Diversification — why spreading your money across bets that don't move together steadies the ride without costing you the reward

Personal Money 5 min 80 sources

Hold many things that fail at different times and their swings offset, so a basket is calmer than any single holding — as long as the bets aren't secretly the same bet.

Key takeaways

Spreading money across investments that don't move together lowers how much your total lurches — without lowering the return you expect. That's why diversification is called finance's only free lunch.
The free lunch depends entirely on correlation: bets that rise and fall in step cancel nothing, so thirty tech stocks is really one bet owned thirty times.
The most dangerous risks are the ones you can't see are linked — an index fund that's secretly a tech fund, or savings held in the same company that pays your salary.

Almost everyone has heard “don’t put all your eggs in one basket.” [5] It is the oldest advice in money, and it sounds like caution — play it safe, expect less. What almost no one is told is the strange part: done right, spreading your money out lowers the risk of the ride without lowering the reward you expect. [1] That combination is so rare in finance that people call diversification the closest thing to a free lunch the field has. [6] This edition explains how that free lunch actually works — and how easy it is to think you have it when you don’t.

The mechanism: swings that offset

Diversification is owning a wide enough mix of investments that a bad day for one is often a fine day for another. [1] The engine underneath is a single idea: correlation — whether two things tend to move together. When two holdings rise and fall in step, owning both changes nothing. When they move independently, or in opposite directions, their ups and downs partly cancel, and the combined value lurches less than either piece did alone. [2]

Work it with two holdings, each of which swings about 20% in a typical year. Put half your money in each. If the two move in perfect lockstep, the basket still swings the full 20% — you gained nothing. If they move completely independently of each other, the arithmetic of combining them pulls the basket’s swing down to about 14% — roughly a 30% smaller ride, for free. If they moved in exact opposite directions, the swings would cancel entirely and the basket would barely move at all. The reward you expect from the basket, meanwhile, is just the average of the two — it does not fall when the swing does. That gap is the free lunch: the same expected return, a steadier ride.

Two kinds of risk

The reason diversification can’t take the swing all the way to zero is that not all risk is the kind that cancels. [2] Part of any stock’s risk is specific to that company — a factory fire, a bad product, a fraud. Spread across enough companies, those private disasters land on different names at different times and largely wash out. But another part is shared: a recession, a rate rise, a market-wide panic drags almost everything down at once. That shared part can’t be diversified away by owning more stocks, because owning more stocks doesn’t escape it. [1] Diversification removes the risk that is unique to each bet; it cannot remove the risk they all sit inside.

How much is enough

An old rule of thumb says thirty stocks captures almost all the benefit. It traces to a 1970 study that found a randomly chosen 32-stock portfolio cut the scatter of returns by about 95% compared with holding the entire New York Stock Exchange. [11] But “95% of the way to the market’s risk” is not the same as fully diversified — that figure was measured against US stocks alone, and a truly wide mix reaches across sectors, company sizes, and countries too. [11] The point isn’t a magic number. It’s that the benefit comes fast at first — going from one holding to ten helps enormously — then flattens, and the last mile is about reaching risks that don’t move with the ones you already own. [9]

The trap: bets that only look different

Here is where the free lunch quietly disappears. Diversification only works if your holdings are genuinely different bets. Own thirty tech companies and you own one bet thirty times — they rise and fall together, so their correlation is near the lockstep case, and the swing barely shrinks. [11] The danger is correlation you didn’t see.

Right now that trap is wide open. A handful of giant technology firms — often called the “Magnificent Seven” — have grown so large that the top ten US companies now make up roughly 35% of the entire US market, up from about 18% a decade ago. [21] Market concentration in 2025 passed even its 1932 peak. [16] Because a standard index fund holds companies in proportion to their size, a plain “broad market” fund is now, underneath, heavily a bet on those few similar businesses — a tech fund wearing a diversified label. [36] One chip-maker alone has grown large enough that its swings noticeably move whole retirement funds. [30] Investors holding several such funds can own the same giants several times over and feel diversified while being anything but. [21]

The most personal version

The sharpest hidden correlation is closer to home: your own employer. [37] Many people hold a big slice of their retirement savings in their company’s own stock, often because the employer matched their contributions in shares. [20] But their salary already rides on that company. If it fails, they can lose the job and the savings in the same week — two bets that looked separate but were the same bet all along. [37] Regulators and planners flag concentrated company stock as one of the most common and costly diversification failures precisely because the correlation is invisible until the day it isn’t. [37]

That is the whole lesson in one picture. Diversification is not about owning more — it is about owning things that fail at different times. The free lunch is real, but it is only served to people who check whether their eggs are truly in different baskets, or just in baskets that happen to be sitting on the same shelf.

