Lesson 4 of 7
Concentration, and what it costs
Weigh knowing fewer companies well against the cost of one mistake landing harder in a smaller portfolio.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- Fewer holdings let you research each one in more depth. More holdings spread the damage any single mistake can do.
- In an equally weighted portfolio, one total loss costs about four times as much at 6 holdings as it does at 25.
- Graham told the defensive investor to hold ten to about thirty issues. Buffett's 1996 letter sends most investors to a low-cost index fund, and tells those who pick their own to stay inside a circle of competence whose edges they know.
Concentration means holding fewer companies, each a larger share of the portfolio. The case for it is depth: you can know a handful of businesses far better than you can know fifty. The cost is that any one of the handful failing costs far more.
What one total loss costs at two portfolio sizes
- 1 of 25 equal holdings fails completely4.0%
- 1 of 6 equal holdings fails completely16.7%
Graham's case for spreading it
In chapter 5 of The Intelligent Investor, Graham told the defensive investor to hold at least ten issues and no more than about thirty. That kept any single misjudgment from doing much damage. The defensive investor, in Graham's own framing, is someone who wants safety with minimum effort, and spreading the list is the trade he makes for it.
The case for concentration
Buffett's 1996 letter starts somewhere else. Most investors, he writes, are best served by a low-cost index fund. For those who build their own portfolio, you need to judge only the companies inside your circle of competence. Its size matters less than knowing where it ends. Fewer businesses, understood well, can be researched far more deeply than many.
A broad portfolio
- Spreads any one mistake thin.
- Cannot be researched to the same depth, name for name.
- Graham's advice to the defensive investor called for this.
A concentrated portfolio
- Allows deeper research into fewer businesses.
- One mistake costs far more of the whole.
- Depends on staying inside a well known circle of competence.
Neither trade is free. Each gives something up for what it gains.
What each gives up
Neither choice removes the risk of being wrong about a company. It only changes how that risk is carried. A broad portfolio carries a small piece of many mistakes. A concentrated one carries a large piece of few.
Check yourself
In an equally weighted portfolio of 25 holdings, what does one total loss cost the whole portfolio?
- 100%
- 4.0%
- 25%
Show the answer
Answer: B.
Each holding is 1 divided by 25, or 4 percent, of the portfolio. A total loss on one holding costs that 4 percent share.Graham's advice to the defensive investor, in chapter 5, called for roughly what kind of holding list?
- 1 to 3 stocks
- A broad list, of ten to about thirty issues
- More than 200 stocks
Show the answer
Answer: B.
Graham told the defensive investor to hold ten to about thirty issues, spread broadly, to keep any one misjudgment from doing much damage.What does a concentrated portfolio trade for the depth of research it allows?
- Lower fees
- How much of the portfolio one mistake can reach
- A guarantee of a higher return
Show the answer
Answer: B.
Fewer, larger positions mean any single mistake reaches a bigger share of the whole. Depth of research does not remove that arithmetic.Sources
- Graham, The Intelligent Investor, revised edition 1973, chapter 5, on advice to the defensive investor to hold a broadly diversified list of common stocks.
- Berkshire Hathaway shareholder letter, 1996, on the circle of competence and knowing its boundaries.
What this leaves out
- The 25 and 6 holding examples are illustrative and equally weighted. Real portfolios are rarely weighted equally.
- The number of holdings alone says nothing about the quality of the research behind each one.
- Neither Graham's list nor a concentrated one removes the chance of being wrong about a company, only how much of the portfolio that wrong call can reach.
- Nothing here is advice about how many companies any reader should own.