What is diversification?

"Don’t put all your eggs in one basket" is diversification in a sentence. But the reason it works — and the limits of how much it can do — is more interesting than the proverb, and it’s one of the most useful ideas a new investor can grasp.

Why diversification works

When one asset in your portfolio falls, another may rise or simply hold steady. If two assets always moved together perfectly, holding both would do nothing for you. But in the real world, different assets respond differently to the same news — so combining them smooths out the ride. That’s diversification: risk reduction through combination, and it builds directly on the idea of risk and return.

Correlation: the key number

The concept that makes diversification precise is correlation — a number between −1 and +1 describing how two assets move relative to each other:

• +1 — they move in perfect lockstep. No diversification benefit at all.
• 0 — their movements are unrelated. Combining them reduces risk.
• −1 — they move in exact opposite directions. A perfect hedge.

Most real-world assets sit somewhere between 0.2 and 0.8 — partially related, which still delivers a meaningful (if imperfect) diversification benefit.

The risk you can’t diversify away

Here’s the crucial limit. Risk comes in two flavours:

• Unsystematic risk is specific to one company or sector — a scandal, a product recall. Hold enough different companies and this risk largely cancels out.
• Systematic risk (or market risk) affects everything at once — recessions, rate shocks, panics. No amount of diversification removes it.

Diversification only eliminates the first kind. The market risk that comes with investing at all is the price of admission, and it stays with you.

How many holdings do you need?

You might assume more is always better, but the maths says otherwise. Studies show most of the diversification benefit arrives with roughly 20 to 30 uncorrelated holdings. Beyond that, adding more offers rapidly diminishing returns. Owning 500 stocks doesn’t meaningfully reduce systematic risk compared to 30 well-chosen ones — it just adds complexity and can dilute your best ideas.