Why portfolio diversification matters more than ever
- Robin Powell
- 1 day ago
- 8 min read
Harry Markowitz called portfolio diversification "the only free lunch in investing." William Sharpe declared it the golden rule. Yet with seven technology stocks representing over a third of the S&P 500, concentration risk hasn't been this extreme in half a century. Here's what seven decades of evidence tells us about protecting your portfolio from the next crash.
If you own a global tracker fund, you probably think you're well diversified. Thousands of companies across dozens of countries. But the "Magnificent Seven" — Apple, Microsoft, Nvidia, Alphabet, Amazon, Meta and Tesla — now account for roughly 35% of the S&P 500's market capitalisation, nearly three times their weighting a decade ago. Because global trackers weight by market value, your supposedly diversified portfolio may have a third of its equity allocation riding on seven correlated American technology companies. Portfolio diversification has never been more relevant.
A recent survey by Bank of America found that 45% of fund managers view an "AI bubble" as the biggest tail risk in markets. The Bank of England and IMF have both warned of potential "sharp correction."
Whether these warnings prove prescient, nobody knows. But we know what happened the last two times markets became this concentrated. When the dotcom bubble burst, the NASDAQ fell 78%. When the global financial crisis struck, investors in financial stocks watched holdings collapse by 80% or more. In both cases, diversified investors suffered far smaller losses and recovered faster. Many concentrated investors never recovered at all.
What Markowitz discovered
Before 1952, portfolio construction was guesswork dressed up as expertise. The prevailing wisdom: identify securities with the best prospects. Pick winners, avoid losers.
Harry Markowitz saw a problem. If maximising expected returns were the only goal, investors should put everything into whichever single security offered the highest anticipated return. Yet nobody did this. Portfolio diversification was "both observed and sensible," as he put it. A theory that couldn't explain why people diversified was no theory at all.
His 1952 paper, Portfolio Selection, offered a mathematical alternative. Markowitz demonstrated that combining assets with low correlations could reduce a portfolio's variance without proportionally reducing expected return. You can reduce risk without giving up return by holding assets that don't move in lockstep.
He was precise about what this meant. "It is not enough to invest in many securities," he wrote. "It is necessary to avoid investing in securities with high covariances among themselves." A portfolio of 60 railway stocks would not be as well diversified as a smaller portfolio spread across railways, utilities, mining and manufacturing.
This explains why owning shares in all seven of today's tech giants isn't diversification. These stocks are concentrated in a single sector, driven by overlapping themes, and highly correlated. When technology sentiment shifts, they move together.
Markowitz's framework also clarifies what diversification cannot do. Spreading investments eliminates "idiosyncratic" risk — the danger that any single company or sector collapses. Enron goes bankrupt? A diversified investor barely notices. But when the entire stock market falls, all equities decline together. There is no hiding within equities alone — which is why diversification across asset classes, particularly into bonds that typically move opposite to stocks during stress, matters so much.
Concentration risk pays no premium. You don't earn higher expected returns for betting everything on technology rather than owning the whole market. You accept more variance for the same expected outcome.
Lessons from the dotcom crash
The technology bubble of the late 1990s offered a masterclass in concentration risk.
From March 2000 to October 2002, the NASDAQ fell roughly 78%. The S&P 500, with broader sector exposure, declined around 47%. Analysis by Jack Bogle found the average large-cap technology mutual fund lost approximately 71%.
Consider what recovery requires. A 50% decline needs a 100% gain to break even. A 78% decline requires 355%. The NASDAQ didn't reclaim its March 2000 peak until 2015.
Diversified investors saw drawdowns of 20% to 50% — uncomfortable, but fundamentally different. Most striking: the Russell 2000 Value index posted a small positive return over the bear market, even as the Russell 2000 Growth index collapsed by more than 50%.
UK investors concentrated in domestic technology darlings fared no better. Marconi, once valued at over £30 billion, saw its shares collapse from over £12 to pennies. The FTSE 100 fell 52% peak to trough, but globally diversified portfolios with value exposure experienced shallower losses and faster recoveries.
Technology investors in 1999 had compelling narratives. The internet was transforming commerce. Some narratives proved correct: Amazon did change the world. But Amazon fell 94% from its 1999 peak to its 2001 trough. Being right about the future didn't protect concentrated investors from the present.
Lessons from the global financial crisis
The 2007-2009 crisis demonstrated that concentration risk can strike any sector investors consider "safe".
Banks were not speculative bets in 2007. They were dividend-paying blue chips with centuries of history. The FTSE 100 was heavily weighted toward financials, and many UK investors held even more through income-focused funds.
