Diversification: How It Reduces Risk and Where It Stops Working

Diversification reduces risk by spreading exposure across things that do not move together. Ten customers are safer than one only if their orders rise and fall for different reasons. The benefit is largest at the first few additions and shrinks with each one after that. Where the exposures share one underlying driver, spreading stops working entirely: they all move together exactly when it matters.

Concentration risk measures how exposure clusters. Diversification is the cure, and the cure carries fine print: what it actually spreads, why the first few moves do almost all the work, when to stop adding, and how to spot the diversification that exists only on paper.

What is diversification, and what exactly does it spread?

Watch two vendors on the same street. One sells umbrellas, the other sells ice cream. A rainy week is a great week for the first and a terrible one for the second; a sunny week reverses it. Now imagine one person running both carts. Their combined income barely notices the weather, not because either business became safer, but because the two swing for opposite reasons and the swings partly cancel.

The cancelling is diversification, and notice precisely what got spread: not effort, not attention, but exposureAny single point that a material share of a holder's money, revenue or value flows through. Defined fully under concentration risk., the dependence of income on one thing happening. A shop owner who works in two locations has spread effort. An investor whose portfolioThe full collection of investments a person or fund holds, viewed as one basket. holds businesses that prosper in different weathers has spread exposure. Only the second is diversification.

Why does spreading work at all?

Because separate reasons to move means the moves partly cancel. In any given season some exposures are up, others are down, and the total swings less than its pieces. The word doing all the work in that sentence is separate. Diversification never makes any single exposure safer; the umbrella cart is exactly as weather-beaten as before. Diversification makes the total steadier, and only for as long as the reasons stay separate.

Why is the benefit largest at the start and smaller with each addition?

The concentration index, run in reverse, shows the same thing. One exposure scores 10,000. Split into two equal exposures: 5,000. The first split removed 5,000 points of concentration. Going from nine exposures to ten, the index moves from 1,111 to 1,000, a removal of 111. The first move did forty-five times the work of the ninth.

The shape of the bars below is one of the most practically useful shapes in finance. Look at it and let it sink in: the move from one customer to two, one product to two, one lender to two is the highest-value diversification any business or investor ever makes. Everything after the fifth addition is fine-tuning.

When does adding more stop being worth it?

Every new exposure has a price tag. A new buyer means onboarding, credit checks, samples, smaller production runs. A new fund in a portfolio means one more thing to monitor. The cost per addition is roughly flat. The benefit collapses along the bars above. Somewhere, usually surprisingly early, the flat cost line crosses above the falling benefit bars, and past that crossing each new name costs more than the risk it removes.

The far end of this curve is familiar enough: the investor who holds twenty-three mutual funds because each one felt like prudence. The twenty-third fund removed almost nothing, costs fees and attention, and mostly holds the same underlying companies as the other twenty-two. Shared holdings of that kind are where diversification stops working, set out below.

Where does diversification stop working entirely?

The fine print governs everything above it. Diversification counts reasons, not names. When the exposures added share one underlying driver, the paperwork has been diversified and the risk left exactly where it was. Everything will move together on the one day togetherness hurts most.

The test is brutal and simple: describe the single event that hurts all of them at once. If one plausible event can be named, the names are one exposure. The two maps below have identical customer counts; only one of them has actually spread anything.

Is Tessora Weaves diversified with three customers?

Run both answers. They disagree, and the disagreement is the lesson. By names: Tessora Weaves' 60/25/15 split scores 4,450 on the Herfindahl-Hirschman Index (HHI)The Herfindahl-Hirschman Index: square each exposure's percentage share and add them. 0 means fully spread, 10,000 means everything on one name. Taught fully under concentration risk.. Equal thirds would score 3,333; ten equal buyers would score 1,000. So there is real room to diversify by names, and the table shows what each step is worth.

Now the second answer. Meridian Retail Group sells to American shoppers, Nordhaven Stores to European ones, Calluna Home to British ones: three names, three countries, and one end marketThe final buyers a product ultimately reaches. Many direct customers can all depend on the same end market. in any season that matters, because western retail demand rises and falls as one weather system. Measured by names, 4,450. Measured by driver, close to 10,000. Tessora Weaves is a one-exposure business holding three contracts.

What does genuine diversification look like for a household saver?

The one-event test applies to two savings patterns. Saver A holds twenty-three equity mutual funds. Saver B holds a fixed deposit, some gold, a share of an ancestral property, and one equity fund. The question is the one established above: what single event hurts everything at once? For Saver A, one bad year in the stock market marks down all twenty-three folios together: underneath the twenty-three names sits one driver. For Saver B, the same bad market year touches one holding of four; the deposit, the gold and the property answer to different forces entirely, interest rates, global fear, the local land market.

No measurement says what either saver should buy, and that judgement depends on circumstances particular to the saver. The measurement is what stands: count the drivers under the names, and a portfolio's diversification is the count of drivers, never the count of folios.

Tessora Weaves Private Limited, Meridian Retail Group, Nordhaven Stores and Calluna Home are invented.
Educational material. Not advice on any investment, tax, budget or market position.