Margin of safety: what it is and how we measure it

The idea

Any estimate of what a business is worth will be wrong by some amount. A margin of safety is the room you leave for that error: paying enough below your estimate that the purchase can still work out when the estimate turns out too high.

It is a statement about the gap, not about the company. A wonderful business bought at its full value has no margin of safety; an ordinary one bought well below a careful estimate has a wide one.

How we measure it

StockValuer measures the margin as a share of the value: (value − price) ÷ value. A $100 price against a $200 value is a 50% margin of safety, because the price could rise by half of the value before the cushion is gone.

Measured the other way round, against the price, the same pair reads as “100% upside”. That number grows without limit as the price falls and flatters every estimate, which is why we do not use it. A margin can never exceed 100%, and a price above the value gives a negative margin, shown as a percentage above value.

Against which value?

A single value hides how much of it rests on forecasts. So a company's Overview shows the price against three values, from the most cautious up:

• The bear case: the researched forecast in which things go poorly.
• Earnings power: what today's normalized profit is worth if the business never grows at all.
• The base case: the central forecast.

Reading the three together

A margin that survives against earnings power does not depend on any growth forecast being right. A margin that exists only against the base case depends on the forecast arriving. Neither is a reason to act; together they show how much of the cushion is made of assumptions.

The screener can rank and filter by any of the three, so a reader can ask for, say, every company priced at least 30% below its earnings power.