In 1988, Warren Buffett bought $1 billion worth of Coca-Cola stock. It was the largest investment Berkshire Hathaway had made at that point. When asked later what model he used to value the company, he didn't describe a five-year forecast with terminal value assumptions and a WACC calculation.
He described a mental framework for understanding what kind of business Coca-Cola was, what made its earnings durable, and whether the price he was paying was obviously less than what the business would produce over time.
This isn't a story about Buffett being unsophisticated. He understands DCF theory — he studied under Benjamin Graham and collaborated with Charlie Munger, one of the most analytically rigorous investors in history. His explanation for why he doesn't build formal DCFs is the interesting part: they give false precision to inherently uncertain estimates, and the businesses worth owning are usually the ones where the math is so obviously favorable that a precise model isn't necessary.
This article explains the specific framework Buffett uses instead — not as a rejection of financial analysis, but as a more honest accounting of what valuation actually requires.
The Problem With DCF That Buffett Identified Decades Before Everyone Else
DCF analysis has a structural problem that practitioners rarely discuss openly: the output is exquisitely sensitive to the terminal value assumption, which is itself based on a growth rate projection about what happens after the explicit forecast period ends.
In most five-year DCFs, 60-75% of the enterprise value sits in the terminal value. That terminal value is computed from an assumption about what the business earns in Year 6 and beyond — projected from a model that can barely see past Year 2 with any reliability.
The practical consequence: two analysts can build equally rigorous, equally defensible DCF models for the same company and arrive at enterprise values that differ by 50-100%. Not because one made errors — because the inputs that drive the output most significantly (terminal growth rate, WACC, long-run margin profile) are genuinely unknowable with precision.
Buffett's actual objection isn't to discounted cash flow math — it's to false precision:
"I have never seen a spreadsheet that I couldn't manipulate to justify any price I wanted to pay." The DCF framework produces a number that looks authoritative. That number reflects the analyst's assumptions more than it reflects the business's future.
What Buffett replaced it with is a set of questions that force the analyst to develop genuine conviction about business quality before ever worrying about what price to pay.
What Buffett Actually Does: The Four-Question Framework
Buffett's investment process can be reconstructed from shareholder letters, interviews, and Berkshire annual reports. It does not involve a spreadsheet as the primary analytical tool. It involves four questions answered sequentially — and if any answer is unsatisfactory, the analysis stops.
Question 1: Do I understand this business?
Buffett's concept of the "circle of competence" is not humility theater — it's an analytical prerequisite. A business you genuinely understand is one where you can make reliable predictions about how it will perform under different conditions. A business you don't understand produces unpredictable outcomes regardless of how sophisticated your model is.
This is why Buffett famously passed on technology companies during the dot-com bubble. Not because technology businesses can't be valuable — because he didn't have a reliable model for predicting which would win and which would lose. The circle of competence boundary is where analytical confidence ends.
Question 2: Does it have durable competitive advantages?
Buffett calls this the "moat" — the structural reason a business earns returns above the cost of capital and can continue to do so for years or decades. The moat question is more important than any single year's financials, because the moat is what determines whether today's profitability persists.
He articulated this clearly in the 1995 shareholder letter: "The key to investing is not assessing how much an industry is going to affect society, or how much it will grow, but rather determining the competitive advantage of any given company and, above all, the durability of that advantage."
Question 3: Is the management trustworthy and competent?
Financial analysis can tell you what a business has produced. Management quality determines whether that production continues and compounds. Buffett spends significant time evaluating whether management allocates capital intelligently, communicates honestly with shareholders, and has interests aligned with long-term value creation rather than short-term metrics.
Question 4: Is the price sensible?
Only after satisfactory answers to the first three questions does Buffett consider price — and his approach to price is qualitative rather than model-driven. He's looking for situations where the business quality is sufficiently high and the price sufficiently reasonable that a precise model isn't required to conclude the investment makes sense.
He described this as looking for "fat pitches" — situations where the odds are so obviously in your favor that you don't need to know exactly what the ball is going to do.
The Moat: The Concept That Replaces Most of the DCF
If DCF models try to forecast cash flows over time, Buffett's moat framework tries to answer a more fundamental question: what structural reason exists for this business to remain profitable?
The distinction matters enormously. A DCF model can project stable or growing cash flows for a company that actually has no structural protection against competition — and that model will be wrong the moment a well-capitalized competitor decides to enter the market.
The five moat types Buffett and Munger have identified:
1. Intangible assets — brands, patents, licenses
Coca-Cola's brand is worth more than its physical assets. A consumer choosing between an unfamiliar cola and Coke doesn't make a purely rational price/taste calculation — they're buying certainty, familiarity, and the social signal of a known brand. That behavioral reality protects margins and pricing power for decades.
