What Is Diversification? (And Why Owning More Doesn't Mean Owning Safer)

"Don't put all your eggs in one basket" is the closest thing investing has to a commandment, and it points at a real idea. Diversification is spreading your money across many investments so that no single one can sink you — the losses of some are offset by the gains of others. That much is true, useful, and worth doing.

But the slogan hides a second belief that rides along with it: that more diversification is always more safety, so the responsible investor keeps adding positions. I want to argue that this second belief is wrong past a surprisingly low number — and that for a lot of people, "I'm diversified" has quietly become a way to avoid ever having to say why they own something.

First, steelman it: diversification really is close to a free lunch

Let me make the mainstream case as strongly as I can, because the weak version of my argument would just be reckless.

Every holding carries two kinds of risk. Idiosyncratic risk (also called unsystematic risk) is the danger specific to one company — a factory fire, a fraud, a failed product. Systematic risk is the danger that hits the whole market at once — a recession, a rate shock. Diversification's genuine magic is that it cancels out idiosyncratic risk almost for free: when you own many companies whose specific fortunes aren't linked, one company's disaster is another's irrelevant Tuesday, and the bumps smooth out without lowering your expected return. That is a rare free lunch, and anyone who tells you to bet everything on one stock is selling you a lottery ticket, not an edge.

So concede the strong form: if you own one stock, going to twenty is an enormous, almost costless improvement. I'm not disputing that. I'm disputing what happens after twenty.

The counter-thesis: the free lunch runs out fast

Here's the part the slogan never mentions. The risk-reduction from diversification isn't linear — it's a curve with a steep start and a flat, boring tail.

By roughly 20–30 uncorrelated holdings, you've eaten nearly the whole lunch. What's left — systematic, whole-market risk — diversification cannot remove, because every stock falls together in a real crash. Adding your 47th and 88th position doesn't lower risk in any meaningful way. It just means you now own a slightly worse, more expensive version of the index.

There's even a name for the failure mode: diworsification (Peter Lynch's coined term) — adding holdings until your portfolio is so spread out that your good ideas are diluted into irrelevance, while your costs, your attention, and your ability to follow anything all degrade.

The real cost isn't fees — it's that you stop knowing why

The fee math is the small problem. The big one is what over-diversifying does to your thinking.

You can hold a genuine, written [investment thesis](what-is-an-investment-thesis.md) — a specific argument for why a holding will pay off — on maybe five or ten things. Nobody is meaningfully tracking the thesis on their 90th position. So past a point, every position you add is one you can't actually defend. The portfolio gets wider and your understanding of it gets thinner, until "I'm diversified" is doing the job that "I have no idea what I own" should be doing.

That's the sleight of hand worth catching: over-diversification can be real risk management, or it can be ignorance wearing a respectable costume. From the outside they look identical — a long list of holdings. From the inside, one investor can tell you the thesis and disconfirmer for each; the other is hiding behind the count.

A concrete example

Two investors each put in $100,000.

Anika owns 8 companies. She can tell you, in two sentences each, why she owns every one and what would make her sell. When one of them stumbles on bad news, she checks it against her thesis and decides on purpose — buy more, hold, or exit.

Ben owns 140 companies, assembled from tips, screeners, and a vague sense that "more is safer." When one drops 40% on bad news, he doesn't notice for a week, and when he does, he has nothing to check it against — no reason he bought it, so no way to judge whether to sell. He feels safe because the list is long. But he's carrying the same whole-market risk Anika is, paying more to hold it, and he can't make a single deliberate decision because he never had a reason in the first place.

Ben isn't more diversified in any way that protects him. He's just less accountable to himself.

Why this matters

Diversification's real job is to make sure no single mistake can ruin you — and a dozen-ish uncorrelated holdings accomplish that. Stretching far beyond it doesn't buy more safety; it buys the illusion of safety while quietly taxing the one thing that actually compounds: your ability to understand and defend what you own. Spreading thin is what you do instead of knowing things. Knowing things is the harder, better lever.

The honest caveat — because a contrarian who won't concede is just a crank: if you have no interest in researching individual companies, then maximal diversification through a single broad index fund is not the trap — it's exactly right for you, and it beats a concentrated portfolio you can't justify. The critique here is aimed squarely at the active researcher who keeps adding names to feel responsible. If you're deliberately choosing to own everything and think about none of it, that's a coherent, often excellent strategy. Owning 140 names you picked one at a time and can't explain is not.

Try this

Open your portfolio and, next to each holding, write the one-sentence reason you own it and the one thing that would make you sell. Two outcomes, both useful: the names you can fill in are your real portfolio, and the ones you can't are the diworsification — positions you added for the feeling of safety, not a reason. That list of blanks is your homework: defend them or trim them.

Keep those one-liners where you'll re-read them when a price moves and your nerve is being tested. In JustJot.ai, jot the thesis for each position as a linked note; later you can ask your library in plain language — "why do I own this again?" — and get your own frozen reasoning back. Once you're writing a reason per holding, the natural next step is logging how those calls turn out: that's the [investing decision journal](the-investing-decision-journal.md), where "I'm diversified" gives way to the better question — do I actually know what I own?