The label "value investor" gets applied to people who would never buy each other's stocks. Benjamin Graham's defensive investor would not own most of Warren Buffett's portfolio. Buffett today would not own most of what Peter Lynch made famous at Fidelity Magellan. Each framework solves a different problem, optimises for a different risk, and produces a different portfolio.
This is the head-to-head comparison. Criteria, resulting portfolios, historical returns, and when each one is the right tool.
The three frameworks in 60 seconds each
Benjamin Graham (defensive value, 1930s-1976). Built after the 1929 crash. Designed for the investor who fears permanent capital loss above all else. Criteria: P/E under 15, P/B under 1.5, current ratio above 2x, debt under net working capital, dividends paid 20 years uninterrupted, EPS growth 33+ percent over 10 years. The result: a portfolio of unloved cyclical industrials, utilities, banks, and conglomerates trading near or below tangible book value.
Warren Buffett (quality compounding, 1960s-today). Built on the foundation of Graham (Buffett was Graham's student at Columbia) but evolved into something different. Criteria: economic moat with sustained ROIC above 15 percent, durable earnings, owner-friendly management, fortress balance sheet, fair-or-better valuation. The result: a portfolio of branded consumer compounders (Coca-Cola, See's Candies, Apple), payment networks (Visa, Mastercard via American Express), and insurers with float (GEICO, Berkshire's own reinsurance).
Peter Lynch (growth at a reasonable price, 1977-1990 at Fidelity Magellan). Made famous by his 29 percent annualised return at Magellan. Criteria: PEG ratio (P/E divided by earnings growth) under 1, mid-cap (avoiding mega-cap institutional crowding), business model that "a 12-year-old can understand", insider buying, growth visible in your own daily life. The result: a portfolio of mid-cap growth-quality businesses with story-driven coverage potential.
Where the frameworks agree
A few core principles are shared:
- All three avoid leverage. None recommends margin debt as a regular practice.
- All three emphasise reading 10-K filings, not relying on stock tips.
- All three insist on a margin of safety (buying below intrinsic value), though they calculate intrinsic value differently.
- All three reject short-term trading as a fool's errand.
The disagreements live in the details - which specific criteria define a "good" business, how heavily to weight current cheapness versus future quality, and how much qualitative judgement to allow.
Where the frameworks disagree
Disagreement 1: How much current cheapness matters
Graham: cheapness is everything. A stock at P/E 8 with a fortress balance sheet beats a stock at P/E 18 with a wider moat, almost always. Graham's "Defensive Investor" portfolio would hold cyclicals, utilities, regional banks - businesses with no growth story but with a measurable cushion against permanent loss.
Buffett: cheapness matters but quality matters more. Buffett evolved from Graham's strict cheapness criteria after watching what Charlie Munger taught him: a wonderful business at a fair price compounds for decades. A cigar-butt cheap stock pays one puff and then you have to find another. Buffett's portfolio today would mostly fail Graham's strict P/E and P/B screens.
Lynch: cheapness adjusted for growth is the real metric. PEG below 1 means you're getting growth cheap. A stock at P/E 25 with 30 percent growth is cheaper than a stock at P/E 10 with 5 percent growth on this view. Lynch was willing to pay up for visible growth in a comprehensible business.
Disagreement 2: How wide the universe should be
Graham: large-cap mostly. The defensive investor framework requires 5+ years of consistent dividends, stable earnings, and substantial size - which excludes most small-caps. Graham's portfolio was concentrated in 30-50 mid-and-large-cap industrials.
Buffett: limited by Berkshire's size, but historically open to small-caps when his AUM was small enough. Today's Berkshire is too large to buy anything below a $20 billion market cap; smaller individual investors can still apply the framework to mid-caps.
Lynch: mid-cap is the sweet spot, explicitly. Lynch wrote in One Up on Wall Street that mid-caps offered the highest information edge - too small for institutional research coverage, too large for the survivorship risk of micro-caps. He'd happily own 1,000+ positions in Magellan, with mid-caps over-weighted.
Disagreement 3: Qualitative vs quantitative judgement
Graham: ruthlessly quantitative. The Defensive Investor screening criteria were designed to be applied mechanically - Graham didn't trust the average investor (or himself) to override the numbers with qualitative judgement.