The non-obvious test for brand moat: does the brand allow the company to charge a premium, or does it simply create customer recognition without pricing power? Brand recognition alone is not a moat. Brand recognition that produces premium pricing or preferential distribution is.
2. Switching costs
A business whose customers face high costs — financial, operational, or psychological — to switch to a competitor has structural protection from competition even without the best product. SAP's enterprise software runs critical business processes at large companies. Replacing it means multi-year implementation projects, extensive staff retraining, and significant operational risk. That switching cost keeps customers renewing regardless of incremental product quality improvements by competitors.
3. Cost advantages
A company that can produce the same product as competitors at structurally lower cost will earn higher margins at the same price — or undercut competitors on price and still earn acceptable returns. This moat is most durable when the cost advantage comes from scale, geography, or proprietary process rather than from temporary factors like labor arbitrage.
4. Network effects
A business whose value increases as more users adopt it has a network effect moat. American Express's payment network is more valuable to merchants because more consumers carry it, and more valuable to consumers because more merchants accept it. This reinforcing dynamic creates a competitive barrier that's very difficult to overcome — the competitor must somehow build the network before they can demonstrate its value.
5. Efficient scale
Some markets are only large enough to support one or two players profitably. A new entrant would have to capture enough market share to generate returns — but in doing so, they'd drive down industry-wide returns to the point where the market is unattractive for everyone. This structural dynamic protects incumbents without requiring them to do anything actively defensive.
Owner Earnings: Buffett's Alternative to Free Cash Flow
Buffett introduced the concept of "owner earnings" in Berkshire's 1986 shareholder letter. It's his answer to the question: what does this business actually produce for its owners, net of everything required to maintain its competitive position?
The formula:
Owner Earnings = Net Income + Depreciation and Amortization − Capital Expenditures required to maintain competitive position − Working Capital changes
The critical word: "required."
Standard free cash flow uses total capital expenditure. Owner earnings uses only the maintenance capital expenditure — the spending necessary to keep the business operating at its current competitive level. Growth capital expenditure (spending to expand capacity or enter new markets) is a choice, not a necessity.
Why this matters in practice:
A capital-light business like See's Candies (owned by Berkshire) requires minimal ongoing capital investment to maintain its competitive position. Its owner earnings are very close to its reported earnings. A capital-intensive business like a steel manufacturer requires heavy ongoing investment just to maintain existing production capacity. Its owner earnings may be significantly below reported earnings.
Two businesses with identical reported earnings can have dramatically different owner earnings — and therefore dramatically different intrinsic values — based purely on capital intensity.
The real-world application:
Buffett bought See's Candies in 1972 for $25 million. It had earnings of approximately $4 million at the time — a 6x earnings multiple that seemed reasonable but not spectacular.
What Buffett recognized was the capital intensity characteristic: See's required minimal capital reinvestment to grow. Over the following decades, See's generated over $2 billion in owner earnings on that $25 million purchase. The capital-light quality of the business meant that almost every dollar of earnings translated directly to cash that Berkshire could deploy elsewhere.
The DCF of See's in 1972, built on visible near-term earnings, would have missed this entirely. The moat and capital intensity analysis would have revealed it.
The Margin of Safety: Price as the Last Consideration
Benjamin Graham taught Buffett that buying at a discount to intrinsic value — the margin of safety — is the central principle of intelligent investing. Buffett kept this principle but applied it to a different conception of value.
Graham's margin of safety was primarily quantitative: buy assets at a discount to liquidation value. Buy $1 of assets for $0.60.
Buffett's margin of safety is primarily qualitative: be so certain about the quality and durability of the business that the price represents an obvious discount to what the business will produce over time.
The practical difference:
A Graham-style investor might buy a deteriorating business at a 40% discount to book value, expecting to profit from the gap between price and asset value.
Buffett's approach: buy an excellent business at a reasonable price rather than a mediocre business at an excellent price. The quality of the business creates a different kind of margin of safety — not from the discount to current assets, but from the reliability of future earnings.
How this applies to the Coca-Cola purchase:
In 1988, Buffett paid approximately 15x earnings for Coca-Cola. That was not a cheap stock by Graham's standards. The margin of safety in Buffett's framework came from his conviction about:
- The global growth runway for Coke's brand in underpenetrated international markets
- The capital-light economics that would translate growing sales into growing owner earnings
- The pricing power that came from a brand consumers trusted globally
- The distribution network that competitors would spend decades trying to replicate
The price was not obviously cheap. The business was obviously excellent. That combination — certainty of business quality at a reasonable price — was the margin of safety.
