There's a well-traveled piece of investment advice: never put all your eggs in one basket. Diversify. Spread risk across many holdings. Buy index funds.
There's a lesser-known piece of investment advice that runs in the other direction: concentration is how the great investors actually got rich. Buffett's famous quote on this: "Diversification is protection against ignorance. It makes little sense for those who know what they're doing."
Both pieces of advice are true. They apply to different people in different situations. The interesting question is where the line is, and how to think about it for your own portfolio.
This is a piece on the actual math of diversification, what it does (and doesn't), and why 20 stocks is the sweet spot for most thoughtful equity portfolios.
What diversification actually does
Diversification reduces a specific kind of risk: the risk that any single company has something specific to it go wrong.
If you own 100% of one stock and that company has an accounting scandal, you lose most or all of your investment. If you own 5% of 20 different stocks and one has an accounting scandal, you lose 5%. Painful but recoverable.
This is "idiosyncratic risk." Risk specific to individual companies. Diversification eliminates almost all of it.
What diversification doesn't eliminate is "systematic risk." That's the risk that the entire market goes down together (recessions, rate shocks, geopolitical events). If you own 100 stocks and the market drops 30%, your portfolio drops about 30%. Diversification doesn't help with this.
The math of diversification reduces idiosyncratic risk fast at first, then slowly. The first 10 stocks eliminate maybe 80% of idiosyncratic risk. The next 10 add another 10%. Going from 20 to 50 stocks adds maybe another 5%. Going from 50 to 500 adds the last 5%, and most of it is at the margin.
This means there's a clear sweet spot. Twenty well-chosen stocks captures roughly 90% of the diversification benefit. Buying more than that adds diminishing returns and dilutes the conviction of your best ideas.
The case for concentration
If diversification reduces risk, why would anyone want concentration?
Because concentration is where alpha lives.
The math here is the inverse of the diversification math. Concentration amplifies the impact of your high-conviction ideas. If you can identify a stock that's going to outperform and you own 5% of it, the outperformance moves the portfolio. If you own 0.5% of it (because you also own 199 other stocks), the same outperformance barely registers.
Studies of professional fund managers find that their best ideas (typically the top 5 to 10 positions in their portfolios) outperform the rest of their holdings by significant margins. The "long tail" of small positions in most funds underperforms the top conviction names. If those funds owned only their top 10 to 20 positions, they'd outperform their actual portfolios by meaningful amounts.
The S&P 500 is the extreme version of diversification: 500 stocks, each carrying tiny weight. The result is a market-tracking return with no excess return. You get exactly the market's beta, no alpha. That's fine for someone who wants the market's return at minimum cost (which is most retirement savers), but it's not a strategy for outperformance.
Why 20 is the sweet spot
The interesting sweet spot is around 20 to 25 stocks. Here's why.
You capture 90%+ of diversification benefits. As discussed, going from 20 to 50 adds maybe 5% more diversification. The first 20 do the heavy lifting.
You retain enough conviction for outperformance. With 20 holdings, each at roughly 5% weight, your top picks can move the portfolio. A great year on a single name produces real impact. With 200 holdings, no individual decision matters much.
You can actually understand what you own. Twenty companies are knowable. You can read the 10-K. You can follow the news. You can have a sense for whether each holding still belongs in the portfolio. Two hundred companies is too many for any individual investor to track meaningfully.
It's executable for individual investors. Twenty positions can be managed in a personal brokerage account with reasonable transaction costs. Two hundred positions becomes a logistical project even with commission-free trading, and rebalancing is genuinely difficult.
It's the size where signal beats noise. With 5 stocks, one bad pick can sink the portfolio. With 200 stocks, no single pick matters. With 20, the portfolio aggregates the conviction of multiple positions while no single name can break it.
This is also why most actively managed equity funds with strong long-term track records hold somewhere between 20 and 60 names. They want enough diversification to absorb mistakes and enough concentration to produce alpha. Twenty to thirty is the typical answer.
The case against owning every stock
Some investors swear by index funds for everything. The argument is that you can't reliably pick winners, fees compound, and the market's overall return is enough.
For investors who genuinely believe they have no insight (and most don't), this is the right answer. A 0.03% expense ratio S&P 500 index fund will deliver the market return for most of an investing career. That's not nothing. It's actually a fantastic outcome compared to most actively managed alternatives.
But there's a cost to over-diversification that doesn't get talked about. By owning 500 stocks, you're explicitly buying companies you'd never consciously invest in. You're buying mediocre businesses alongside great ones. You're buying companies whose strategies are failing alongside companies executing brilliantly. The market average is a blend of all of it. Some of what you own is genuinely bad.
A thoughtful 20-stock portfolio lets you own only the businesses that pass a quality screen. The screen can be different things (12-figure track records, multiple insider buys, overlap of disciplined institutional holdings, factor-based criteria), but the result is the same: every position is there for a reason. You're not just buying everything because you can't decide.
The role of methodology
Concentration only works if the methodology for choosing the holdings is sound. A 20-stock portfolio of randomly chosen names will dramatically underperform an index. A 20-stock portfolio chosen using a defensible process should outperform on a risk-adjusted basis over long periods.
The best 20-stock portfolios usually share characteristics:
Quality bias. Companies with strong balance sheets, durable competitive advantages, and rational management teams. The screen excludes obvious failures.
Long holding periods. Turnover is low because the underlying logic doesn't change quarter to quarter. Holdings might stay in the portfolio for years.
Conviction-driven sizing. The 20 names aren't equal in quality. Position sizing reflects relative conviction (or, in equal-weight portfolios, the discipline of letting time and rebalancing reveal which names earned their weight).
Defensible exit criteria. A position only leaves when something specific changes (the fundamental thesis breaks, the company changes character, the institutional consensus moves on). Not on price moves alone.
These characteristics describe most successful long-term concentrated portfolios. They're absent from many "active" mutual funds that own 200 names with no clear discipline, which is why those funds rarely outperform.
Where Advising Alpha sits
Core 20 is built on this thinking. Twenty stocks at roughly 5% each. Selection driven by where multiple disciplined institutional investors overlap. Quarterly rebalance. Sell buffer to keep turnover low. The whole point is to capture the alpha of concentration while maintaining the durability of diversification.
Market Masters runs the same logic with more concentration (19 names, more growth-tilted, higher conviction expressed as bigger positioning). Tepper Tactical is even more concentrated, modeled on a single legendary investor's portfolio with risk overlays.
All three sit in the "thoughtfully concentrated" zone, between index-fund diversification and single-stock-bet concentration. The exact number of holdings varies. The principle is the same: enough names to survive any individual mistake, few enough names to let the conviction matter.
The bottom line
Diversification is protection against ignorance, as Buffett said, and most investors should accept ignorance and own broad index funds. That's the right answer for the average retirement saver.
For investors who want active outperformance, concentration is necessary. You can't beat the market by owning the market. The math doesn't allow it.
The sweet spot for thoughtful concentration is around 20 stocks. Enough to absorb mistakes, few enough to let conviction matter. The methodology has to be sound (random concentration is worse than diversification), but a 20-stock portfolio chosen by a defensible process can outperform indexes meaningfully over long periods.
If you're going to be active, be actively concentrated. If you're not going to be active, be passively diversified. The dangerous middle ground is being actively diversified, paying active fees for a portfolio that mostly tracks the index. That's the worst of both approaches.