Tesla: A Long-Term Owner's Perspective

Of course, I know nobody values Tesla on its 2025 earnings. The market stopped treating it as a carmaker long ago. What the market is buying is Robotaxi, FSD, Optimus, the AI money-printing machine still sitting unbuilt inside Musk's head.

But that's exactly the problem.

If you're like me, an old grump who bets by reading the ledger, your first reaction to a 354x P/E is not "this company's future must be incredible."

Your first reaction is: how incredible would it have to be to deserve 354x?

This piece is here to answer that question. I'm treating Tesla as a business and going through it from the plainest angle there is: if I buy in today, how much real cash can I take out of this business over the next ten years? Let me say the conclusion up front, and it's a buzzkill. At its current price, my rating on TSLA is Avoid.

That's not to say Tesla is a bad company.

Quite the opposite, Tesla is a good company.

What's bad is the price.

……

Let me put the least pleasant part first.

First, Tesla's cars themselves are not as remarkable as the company's own narrative makes them out to be. In 2025, automotive sales revenue was $65.821 billion, down 9% year over year. One of the reasons the company gives in its own annual report is "lower ASP and higher customer incentives."

ASP stands for Average Selling Price. "Lower ASP" in plain language means the cars sold cheaper. "Higher customer incentives" in plain language means the discounts got more aggressive. Stack the two together and it reads: price cuts and promotions.

Carmakers that cut prices and run promotions are everywhere. No one has ever handed a discounting carmaker a 354x P/E.

Second, of that $3.794 billion in 2025 net income attributable to shareholders, regulatory credit revenue accounted for $1.993 billion, roughly 53%.

Regulatory credits are quotas that other automakers buy from Tesla to avoid environmental penalties. They have nothing to do with "how good Tesla's cars are" and everything to do with "how slow other U.S. automakers are on electrification." This money will shrink sooner or later, because other carmakers' electrification will eventually catch up; this money could also shrink overnight if policy changes with a stroke of the pen.

Set that $1.993 billion aside and look at the $3.794 billion again, and you'll find that Tesla, a company with a $1.36 trillion market cap, made only a little over $1.8 billion in "clean profit" in 2025.

$1.8 billion holding up $1.36 trillion. I can no longer tell whether this is a P/S or a P/E.

Third, the company expects 2026 capital expenditures to exceed $25 billion.

For reference, actual 2025 capex was $8.527 billion.

In other words, Tesla is telling you that next year it plans to spend roughly three times what it spent this year. That money goes into AI compute, data centers, capacity expansion, and R&D production lines.

Let me tell you, this is neither bad nor good.

Not bad, because Tesla currently has $44 billion in cash sitting on its balance sheet, so it can afford to spend $25 billion; and because Tesla has historically proven that its efficiency at burning cash really isn't poor.

Not good, because after that $25 billion goes in, the shadow of a return hasn't appeared yet. Robotaxi has only just started running unsupervised in two or three cities in Texas, Optimus is still moving boxes around in the factory, and the AI chip doesn't even have a competing product out yet.

The vision is grand. The books haven't been settled.

But the share price has already been set as if the books were settled.

……

At this point, I need to talk about what Tesla's business actually is. A lot of people skip this step, because everyone thinks they already understand it. People who drive have ridden in a Tesla, same as a cab driver; people who follow the news can chat for a couple of minutes if you toss out the three words FSD, Robotaxi, Optimus.

But if you really take it apart, you'll find Tesla is two companies tangled together.

The first is the Tesla that already exists. It sells cars, sells energy storage, sells services, sells software subscriptions. In 2025 the revenue mix looked like this: automotive sales $65.821 billion, regulatory credits $1.993 billion, automotive leasing $1.712 billion, services and other $12.530 billion, energy generation and storage $12.771 billion.

This company you can see clearly. It's a heavy-asset manufacturer with gross margins around 18%, roughly equivalent to "a carmaker the industry prefers plus a decent energy business."

The second is the Tesla that doesn't exist. Robotaxi, FSD, Optimus, AI infrastructure, the platform-type cash flows of the next ten years.

This company you can't see at all. It lives in the slide decks, in Musk's tweets, in analysts' models for 2030. Its revenue today is 0, its revenue over the next ten years is X, and what X equals nobody knows.

Of the $1.36 trillion valuation the market assigns to Tesla, the overwhelming majority is set on the second company.

