Diversification: The Closest Thing to a Free Lunch
The one rule investing lets you bend
The risk and return lesson ended on a hard law: you can’t get a high expected return without accepting a wide range of outcomes, because that range is what you’re being paid to bear. No asset is both high-returning and safe — if it were, everyone would pile in until it wasn’t.
That law still holds for any single asset. But there’s a loophole, and it’s the most important idea in all of investing: you don’t have to hold a single asset. When you spread money across several whose fortunes don’t rise and fall together, something almost magical happens to the combination — its range of outcomes shrinks, while its expected return stays put. You get a calmer ride for the same destination. That’s diversification, and it’s the closest thing investing has to a free lunch.
Why a blend is calmer than its pieces
Picture two assets. One year stocks have a rough stretch while bonds hold steady; another year bonds sag while stocks run. If you own both, each one’s bad patch lands on a year the other is doing fine — so the blend never swings as far in either direction as the wild one does alone. The dips don’t disappear, but they partly cancel.
Here’s the part that feels like cheating: the blend’s expected return is just the weighted average of the two assets’ returns — mixing in some bonds doesn’t magically raise it. But the blend’s risk is less than the weighted average of the two risks, because some of the bouncing offsets itself. Lower risk, same return. The gap between “what you’d expect the risk to be” and “what it actually is” is the free lunch, and it comes from one thing only: the two assets not moving in lockstep.
The bold cone is a diversified stocks-and-bonds blend. The faint, wider cone behind it is the future that same blend would have if the two assets moved in perfect lockstep — the no-diversification world, where nothing cancels. The shaded gap between them is what diversification buys you: a higher floor (a better bad case) and a tighter range, for the same typical outcome in the middle.
Things worth trying
- Drag “Move together” from high down to low. This is the correlation between the two assets — how much they tend to rise and fall as one. Up near the top, the bold cone swells to fill the faint one: assets that move in lockstep can’t cushion each other, so there’s nothing to gain. Pull it down and the bold cone visibly pulls inward — the lower the correlation, the more the dips cancel, and the bigger the free lunch. Correlation, not the number of holdings, is the engine.
- Watch the “Typical outcome” card while you do it. It barely moves. That’s the whole point: you’re narrowing the range of outcomes without trading away expected return. Compare that to the “Downside saved” and “Range narrowed” cards, which swing a lot. Risk changes; the typical destination doesn’t.
- Slide “Share in stocks” to 0% or 100%. The gap collapses to nothing — both cones land exactly on top of each other, and “Downside saved” reads $0. With every dollar in one asset there’s nothing to diversify against. Diversification is a benefit you only get by splitting the money; a concentrated bet forfeits it, no matter how good the single asset looks.
- Find the sweet spot. Nudge the stock share down from 100% toward a mix and watch the bad case improve. For two assets this different, adding a slice of the steadier one to an all-stock portfolio can cut the downside sharply while costing only a little expected return — often a trade well worth making.
Correlation is the secret ingredient
The reason “how many stocks do you own?” is the wrong question is that owning fifty stocks that all crash together is barely more diversified than owning one. What matters is correlation — the tendency of two assets to move in the same direction at the same time:
- Correlation near +1 (move together): the assets are practically the same bet. Their dips line up, nothing cancels, and combining them does almost nothing. This is why a portfolio of fifty tech stocks isn’t truly diversified — in a tech downturn, they all fall at once.
- Correlation near 0 (move independently): the assets wander on their own schedules. One’s bad year often lands on another’s flat or good year, so the blend’s swings shrink. This is the realistic, valuable case — stocks and high-quality bonds have historically had low correlation.
- Negative correlation (one zigs as the other zags): the holy grail, and rare. When one reliably rises as the other falls, the cancellation is strongest of all. True negative correlation is hard to find and doesn’t always last, but even low correlation delivers most of the benefit.
The simulator’s “Move together” slider is exactly this dial. Drag it and you’re not changing the assets’ individual returns or risks at all — only how synchronized they are — and yet the blend’s entire range of outcomes breathes in and out. That’s correlation doing the work.
Why it really is “free”
Most good things in investing cost something. Higher expected return costs you a wider range of outcomes. Liquidity costs you yield. Diversification is the rare exception, and it’s worth being precise about why.
When you mix two assets, the math splits cleanly. The return of the blend is the plain weighted average of the parts — no bonus, no penalty. But the risk of the blend is not the weighted average of the parts; it’s less, by an amount that grows as the correlation falls. So you’re getting a reduction in risk that you never paid for in lost return. Nothing else on the menu works like that — which is why Nobel-winning finance calls diversification the only free lunch. You don’t have to forecast which asset will win. You don’t have to time anything. You just have to not concentrate.
The catch — and there’s always a catch — is that diversification only smooths the bumps that are specific to individual assets (one company’s scandal, one sector’s slump). It can’t erase the risk that hits everything at once (a global recession, a market-wide panic). That shared, undiversifiable piece is market risk, and it’s the part you genuinely get paid the risk premium to hold. Diversification’s job is to clear away the unpaid risk — the asset-specific noise you bear for no extra return — so that what’s left is mostly the risk worth being compensated for.
What you actually do with this
- Hold things that don’t all move together. A single broad index fund already diversifies across hundreds of companies in one purchase — that’s most of the free lunch claimed in one step. Adding an asset class with low correlation to stocks (high-quality bonds is the classic) claims more of it.
- Don’t mistake “a lot of holdings” for “diversified.” Ten funds that all track the same market are one bet wearing ten name tags. Diversification lives in low correlation between your holdings, not in their count.
- Treat concentration as a choice with a cost. Putting everything in one stock — even a great one, even your employer’s — means deliberately giving up a protection that was free. Sometimes there’s a reason to concentrate, but know that you’re leaving the free lunch on the table.
- This is the bridge to building a portfolio. Risk and return told you how much risk to take. Diversification tells you how to take it efficiently — getting the most expected return available for whatever range of outcomes you can live with. From here, the question becomes which assets to blend and in what proportions, which is the work of asset allocation.
The habit to keep
Before you add anything to a portfolio, ask not just “is this a good asset?” but “does it move differently from what I already own?” A merely-good asset that zigs when your portfolio zags is often worth more to the whole than a great asset that moves in lockstep with everything else. The goal isn’t to own the best single thing — it’s to own a combination whose pieces cover for each other. Get that right and you’ve claimed the one advantage the market hands out for free.
Key terms
- Diversification — spreading money across assets whose returns don’t move in lockstep, so their asset-specific ups and downs partly cancel and the blend’s range of outcomes narrows without lowering its expected return.
- Correlation — how strongly two assets tend to move in the same direction at the same time, from +1 (perfect lockstep) through 0 (independent) to −1 (perfect opposites). The lower it is, the bigger the diversification benefit.
- Market (systematic) risk — the risk that hits all assets at once and can’t be diversified away. It’s the part you’re paid the risk premium to bear.
- Specific (idiosyncratic) risk — the risk unique to one asset or sector. Diversification’s whole job is to cancel this out, because you aren’t compensated for bearing it.
- The only free lunch — the nickname for diversification: a reduction in risk that costs nothing in expected return, available to anyone willing to not concentrate.
Risk and return set the menu; diversification lets you order more cheaply than the prices suggest. Next comes the work of choosing the actual ingredients and their proportions — asset allocation — where these ideas turn into a real, buildable portfolio.