When Buffett Does Think About Numbers: The Back-of-the-Envelope Test
Buffett does perform numerical analysis — he just does it at a different level of precision and in a different sequence than formal DCF modeling.
His numerical test, as described across various interviews, is essentially:
What is this business likely to earn in 10 years? Is what I'm paying today obviously less than that future stream, discounted at a rate I'm comfortable with?
This is mathematically equivalent to DCF, but applied with wide bounds rather than false precision.
Practical application to a hypothetical:
A consumer staples company earns $500 million in owner earnings today with a clear track record of 7-8% annual growth sustained for 20+ years. The market capitalization is $8 billion — 16x current owner earnings.
Buffett's rough calculation: if earnings grow at 6% (below historical, providing conservatism) for 10 years, owner earnings will be approximately $895 million in Year 10. At a 15x multiple (conservative for a proven compounder), the business is worth approximately $13.4 billion in Year 10.
That's a 67% increase from today's $8 billion market cap over 10 years, before dividends. Not spectacular. The relevant question becomes: is the moat strong enough that the 6% growth assumption is reliable? And is there additional upside from international expansion that the conservative base case excludes?
The numbers serve the qualitative analysis, not the other way around.
The difference in practice:
A DCF analyst builds the model and reads the output: the model says fair value is $X, and the current price is Y% above or below that.
Buffett asks the business quality questions first, forms a conviction about what the business will produce, and then checks whether the price he's being asked to pay makes the return obviously attractive. If the answer requires a precise model, the conviction isn't sufficient.
What This Framework Misses — And When DCF Is Actually Necessary
Buffett's framework is not universally superior. It has specific limitations that practitioners must understand.
Where Buffett's approach struggles:
Capital allocation-intensive businesses: For businesses where the value is heavily driven by how management deploys capital over time — infrastructure companies, conglomerates, financial institutions — the moat framework is insufficient. You need to model the capital deployment explicitly.
Cyclical industries: Buffett largely avoids deeply cyclical businesses. His framework, which focuses on normalized earnings power, can produce dangerously misleading conclusions at cycle peaks. A steel company's earnings at the top of a commodity cycle look like a high-quality business when normalized.
Turnarounds and restructurings: When a business is changing — exiting divisions, restructuring debt, implementing new management — current earnings are a poor guide to normalized power. This is exactly the terrain where careful DCF modeling produces more insight than a moat assessment of the current business.
Acquisition pricing in investment banking: In M&A contexts, a formal DCF is necessary not just for valuation but for client credibility, regulatory review, and fairness opinion requirements. Telling a board of directors you didn't build a DCF because the math was obviously favorable would not go well.
The integration of both approaches:
Sophisticated practitioners use both. The moat framework to assess whether a business deserves a premium valuation and whether its earnings are likely to persist. The DCF to test whether a specific price is consistent with the range of plausible outcomes. Each approach checks the other's blind spots.
Closing: Buffett's Framework Is One Lens in a Complete Valuation Toolkit
Understanding how Buffett thinks about business quality, moats, owner earnings, and the margin of safety reframes valuation from a modeling exercise to a judgment exercise. That reframing is genuinely useful — it identifies the questions that DCF models answer poorly (will this business maintain its competitive advantage?) and the questions where qualitative conviction matters more than numerical precision.
But this framework is one component of a complete analytical toolkit that practitioners need to build.
After internalizing Buffett's approach, the natural questions that follow are: How do you build the comparable company analysis that tests whether your qualitative conviction about a business is reflected in how the market prices similar businesses? How do you construct and walk through a DCF in an investment banking context where the model must support a client recommendation, not just your personal investment thesis? How do you apply these valuation frameworks to real case studies — Amazon in 2001, Apple in 2016, a pre-IPO fintech today — where the moat assessment is genuinely contested and the numbers are genuinely uncertain?
These are the questions where conceptual frameworks meet applied judgment developed through practice on real problems.
At Meritshot, the Investment Banking program is built around exactly this progression — starting with valuation fundamentals, moving through moat analysis and owner earnings frameworks like those described in this article, building DCF and comparable company models on real deal cases, and developing the judgment that distinguishes mechanical model-building from genuine analytical insight. Students work through real acquisition case studies where the qualitative moat assessment conflicts with the DCF output, and learn how practitioners reconcile those tensions in actual client recommendations. Instruction comes from practitioners who have applied these frameworks on live transactions.
If this article changed how you think about what valuation actually requires — less model, more judgment about business quality — the next step is developing both skills together, because the most effective analysts are the ones who can do both.
The best valuation isn't the most precise model. It's the clearest thinking about what makes a business worth owning.
