Fluence-like) to 12x (energy platform).
FY2025 energy generation and storage revenue: $12.8B, up 27% YoY. Deployed 46.7 GWh (vs 10.4 GWh in 2024). Gross margin 28-31%. Tesla has ~80% US grid storage market share. Revenue per GWh declining ($273M) as cell prices fall. Source: Tesla 10-K FY2025.
15x EV/Revenue for FSD software. Above SaaS median (6-7x) due to 100% subscriber growth. Below top-quartile SaaS (13-14x) due to hardware dependency and regulatory risk. Range: 5x (feature-like, low stickiness) to 25x (platform-like, high growth).
FSD annualized recurring revenue: ~$1.3B from 1.1M subscribers at $99/month. Disclosed for first time in Q4 2025 earnings. Subscriber count roughly doubled during 2025. Transitioning to subscription-only model (sunsetting one-time purchases). FSD deferred revenue balance: $3.6B. Source: Tesla Q4 2025 earnings.
Does HW4 have enough compute for real-time L4? MARKET-IMPLIED: 55%. HW4 is 5x HW3 compute, but HW3 already insufficient (stuck on v12, HW4 gets v13+). Tesla building TeraFab chip factory suggests they know current silicon may not suffice. If HW4 insufficient, only future vehicles with HW5+ can be robotaxis -- losing the 'every Tesla is a robotaxi' narrative.
Net cash = $44.1B cash/investments minus $8.4B total debt = $35.7B. SEC-audited from FY2025 10-K. Cash increased $7.5B in 2025. BUT: $20B+ capex planned for 2026, potential capital raise for Terafab. FSD deferred revenue of $3.6B is a liability, not included here.
Early commercial price ~$100K. Musk targets $20-25K at mass scale. Current cost ~$50-100K per unit. Comps: BD Atlas ~$420K, Unitree G1 $16-74K. Figure AI estimated $30-150K.
increased from 15% to 20%. At 200K units, Tesla achieves manufacturing scale efficiencies. Vertical integration in batteries and motors gives cost advantage vs competitors. Robot-as-a-Service (RaaS) model could add recurring software revenue. Labor economics support premium pricing: at $50K/robot, hourly cost is $3-5/hr vs warehouse worker $18-25/hr, implying strong willingness to pay. Range: 5% (R&D-heavy ramp) to 30% (mature with RaaS).
20x OI multiple for growth hardware/robotics business. Industrial robot companies trade at 15-25x. 20x reflects growth-stage with TAM optionality. Could be higher if software services (Robot-as-a-Service) dominate.
P(Optimus reaches commercial scale by 2030) = 20%. v5.1: probability UNCHANGED despite conditional value increase. Zero useful work done as of Q4 2025 (Musk admission). Gen 3 production commenced Jan 2026 at Fremont. Competitors ahead on deployment: Figure AI (30K BMW vehicles), Unitree (5.5K units shipped, world leader in actual sales). Goldman Sachs upgraded forecast 6x to $38B market by 2035 with 1.4M units. 20% reflects real progress + extreme timeline skepticism + Musk credibility 1.5/10.
increased from 100K to 200K units. How many Optimus robots will Tesla ship annually by ~2030? Goldman Sachs base case: 1.4M units by 2035, implying ~250K by 2030. Tesla capturing 80% of addressable market (up from 40%) reflects manufacturing scale advantage -- Tesla is converting Gigafactory lines and targeting 50-100K initial capacity. Musk targets 1M/yr (credibility 1.5/10). 200K = Goldman's 250K global minus 50K for competitors (Unitree, Figure, Apptronik all shipping/deploying).
Will regulators approve unsupervised robotaxis in 25%+ of the US within 5 years? MARKET-IMPLIED: 60%. Mixed signals: pro-AV administration and NHTSA FMVSS revision proposals are positive. But active NHTSA EA26002 investigation, $243M verdict upheld, class action certified, Cybercab FMVSS exemption NOT FILED despite April 2026 production target. Waymo proving the regulatory path IS possible (expanding to 20+ cities in 2026).
Revenue per autonomous vehicle per year. Based on ~$0.95/mile, ~100 miles/day, ~95% uptime = ~$34,660. Waymo actual: ~$142K/vehicle/year at small scale (premium pricing, dense cities). At 7.5M fleet, utilization drops significantly and pricing pressure increases. Tesla's 19% availability rate is far below 95% assumed. Uber driver car generates ~$30-40K gross/year.
MARKET-IMPLIED fleet to justify $370: 7.7 MILLION autonomous vehicles. v5.1 UPDATE: increased from 7.5M to 7.7M because reducing auto multiple to 1.5x pushed more value into the robotaxi residual. For context: Waymo has 2,500 vehicles after 15 years and $30B+ invested. Tesla has 4 unsupervised cars. The implied fleet is a 1.925 MILLION times scale-up from Tesla's current unsupervised fleet. The entire US taxi/rideshare fleet is ~1M vehicles. ARK Invest projects 5M+ by 2030 (bull case). Morgan Stanley expects 1,000 in 2026, 1M by 2035. NOTE: some of this residual may be unattributable premium (brand, momentum, speculative froth) rather than pure robotaxi belief.
Net margin for robotaxi operations: 40%. Key advantage: no driver cost (Uber drivers take 70%+ of revenue). Key costs: vehicle depreciation, electricity, insurance, maintenance, cleaning, remote operations. Waymo reportedly spends heavily on operations even with a small fleet. Insurance for AVs is expensive due to limited actuarial data.
30x operating income multiple for robotaxi platform. Uber trades at ~30-40x forward earnings. A driverless ride-hailing network with no driver costs would have fundamentally superior unit economics. Premium for network effects (more vehicles = shorter waits = more riders). But competition from Waymo, Uber-AV partnerships prevents monopoly economics.
1.0x EV/Revenue for services and other. Mixed business: Supercharging (high-margin, growing), insurance (scaling), used cars (low-margin), parts/service (stable). 1.0x is conservative -- Supercharging network alone may justify a premium as NACS becomes industry standard.
FY2025 services and other revenue: ~$12.5B, up 19% YoY. Includes Supercharging, insurance, used vehicle sales, parts, body shop, retail merchandise. Margins improving as Supercharging and insurance scale. NACS (North American Charging Standard) adoption by other OEMs creates Supercharging network moat. Source: Tesla 10-K.
Diluted shares outstanding ~3.53B as of FY2025. Up 0.86% YoY from SBC dilution. NOTE: Tesla may issue secondary offering (first since 2020) to fund Terafab $25B+ chip factory, which could add 1-2% dilution. 10-K states company 'may decide it is best to raise additional capital.'
Methodology
Price-Implied Expectations (PIE) framework based on Mauboussin & Rappaport's "Expectations Investing." Segments valued using comparable company multiples (Layer 2), with residual allocated to probability-weighted speculative businesses (Layer 4). Evidence sourced from SEC filings, earnings calls, and public reports.
PIE Model • v5.1 • Last updated: 3/26/2026