A client asked a fair question this week. The research says own 300 stocks. The biggest tier of the Capital Wealth Growth Portfolio holds 56. Who is right? Instead of arguing, we measured it on our own holdings.
The expired rule
The textbook said 30 stocks was plenty. That rule dates to 1968, when every trade cost real commission money, so you stopped buying diversification early.
Then commissions fell to zero, and the rule’s own author checked his math. In 2004, Meir Statman reran the numbers and found the right count is over 300. The rule was not wrong. It expired.
The market also got noisier. A Journal of Finance study found single stocks roughly doubled in volatility — how much a price bounces around — from the 1960s to the 1990s. It now takes about 50 names to do the smoothing 20 once did.
The real risk
Here is the finding that flips the whole debate. Since 1926, just 4% of companies created all of the market’s wealth above Treasury bills. The median stock, over its lifetime, lost to T-bills.
So a 2% position that dies costs you two cents on the dollar. Missing one of the 4% costs you the decade. A short list is a standing bet that the next great compounder is already on it.
Peter Lynch coined “diworsification,” and short-list fans quote him constantly. Lynch ran 1,400 positions and beat the market thirteen straight years. He was warning against owning things for no reason — not against owning many things.
What we measured
On July 4, 2026 we widened the $500K tier: 37 positions became 54, and the top single-name weight dropped from 4% to 2.5%. Philip Morris and British American later took it to 56. On August 11, Chubb, Aflac, and CME Group joined the $50K, $100K, and $250K tiers. The five-year test:
| Through July 2, 2026 | 37-name portfolio | 54-name portfolio | S&P 500 |
|---|---|---|---|
| YTD return | +10.3% | +11.3% | +9.8% |
| 5-yr annualized | +8.8% | +12.3% | +13.0% |
| Sharpe ratio | 0.40 | 0.71 | 0.52 |
| Volatility | 12.0% | 11.7% | 17.2% |
| Max drawdown | −11.7% | −12.4% | −24.5% |
| Beta vs S&P (daily, 5-yr) | 0.41 | 0.50 | 1.00 |
The Sharpe ratio — return earned per unit of risk — nearly doubled. Then we stretched the test to nineteen years: monthly data, dividends reinvested, straight through 2008, 2020, and 2022.
One honest warning first. This tests today’s roster backward, so it carries survivorship bias — today’s survivors flatter the past. Read it as a test of shape, not a promise.
| Sept 2007 – July 2026 (monthly) | Growth Portfolio | S&P 500 |
|---|---|---|
| Annualized return | +12.0% | +11.0% |
| $10,000 became | $85,300 | $71,800 |
| Volatility | 11.6% | 15.6% |
| Correlation to S&P | 0.80 | 1.00 |
| Beta (monthly, 19-yr) | 0.59 | 1.00 |
| Worst drawdown (2008–09) | −28.7% | −50.8% |
| 2008 calendar year | −14.4% | −36.8% |
| 2022 calendar year | +5.3% | −18.2% |
Two numbers explain everything. A 0.80 correlation — how often two things move together — means the portfolio moves with the market most months. A 0.59 beta — how far it moves when the market moves — means the waves are six-tenths the size. Same direction, smaller waves: that turned 2022’s −18.2% into +5.3%.
Finally, the client’s actual question: is 56 better than 100? We drew 400 random subsets of our own holdings at each size and measured the volatility:
| Names held | Portfolio volatility | Gained by the last 5 names |
|---|---|---|
| 5 | 15.24% | — |
| 10 | 13.98% | −1.26 pts |
| 20 | 13.33% | −0.33 |
| 30 | 13.06% | −0.13 |
| 40 | 12.99% | −0.04 |
| 53 | 12.89% | −0.03 |
The honest answer: 56 versus 100 is a statistical tie. The risk curve stops paying around 30 names. Anyone selling a 56-name portfolio as safer than a 100-name one is selling a preference as a finding.
So what sets the count? Account size first: 100 names in a $50,000 account is $500 a position — too small to rebalance or trim. That is why the tiers hold 30, 34, 39, and 56 names. Mandate second: every seat must pay a dividend covered by free cash flow, and that universe runs to dozens, not hundreds. Monitoring third: 56 written theses is a job. Three hundred is a filing cabinet.