——
Level — · Streak —
— XP
Mastery—weighted
Quizzes 30d—last 30 days
Cert progress—active level
Rank—global
Loading...
The page behind this dialog is live. Create a free account or sign in and you'll land right back on it.
The professional's preferred metrics
P/E is the most popular valuation multiple, but it's not the most rigorous. Professional analysts and investment bankers rely heavily on for cross-company comparisons because it normalizes for capital-structure and tax differences that distort P/E. Other multiples — Price-to-Sales, Price-to-Book, Price-to-FCF — fit specific business types better than P/E. Knowing which multiple to use when is one of the highest-leverage analytical skills.
Enterprise Value (covered in m0_l3) = Market Cap + Total Debt − Cash. It captures the total price an acquirer would pay to take the business private. EBITDA = Earnings Before Interest, Taxes, Depreciation, Amortization — operating income before financing decisions and non-cash items. EV/EBITDA expresses how much an acquirer is paying per dollar of operating cash flow proxy. Because the metric strips out interest expense (financing) and taxes (jurisdiction-dependent), it makes companies with different debt levels and different tax rates more comparable. Two companies with identical operations but one carrying \$10B more debt will have very different P/Es but similar EV/EBITDA — and the EV/EBITDA is the more economically meaningful comparison.
Different multiples fit different business types. works for profitable, stable businesses. is the cross-comparison default, especially when peers have different leverage. suits unprofitable growth companies (early SaaS, biotech). fits banks, REITs, insurance, and other asset-heavy financial businesses. is arguably the most rigorous single multiple for cash-generative businesses. Match the multiple to the business; using the wrong one produces nonsense.
EBITDA has one important critic worth knowing about. Charlie Munger has repeatedly said that EBITDA is 'BS earnings' because it adds back depreciation and amortization — real costs that companies must eventually replace. A factory wears out; an acquired patent loses value; capitalized software ages out. Treating these as 'non-cash' and adding them back to compute EBITDA flatters profitability for capital-intensive businesses. The correction: in capital-intensive industries (industrials, telecom, utilities), use FCF or EBITDA-minus-Capex rather than raw EBITDA. EBITDA is most useful for asset-light businesses where capex is genuinely small relative to D&A; it's actively misleading for capital-heavy businesses where capex is structurally large. Apply Munger's filter when the multiple is being used in industries where physical assets dominate.
Different multiples sometimes tell different stories. The Valuation tab shows all of them so you can triangulate. When multiple metrics agree, confidence is high. When they disagree, the disagreement is itself information.
| Metric | NVDA (FY2025 ref) | Sector avg | S&P 500 avg | Assessment |
|---|---|---|---|---|
| Forward P/E | ~35x | ~25x | ~20-22x | Premium for growth |
| EV/EBITDA | ~40x | ~18x | ~14x | Expensive vs sector |
| P/FCF | ~45x | ~22x | ~18x | Premium for FCF growth |
| PEG (P/E ÷ growth) | ~1.4 | ~1.8 | ~2.0 | Reasonable for growth rate |
| FCF Yield | ~2.2% | ~4.5% | ~5.5% | Low — priced for high growth |
Private-equity firms transacting on hundreds of buyouts per year use EV/EBITDA as their primary valuation language. The empirical record from buyout transaction databases (PitchBook, Pitchbook-LCD, S&P LCD) shows: large-cap buyouts have transacted at ~10-12x EV/EBITDA in normal markets, dropping to 7-9x in tight credit markets and rising to 13-15x at cycle peaks. Mid-cap deals run 1-2x lower than large-cap. By industry: software has transacted at the highest multiples (15-25x), industrials and consumer at the middle (8-12x), commodity-cyclical at the lowest (5-8x). When a public company trades at an EV/EBITDA materially below its industry's typical buyout multiple, private-equity buyers often start sniffing. The reverse is also true: when public valuations exceed buyout multiples, IPO and SPAC activity peaks (because public markets are willing to pay more than private buyers). Source: PitchBook leveraged buyout data; S&P LCD M&A research; aggregated transaction multiples by industry.
The Valuation tab on every stock page shows P/E, EV/EBITDA, P/Sales, P/Book, P/FCF, FCF Yield, and PEG side-by-side with peer-group and sector averages. The KPIs tab plots each multiple over 10 years so you can see whether current levels are at the high end, low end, or middle of historical range. The /screener page lets you filter by any combination of multiples, useful for cross-multiple consistency checks.
First: using P/E for capital-intensive businesses. Different debt levels and depreciation regimes distort P/E across peers; EV/EBITDA is more comparable but Munger's caveat applies (subtract capex). For real comparison, use EV/(EBITDA − Capex) or P/FCF. Second: using P/Sales without checking gross margin. A company at 5x P/Sales with 70% gross margin is structurally cheaper than one at 5x P/Sales with 30% gross margin. Always pair P/Sales with profitability metrics. Third: using a single multiple to make a decision. The discipline is triangulation across multiple multiples; when they disagree, investigate why before forming a thesis.
I think every time you see the word 'EBITDA,' you should substitute the words 'BS earnings.' What does it really mean? It means that we're not going to count the cost of the depreciating assets. It would be sensible to say EBITDA stands for 'earnings before interest, taxes, and BS earnings.' That's the right way to think about it.