Skip to content
Menu
Finabase Learn

Value a company Lesson 10 of 10

Fundamentals · Valuation

Margin of safety: Benjamin Graham’s rule for valuation errors

What a margin of safety is, how to calculate one, why valuations are ranges rather than points, and how DCF sensitivity and position sizing feed into it.

On this page
  1. What a margin of safety is
  2. Why valuations are ranges, not points
  3. How big should the margin be?
  4. Margin of safety in other forms
  5. How it connects to position sizing
  6. Common misconceptions
  7. Margin of safety compared with related ideas
  8. How Finabase does it
  9. Frequently asked questions
  10. Summary
  11. Check your understanding
  12. Exercise

A margin of safety is the gap between what you estimate a stock is worth and the price you pay for it. If you think a share is worth $100 and you will only buy it at $75 or less, your margin of safety is 25%. The idea is that your estimate might be wrong, the future might be worse than you expect, or both, and a price well below your estimate leaves room for those errors before you lose money.

Benjamin Graham, the Columbia professor whose books shaped value investing, developed the idea in Security Analysis (1934, with David Dodd) and made it the centerpiece of The Intelligent Investor (1949). In the closing chapter of the book's later editions, he condensed sound investing into a three-word motto, writing that "we venture the following motto, MARGIN OF SAFETY." This guide explains what the concept means in practice, how to calculate it, why it depends on treating valuation as a range rather than a single number, and how it connects to the DCF sensitivity table and to position sizing.

What a margin of safety is

Graham's insight was that investors do not need to know the exact value of a business. They need to know, with reasonable confidence, that it is worth substantially more than they pay. An engineer designing a bridge for 10-ton trucks builds it to carry far more, not because the trucks will weigh more but because the calculations, materials and future loads are uncertain. The margin of safety is the investor's version of that over-engineering.

Margin of safety: buying well below a cautious estimate of value Range of estimated value central estimate $100 buy-below price $75 (25% margin of safety) market price $130 Illustrative numbers. The width of the band, not the central number, is the honest output of a valuation.
Illustrative numbers: a 25% margin of safety below a $100 central estimate puts the buy-below price at $75; a $130 market price sits outside the range entirely.

The diagram makes one more point: the central estimate sits inside a band. The honest output of a valuation is the band. The margin of safety is how far below that band, or below its lower part, you want the price to sit.

Why valuations are ranges, not points

Every valuation method rests on forecasts and on a required return, and small changes in either move the answer a lot. A Discounted cash flow (DCF): a step-by-step valuation on Apple model built on Apple Inc.'s fiscal 2025 free cash flow of $98.8 billion (the year ended September 27, 2025) shows how much.

The wider the plausible range, the larger the margin you need before your estimate tells you anything. A stable business with predictable cash flows, perhaps protected by an economic moat, produces a narrower range, so a smaller margin may do. A young, cyclical or heavily indebted company produces a wide one, and a 25% margin may not be enough.

How big should the margin be?

There is no official number, and the right size depends on how uncertain the estimate is. Value investors often talk about margins of a quarter to a half of estimated value, but those are conventions, not rules. Factors that argue for a bigger margin:

  • Forecast uncertainty: new products, cyclical demand, customer concentration.
  • Balance-sheet risk: high debt can turn a bad year into a permanent loss.
  • Model risk: a valuation dominated by its terminal value, as most DCFs are.
  • Your own track record: if your past estimates have been optimistic, assume they still are.

Factors that can justify a smaller one: long, stable operating history, a strong balance sheet, and a valuation that holds up across several methods and most of its sensitivity table.

Margin of safety in other forms

The idea extends beyond price versus value.

  • Balance-sheet margin. Graham's early work looked for companies trading below their net current assets, so that the balance sheet itself provided the cushion.
  • Earnings margin. For a bond or a preferred stock, the margin is how many times earnings cover the interest or dividend.
  • Assumption margin. Using deliberately conservative inputs, such as lower growth and a higher discount rate, builds the cushion into the estimate rather than the price. Be careful not to double-count: a conservative estimate plus a large required discount can mean you never buy anything.

How it connects to position sizing

A margin of safety reduces the chance and size of a loss on any one investment, but it cannot eliminate it. Forecasts can be wrong in ways no model anticipated. That is why it works alongside Diversification: why spreading your investments reduces risk and Position sizing: how many shares to buy for a set risk: spreading money across holdings and limiting how much any single position can cost you. A position with a thin margin of safety or a wide valuation range is one where a smaller position size limits the damage if the estimate proves wrong. The SEC's Investor.gov guidance on asset allocation makes the same general point about diversification: owning a range of investments reduces the impact of any one of them going badly.

