Portfolio Diversification: 6 Rules for Spreading Risk (and the Risk It Can't Remove)

Portfolio diversification is the closest thing investing has to a free lunch, and it's also the most misunderstood idea in the beginner playbook.

Spreading your money across many investments really does lower your risk without lowering what you can expect to earn. But it lowers one specific kind of risk, the kind tied to a single company or industry. It does very little about the risk that the whole market drops, and in some years even bonds have fallen right alongside stocks.

Below is what diversification fixes, where it falls short, what 40 years of single-stock data shows, and six rules for building a diversified portfolio.

What portfolio diversification does (and what it can't do)

Every stock you own carries two kinds of risk.

The first is company risk, which textbooks call unsystematic risk. A CEO makes a bad bet, a drug trial fails, a competitor eats the market. That risk belongs to one company or one industry, and you can shrink it to almost nothing by owning lots of companies. When one of them stumbles, it's a small slice of your money.

The second is market risk, or systematic risk. Recessions, interest-rate shocks and pandemics hit nearly every company at once. Owning more stocks won't get rid of it, because it's the risk of owning stocks at all.

Your main lever against it is how much of your money sits in stocks versus bonds and cash. That split is your asset allocation.

How many stocks does it take to wash out company risk? More than most people think. Meir Statman's study of how many stocks make a diversified portfolio, published in 1987, found you need at least 30 to 40 randomly chosen stocks, not the 10 or so that was the accepted wisdom at the time. A single total-market index fund holds thousands, so it does the job in one purchase.

Portfolio diversification cuts the odds of a disaster. It doesn't cut out losses. The S&P 500, which holds 500 companies, still lost 36.6% in 2008 with dividends included, according to NYU Stern professor Aswath Damodaran's historical returns data.

The cost of putting too much in one stock

Concentration feels brilliant while it's working. The long-run numbers are less kind.

J.P. Morgan's Agony and Ecstasy study of concentrated stock positions, published in March 2021, tracked every company that was ever in the Russell 3000 index (roughly the 3,000 largest US companies) from 1980 to 2020. The results, by share of companies:

• 44% suffered a "catastrophic loss," a drop of 70% or more from their peak that never recovered.
• 42% had negative returns over their lifetime, so plain cash would have beaten them.
• 66% did worse than the index as a whole.
• Only 10% became "megawinners," beating the index by 500% or more.

The index did fine because a small group of huge winners carried it. If you own the index, you own those winners automatically. If you own five stocks, you're betting you picked them. On those odds, about two of five randomly picked stocks would eventually hit a catastrophic loss.

You can't reliably spot the losers in advance, either. J.P. Morgan's October 2024 update found that 54% of recent catastrophic decliners were profitable at their peak, and most traded at "reasonable" valuations.

Do stocks and bonds really move in opposite directions?

Bonds are a different asset class from stocks, and in most of the bad years since 2000 they did their job. Damodaran's data shows the S&P 500 lost 36.6% in 2008 while 10-year Treasury bonds gained 20.1%. In 2002, stocks fell 22.0% and Treasuries rose 15.1%. By our calculation, a simple mix of 60% stocks and 40% Treasuries, rebalanced each year, would have lost about 13.9% in 2008 instead of 36.6%.

The 2020 crash followed a similar script. The Fed's May 2020 Financial Stability Report described equity prices plunging while yields on longer-dated Treasuries fell to very low levels. Falling yields mean rising bond prices. Corporate bonds helped less, as their spreads over Treasuries jumped to the widest since 2009.

Then came 2022.

Inflation surged and the Fed raised rates at its fastest pace in decades. US stocks fell 18.0% for the year and 10-year Treasuries fell 17.8%. That same 60/40 mix lost about 18%, barely better than owning stocks alone.

2022 wasn't a glitch. Correlation measures how closely two investments move together, and it shifts over time. J.P. Morgan's March 2024 analysis of stock-bond correlation found that, going back to 1940, stocks and bonds moved in opposite directions only 38% of the time.

The low-inflation 2000s and 2010s were the unusual stretch. When inflation runs hot, both can fall together.

Does that make bonds pointless? I don't think so.

In five of the six down years in the chart, Treasuries rose or held flat, and they're still the main way to dampen volatility for money you'll need within a few years. Notice the cash bars in the chart, too. Cash in 3-month Treasury bills earned a positive return in every one of those down years.

Why international stocks belong in the mix

A US-only portfolio spreads your money across companies, but it's still a bet on one country.

