Tesla’s promises need a price tag

A Tesla valuation can begin with a car and end with a robot workforce. Somewhere in the middle, the investor needs to write down how much money each business could make.

That step is easy to skip. A large addressable market sounds impressive. A product demonstration is memorable. Neither tells you what a shareholder earns after operating costs, investment, competition and the years spent waiting.

Our view: a serious Tesla bull case should put separate numbers on the businesses it expects to succeed. If the return only works when every ambitious project works on schedule, the investor should know that before buying.

In Theory A, a multiple is an assumption you can move. Set a lower P/E in the chart’s multiplier controls and see how much the earnings line moves away from market value. Then raise it. Neither setting proves Tesla is cheap or expensive; the exercise makes the price assigned to future growth visible.

Separate the reported history from the analyst-estimate portion of the chart. Estimated earnings are one possible path. The robotaxi case still has to explain what could make earnings exceed that path, and what investment would be needed.

Start with the business that files an income statement

Tesla reported $94.827 billion in revenue for 2025. Automotive revenue was $69.526 billion; energy generation and storage contributed $12.771 billion. The latter grew while automotive revenue declined. Those are different businesses moving at different speeds. Tesla’s annual filing provides the starting point.

It would also be wrong to describe all of Tesla’s newer businesses as imaginary. Its Q3 2026 filing describes a Robotaxi service launched in June 2025 and a Cybercab rollout in September 2026. A service launch establishes that something exists. It leaves open the harder valuation questions: utilization, cost per mile, fleet investment and sustainable profit.

For the established operations, build a range around vehicle volumes, pricing, margins and storage demand. For autonomy and robotics, make the uncertain assumptions visible. Separate the value of an operating business from the value assigned to a business still proving its economics.

What earnings would justify the price you pay?

Consider a hypothetical purchase at 100 times annual earnings. Assume earnings per share grow 25% a year for five years, and investors pay 30 times earnings at the end. Earnings more than triple. The share price still ends about 8% below the purchase price.

The arithmetic is simple: 1.25⁵ × 30 ÷ 100 ≈ 0.916. Those multiples are illustrative, not a statement of Tesla’s current valuation. The example shows how a business can make enormous progress while its shareholders wait for the earnings to catch up with what they paid.

At those entry and exit multiples, earning a 10% annual price return requires roughly 40% annual EPS growth for five years. That is the hurdle created by the assumptions. Decide which businesses could supply those earnings, and how much financing they would need along the way.

Give the robotaxi thesis a unit of measurement

A useful autonomy model can begin with paid miles per vehicle, revenue per mile and the costs of delivering those miles. Add insurance, cleaning, maintenance, charging, remote assistance and depreciation. Then decide who owns the fleet and who pays to expand it.

Fleet ownership changes the economics. Selling software to someone else’s vehicles and financing a large vehicle fleet can produce similar-looking revenue forecasts with very different capital requirements. A valuation should specify which version it assumes.

Do the same for robotics: units, price, service costs, margins and the time needed to reach volume. An attractive outcome deserves a probability, not an automatic place in the base case. Avoid counting the same economic benefit twice across vehicles, software and the network.

The bull case deserves more than a car-company multiple

The strongest objection is fair: valuing Tesla solely on today’s vehicle earnings could miss a successful transition into much larger profit pools. A new business can look immaterial in historical accounts immediately before it becomes important. Historical P/E alone cannot settle that debate.

Nor is 30 times earnings a rule for Tesla’s future. It is one assumption to challenge. A durable, fast-growing stream of profits might warrant a higher multiple. The burden is to explain its durability and growth without assuming the same exceptional growth forever.

This is why a single precise fair-value number can give more comfort than insight. A range makes disagreements useful: one investor expects higher fleet utilization; another expects lower margins or a slower rollout. Both can see what those differences do to value.

What would make the promises easier to underwrite?

Repeated evidence of profitable paid usage, economics that remain attractive outside the first deployment area, and operating cash generation after the investment needed to grow would strengthen the case. Expansion that requires more capital than expected, persistent delays or dilution that absorbs the earnings growth would weaken it.

The practical question behind “is Tesla overvalued?” is whether a reasonable set of business outcomes can deliver your required return from your purchase price. Keep the ambition in the model. Give the costs, time and failure cases a place beside it.

For the other side of this problem, read how an excellent established business can still disappoint at the wrong entry price.