The first is falsifiable, the second is arguable.

That's how arguable things are. You say it can be worth $5 trillion, I say it can be worth $50 billion, and neither of us can pin the other down. All you can do is wait for time to pass and see whose script turns out right.

This is the essential contradiction of Tesla's business: what the 【fact】 tells you and what the 【opinion】 tells you differ by more than tenfold.

My own judgment is that the business is 3/5 on comprehensibility. The first half can be understood, the second half cannot.

……

At this point, there's no getting around those two words: "the moat."

The moat is Buffett's favorite topic. He bought Coca-Cola because Coca-Cola has a brand moat; no one in the world can build another Coca-Cola. He bought American Express because the Amex card has a network-effect moat.

So does Tesla have a moat?

It does, but not the kind that crushes on a single point. Tesla doesn't have a "one trick wins it all" moat; its moat is the kind you "accumulate and use together."

The brand counts for something, one of the strongest EV brands in the world, though the premium has been ground down a notch by the price war. Scale counts for something, with 1.6361 million vehicles delivered in 2025 and 9.2 million cumulatively. Vertical integration counts for something, with batteries, in-house chips, and the operating system all done internally. The charging network counts for something, with 8,463 Supercharger stations and 79,918 charging connectors, which BYD does not have. Real-world driving data counts for something, with roughly 1.28 million active FSD subscriptions generating an enormous volume of video and road data every day.

Pile these five or six things together, and they really do form a composite barrier that's "not easy to replicate."

But there's one thing to watch.

The core automotive moat is narrowing, while the optionality moat in data, charging, software, and robotics is still widening.

These two things are happening at the same time.

The evidence that the automotive moat is narrowing was already mentioned: automotive sales revenue fell 9% in 2025.

The evidence that the new moat is widening goes like this. In 2025, services and other revenue grew 19%, and the energy business grew 27%. By the first quarter of 2026, services revenue grew 42% year over year, and the energy business gross margin worked out to nearly 39.5%.

You see, the old engine is sinking and the new engine is lifting.

The problem is, nobody knows how far the new engine is from running at full speed. It can take off now because the base is still very low; services revenue is $12.5 billion a year and energy is $12.7 billion, $25 billion combined, but Tesla's total revenue is $94.8 billion. The fast-growing part is not yet big enough to support a $1.36 trillion valuation.

It's like a 50-year-old boxer whose left hand used to be powerful but is now shot, and whose right hand he's only just started training, looking talented, but whether he can fight a championship bout won't be known until he's 60.

Whether you're willing to bet right now on his championship purse at 60 is a matter of personal taste.

But you have to know what you're betting on.

Moat strength score, 3/5.

……

Before talking about management, let me say one thing.

My view of Musk is probably different from a lot of people's.

I'm neither a Musk fan nor a Musk hater. My judgment of him is that he's one of the rarest figures in human history over the past twenty years, an engineer plus a salesman who can build rockets and electric cars at the same time.

The engineer part is real skill. SpaceX's reusable rockets, Tesla's vehicle manufacturing, Starlink's operation, put any other CEO in charge of these and they most likely couldn't pull them off. This is the most valuable asset at the company called Musk.

The salesman part is also real skill. One tweet can blow the share price up, one launch event can make investors forget about capex, and that's an ability only 1% of CEOs have.

But as an investor, the most uncomfortable part is the other side.

The 【fact】 is that Tesla's 2025 CEO Performance Award granted Musk 423,743,904 restricted shares, with a disclosed maximum grant-date fair value as high as $132.299 billion.

$132.299 billion in equity compensation.

This number is so large you need to stop and read it character by character.

It means that if Tesla hits a series of performance targets in the future, the company will hand out $132.299 billion worth of new stock to one person, Musk. Where does that stock come from? From the pockets of all shareholders, because every new share issued dilutes existing holders' ownership by a notch.

Add to that the 303,960,630 shares tied to the 2018 CEO award, fully achieved and exercisable. This is the largest equity grant in history, bar none.

And there's more. In the first quarter of 2026, Tesla invested $2 billion in SpaceX common stock, and in the same quarter Tesla sold $87 million of Megapack to SpaceX. The two companies share the same boss. $2 billion goes out, $87 million comes back. Whether the flow of money in between is "strategic cooperation" or "a related-party transaction" depends on which side you stand on.