Over the past decade that bet paid off. In Vanguard's October 2025 discussion of international diversification, its economists note that US stocks returned about 13.1% a year from 2014 to 2024, beating international markets by 7.7 percentage points a year in dollar terms. They also point out that US stocks lagged other developed markets in three of the last five decades: the 1970s, the 1980s and the 2000s.

The 2000s are the warning. Using Damodaran's data, the S&P 500 lost 9.1% in total from 2000 through 2009, dividends included. Someone who put a lump sum into US stocks on the first day of 2000 had less money ten years later.

Vanguard's projections, as of October 2025, put international stocks ahead over the next decade. Treat that as a forecast, not a promise.

One honest caveat. International stocks help most over long stretches. In a sudden panic like early 2020, markets around the world tend to fall at the same time.

How to diversify your portfolio: 6 rules

1. Own the whole market with index funds

An index fund (many are ETFs) buys every company in an index, so you get hundreds or thousands of stocks in one purchase at a low expense ratio. A total-market fund goes further than an S&P 500 fund by adding mid-size and small companies. Our index investing guide for beginners covers the basics.

2. Go global

You can own a single whole-world fund, or pair a US fund with an international one. We compared the two approaches in whole-world vs US-only index funds.

3. Add bonds based on when you need the money

Stocks can drop a third in a single year, which is a real problem for money you need in the next two to five years. Money you won't touch for 30 years can carry far more stock risk, if your risk tolerance lets you sleep through a 35% drop. Our guide to asset allocation by age walks through how the stock-bond split usually shifts over time, and our explainer on the six types of bonds every investor should understand covers what goes in the bond side.

4. Check your funds for overlap

Owning more funds doesn't mean owning more variety. FINRA's warning on concentration risk flags overlapping tech holdings across individual stocks, mutual funds and index funds as a common blind spot. Three US large-company funds can end up as one fund with three fees. Check each fund's top 10 holdings on the provider's website before you add it.

5. Keep any single stock small, including your employer's

My rule of thumb is to keep any one company at 5% of your portfolio or less. Employer stock deserves extra care, because your paycheck already depends on that company. If it struggles, you could lose your job and your savings in the same month.

6. Rebalance on a schedule

Markets move your mix around without asking. A few strong years for stocks can turn a 70/30 mix into 80/20, and your risk rises with it. FINRA's guide to asset allocation and diversification suggests a yearly review, and you can rebalance by steering new contributions to whatever's behind before you sell anything. Our guide to portfolio rebalancing covers the calendar and threshold methods.

What about real estate, gold or crypto? They're optional extras, not the foundation. Gold rose 25.6% in 2002 but only 0.6% in 2022, per Damodaran's data. If you add any of them, keep the slice small.

A worked example: Leah's "diversified" portfolio

Leah is 27 and has $40,000 invested. She owns four things, so she figures she's diversified:

• $12,000 in her employer's stock, a mid-size tech company
• $10,000 in an S&P 500 index fund
• $10,000 in a Nasdaq-100 fund
• $8,000 in a technology sector fund

Look closer and it's one bet in four wrappers. It's all US stocks, with no bonds and no international exposure. The biggest tech companies sit near the top of three of her funds. And 30% of her money rides on the company that also pays her salary.

Now say her employer's stock has a catastrophic year and falls 70%. Her $12,000 becomes $3,600. That $8,400 hit wipes out 21% of her whole portfolio from a single company.

One way she could restructure it is to put $34,000 in a whole-world index fund, $4,000 in a bond fund and keep $2,000 in employer stock (5%). The same 70% drop in her employer's stock would now cost $1,400, or 3.5% of her portfolio. She still owns plenty of tech, just at its market weight across thousands of companies.

This is an illustration, not a recommendation. Her right mix depends on her goals and timeline. But the difference between those two outcomes is what portfolio diversification is for.

The short version

• Portfolio diversification removes company risk. It can't remove market risk.
• Owning one stock is a long shot. From 1980 to 2020, 44% of US companies suffered a catastrophic loss and two-thirds trailed the index.
• Bonds cushioned stocks in 2002, 2008 and 2020, but fell with them in 2022. Cash stayed positive in every down year.
• Other developed markets beat US stocks in three of the last five decades.
• Index funds, a global mix, bonds by time horizon, no overlap, small single-stock bets and a yearly rebalance cover most of what a beginner needs.

You can't control what the market does next year. You can control how much damage any one company is allowed to do.