Margin of safety: how we calculate the number
Margin of safety is the gap between what a business is worth and what you pay for it. We calculate it as a percentage discount to a conservative estimate of intrinsic value — a range, not a single point. We demand a wider discount when a business is harder to forecast and a narrower one when it's predictable, so the size of the margin is set by the quality of the company, not a fixed rule. The number isn't there to predict the upside; it's there to protect you when you're wrong.
The phrase gets quoted constantly and calculated almost never. "Buy with a margin of safety" sounds like wisdom, but it means nothing until you can put a number on it — a discount you'd actually require before buying, derived from a value you can defend. This piece shows how we get to that number.
It builds on our pillar on what the 167-point framework actually measures and pairs with our work on quality versus value. Value tells you what a business is worth. Margin of safety tells you how much room you're demanding between that figure and the price — your buffer against the fact that every valuation is an estimate made with incomplete information.
What the margin of safety actually protects against
The concept comes from Benjamin Graham, and the point people miss is that it isn't an upside calculation. It's an error calculation. You build a margin of safety because your estimate of intrinsic value will be wrong — the only question is by how much and in which direction. The discount is the cushion that lets you be wrong and still not lose money.
That reframing changes how the number is set. If the margin existed to maximize return, you'd want it as wide as possible on every stock. But it exists to absorb forecasting error, so the right width depends on how much error a given business invites. A regulated utility with contracted cash flows needs a thinner cushion than a cyclical commodity producer, because there's simply less to get wrong. The margin scales with uncertainty.
How we calculate the number
We start with a conservative intrinsic value — emphasis on conservative. We don't use the optimistic case. We use assumptions a skeptic would accept: modest growth, normalized margins, a discount rate that doesn't flatter the business. That gives us a value range, not a single figure, because pretending to three-decimal precision on a number built from assumptions is the most common way investors fool themselves.
The margin of safety is then the discount of price to the low-to-middle end of that range:
Margin of safety = (Intrinsic value − Price) ÷ Intrinsic value
A stock we value conservatively at $100 and can buy at $70 carries a 30% margin of safety. The discipline is that we anchor to the conservative end of the value range, not the optimistic one — so the discount we report is already stress-tested.
Why the required margin changes by company
Here's where most rules-of-thumb fall apart. A flat "always demand 30%" treats a predictable compounder and a turnaround speculation as if they carry the same forecasting risk. They don't. We set the required margin against the durability and predictability of the business:
- A wide-moat business with stable cash flows and a long track record is forecastable. We can demand a smaller discount — often in the 15–25% range — because the value estimate is more reliable.
- A decent business in a cyclical or fast-changing industry needs more — frequently 30–40% — because the honest value range is wider, and a wider range means more room to be wrong.
- A low-quality or hard-to-predict business needs the widest margin of all, and often the right answer is to demand a discount so large it never appears — which is simply our way of saying we should pass. A cheap price doesn't rescue a business we can't forecast.
This is the part the formula alone won't tell you: the quality work and the margin-of-safety work are the same work. The more confident the quality assessment, the thinner the cushion you need — which is exactly why we score the business before we ever price it.
The mistakes that hollow out the number
A margin of safety built on a bad intrinsic value is worse than none, because it gives false confidence. The errors cluster in a few places. Extrapolating a peak-cycle year as if it were normal inflates value and erases the cushion you think you have. Using a discount rate that quietly assumes the business is safer than it is does the same. And treating a low headline multiple as a margin of safety confuses cheap with undervalued — a stock can be statistically cheap and still worth less than its price if the business is eroding. The margin only means something when the value beneath it is conservative and honest.
The Claremont Street framework score
Margin of safety (part of the X / 167 total)
| What we report | What it tells you |
|---|---|
| Conservative intrinsic value range | Low to high, on a skeptic's assumptions |
| Current discount to value | The live margin at today's price |
| Required margin | Set by the quality and predictability score |
| Verdict | Does the current discount clear the required bar? |
The full report shows the conservative value range, the discount today's price offers, and whether that discount is wide enough for this specific business. See how we calculate a real company's margin of safety →
FAQ
What is a good margin of safety?
There's no universal number. For a predictable, wide-moat business, 15–25% can be enough. For a cyclical or harder-to-forecast company, you'd want 30–40% or more. The right margin is set by how reliable the underlying value estimate is — higher quality earns a thinner cushion.
How do you calculate margin of safety?
Subtract the price from a conservative estimate of intrinsic value, then divide by intrinsic value. A $100 conservative value bought at $70 is a 30% margin of safety. The discipline is in the word "conservative" — the calculation is only as good as the value you anchor to.
Is a low P/E the same as a margin of safety?
No. A low multiple can signal a margin of safety or a business in decline. Margin of safety is a discount to what a company is genuinely worth; a low P/E is just a low price relative to recent earnings. They overlap only when the earnings are durable.
Why require a bigger discount for some stocks than others?
Because the margin of safety protects against forecasting error, and some businesses are far harder to forecast than others. The less predictable the cash flows, the wider the range of plausible values — and the wider the discount you should demand before risking capital.
Does a margin of safety guarantee you won't lose money?
No. It improves your odds and limits the damage when you're wrong, but it's a buffer, not a guarantee. Any single estimate can be off by more than the cushion. It works as a discipline applied consistently across many decisions, not as insurance on any one.
Related reading
- How we think about moats — quality, and why we score the business before we price it
- Margin of safety for the long-term investor — the behavior a real margin buys you
- Want the full 167-point breakdown on any ticker? →
This analysis is for informational and educational purposes only and is not investment advice. Claremont Street is not a registered investment advisor. Do your own research.
Data sources: SEC EDGAR, Financial Modeling Prep (FMP), Quartr.
Patient, AI-native investing — built for the long term.
Run any stock through the 167-point framework. Free to start, no card required.