# Relative valuation: multiples, comps, and market signals
On January 20, 2022, Netflix closed at $508 per share, trading at roughly 8.5x forward sales and 45x forward earnings. Six months later, it traded at $174, a 66% drawdown that eventually bottomed near $166 in May. The company's revenue had grown. Its subscriber base, after one bad quarter, was recovering. EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → margins were intact. Yet $200 billion of market capitalization had evaporated.
Nothing fundamental had broken. What broke was the *multiple*. And if you can't explain to your board why a healthy business can lose two-thirds of its value in a quarter without missing a number, you don't actually understand relative valuation, you just know the formulas.
This lesson is about closing that gap.
Discounted cash flow models are intellectually honest and operationally useless in a deal room. When a banker calls at 7 a.m. about a take-private opportunity, no one builds a 10-year before lunch. They pull comps. They benchmark. They speak in multiples, because multiples compress an enormous amount of information (growth, margins, capital intensity, risk, sentiment) into a single number that can be argued about in a 20-minute meeting.
The three workhorse multiples every CFO must master:
A clean multiple requires a clean numerator and denominator. Most aren't. Here's what a serious CFO checks before quoting a comp:
1. Lease-adjusted EV. Post-IFRS 16, operating leases sit on the balance sheet as right-of-use assets and lease liabilities. Some analysts include lease liabilities in EV; some don't. If your comp set mixes the two, your multiple is garbage. For retailers and airlines, this can move EV/EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → by 2-3 turns.
2. Stock-based compensation. Pre-2022, tech analysts cheerfully added back SBC to compute "adjusted EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète →." Post-rerating, the buy side largely stopped. When Snowflake or Palantir quote adjusted figures, institutional investors now haircut them. As a CFO, decide which convention you're using and disclose it.
3. Pillar Two distortions on P/E. Companies that historically benefited from low-tax jurisdictions (Irish-domiciled pharma, Singapore-routed tech) are now seeing effective tax rates normalize toward 15%+. Trailing P/E looks artificially low; forward P/E tells a different story. Always quote forward.
Return to Netflix. In late 2021, the stock traded at roughly 45x forward earnings. The 10-year Treasury yielded 1.5%. By October 2022, the 10-year hit 4.2%. Mechanically, the present value of distant cash flows, and Netflix was very much a "distant cash flow" story, collapsed.
Here's the math executives often miss: a stock trading at 45x earnings is implicitly saying that earnings growth justifies discounting those earnings at a low rate. Raise the discount rate by 250 basis points, and the appropriate multiple doesn't fall by 250 basis points, it can fall by half or more, because the duration of those cash flows acts like bond duration. Long-duration equities behave like 30-year zero-coupon bonds. They got crushed exactly the same way.
Netflix re-rated from ~45x to ~17x forward earnings at the trough. That compression, not earnings deterioration, accounted for the bulk of the drawdown. Spencer Neumann, Netflix's CFO, didn't miss a single guidance number that mattered structurally. The market simply changed the price it was willing to pay for the same dollar of future profit.
The lesson for CFOs: your multiple is a function of macro conditions you don't control, peer behavior you partly influence, and a narrative you absolutely must manage. Confusing these three is how value gets destroyed in IR meetings.
The single most common mistake, and the one bankers exploit when they're selling you a deal, is a lazy comp set. The rules:
1. Business model congruence beats industry labels. Spotify is not a comp for Warner Music, even though both are "music." One is a distribution platform with negative working capitalworking capitalWorking capital is the difference between a company's current assets and current liabilities, measuring short-term liquidity and the funds available to run daily operations.Voir la définition complète → and content licensing costs; the other owns IP with recurring royalties. Their multiples should diverge, and do.
2. Size matters more than you think. Sub-$1B market cap companies trade at structural discounts to large caps in the same sector, often 20-30%. Don't blend them.
3. Growth-adjusted multiples beat raw multiples. EV/EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → divided by growth (the "EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → PEG") normalizes for the fact that a 4% grower and a 25% grower should never trade at the same multiple. If your comp set's average EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → growth is 8% and you're growing at 18%, you deserve a premium, quantify it.
4. Always show median, not mean. One outlier (an acquisition target trading on rumor, or a distressed comp) can shift a mean multiple by 3+ turns.