Let me tell you, this is not the question of "whether Musk is a good boss," this is the question of "whether Tesla's governance structure is friendly to minority shareholders."

When Buffett buys a company, one implicit premise is that the CEO treats shareholders as partners. Munger said something pretty harsh in an interview once: "Assume every CEO wants to steal from you, and your investing gets a lot safer."

I'm not saying Musk is stealing. I'm saying Tesla's governance structure is not consistent with what conservative value investors prefer: "simple, transparent, few related-party transactions, little share dilution."

Management and capital allocation score, 2/5.

……

On the financials, let me put the good news first.

Tesla's balance sheet is very healthy.

As of the end of 2025, cash, cash equivalents, and short-term investments totaled $44.059 billion, total debt and finance leases came to roughly $8.376 billion, and net cash was about $35.7 billion. By the first quarter of 2026, cash and short-term investments had risen further to $44.743 billion.

So Tesla's ability to survive an industry downturn is clearly stronger than most of its peers. Ford and GM don't necessarily have this much net cash.

This really is strong.

But the good news is over.

Below is the bad news, or you could call it "the easily overlooked detail."

If you only look at the P/E, you'd think Tesla's profit decline was a one-year affair in 2025. Look at the actual data, and it goes like this.

2022 operating income $13.656 billion, operating margin 16.8%. 2023 $8.891 billion, operating margin 9.2%. 2024 $7.076 billion, operating margin 7.2%. 2025 $4.355 billion, operating margin 4.6%. Q1 2026 $941 million, operating margin 4.2%.

Operating margin slid all the way from 16.8% to 4.2%.

This isn't a one-year fluke, it's a four-year trend in a row.

And what kind of level is a 4.2% operating margin? It's a fraction of Toyota's. Toyota's operating margin in fiscal 2024 was around 11%, BYD's around 6%, Honda's 6%–7%.

Tesla's operating margin is now below the industry average.

This is a "once high-margin growth stock" turning into a "below-average-margin carmaker."

But the P/E is still 354x.

Can you feel that something's off?

Earnings quality also has to be taken apart. In 2024, net income attributable to shareholders was $7.091 billion, down sharply from 2023. But that high 2023 profit of $14.997 billion included $6.54 billion from the release of a U.S. deferred income tax valuation allowance. That release is a one-time accounting move with nothing to do with operations. Strip out that $6.54 billion, and 2023's actual operating net income was roughly $8.5 billion.

That's a fair comparison baseline.

ROE tells the same story. In 2022, ROE peaked around 30%; by 2025 it had fallen to the mid single digits. ROIC declined about the same. The marked retreat in return on capital is the most important fact of the past few years, and the one the share price has masked.

Finally, look at cash flow.

Tesla's 2025 operating cash flow was $14.747 billion, capex $8.527 billion, free cash flow $6.220 billion.

$6.2 billion against a $1.36 trillion market cap is a free cash flow yield of about 0.45%.

If you then add stock-based compensation back as a real cost, 2025 SBC was $2.825 billion. "Conservative owner earnings" = 14.747 − 8.527 − 2.825 = $3.395 billion.

Against a $1.36 trillion market cap, that's roughly a 400x P/E on the conservative basis.

If you put $1 into Tesla today and run it honestly, your first-year return is 0.25 cents.

Not 2.5 cents. 0.25 cents.

And the yield on the U.S. 10-year Treasury on 2026-05-15 was 4.59%.

So you don't need to buy Tesla. Just buy U.S. Treasuries, and your first year gets you 4.59 cents.

Buying Tesla is betting that ten years from now those 0.25 cents can rise above 4.59 cents.

Whether that bet is right, weigh it in your own mind.

……

On valuation, you need something to compare against.

Looking at a 354x P/E in isolation is actually meaningless, because Tesla is a unique species. But compare it with peers and you can feel just how strange the number is.

BYD's current forward P/E is about 20.55x, P/S about 1.06x, P/B about 3.05x. GM's P/E is about 27.3x. Ford's P/E is negative, it's running a loss.

Tesla's P/E is 354x, P/S about 14.4x, P/B about 16.6x, P/FCF about 220x, EV/adjusted EBITDA about 91x.

Put these companies in one table, and the valuation gap between Tesla and BYD isn't 10x or 20x, it's a 17x P/E versus a 354x P/E, a 1x P/S versus a 14x P/S.