Buffett: hybrid. Quantitative pillars (ROIC, FCF, margins) set the universe; qualitative judgement about moats, management, and competitive dynamics determines the final selection. Buffett famously said you can't quantify everything that matters: "It is better to be approximately right than precisely wrong."
Lynch: heavily qualitative. Lynch's famous "buy what you know" advice was about using the qualitative information you collect in everyday life (which restaurant chain is packed, which retailer's stores feel busy) as a leading indicator before the financials show it. He insisted that retail investors had an information edge over Wall Street precisely because they could see consumer trends earlier.
Disagreement 4: How long to hold
Graham: hold until intrinsic value is reached, then sell. Typical holding period 12-24 months. Graham was not a buy-and-hold investor; he was a buy-cheap-and-sell-when-fair investor.
Buffett: forever, or as close to it as the business stays wonderful. Buffett famously holds Coca-Cola (since 1988), American Express (since 1991), Moody's (since 2000). Selling triggers tax and gives up compounding.
Lynch: until the story changes. Lynch was a "story stock" investor - he'd sell when the underlying business thesis broke (saturation, management change, etc.), regardless of current valuation. Average holding at Magellan was 2-4 years.
Historical returns by framework
A few replicated backtests give a rough picture:
Graham defensive (1976-2024): ~13-14 percent annualised, beating the S&P 500 by ~2-3 percentage points. Drawdowns in growth-led markets (1995-1999, 2015-2019) when defensive cyclicals lag.
Buffett quality (Berkshire 1965-2024): ~20 percent annualised, beating the S&P 500 by ~10 percentage points. Drawdowns relatively shallow but real (1999, 2009, 2022).
Lynch GARP (Magellan 1977-1990): 29 percent annualised, beating the S&P 500 by ~14 percentage points. Note: this is the in-fund return; replication backtests of Lynch's published criteria on subsequent decades show ~12-14 percent annualised, materially lower than Magellan's actual record. The Magellan outperformance was partly the framework and partly Lynch himself.
Multi-framework consensus (5+ of 7 frameworks, invest-like.com backtest 2019-2024): outperforms the S&P 500 by 5-7 percentage points annualised in our internal backtest, with a smaller drawdown profile than any single framework. The intersection-of-disciplines effect: stocks that pass multiple value-investing screens simultaneously tend to be genuinely mispriced rather than passing one framework by accident.
When each framework wins
Graham wins in deep recessions and credit-crisis markets. When P/E ratios collapse and book values look meaningful again, Graham's defensive screen identifies genuinely cheap businesses with bankruptcy-proof balance sheets. 2008-2009 was a Graham moment.
Buffett wins in normal bull markets and slow-growth regimes. When growth is steady and quality is being rewarded, the wonderful-business-at-fair-price compounder wins decade after decade. 2010-2024 has been mostly Buffett's era.
Lynch wins in retail-led mid-cap rallies. When mid-cap growth is being mispriced relative to mega-caps, Lynch's GARP approach captures alpha. 1985-1990 was the classic Lynch era; 2003-2007 was a smaller Lynch revival.
No framework wins in pure speculative bubbles (1998-1999, 2020 SPAC mania, certain crypto cycles). Anyone applying a value framework in those periods looks foolish for a year or two before the bubble pops. This is unavoidable; the alternative (abandoning the framework during bubbles) is worse.
How to use the comparison practically
Three honest takeaways:
1. Pick a framework that matches your psychology. If you can't sleep when a stock you hold drops 30 percent, Graham's defensive framework is the right fit (less drawdown, more cyclical re-pricing). If you don't mind volatility but want compounding, Buffett's quality framework works. If you genuinely enjoy researching consumer trends, Lynch's GARP framework leverages your natural curiosity.
2. Don't try to mix mid-position. Many retail investors hold a Buffett-style core, panic-sell during a Graham-style drawdown, and buy back at the wrong moment. Pick your discipline and stick with it.
3. Multi-framework consensus is structurally robust. Stocks that pass Buffett AND Graham AND Lynch simultaneously are rare and tend to be genuinely mispriced. The invest-like.com /consensus/ page surfaces this intersection live.
Where invest-like.com applies each
The seven-framework consensus on invest-like.com includes all three plus Fisher, Greenblatt, Munger, and Smith. The per-strategy pages:
The methodology for each framework is published openly. You can see exactly why a given stock passes Buffett but fails Graham, or vice versa, by reading the per-pillar breakdown on the per-stock page.
Further reading
Educational only. Not investment advice.