The evolution of value investing

What does "the evolution of value investing" mean?

Value investing has not changed its definition — pay less than a business is worth — but the way practitioners measure worth has moved through three eras. Call it 1.0, the Graham era: buy statistical bargains, companies trading below net working capital or book value, on low price-to-earnings and low price-to-book multiples, with little regard for business quality. Call it 2.0, the Buffett–Munger era: stop buying cheap mediocrity and start paying a fair price for durable quality — a wide moat, high returns on capital, and a long reinvestment runway. Call it 3.0, the era we are in now: the best businesses are light on tangible assets and heavy on intangible ones — software, networks, brands, data — and accounting was built for the factory, not the platform. The constant across all three is intrinsic value versus price. What evolved is where the value hides and how you have to look for it.

It is tempting to treat each era as a refutation of the last. It is better read as a widening lens. Graham gave the field its first principle — margin of safety — and a method for a world of cheap, asset-heavy, post-Depression businesses. Buffett and Munger kept the principle and changed the object, from cheap assets to durable earnings. The current generation keeps both and adapts to a world where the most valuable assets are not on the balance sheet at all.

Why did Buffett and Munger move on from cigar-butt investing?

Buffett learned investing from Benjamin Graham, and Graham's method was to buy "cigar butts" — beaten-down companies with one last free puff of value left in them, bought so cheaply that even a mediocre outcome made money. Charlie Munger pushed him off it. Munger's argument was that "it's far better to buy a wonderful company at a fair price than a fair company at a wonderful price." A cheap, declining business hands you a one-time gain and then a problem; a wonderful business compounds for decades and lets the holding period do the work. See's Candies in 1972 was the hinge — a business worth far more than its book value because of pricing power and brand, not because of the assets on its balance sheet. By the time Buffett put $1 billion into Coca-Cola in 1988 at roughly 15× earnings, he was paying a fair price for unassailable quality, not a wonderful price for a fair company. Two practical forces reinforced the shift: the supply of true statistical bargains shrank as markets got more efficient, and Berkshire grew too large to move the needle with tiny deep-value positions.

It’s far better to buy a wonderful company at a fair price than a fair company at a wonderful price.
— Charlie Munger

From book value and low multiples to durable quality — what changed?

The screen changed. Value investing 1.0 ranked the world by cheapness: low price-to-book, low price-to-earnings, assets you could count and liquidate. The trouble is that a low multiple is often a fair price for a business whose earnings are about to shrink — the value trap. Value investing 2.0 replaced the cheapness screen with a quality screen: a moat that protects margins from being competed away, a high return on invested capital, and the ability to reinvest retained earnings at that same high return for years. The math is ROIC × reinvestment rate, and it compounds. A business earning 30% on capital and reinvesting most of its profit is worth a premium multiple precisely because the multiple is the entry tax for a holding period in which the math runs in your favour. Cheapness became a secondary consideration; durability of returns became the primary one.

From tangibles to intangibles — why does it break old metrics?

Accounting capitalises a factory and depreciates it slowly, but it expenses the things that build a modern moat — research and development, sales and marketing, software engineering, brand-building, stock-based compensation — in the period they are incurred. The result is that the income statement understates earnings for a company investing heavily in intangibles, and the balance sheet understates the capital actually at work. Reported return on invested capital looks distorted in both directions: the denominator misses the intangible investment, while the numerator is depressed by expensing it. Price-to-book becomes almost meaningless when the most valuable asset — the codebase, the network, the brand — never appears on the books. A value investor working from unadjusted screens will read a great compounder as expensive and a melting ice cube as cheap. The fix is to think like an owner: capitalise the intangible spending mentally, separate maintenance cost from growth investment, and judge the return the business earns on the capital it actually deploys.

The owner’s adjustments worth making before judging a modern business:

• Capitalise the growth portion of R&D, sales and marketing rather than expensing all of it as a cost of the current period.
• Separate maintenance spending from investment — only the first is a true cost of running the business as it is.
• Measure return on the capital actually deployed, including the intangible investment accounting leaves off the books.
• Treat price-to-book with suspicion when the codebase, the network or the brand is the real asset.

How does Amazon illustrate Value Investing 3.0?

Amazon is the canonical case. It began as a book retailer with thin reported profits and, for years, looked overpriced on every traditional value metric — negligible earnings, no dividend, a sky-high multiple on what little profit it showed. But the low reported earnings were a choice: Amazon was expensing enormous forward-looking investment — fulfilment, logistics, and the technology that became Amazon Web Services — straight through the income statement. That was deferred gratification in action, the willingness to suppress today's reported profit to build tomorrow's franchise. "Your margin is my opportunity" was a strategy, not a slogan. A 1.0 investor screening on P/E would never have touched it; a 3.0 investor who capitalised the investment, recognised the moat forming in cloud and marketplace, and underwrote the reinvestment runway could see a business compounding intrinsic value far faster than reported earnings suggested. The lesson is not that price does not matter — it is that reported earnings are an output of accounting policy, and the value is in what the business is building.

Is this still value investing?

Yes. Munger's line settles it: "all investing is value investing — what else is there?" The discipline has always been to estimate what a business is worth and pay less than that. Buying a high-ROIC software compounder on a forward view of its cash flows is not a different activity from buying a railroad below book value; it is the same activity with the inputs moved forward in time and onto the parts of the business the balance sheet cannot see. Growth and value were never opposites — growth is simply one of the inputs to value, and for an intangible-heavy business it is often the dominant one. What separates the value investor from the speculator is unchanged: a written thesis about why the business will be worth more than the price implies, a moat that makes the thesis durable, and a margin of safety against being wrong. The 3.0 investor demands all three; they just apply them to companies Graham would not have recognised.

All investing is value investing — what else is there?
— Charlie Munger

What does Value Investing 3.0 ask of the analyst?

It asks for forward-looking underwriting without abandoning discipline. Start with a thesis-first underwrite: a specific, falsifiable claim about why the business will be worth more in the future than the current price implies — the moat type, the growth driver, the kill criterion. Capitalise the intangible investment that accounting expenses, and judge the return the business earns on incremental capital, not just the depreciated-asset base. Weigh optionality — the company's ability to reinvest into new high-return businesses, as Amazon did with cloud — as a real, if uncertain, component of value. Use a reverse-DCF to check what growth the current price already assumes, so quality does not become an excuse for any price. And keep the margin of safety, applied to the thesis rather than to the book value: the buffer that protects you against the inevitable error in any forward estimate. Quality at a fair price, underwritten forward, with the discipline to exit when the thesis breaks — that is value investing 3.0.

None of this lowers the bar. Forward-looking underwriting is harder than reading a low multiple off a screen, and the room for error is larger, which is exactly why the margin of safety and a written kill criterion matter more, not less, the further out the thesis reaches. The evolution of value investing is not a softening of Graham’s discipline. It is the same discipline pointed at where value now lives.