This isn't "Tesla is more expensive than BYD," this is "Tesla and BYD simply aren't the same species."

The market is valuing it as the OpenAI type of company. But OpenAI at least has ChatGPT, with 500 million monthly active users. And Tesla's Robotaxi monthly actives? Austin, Dallas, Houston, three cities, currently in pilot.

You might say, valuation just reflects the future.

Fine.

But how much future it reflects, and how far into the future, has limits.

I ran a simple DCF. Three scenarios.

Conservative scenario. Starting owner earnings of $3.5 billion, 12% compound growth over the first ten years, a 10% discount rate, 3% terminal growth. The result is a fair value of roughly $35–60 per share.

Neutral scenario. Starting owner earnings of $5 billion, 18% compound growth over the first ten years, a 10% discount rate, 3% terminal growth. The result is a fair value of roughly $70–140 per share.

Optimistic scenario. Starting owner earnings of $8 billion, 25% compound growth over the first ten years, a 10% discount rate, 4% terminal growth. The result is a fair value of roughly $220–350 per share.

Note that 25% annualized growth over the next ten years in the optimistic scenario is already a very bold assumption. Fewer than 1% of public companies compound at 25% for ten years straight.

But even in this bold optimistic scenario, the upper edge of the corresponding fair value only reaches $350. At the current $422, it's still 20% higher.

To make $422 reasonable, you'd need Tesla to compound at 30%–35% or more over the next ten years, and ultimately deliver it as a high-quality, sustainable, high-margin cash flow platform.

In the end it comes down to this: Robotaxi and Optimus not only succeed, but succeed enormously.

Can this bet win? It can.

But value investors don't bet at these odds.

Ideal buy price range, $70–140 per share.

Current price, $422.

Margin of safety, none.

That's the conclusion.

……

Having finished valuation, let me talk about risk.

Tesla's biggest risk isn't "how many cars it sells next year." Tesla's biggest risk is the long-term gap between the narrative and the real cash flow.

If Robotaxi doesn't pan out in three years, FSD penetration stalls, and Optimus is still moving boxes, then the $25 billion of capex in 2026 goes in and you hear no sound. The market will re-grade Tesla.

On that day, Tesla won't turn into a bad company. Tesla will turn into a good carmaker plus a decent energy company.

But its corresponding fair valuation will drop from $1.36 trillion to, perhaps, $300 billion to $500 billion.

What does $300 billion to $500 billion feel like? It's a 60% to 80% drop from the current share price.

I'm not talking about a black swan. I'm talking about that one thing: "the market re-rates."

In March 2026, NHTSA upgraded its investigation into Tesla's FSD to an engineering analysis, covering roughly 3.2 million Teslas equipped with FSD. Once a major accident, recall, or expansion blockage occurs, the most expensive part of the valuation gets hit first.

Add the dilution mentioned earlier. The $132.299 billion equity grant is a chronic bleed. Add SpaceX's related-party transactions and the $25 billion of capex in 2026.

Each of these things taken alone is not the end of the world.

Taken together, it's a cup of coffee with three spoons of sugar, two spoons of salt, and one spoon of chili. Each spoon is normal; mixed together it's undrinkable.

……

At this point, I need to look at it from a different angle.

Let me tell you about a judgment I've always thought is very important, called "opportunity cost."

Buffett has said this repeatedly, more or less: in every investment decision, your opponent isn't the market, it's your next best alternative.

If you have $1 million on hand right now and you buy Tesla, you aren't buying Tesla, you're buying "Tesla minus everything else you could have bought."

So let's look at the other things you could buy.

Option one, the U.S. 10-year Treasury. A yield of 4.59%, risk-free, the money is right there. 4.59 cents a year.

Option two, an S&P 500 index ETF. Long-term annualized returns of roughly 8%–10%, manageable volatility, diversified, no thinking required. About 8–10 cents a year.

Option three, BYD. A forward P/E a touch above 20x, 2025 net income of 32.6 billion RMB, and global EV shipments that already exceed Tesla's. Of course BYD has its own troubles too: the price war, declining margins, overseas policy risk. But a 20x P/E has already priced all of that in. A reasonably valued industry leader.

