Margin of Safety: Approximately Right Investing
Better to be approximately right than precisely wrong.
- Difficulty
- Moderate
- Time to result
- ~ongoing to results
- Steps
- 5
- Confidence
- 75%
Ben Graham's margin-of-safety principle, as applied by Buffett, accepts that no one can calculate a company's exact intrinsic value. Instead of chasing false precision, you build in enough cushion between price and your conservative estimate of value that you can be directionally right even if your numbers are off. Buffett's purchase of 5% of Disney in 1966 — valuing the entire company using only the future re-release revenue of Mary Poppins and assuming everything else was worth zero — is the clearest illustration of anchoring to one conservative, defensible cash flow stream rather than trying to model the whole business precisely.
Origin
Extracted from Acquired
Core principles
- 01Precise intrinsic value is unknowable; approximate value with a cushion is usable
- 02Anchor valuation to a conservative, provable cash flow stream
- 03The same imprecision applies to exit decisions, not just entry
- 04A big enough margin of safety protects you even if your estimate is wrong
How to run it
- 1
Accept You'll Never Know Exact Intrinsic Value
Start from the premise that precise valuation is impossible; the goal is to be approximately right, not precisely wrong.
- 2
Build In a Margin of Safety
Only invest when price is low enough relative to your conservative value estimate that you have real room for error.
- 3
Value Off a Single Conservative Anchor
When possible, value the business off one clearly identifiable, durable cash flow stream and assume everything else is worth zero, as Buffett did valuing all of Disney off Mary Poppins alone.
- 4
Apply the Same Discipline on the Way Out
Remember exit price is just as imprecise as entry price — a modest gain doesn't necessarily mean you've captured the business's full value.
- 5
Revisit the Thesis as the Story Develops
Periodically re-examine whether your original conservative anchor still undervalues the business rather than mechanically selling at a preset target.
In the wild
Buffett valued the entirety of Disney's roughly $80M market cap against just the expected future re-release revenue of one film, Mary Poppins, assuming the theme parks and film library were worth zero.
→ He bought 5% of Disney for $4M and sold within about a year for a 50% gain, missing decades of subsequent compounding in the business.
Common mistakes
Confusing 'Approximately Right' With 'Good Enough'
Buffett used margin-of-safety logic to justify exiting Disney after a quick 50% gain, when the underlying business was worth vastly more over subsequent decades.
Ignoring Optionality Beyond the Anchor
Valuing Disney solely off Mary Poppins ignored the theme parks and film vault — useful for setting a conservative entry price, but dangerous if it also caps how you think about upside.
Is it for you?
Best for
Investors who need a decisive way to act on incomplete information without false precision.
Not ideal for
Situations requiring exact, model-driven valuation, such as pricing a merger or arbitrage spread.
From the transcript
“better to be approximately right than precisely wrong”
“there's no way that you're gonna be able to exactly know the intrinsic value of the company. So you should you will never know exactly…”
From the episode
Berkshire Hathaway Part II