Vérification des acquis
1. On January 20, 2022, Netflix traded at approximately what forward sales multiple before its 66% drawdown over the following months?
2. Why does the lesson argue that EV/EBITDA has become particularly important under the current global tax regime?
3. According to the lesson, why are DCF models described as 'operationally useless in a deal room'?
4. Select ALL correct answers about the limitations of the P/E multiple as described or implied in the lesson:
Sélectionnez toutes les réponses correctes.
5. Select ALL correct answers about what the Netflix 2022 episode illustrates about relative valuation:
Sélectionnez toutes les réponses correctes.
A multiple is a price. A price is information. The advanced practice of relative valuation is treating the multiple itself as a *signal*, about market expectations, peer positioningpositioningThe mental space you want your brand to occupy in your target customer's mind relative to alternatives.Voir la définition complète →, and your own credibility.
Signal 1: The multiple gap to your closest peer. If you're a software CFO and your nearest comparable trades at 12x revenue while you trade at 7x, the market is telling you something. It's either a growth discount, a margin discount, a governance discount, or a narrative discount. Adobe traded at a persistent premium to Salesforce throughout 2023-2024 not because its growth was higher (it wasn't) but because its FCFFCFFree Cash Flow is the cash a company generates from operations after funding the capital expenditures needed to maintain and grow its asset base.Voir la définition complète → conversion was cleaner. Marc Garrett, finance executives across the SaaS sector took notes.
Signal 2: Multiple dispersion within your sector. When the gap between the highest and lowest multiple in your peer group widens, the market is differentiating. When it compresses, the market is treating the sector as a monolith. In 2022-2023, semiconductor multiples dispersed dramatically: Nvidia detached from the pack and traded at 35x+ forward earnings while Intel sat at 15x. By 2025, with AI infrastructure spend normalizing, dispersion narrowed slightly. The signal: when dispersion widens, idiosyncratic execution matters most. When it compresses, macro dominates.
Signal 3: Implied vs. trading multiple. Reverse-engineer what your DCFDCFDiscounted Cash Flow (DCF) is a valuation method that estimates an asset's value by projecting future cash flows and discounting them to present value using a required rate of return.Voir la définition complète → says your multiple *should* be, given your plan. If your plan implies 15x EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → and you trade at 9x, either your plan is wrong or the market is wrong. Both are CFO problems.
Precedent transaction multiples almost always exceed trading multiples, by 20% to 40%, depending on sector and cycle. This is the "control premium," and it bundles together synergies, scarcity, strategic value, and auction dynamics.
A practical example: when Microsoft acquired Activision in 2023 for $68.7 billion, the deal valued Activision at roughly 19x trailing EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète →. Activision had been trading at ~14x. The 5-turn premium reflected gaming IP scarcity, mobile (King) cross-sell, and Game Pass synergies. As a CFO evaluating a similar transaction, you don't quote the 19x. You quote the *unaffected* trading multiple plus a defensible synergy bridge, and you discount the seller's claim that "comparable deals went for 20x."
A second example, closer to the current cycle: in 2025, Synopsys closed its acquisition of Ansys at approximately 30x EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète →, a multiple that horrified some analysts and was justified by EDA-simulation convergence and AI chip design tailwinds. Sassine Ghazi, Synopsys's CEO, sold the deal to the Street on strategic logic, not multiple discipline. Whether that bet pays off will be the case study of 2027.
Theory is cheap. Here's what you actually do.
Build a one-page dashboard, updated weekly, that shows:
Review this with your CEO monthly. Review it with your board quarterly. When activists or acquirers show up, you'll know exactly where you stand, and where you're vulnerable.
If you trade at a discount, you have three choices: change the business, change the disclosure, or change the comp set the market uses. The third is underrated. When SAP shifted its IR messaging in 2023-2024 to emphasize cloud ARRARRAnnual Recurring Revenue (ARR) is the normalized, predictable revenue a subscription business expects to earn from active contracts over a single year.Voir la définition complète → and RPO, the metrics Salesforce uses, its multiple started compressing toward Salesforce's, not legacy enterprise software's. Dominik Asam, SAP's CFO, didn't just report better numbers; he reframed which numbers mattered.
When your multiple compresses, distinguish for your board between:
These have completely different responses. Sector compression: ride it out, maybe buy back stock