Common misconceptions

  • "A margin of safety means I can't lose money." It lowers the odds and the severity; it does not remove them. A business can deteriorate after you buy.
  • "A low P/E is a margin of safety." A low multiple is cheap only relative to the earnings; if those earnings are about to fall, the stock may be expensive.
  • "The margin is measured from the price I paid." It is measured from your estimate of value, which should be revisited as facts change.
  • "Bigger is always better." An extreme margin requirement can keep you out of every investment; the cost of that is real too.
Idea What it measures Used for
Margin of safety Discount of price to a cautious value estimate Deciding whether a price leaves room for error
Sensitivity table How value changes as each input changes Measuring how wide the value range is
Reverse DCF The growth the current price implies Judging whether the market's assumptions are plausible
Diversification Spread of holdings across companies and sectors Limiting the impact of any one mistake
Position sizing How much to commit to one holding Capping the loss from any one position

How Finabase does it

Finabase does not compute an intrinsic value, a buy-below price or a margin of safety. It gives you the data to build your own range. The Financials tab has the operating cash flow, capital expenditure, cash, investments and debt a DCF needs, with every row chartable over time so you can see how variable the inputs have been. The Valuation tab charts ten fiscal years of the company's own multiples, such as P/E, P/FCF and EV/EBITDA, which helps you judge whether today's price sits high or low against its own history. Its "Financial health" card adds Altman Z-Score and Piotroski F-Score readings from the data provider, useful context for the balance-sheet side of the margin.

Frequently asked questions

What is a margin of safety in investing?

It is the difference between your estimate of a stock's value and the price you pay, expressed as a percentage of the estimate. Buying at $75 something you estimate is worth $100 gives a 25% margin of safety.

How do you calculate margin of safety?

Subtract the price from your estimated value and divide by the estimated value: (value − price) ÷ value. To find the price that delivers a required margin, multiply the estimate by one minus that margin.

What is a good margin of safety?

There is no standard figure; it should grow with the uncertainty of your estimate. Value investors often talk about a quarter to a half of estimated value, with larger margins for less predictable or more indebted businesses.

Who came up with the margin of safety?

Benjamin Graham developed the concept in Security Analysis (1934, with David Dodd) and made it the central idea of The Intelligent Investor (1949). Warren Buffett, his student, later popularized it further.

Summary

A margin of safety is the cushion between a cautious estimate of value and the price paid, calculated as (value − price) ÷ value. Graham made it the core of sound investing because every estimate is uncertain. Treat the output of any valuation as a range: an illustrative DCF on Apple's fiscal 2025 cash flow spans about $105 to $189 a share on modest input changes, so the margin you demand should reflect how wide your own range is. A margin of safety reduces, but never removes, the risk of loss, which is why it belongs alongside diversification and sensible position sizing.

Check your understanding

4 questions on this lesson. Each answer comes with the reason, and nothing is scored.

  1. Your central estimate of a share’s value is $80 and you require a 30% margin of safety. What is your buy-below price?

    1. $61.54
    2. $56.00
    3. $24.00
    Show the answer

    B. $56.00 Buy-below price = estimated value × (1 − required margin) = $80 × 0.70 = $56. The $24 is the size of the discount, not the price, and dividing $80 by 1.30 measures the margin from the price instead of from the estimate.

  2. You estimate a share is worth $60 and it trades at $75. What is the margin of safety?

    1. −25%: the price is a quarter above your estimate
    2. −20%: the estimate is a fifth below the price
    3. 25%: the gap is a quarter of your estimate
    Show the answer

    A. −25%: the price is a quarter above your estimate Margin = (estimated value − price) ÷ estimated value = ($60 − $75) ÷ $60 = −25%. It is measured from your estimate, not from the price, and a negative result means that on your own numbers there is no margin at all.

  3. One company has a long, stable history and a valuation range of $90 to $110. Another is young and heavily indebted, with a range of $50 to $150. Both have a central estimate of $100. What follows?

    1. The same margin suits both, because the central estimates are equal
    2. The stable company needs the larger margin, because its upside is smaller
    3. The second company needs the larger margin, because its estimate is far less certain
    Show the answer

    C. The second company needs the larger margin, because its estimate is far less certain The honest output of a valuation is a range, and the width of the range should set the size of the margin. A wide range, high debt and a short history all argue for a bigger cushion; a narrow range can justify a smaller one.

  4. An investor builds a DCF with deliberately low growth and a high discount rate, then also demands a 50% discount to the result. What does the lesson warn about this?

    1. Nothing: stacking both cushions is always the safest course
    2. It double-counts the cushion and can mean never buying anything, which has a real cost too
    3. The conservative inputs cancel out the discount
    Show the answer

    B. It double-counts the cushion and can mean never buying anything, which has a real cost too Conservative inputs already build the margin into the estimate. Adding a large required discount on top counts the same caution twice. Bigger is not always better: an extreme requirement keeps you out of every investment.

Exercise

One task to do in Finabase. Nothing is graded or saved.

Open Coca-Cola’s Analysts tab and read the Analyst view: Target (low), Target (mean), Target (high) and the current price. Treat the mean target as one estimate of value and apply the lesson’s formula: (estimate − price) ÷ estimate. Then divide the high target by the low one. Is the margin positive or negative, and how wide is the range of estimates?

Open Coca-Cola’s Analysts tab
What you should find

A price above the mean target gives a negative margin: on that estimate there is no cushion at all. The high-to-low ratio shows how far informed estimates differ, and the wider it is, the larger the margin an estimate needs before it tells you anything. Analyst targets are other people’s opinions, not your own cautious estimate, so this is practice with the arithmetic, not a signal.

Sources and further reading

  1. Graham, B., The Intelligent Investor (Harper & Brothers, 1949; revised editions to 1973)
  2. Graham, B. and Dodd, D., Security Analysis (McGraw-Hill, 1934)
  3. Apple Inc. Form 10-K for the fiscal year ended September 27, 2025 (SEC EDGAR)
  4. U.S. Securities and Exchange Commission, Investor.gov — Asset allocation and diversification