The classic answer is wrong
The advice you'll get from most financial websites: "Diversify. Don't put all your eggs in one basket. Own as many stocks as possible."
That's half-right. The half that's wrong is the implication that more is always better. It isn't. There's a number where adding more stocks stops meaningfully reducing risk and starts diluting the impact of your best picks.
The math has been studied for decades. Let's go through it.
The free lunch, and where it ends
Here's what diversification actually buys you. When you own a single stock, your portfolio has whatever volatility that stock has. Roughly 30-50% annualized volatility for a typical large-cap, much higher for small- or mid-caps.
When you add a second stock, the combined portfolio's volatility usually drops, because the two stocks don't move in perfect lockstep. When one is down on a given day, the other is sometimes up, which dampens the overall swing.
Adding more stocks continues this effect, but the marginal benefit shrinks fast.
The rough numbers (from decades of academic studies):
- 1 stock: full single-stock volatility (~40% annualized)
- 5 stocks: drops to ~25%
- 10 stocks: drops to ~22%
- 20 stocks: drops to ~20%
- 30 stocks: drops to ~19%
- 100 stocks: drops to ~18%
- The whole market: ~16%
Notice the pattern: the first 5 stocks cut your volatility almost in half. The next 15 give you another 5 points. After that, you're down to fractional improvements.
The implication: 20 carefully chosen stocks captures essentially all the diversification benefit available to a stock-only portfolio. Adding stocks 21 through 500 does almost nothing for your volatility, but it does dilute the impact of your best picks.
What the academic literature says
The seminal paper here is Evans and Archer (1968), which first demonstrated the diminishing-returns curve. Subsequent work by Statman (1987) refined the number. Newer research using more decades of data and broader universes has generally confirmed the conclusion: somewhere between 15 and 30 stocks captures the bulk of the diversification benefit.
The exact number depends on:
- Sector spread. 20 stocks all in tech aren't really 20 stocks for diversification purposes; they're closer to 5 or 6 because they all move together.
- Size and style spread. Mixing large and small, growth and value, gives you more diversification per stock than concentrating in one box.
- International exposure. US-only portfolios benefit from adding international names, but the marginal benefit also flattens past a point.
Why "more = better" is wrong
There's an important point that almost never gets made: more stocks doesn't just stop helping. It actively hurts your ability to outperform.
Here's the math. Suppose you have a portfolio of 20 stocks, and the average pick beats the market by 1% per year. Your portfolio beats the market by 1% per year. Simple.
Now suppose you add 480 more stocks, equal-weighted, that you don't have particular conviction in. You've gone from 20 to 500 holdings. Your portfolio now looks much more like the index. Your alpha gets diluted toward zero.
This is why index funds, by definition, return the index. They own everything; they have no concentrated bets. They charge low fees because they're not trying to do anything but match the market.
If your goal is to beat the market, you need concentration. Not extreme concentration (5 stocks is too few; you can be wiped out by one bad pick). But concentration in the 15-30 range, with thoughtful picks, gives you:
- Most of the available diversification benefit
- Enough room for individual picks to matter
- A portfolio that's easy to understand, monitor, and rebalance
This is exactly why our flagship portfolio is called Core 20. Twenty equally-weighted, well-researched names. Not five (too risky). Not five hundred (too diluted).
The objections, addressed
"What if one of your 20 stocks goes to zero?"
In an equal-weighted 20-stock portfolio, each holding is 5%. A complete blow-up means a 5% portfolio loss, uncomfortable but recoverable. In our 25-year backtest of Core 20, no single holding has gone to zero, and even our worst-performing holdings have rarely cost more than 2-3% of total portfolio value when they ultimately got swapped out at rebalance.
"What about black swans?"
A 20-stock portfolio is more exposed to a single-stock catastrophe than a 500-stock index. That's true. But over the long run, the 20-stock portfolio's higher expected return compensates for that idiosyncratic risk by a significant margin. The math on this has been worked out repeatedly.
"Why not just own an index fund?"
Index funds are great. We own them ourselves. The right answer for most people is a mix: an index core for the bulk of your wealth, plus an active sleeve of high-conviction holdings (15-30 stocks) for the part of your portfolio where you want to try to beat the market.
The active sleeve is what we focus on. The index part you can do at Vanguard or Fidelity for free.
Sector diversification matters more than count
A 20-stock portfolio that's all in software stocks is not really diversified. When the software sector rolls over, all 20 names go down together.
A 20-stock portfolio that spans technology, healthcare, financials, energy, consumer staples, and industrials gives you the actual benefit of diversification, different sectors march to different drums.
When we construct portfolios at Advising Alpha, sector spread is one of the things we explicitly track. Looking at the Core 20 holdings page, you'll see roughly 6-8 sectors represented, each with 2-4 holdings. That's enough sector spread to get the full diversification benefit.
Equal weight vs market weight
A subtler question: once you've picked 20 stocks, how should you weight them?
Market-cap weighting (each stock weighted by its market cap) is what an index fund does. The biggest companies dominate.
Equal weighting (each stock 1/N of the portfolio) treats every pick the same. The benefit: it forces you to rebalance, selling whichever names have grown over their target weight and buying whichever have shrunk. That's a built-in "sell high, buy low" mechanism that's been shown to add a small but persistent amount of return over time.
We use equal-weighting on Core 20 (5% per name) and Market Masters. Tepper Tactical uses a slight tilt toward conviction-weighted positions. Both approaches work; equal-weight is simpler and easier to maintain.
What this means for you
If you want to beat the market through stock picking:
- Aim for 15-30 well-chosen stocks, not 5 (too risky) and not 100+ (too diluted)
- Spread across sectors: don't concentrate in one industry
- Equal-weight or near-equal-weight for built-in rebalancing benefits
- Mix with index funds for the part of your wealth that doesn't need to outperform
Most retail investors who try to outperform fail because they either own too few stocks (concentrated bets that go wrong) or too many (de-facto index fund with extra fees). The middle path is where the math works.