How Many Stocks Do You Actually Need Before Adding One More Stops Helping?
6 min read
Loading...
6 min read
You own six stocks. Maybe eight. Someone tells you that is not diversified enough, you need at least fifteen, or twenty, or "more." So how many stocks do you actually need to diversify a portfolio, and is there a point where adding one more stops helping at all?
The honest answer moved more than once as researchers asked sharper questions. The short version: the "15 stocks is enough" number you have probably heard is older and shakier than it sounds, and the real point of diminishing returns sits higher, especially in a market like Indonesia's.
The "15 to 20 stocks is diversified enough" advice traces back to a 1968 study by researchers John Evans and Stephen Archer. They built random portfolios of different sizes out of real stocks and measured how much each portfolio's volatility, how much its value swings around, shrank as they added more holdings. Their finding: once a portfolio held around 10 stocks, its risk already looked close to the risk of the whole market.
Later textbooks rounded that up into a cleaner rule of thumb, one you have probably seen repeated: a portfolio of about 15 stocks removes roughly 85% to 90% of "unsystematic risk," the risk specific to one company, a bad earnings report, a lawsuit, a factory fire, rather than the market as a whole. Twenty stocks reportedly gets you to about 95%. That is where the "15 to 20 stocks" advice still common in personal-finance content comes from.
That still leaves a real chunk of risk on the table. "Systematic risk," the kind tied to interest rates, currency moves or a market-wide selloff, cannot be removed by owning more stocks. Diversification never gets you to zero risk. It only removes the part that is specific to picking the wrong company.
Evans and Archer's own number, 10, and the "15 to 20" rule that followed it, held up as conventional wisdom for two decades. Then researchers went back and asked a sharper question: not "when does average risk stop falling," but "how many stocks do you need before you can be confident, not just on average, that you are actually diversified."
Meir Statman revisited the question in 1987 and found the honest answer was already higher: a well-diversified portfolio of randomly chosen stocks needed at least 30 stocks for an investor using borrowed money, and 40 for one who was not. He came back to it again in 2004, in a paper literally titled "The Diversification Puzzle," arguing the number had grown further still since the 1960s, because individual stocks had become more volatile relative to the market than they used to be.
A separate 2007 study, by Dale Domian, David Louton and Marie Racine, tested something closer to what actually worries a retail investor: not average risk, but the chance of a genuinely bad long-run outcome. Tracking 20-year holding periods across 1,000 large U.S. stocks, they found investors needed more than 60 stocks just to avoid a meaningful "shortfall risk" gap against a 100-stock benchmark, and that the benefit of adding more stocks kept showing up even past 100 names. Their paper's title says it plainly: "100 Stocks Are Not Enough."
Almost all of the research above was done on U.S. stocks. A 2021 study by Rajan Raju and Sobhesh Agarwalla focused specifically on India, an emerging market with a profile closer to Indonesia's than the U.S. one, more volatile individual stocks, thinner liquidity than a developed market. It found that cutting diversifiable risk by 90%, at 90% confidence, took 40 to 50 stocks. Their conclusion was blunt: the traditional 15-to-20-stock rule of thumb "is likely inadequate."
That is a closer real-world proxy for the IDX than any U.S. study. The Indonesia Stock Exchange listed more than 950 companies as of early 2026, per Kompas Data's tracking of BEI's own count, so choice is not the problem for most retail investors. A portfolio built only from the biggest, most familiar names, BBCA, BBRI, BMRI, can look diversified by ticker count while sitting almost entirely in one sector, banking.
Put the four studies side by side and a shape emerges, even without averaging four different methodologies into one fake single number:
| Study | Year | Market | Stocks needed | What they were solving for |
|---|---|---|---|---|
| Evans & Archer | 1968 | U.S. | About 10 | Match the market's average risk |
| Statman | 1987 | U.S. | 30 to 40 | Diversify net of borrowing/lending costs |
| Domian, Louton & Racine | 2007 | U.S. | 60+, still falling past 100 | Avoid shortfall risk over a 20-year hold |
| Raju & Agarwalla | 2021 | India | 40 to 50 | Cut diversifiable risk 90%, at 90% confidence |
The steepest drop happens early: going from 3 or 4 stocks to somewhere around 15 to 20 removes most of the risk that is cheap to remove. That is also the range where concentration risk does the most damage, a portfolio with only 4 or 5 holdings can easily have 30% or 40% sitting in a single name without anyone deciding that on purpose.
Past 20 stocks, the curve keeps falling, just more slowly, and the research above suggests it keeps falling out to 40, 60, even past 100 names before the payoff flattens for good. Past that point, you have mostly rebuilt an index fund by hand, with more effort and more fees than buying one.
Count alone does not settle it, either. Correlation matters as much as how many names you hold: four bank stocks that all fall together on the same BI rate hike do less diversifying work than four stocks spread across four different sectors, whichever side of the 15-versus-50 debate you land on.
You do not need all 950-plus names on the IDX, and you do not need to force your holding count to exactly 40 or 50 either, that is a research finding about risk reduction, not a shopping list. What is worth checking is whether your current holdings are clustered in one sector the way the IHSG itself is concentrated in banks, and whether one more name from a different sector or market would actually reduce your risk, or just add a ticker you already have exposure to under a different name.
Use Explore to screen stocks, ETFs and crypto by market and risk level, and see what is missing from your spread before deciding whether "one more stock" is worth adding.