Tesla increases investment in embodied AI, but near-term upside is limited

Overview

This report focuses on Tesla's post-1Q26 investment cycle, progress in autonomous driving and embodied AI, the pace of its energy business, and valuation changes. Morgan Stanley believes Tesla is entering a phase of significantly higher capital expenditure, expanding energy, automotive, semiconductors, compute capacity, and next-generation embodied AI infrastructure. This investment is necessary to establish long-term leadership, but in the short term it will lead to free cash flow consumption, while projects such as robotaxi, Unsupervised FSD, and Optimus are advancing in a more cautious, validation-driven manner, meaning commercialization could proceed more slowly than the market expects.

Core views

The core view is that 'the long-term conviction remains intact, while near-term risk-reward is balanced.' The report continues to believe in Tesla's long-term path in autonomous mobility, renewable energy, robotics, and embodied AI, but sees limited near-term upside in the stock price. 1Q26 automotive and energy gross margins were better than expected, but included one-off factors such as warranty reserve releases and tariff refunds; the company also highlighted margin pressure from competition and tariff costs. Capex guidance was raised to >$25bn, implying FCF will remain under pressure this year. Robotaxi will continue city expansion, but the key constraint is strict safety validation; HW3 cannot be upgraded to Unsupervised FSD, reducing confidence in the bull-case thesis of rapidly bringing the existing fleet into the robotaxi story.

Analysis framework

The report uses fundamental forecast revisions, sum-of-the-parts SOTP valuation, DCF, scenario analysis, and a risk-reward framework. The analysis starts from 1Q26 earnings variances, 2026/2027 forecast revisions, capex and FCF, robotaxi/FSD progress, ESS shipments, and Terafab semiconductor capability buildout, and then values Auto, Network Services, Tesla Mobility, Energy, and Humanoids separately before summing them into the price target.

Methodology notes

• Valuation methodsSOTP valuation

  sum-of-the-parts valuation

  The base-case price target of $415 is made up of five parts: Auto $45, Network Services $145, Tesla Mobility $125, Energy $40, and Humanoids $60, reflecting Tesla's valuation structure as it expands from an auto manufacturer into software services, autonomous mobility, energy, and robotics platforms.

• Valuation methodsDCF

  discounted cash flow

  Tesla Mobility is valued using DCF, with the base-case assuming about 5mn vehicles by 2040 and about $1.33/mile; Auto valuation uses assumptions such as 2040 unit sales, exit EBIT margin, WACC, and exit EBITDA multiple.

• scenario_analysisbull/base/bear case

  bull, base, and bear scenarios

  The report presents a bull-case target price of $826, a base-case of $415, and a bear-case of $131, with key variables including 2040 vehicle sales, FSD/Network Services attach rate, ARPU, Mobility fleet size, Energy growth, and the probability discount for Humanoids.

• risk_rewardRisk Reward Chart and Options Implied Probabilities

  12-month risk-reward and options-implied probabilities

  The risk-reward chart shows a current price of about $387.51, a target price of $415, a bull-case target of $826, and a bear-case target of $131, while indicating a probability of about 33.1% of exceeding $415 and about 4.1% of falling below $131.

• forecastingMorgan Stanley ModelWare

  internal forecasting model framework

  The report states that unless otherwise noted, key metrics are based on the Morgan Stanley ModelWare framework, and it lists differences between MSe and consensus for 2026/2027 revenue, gross profit, EPS, FCF, and capex.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Tesla Inc (TSLA.US)

  research target

  Strengths

  It has integrated platform potential across EVs, autonomous driving, energy, robotics, and embodied AI infrastructure; Network Services, Tesla Mobility, and Humanoids provide long-term valuation optionality.

  Weaknesses

  Near-term capex is rising and FCF consumption is widening, the rollout pace of robotaxi and Unsupervised FSD is slower than investors expected, and the auto business still faces competition and margin pressure.