02 · Lesson · why it matters

The basket is calmer than the eggs

Bets that fail at different times cancel each other's swings — so a mix is steadier than any piece inside it, unless the pieces are secretly the same bet.

The one free lunch

Almost everything in money is a trade. Want more return? Take more risk. Want your cash the moment you need it? Give up some interest. There is one place where the trade seems to break, and people have noticed it for so long they gave it a name: diversification is the closest thing finance has to a free lunch. You spread your money across many things and the reward you expect stays the same — but the ride gets calmer. Something got better and nothing got worse. That is worth stopping on, because free lunches in money almost never exist.

Why the whole is steadier than the parts

Picture two holdings. Each one, on its own, swings hard — up 20% in a good year, down 20% in a bad one. Now hold half your money in each. You might expect the pair to swing just as hard. It doesn’t. On the days one is falling, the other is often rising, and the two motions partly cancel inside your total. The good years and bad years don’t line up, so the peaks and troughs shave each other down. The basket ends up lurching less than either egg inside it — even though each egg is exactly as risky as it ever was.

Nothing was made safer. The individual bets are untouched. What changed is that you stopped looking at them one at a time. The steadiness lives between the holdings, not inside any of them. This is the shape of the whole idea: a mix has a property none of its parts have. Take the same holdings apart and the calm vanishes — it was never in the pieces, only in how their swings met.

The catch is a single word

All of that hangs on one thing: the bets must move differently. The technical word is correlation — whether two things tend to rise and fall together. And it is the whole game.

If your two holdings move in perfect step, holding both is holding one twice. Their swings don’t cancel; they add. The basket lurches the full 20%, and the free lunch never arrives. If they move independently, the swings partly offset and the ride drops to around 14% — a third calmer, for nothing. If they moved in exact opposition, they would cancel completely. The reward you expect never changes across any of this. Only the swing does, and only correlation decides how much.

So diversification is not about owning more. It is about owning things that fail at different times. Thirty companies that all sell to the same customers, ride the same trend, and sink in the same recession are not thirty bets. They are one bet, owned thirty times, wearing the costume of caution.

The danger you can’t see is the linked one

Here is the part that should make anyone humble. The correlation that hurts you is the one you didn’t know was there.

Consider someone with a plain, sensible index fund — the “whole market,” the definition of not putting your eggs in one basket. Underneath, a handful of giant technology firms have grown so large that they now make up a huge share of that fund. The saver owns the broad market and, without choosing to, owns mostly the same few companies several times over. The label says diversified. The contents say one bet. The eggs were moved into different-looking baskets that all sit on the same shelf.

Or consider someone whose retirement savings sit heavily in their own employer’s stock — often because the company handed them shares. It feels like an investment separate from their paycheck. It isn’t. The paycheck and the shares depend on the same firm. If it fails, both go in the same week. Two things that looked like separate baskets were, all along, resting on one floor. The link was invisible right up until the day it mattered, which is the only day it ever shows itself.

What this leaves you holding

The comforting version of diversification is “own a lot of things and you’re safe.” The true version is harder and quieter: own things whose fortunes are genuinely unhitched from one another, and know that you can rarely be sure they are. Some of what moves your holdings together is visible — the same industry, the same country. Much of it is not. A shared reliance on cheap borrowing, on one government’s policy, on a single mood in the market, can knit apparently unrelated things into one when the pressure comes on. The very moment you most need your baskets to be separate — a crash — is the moment hidden links tend to surface and everything falls together.

None of this argues against spreading your money. It argues for holding the idea with both hands: the calm is real, but it is borrowed from a difference between your bets that you can measure imperfectly and never fully see. You are not standing above your portfolio, arranging tidy independent pieces. You are inside a web of connections most of which are hidden from you, holding a steadiness that depends on links you did not put there and cannot fully trace. The free lunch is real. So is the fine print, and the fine print is written in ink you can’t read until the bill comes.

03 · Lab · your turn

The Correlation Dial

Rehearse how a basket's swing shrinks as its bets move less together, and blows back up when they are secretly the same bet.

04 · Hope · carry this

The one genuine free lunch in money was worked out decades ago and handed to everyone for nothing. Some of the most useful knowledge there is belongs to anyone willing to learn how it works.