Diversification Works Because of Correlation, Not Because of Counting
Owning thirty stocks that all move together is one position expressed thirty ways. The benefit of diversification comes from how holdings relate, not from how many there are.
The Core Idea
Combining assets that do not move identically produces a portfolio less volatile than the average volatility of its components. That result is not a heuristic, it falls directly out of the arithmetic of variance, and it is the closest thing to a free lunch in finance.
The essential input is correlation, which measures how two assets move relative to each other on a scale from negative one to positive one. At positive one they move identically and combining them reduces nothing. Below one, some volatility cancels. At negative one, a specific combination eliminates volatility entirely.
Why Counting Holdings Misleads
The common measure of diversification is the number of positions, which is close to useless on its own.
A portfolio of thirty regional bank stocks holds thirty names and one exposure. They share interest rate sensitivity, credit cycle exposure, deposit funding risk, and regulatory regime. In a banking stress event they fall together, which is precisely when diversification was supposed to help.
A portfolio holding a broad equity index, government bonds, and real assets may contain fewer decisions and far more genuine diversification, because the return drivers differ.
Ask what your holdings have in common rather than how many you own. The shared exposure is the position you actually hold.
The Diminishing Returns
Within a single asset class, most of the available benefit arrives quickly. Classic studies found that the majority of diversifiable risk in a stock portfolio is eliminated with roughly twenty to thirty holdings, with limited improvement beyond.
The reason is that diversification can only remove company specific risk. It cannot remove market risk, which affects everything simultaneously. Once idiosyncratic risk is largely gone, adding names does little except increase administrative burden and dilute conviction.
The Failure Mode
The critical weakness is that correlations are not stable. They are estimated from historical data and they change, generally in the least convenient direction.
During severe stress, correlations across risk assets tend to converge toward one. Assets that appeared unrelated in normal conditions fall together, because the driver stops being their individual fundamentals and becomes a common factor: investors needing liquidity, deleveraging, or a collapse in risk appetite.
This occurred in 2008 and again in March 2020. Investors who believed they held diversified portfolios discovered that the diversification was measured in calm periods and evaluated in a crisis. That is the fundamental limitation, and no amount of historical optimization fixes it.
What Actually Diversifies
Genuine diversification requires assets whose returns depend on structurally different drivers, not merely on different companies.
Government bonds have historically provided this against equities, because they benefit from the rate cuts that typically accompany a downturn. That relationship is not guaranteed, and 2022 demonstrated the exception, when inflation drove both stocks and bonds down together and the standard balanced portfolio suffered its worst year in generations.
The honest conclusion is that diversification reduces risk substantially in most environments and least in the environments that matter most. Recognizing that is more useful than pretending a correlation matrix solves it.
The Bottom Line
Diversification is about correlation rather than count, and correlations rise toward one exactly when you need them low. Build portfolios around distinct return drivers and expect less protection in a genuine crisis than the historical numbers imply.