  Comparison

  MSe's 2026 revenue, deliveries, and FCF are below consensus, but its automotive gross margin forecast is above consensus; risk-reward is relatively balanced, so the Equal-weight rating is maintained.

  Risks

  Execution risk in robotaxi/FSD/humanoids, regulatory risk, China market risk, dilution risk, and valuation risk.

• Tesla Auto

  price target component

  Strengths

  Base-case valuation of $45/share; base-case assumptions are 8.5mn units in 2040, 9.7% exit EBIT margin, and a 10x exit EBITDA multiple.

  Weaknesses

  Auto sales volumes and margins are affected by competition, tariffs, and demand pressure, and MSe 2026 auto deliveries are 9.1% below consensus.

  Comparison

  The bull-case assumes 12.0mn units in 2040 and about 14.5% EBIT margin, corresponding to about $81/share; the bear-case assumes 5.6mn units and 4.5% EBIT margin, corresponding to $20/share.

  Risks

  Competition from traditional OEMs, Chinese players, startups, and major technology companies, as well as underperformance in new model launches and cost milestones.

• Network Services

  price target component

  Strengths

  The report calls it the 'crown jewel,' with a base-case valuation of $145/share, assuming an 80% attach rate and $240/month ARPU in 2040.

  Weaknesses

  Value realization depends on FSD/software services attach rate, users' willingness to pay, and the maturity of autonomous driving functions.

  Comparison

  Bull-case valuation is $251/share, assuming a 90% attach rate and about $250/month ARPU; bear-case valuation is $49/share, assuming a 60% attach rate and $200/month ARPU.

  Risks

  Delays in the rollout of Unsupervised FSD, HW3 technical limitations, regulatory approval, and slow progress in safety validation.

• Tesla Mobility / robotaxi

  price target component

  Strengths

  Base-case valuation of $125/share, with long-term benefit from the robotaxi network, vehicle scale, and revenue-per-mile model.

  Weaknesses

  The pace of city expansion is constrained by strict validation, while investors have high expectations for rollout speed; HW3's inability to support Unsupervised FSD weakens the value of the existing fleet.

  Comparison

  Bull-case valuation is $206/share, assuming an 8.5mn fleet and 30% EBITDA margin in 2040; bear-case valuation is $43/share, assuming a 2mn car fleet and about 40% EBITDA margin.

  Risks

  Slow city rollout, safety data failing to meet standards, NHTSA data performance lagging Waymo, geofencing limitations, and regulatory resistance.

• Tesla Energy / ESS

  price target component

  Strengths

  Base-case valuation of $40/share, with long-term demand still viewed positively, and energy gross margin in 1Q26 significantly better than both MSe and consensus.

  Weaknesses

  Near-term project timing noise is high, and the report lowers 2026 ESS deployments to 55.3 GWh.

  Comparison

  Bull-case valuation is $81/share, assuming 15-year revenue CAGR of 21% and terminal EBIT margin of 21%; bear-case valuation is $18/share, assuming 15-year revenue CAGR of 14% and terminal EBIT margin of 13%.

  Risks

  Timing of large project deliveries, tariff impact, competition, and growth pace falling below the roughly 50% year-over-year growth rate in 2025.

• Humanoids / Optimus

  price target component

  Strengths

  Base-case valuation of $60/share, already including a 50% probability discount; bull-case is $208/share with no probability discount.

  Weaknesses

  Commercialization remains at an early stage, and value is highly dependent on technology maturity, cost reduction, scalability, and validation of use cases.

  Comparison

  In the bear case, the Humanoids valuation is $0/share, showing high uncertainty in the option value of this segment.

  Risks

  Execution risks in R&D, mass production, cost, demand, and safety.

• Terafab / semiconductors

  capability buildout and long-term infrastructure

  Strengths

  Plans to build a $3bn research fab at Giga Texas covering logic, memory, and packaging, and to use Intel 14A process technology to support its compute and embodied AI roadmap.

  Weaknesses

  Requires high capital investment in the short term, with a long payback period.