Option four, Tesla. A 354x P/E. What you're buying is a bet on the next ten years that "Robotaxi must succeed plus Optimus must succeed plus AI must succeed." If all of them succeed, the return could be decent. If even one fails, a 60%–80% drawdown is possible.

Put these four options side by side and tell me which one a value investor should choose.

This is actually a gimme question.

……

At this point, I'm about done.

But I still have to say one stretch about my own feelings, because it bothers me not to.

Before writing this, I hesitated for a long time. Tesla as a stock carries far too much love and hate on the internet. Those who love it think it's the greatest company of the 21st century; those who hate it think it's an epic bubble.

Slow as I am, I'm actually on neither side.

The way I feel about Tesla is that this is a great company, but a great company doesn't equal a great investment.

Coca-Cola is a great company, but Coca-Cola in 1998 was not a great investment. People who bought Coca-Cola in 1998 were underwater for 15 years before breaking even. Cisco is a great company; Cisco in 2000 was not a great investment, and people who bought Cisco in 2000 still haven't broken even to this day.

It's not that the company is no good, it's that the price is no good.

Buffett has another line: "It's far better to buy a wonderful company at a fair price than a fair company at a wonderful price."

Note the words in the middle, "a fair price." Not any price.

Tesla at $422 is not a fair price.

Tesla at $70–140 is a fair price.

Between those two ranges, the difference is 3 to 6 times.

The reason for such a big difference isn't that the company changed, it's that market sentiment changed. Market sentiment can turn the same company from $70 into $422, and it can turn the same company from $422 back into $70.

What a value investor must do is not get swept up by these two tides of sentiment.

Don't chase when it climbs to $422, don't fear when it falls to $70.

Do the math honestly, and wait honestly.

Time won't fix a valuation error on its own. But time will let you see clearly who is swimming naked.

……

Finally, let me lay out the tracking signals, for those who already hold it.

If you're already on board, what you should keep your eyes on isn't the share price, it's these few numbers.

First, automotive gross margin, automotive gross margin excluding regulatory credits. This is the most important. If this number keeps sliding, the story can't hold up.

Second, FSD paid subscriptions and penetration rate. This is the key metric for the software narrative.

Third, Robotaxi's city expansion, paid miles, accident rate, and unit economics. This is the key to the most expensive part of the valuation.

Fourth, energy business revenue growth, segment gross margin, and GWh deployed. This is the key to the new engine.

Fifth, operating cash flow, capex, free cash flow. This is the key to the "real cash."

Sixth, annual share count change, diluted shares, and the scale of SBC. This is the key to "whether the share I'm getting has actually been diluted."

If all of these numbers move in a good direction, your story is still intact; if two or three start to weaken, then it's time to reassess.

……

By the time I'd written this far, I realized I'd already written more than 4,000 words.

Closing the laptop, I actually thought of something else.

Before the internet bubble burst in 2000, there was a company called Cisco.

At the time Cisco was the most important company in the world, the "plumbing and power grid of the internet age." Its share price rose 750x from 1990 to 2000, its market cap briefly broke $550 billion, and its P/E topped 200x.

Everyone said Cisco was the greatest investment of the 21st century.

26 years have passed. Cisco is still around, and Cisco's products are still in use. The internet revolution did indeed happen, and more thoroughly than anyone expected at the time.

But the people who bought Cisco at the 2000 high.

Still haven't broken even to this day.

I say this not to suggest Tesla will become Cisco. Tesla and Cisco are two completely different companies, with different eras, business models, and moats.

I say this because between "an era of change" and "making money in that era," there is always a valuation in the way.

AI will certainly arrive, and autonomous driving will most likely arrive too.

But that has no necessary connection to "buying TSLA at $422 makes money."

Price decides everything.

……

Finally, three ratings.

【Investment Rating】Avoid.

【One-Sentence Thesis】 Tesla is a business with real strategic assets and enormous optionality, but the current price has already prepaid too much for "large-scale success in autonomous driving, Robotaxi, and robotics," and lacks the margin of safety value investing requires.

【Fair Buy Price】 $70–140 per share. The current $422 is clearly overvalued.

If one day Tesla falls below $140, I'll reopen this research and run all the numbers again.

But not today.

Today I wait, and I don't chase the price.

One last word. We can't control the market's emotions; the only thing we can control is the size of our own bets. A person who has thought through how much they're willing to pay always outlives the person led around by the price.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.