Diversification: the only free lunch

The economist Harry Markowitz reportedly called diversification 'the only free lunch in investing.' The phrase has been repeated so often it's lost its meaning, so let's recover it. A free lunch is something you get without paying for it. In markets, almost nothing fits that description — higher expected return usually demands more risk, more effort, or more luck. Diversification is the rare exception: spread your money across enough genuinely different things, and you can reduce your risk without reducing your expected return. You get something for nothing. That's worth understanding precisely, because most people misunderstand both what it gives and what it can't.

Two kinds of risk, and only one you can erase

Every stock carries two distinct flavours of risk, and the whole logic of diversification rests on telling them apart.

• Idiosyncratic risk (also called specific or unsystematic risk) — the danger that belongs to that one company. A factory fire, a fraud, a key client lost, a botched product launch, a CEO who resigns. It's particular, and crucially, it's uncorrelated across companies — your software firm's accounting scandal has nothing to do with whether a cement maker's plant floods.
• Market risk (systematic risk) — the danger that hits everything at once. A recession, an interest-rate shock from the RBI, a global sell-off, a pandemic. When the tide goes out, almost every boat sinks together.

Here is the elegant part. Idiosyncratic risks are random and unrelated, so when you hold many companies they tend to cancel out. In any given year, some of your holdings will be hit by bad luck and some blessed by good luck, and across a large enough basket the surprises wash against each other. The risk doesn't disappear from any single stock — it disappears from your portfolio, because you're no longer exposed to any one company's fate.

Correlation: the hidden ingredient

Diversification's power comes entirely from correlation — the degree to which two holdings move together. If two stocks always rise and fall in lockstep, owning both gives you no diversification at all; you effectively own one position twice. If they move independently, or better still in opposite directions, combining them smooths your ride dramatically.

This is the trap that catches people who think they're diversified. Buy a private bank, a public bank, an NBFC and a housing-finance company, and it feels like four holdings. But they're all lenders, all sensitive to the same interest rates, the same credit cycle, the same RBI policy. They are highly correlated. When the credit cycle turns, all four fall together. You have four names and one bet. Real diversification is about owning things that don't share the same fate, not simply owning a longer list.

How many stocks is 'enough'?

This is the question everyone wants a number for, and the honest answer is a range. The benefit of adding stocks is steep at first and then flattens fast. Going from 1 stock to 10 slashes your idiosyncratic risk enormously. Going from 10 to 20 helps meaningfully. Going from 20 to 30 helps a little. Beyond roughly 25–30 genuinely different holdings, you've removed almost all the company-specific risk you ever could, and each new name barely moves the needle.

So a rough, defensible answer for a stock portfolio is somewhere around 15 to 30 names across different sectors — enough that no single disaster can sink you, few enough that you can actually follow each business. The exact number matters far less than the spread: 15 stocks across ten unrelated industries beats 30 stocks across two.

The other failure: over-diversification and closet indexing

If too few stocks is dangerous, too many is merely pointless — and often expensive. Spread your money across 80 stocks and you've built something that behaves almost exactly like the index, except you're paying more attention (and possibly more fees) to achieve what a cheap index fund would have handed you for a fraction of a percent.

This has a name in the mutual-fund world: closet indexing. A fund charges active-management fees while quietly hugging the benchmark so closely that it can't really underperform — or outperform. You pay for a stock-picker and receive an index in disguise. The lesson cuts both ways: if you're going to diversify so widely that you'll match the index, buy the index on purpose and stop paying for the illusion of selection.

Diversifying across more than just stocks

Spreading across companies handles idiosyncratic risk. To soften market risk, you have to diversify along other dimensions entirely — across asset classes, across geographies, and across time.

• Across sectors — the within-equity spread we've discussed: lenders, consumer, pharma, IT, industrials, energy.
• Across asset classes — equities, debt (bonds, fixed deposits), gold, real estate, cash. These respond to different forces; when stocks fall in a panic, high-quality bonds and gold often hold or rise, cushioning the whole.
• Across geographies — adding some global exposure means you're not betting your entire future on a single country's fortunes (more on this in a moment).
• Across time — spreading your entry so you don't commit everything at one price on one day.

Diversifying across time: the SIP

The cleanest Indian example of time-diversification is the Systematic Investment Plan (SIP) — investing a fixed rupee amount at regular intervals, say monthly, regardless of the market level. This is rupee-cost averaging, and its logic is quietly clever.

Because you invest a fixed rupee amount each month, you automatically buy more units when prices are low and fewer units when prices are high. You don't have to time anything — the mechanism tilts you toward buying more cheaply, on average, than a single lump-sum at a random moment. Just as importantly, it removes the agonising decision of when to invest, which is where most people sabotage themselves by waiting for a 'better' level that never feels like it's arrived.

Home-country bias

Investors everywhere suffer from home-country bias — the tendency to hold far more of their own country's assets than makes sense, simply because it's familiar. An Indian investor with 100% in Indian stocks is making a large, mostly unexamined bet: that the rupee, Indian policy, and the Indian economy will all cooperate over their investing life. India may well do brilliantly. But 'all my eggs in the country I happen to live in' is a concentration, not a strategy.

There's a second, sneakier layer. Your income — your job, your salary, your career — is almost certainly tied to the Indian economy too. So is your home, if you own one. If your human capital and your home are already a giant bet on India, loading your entire investment portfolio onto the same country doubles down on a wager you've already placed without choosing to. A slice of global exposure isn't unpatriotic; it's simply refusing to let one nation's luck decide your entire financial life.