How to Strip a Stock to Its Cash Bones

Filings first. Cash over stories. A repeatable way to read any company.

I don’t buy stories. I buy cash. Tesla is the perfect stress test: price cuts pulled margins into the teens, free cash flow stayed positive but thinned during a capex binge, and the stock still asks you to pay for an autonomy payday that isn’t cash yet. This method exists to catch the gap between what the business is and what the price assumes.

Filings, not marketing

I start with filings, not marketing. The 10-K and 10-Q are what matter: income statement, balance sheet, cash flow, and the footnotes where the truth hides—stock-based comp, leases, segment economics, debt terms. I check diluted share count to see who’s getting paid and how. Investor decks come after the math, never before.

No priors, on purpose. I don’t start bullish or bearish; I start blind. Filings first, same six steps in the same order, no slideware until after the math. I write a one-sentence bull and a one-sentence bear, name the single metric that would flip me, and pre-commit to changing my mind if it prints. That’s how you trade conviction for discipline.

The order that never changes

I run the same order every time: revenue trend, margins, ROIC, free cash flow, balance sheet, dilution, then valuation. If gross margin falls while units are flat, pricing power slipped. If operating margin falls while gross is steady, costs bloated. ROIC tells me whether each dollar invested clears the hurdle; free cash flow tells me whether any of it is real.

I look at revenue year over year and over two years. Flat or down while discounts get steeper isn’t growth; it’s a demand or pricing problem dressed up as “share defense.” Free cash flow is the truth serum—cash from operations minus capex, tracked as a percent of revenue. Profits up while free cash flow falls usually means working-capital strain or a capex binge; fine in a build-out if it later shows up as higher margins and stronger cash conversion. If it doesn’t, it’s waste.

I track operating cash flow and GAAP earnings together to see whether profit and cash are moving in the same direction or if one is faking it for a quarter or two.

Survivability is next: net cash or net leverage, and whether operating income covers interest with room to breathe. I also watch dilution. Stock-based comp is a silent tax; buybacks that only plug the SBC hole don’t create value. Only then do I talk price. I use the metric that fits the model—EV/EBIT or EV/FCF for asset-heavier names, forward P/E for steadier earners, EV/S for early high-growth with a clear path to cash—and I compare it to the company’s own history and to true peers.

Tesla, read straight

Tesla illustrates the read. Price cuts defended units but bled pricing power; gross and operating margins compressed. ROIC is solid for an automaker but nowhere near software. Free cash flow stayed positive but fell as capex stayed heavy. The balance sheet is fine. The multiple, however, isn’t paying for today’s P&L; it’s paying for robotaxis and scaled autonomy. That’s a legitimate option. It just isn’t cash yet.

Scope, by the numbers. I grade what moves revenue and cash today. In 2024, Tesla’s revenue was roughly 78.9% Automotive, 10.3% Energy, 10.8% Services & Other. In Q3 ’25, the quarter split was about 75.5% Automotive, 12.2% Energy, 12.4% Services & Other. On a trailing-twelve-month basis through 9/30/25, it’s near 74.9% Automotive, 12.5% Energy, 12.6% Services & Other. Two engines—autos + energy—drive almost 87–88% of the top line. Most of the “other businesses” live inside Services & Other (paid Supercharging, used vehicles, insurance, parts and service) and don’t move consolidated results yet. Regulatory credits? Call it ~2.8% of 2024 revenue—helpful, fading, not a pillar.

Materiality rule: if it isn’t broken out and it doesn’t move cash yet, it’s optionality, not part of the base case.

One thing most people miss: energy storage is no longer a rounding error. Deployments keep compounding, factory capacity is scaling, and segment margins are firming. If that mix shift holds, energy becomes the cash buffer that lets Tesla defend share in autos without torching the income statement. I treat it as the second pillar—autos defend; energy pays; software is optionality.

Peers keep you honest

I refuse to let a company live in isolation. I line it up against two or three real peers on the same period and currency. With Tesla, the stack is straightforward: U.S. incumbents with low multiples and lower quality, and BYD—a scaled rival with cost advantages in the largest EV market. If you’re best on margin and ROIC but growth is flattening and you trade at the richest multiple, gravity kicks in unless a new engine starts cranking cash. Peer context tells you whether strength is structural or just a moment.

Break your own thesis

Then I tear my own idea apart. I look for concentration anywhere that can hurt—country, customer, product. I ask whether pricing power is rising or eroding. I identify temporary props: subsidies, regulatory credits, one-off benefits, and accounting choices that make reported earnings look better than the cash. I separate maintenance capex from growth capex; maintenance keeps the lights on, growth should lift margins and free cash flow later.

Tesla has obvious fault lines: heavy exposure to China for both production and sales; regulatory credits that help but fade as others catch up; and a gap between reported earnings and cash that must close through real operating leverage, not one-time boosts. None of that kills the long-term story on its own. It just forces you to admit what’s carrying the water today.

Make it falsifiable

I pre-commit to what would flip me. No sermons—just numbers.

Bull side: auto gross margin rises a couple hundred basis points year over year without regulatory crutches, and free-cash-flow margin holds double digits while units grow.
Bear side: units stall and price/mix stays under pressure for two quarters.

If those prints land, I change my mind.

Translate the math to English

Quality is ROIC versus the cost of capital—creator, neutral, or destroyer. Trajectory is the direction of revenue and margins—accelerating, flattening, or compressing. Price is the multiple versus history and peers—cheap, fair, or expensive for what you’re getting.

For Tesla, my read is plain: a very good automaker that built scale and profitability, now dealing with cooled growth and compressed margins while the stock still asks you to pay for an autonomy outcome. The base case is a strong car company with tech optionality whose multiple drifts toward whatever the operating line earns. The bull case requires real autonomy progress and scaled energy storage that show up in margin and ROIC, not in slideware. The bear case is continued pressure from China and lasting price cuts that keep margins stuck while the market eventually demands a lower multiple.

Four outcomes, one gut-check:

• Autonomy disappoints and core auto grinds → multiple drifts toward great-car-company territory.
• Autonomy disappoints but energy scales → mix keeps margins afloat; muddle through at a fair price.
• Autonomy hits while autos stay healthy → you finally earned the premium.
• Both hit at once → you weren’t paying up—you were early.

Bottom line (and your turn)

The edge isn’t a hot take; it’s a repeatable read that turns filings into a yes/no with a clock on it. My call here: a very good automaker with real energy momentum, priced for software-level outcomes.