  Comparison

  Compared with traditional automakers, Tesla is attempting to gain deeper control over AI compute and semiconductor-related infrastructure.

  Risks

  Difficulty of semiconductor manufacturing execution, uncertain investment returns, changes in technology pathways, and cash consumption.

Key data

• RatingEqual-weightThe report reiterates an Equal-weight rating.
• Price target$415.0012-18 month price target, based on SOTP.
• Current share price$387.51Closing price as of Apr 22, 2026.
• Implied upside+7.09%The upside of the target price versus the current price in the risk-reward chart.
• 2026 capex guidance>$25bnThe company raised its full-year capex guidance from >$20bn to >$25bn.
• Morgan Stanley 2026 capex forecast$26.1bnThe report raises its 2026 capex forecast to $26.1bn.
• Morgan Stanley 2026 FCF forecast-$11.6bnThe report expects about $11.6bn of free cash flow consumption in 2026.
• 2026 EPS forecast$1.85MSe 2026 non-GAAP diluted EPS, below consensus of $1.91.
• 2026 Auto deliveries forecast1,522,160MSe is 9.1% below consensus of 1,674,753.
• 2026 Storage deployed forecast55.3 GWhReduced from the previous 59.1 GWh; about +18% year over year.
• Estimated HW3 vehicles~3.5mnAbout ~40% of the existing Tesla car parc.
• Estimated existing Tesla fleet~8.5mn vehiclesUsed to assess the impact of HW3's inability to upgrade on the robotaxi thesis.
• Research fab investment$3bnTesla will build a research fab at Giga Texas involving logic, memory, and packaging, and plans to use Intel 14A process technology.

Impact & implications

The implication for investors is that Tesla's long-term valuation increasingly depends on option value from embodied AI, autonomous driving services, robotaxi networks, and Humanoids, rather than simply auto unit sales. However, these projects have longer commercialization validation cycles and require higher capital expenditures, so near-term financial performance will be affected by FCF consumption and margin pressure. If there are positive breakthroughs in safety data, city expansion, FSD attach rate, and hardware upgrade pathways, the upside to valuation could be substantial; if execution on robotaxi/FSD/Optimus is slower than expected or competition and regulatory pressure intensify, the current valuation may face compression.

Risks

• Rising capex and widening FCF consumption may weigh on near-term valuation and financial flexibility.
• The commercialization pace of robotaxi, Unsupervised FSD, and Optimus may be slower than investors expect.
• HW3's inability to support Unsupervised FSD may reduce the value of the existing fleet in the future of autonomous driving.
• The auto business faces competition from traditional OEMs, Chinese players, startups, and major technology companies.
• Regulatory and safety validation may limit the expansion speed of robotaxi/FSD.
• China market risk, dilution risk, and high-valuation pullback risk.
• The energy storage business is affected by customer project timing, and near-term growth may be slower than previously expected.
• The 1Q26 gross margin improvement included one-off factors and may not be sustainable going forward.

What to watch

• The number and pace of new robotaxi city launches, and whether the directional goal of 'a dozen or so states by YE' is achieved.
• Changes in Tesla's safety metrics in monthly NHTSA data, and comparisons with Waymo's early rollout pace.
• Whether Unsupervised FSD launches as planned in 4Q26 first within specific geofenced operating domains.
• HW3 user compensation or upgrade plans, and the impact on the value of about 3.5mn HW3 vehicles.
• Whether 2026 capex remains above $25bn, and whether FCF burn approaches Morgan Stanley's forecast of $11.6bn.
• Auto gross margin ex-credits, Energy gross margin, and true margins excluding tariffs and one-off items.
• Whether ESS deployments reach the forecast 55.3 GWh, and the pace of large project deliveries.
• Progress in Terafab construction, cooperation on Intel 14A process technology, and support for the AI/semiconductor roadmap.
• FSD attach rate, Network Services ARPU, progress on Optimus, and new models such as Roadster and multivan.