Course → Unit 7: Portfolios and Risk Management → Lesson 7.1

Diversification

If a single bad piece of news could wipe out your entire portfolio overnight, you wouldn't be investing — you'd be gambling, and this lesson is about the one tool that turns the first into the second.

Don't put all your eggs in one basket — but why exactly?

You've heard the saying since you were a kid, and in investing it has a precise, mathematical meaning. Imagine you put your entire $10,000 savings into a single stock — say, a local F&B chain. If that company's central kitchen catches fire, or its biggest supplier collapses, or a food safety scandal breaks, your $10,000 could become $3,000 in a week. Nothing else in your life changes to cushion that fall — your entire financial position moves in lockstep with that one company's fortunes.

Now imagine you'd spread that same $10,000 across twenty different companies in twenty different industries. If one of them has a terrible month, it might drag your portfolio down by a percent or two — annoying, but survivable. This spreading-out strategy is called diversification — holding a variety of investments so that no single one has the power to determine your overall result. The "why exactly" is simple: company-specific disasters are common, but a disaster that simultaneously destroys twenty unrelated companies is vastly rarer. Diversification doesn't eliminate risk; it eliminates the risk of a single point of failure.

Correlation: how assets move relative to each other

Here's the catch: diversification only works if the things you're combining don't all get hit by the same problem at the same time. This is where correlation comes in — a measure of how closely two investments' prices move together. Correlation is scored from +1 to -1:

• +1 (perfectly correlated) — the two assets move in exactly the same direction, by proportionally the same amount, every time.
• 0 (uncorrelated) — the two assets' movements have no predictable relationship to each other at all.
• -1 (perfectly negatively correlated) — when one goes up, the other goes down by a proportional amount, every time.

Owning 20 stocks sounds diversified, but if all 20 are Singapore banks, they're all exposed to the same interest rate decisions, the same regional economic cycle, and the same regulatory environment. Their correlation to each other is very high — likely above +0.8. When one bank stock falls on bad news about Singapore's economy, the other nineteen are likely to fall too, because they share the same underlying driver. You've diversified against company-specific news (a scandal at one specific bank) but not against sector-wide or economy-wide news — which is most of what actually moves markets.

True diversification means combining assets with low or negative correlation — for example, Singapore bank stocks (which do well when the economy is strong and rates are healthy) alongside government bonds (which often do well when the economy weakens and rates fall). When one zigs, the other tends to zag, smoothing out your overall ride.

The power of combining uncorrelated assets

The mathematics behind this is one of the few genuine "free lunches" in finance. Consider two hypothetical investments, each expected to return 8% a year on average, each with the same level of up-and-down volatility on its own. If they are perfectly correlated (+1), combining them in a portfolio gives you the same 8% expected return with the same volatility as holding just one — no benefit at all. But if their returns are uncorrelated or negatively correlated, combining them still gives you the same 8% expected average return, but with meaningfully lower volatility, because their individual ups and downs partially cancel each other out.

This is why professional portfolio managers obsess over correlation, not just individual asset quality. A portfolio of ten "good" investments that are all highly correlated is riskier than a portfolio of ten "good" investments that move independently of each other — even though both portfolios might contain equally strong individual picks.

How adding one bad stock to a diversified portfolio barely affects it

Here's a concrete worked example. Suppose you hold a well-diversified portfolio of 30 stocks across different sectors and countries, each roughly equal-weighted at about 3.3% of your portfolio. Now suppose one of those companies — say, a small manufacturing firm — turns out to be poorly run, and its stock falls 50% over the year while everything else in your portfolio returns a modest 6%.

Instead of the 6% you would have earned without the bad pick, you earned about 4.1% — a real but modest drag, not a catastrophe. Compare that to putting your entire portfolio into that one manufacturing stock: you'd have lost half your money. This is the practical payoff of diversification — it doesn't make you immune to bad picks (everyone makes bad picks eventually), but it makes any single bad pick survivable. This is also why professional fund managers rarely bet more than a small percentage of a portfolio on any one idea, no matter how confident they feel — conviction is not the same as certainty, and diversification is the insurance policy against being wrong.

The lesson compounds across a lifetime of investing: you will be wrong about some of your individual picks. Diversification is what ensures that being wrong sometimes doesn't stop you from being right, on average, over time.