US financial stocks fell more than 55% in 2008. RBS and HBOS lost 80% to 90% of their value and required emergency government intervention. The S&P 500 fell 37%. Consumer staples declined only 15%.
During the panic, correlations between stock markets spiked toward 1.0. Global equity diversification offered little protection. But UK gilt funds returned 11.6% while equities imploded. Investors holding a 60/40 portfolio experienced drawdowns of around 30% to 35%, roughly half the equity-only decline.
A portfolio of 100 stocks, all in financials, offered no protection. A simpler portfolio of one global equity fund and one bond fund would have cut losses nearly in half.
The lost decade
The first decade of the 21st century offered a brutal lesson in why portfolio diversification matters.
The S&P 500 delivered roughly zero total return over those ten years. Yet emerging market equities were the decade's standout performers. The UK's FTSE 250 roughly tripled, vastly outperforming the FTSE 100's 50% return. A balanced 60/40 portfolio delivered solid positive returns.
UK investors who held only FTSE 100 stocks owned a concentrated bet on banks, oil and miners. When financials collapsed and commodity prices crashed, they suffered doubly. Global diversification would have captured gains wherever they occurred.
What recent research tells us
The COVID-19 crash of March 2020 — the S&P 500 dropping 34% in 23 trading days — provided another natural experiment. Academic research published since confirms Markowitz's insights remain robust.
Cryptocurrencies disappointed those who had marketed them as "digital gold." Studies by Abdelmalek (2023) and Colombo and colleagues (2021) found that during normal conditions, crypto showed low correlation with traditional assets. But during the COVID crash, correlations with equities increased sharply. Bitcoin fell alongside stocks, not against them.
Research by Tiwari and colleagues (2022) and Yousaf, Suleman and Demirer (2021) suggests ESG investments behave like conventional counterparts during stress — useful for values-aligned investors, but not a source of crisis protection.
Seven decades after Portfolio Selection, the evidence points the same way.
Why this matters now
History doesn't repeat, but concentration risk follows familiar patterns.
Whether current market concentration reflects rational pricing or speculative excess is debated. The bull case emphasises genuine earnings growth and transformative technology.
The bear case points to valuations untethered from profits and circular financing — Nvidia investing in OpenAI while OpenAI buys Nvidia chips — that resembles patterns seen before previous bubbles burst.
OpenAI's CEO Sam Altman has acknowledged investors are "overexcited". Torsten Slok at Apollo Global Management argues the top ten S&P 500 companies are more overvalued than during the 1990s bubble. A 2025 MIT report found 95% of organisations getting "zero return" from generative AI investments.
Prudent investors don't need to predict whether AI stocks will crash. The question is simpler: does concentrating a third of your equity portfolio in seven correlated companies offer adequate compensation for the risk?
If AI stocks continue rising, diversified investors will capture most of the gains through their index holdings. If they crash, diversified investors suffer a painful but survivable setback rather than a devastating one.
What true diversification looks like
Effective diversification doesn't require complexity. A handful of low-cost index funds can provide broader exposure than most professionally managed portfolios.
Start with asset classes. The most important decision is how to divide your portfolio between equities and bonds. A 60/40 split has historically delivered roughly two-thirds of pure equity returns with substantially less volatility.
Diversify geographically. The FTSE 100 concentrates heavily in financials, energy and mining. A global equity allocation spreads exposure across regions, capturing growth wherever it occurs.
Spread across styles. Growth and value stocks take turns outperforming. Rather than guessing which will win, own both.
Be sceptical of alternatives. Cryptocurrencies have shown increased correlation with equities during crises. For most investors, global equities and quality bonds provide the most reliable foundation.
Rebalance periodically. Check quarterly; trade when allocations drift beyond tolerance bands of 5% to 10%.
The simplest diversified portfolio might contain just two funds: a global equity index and a bond index.
The behavioural challenge
Diversification is simple to understand and difficult to live with.
A properly diversified portfolio always contains something disappointing. When technology stocks soar, your bond allocation drags. When US markets outperform, international holdings look like dead weight. This is diversification working as intended.
In the late 1990s, diversified investors watched technology funds return 30%, 40%, 50% annually while balanced portfolios delivered steady but unspectacular gains. Many abandoned their strategy to chase performance, often just before the crash.
The behavioural finance research is clear. Investors consistently underperform the funds they own through poor timing — buying after prices rise, selling after prices fall. Dalbar has documented this "behaviour gap" at roughly 1.2 percentage points annually. Diversification helps close this gap: portfolios with shallower drawdowns generate less panic.
When colleagues boast about Nvidia gains, holding a diversified portfolio feels like missing out. But regret works asymmetrically. Missing gains stings briefly. Losing half your retirement savings leaves scars.
Portfolio diversification: the only free lunch
Markowitz published Portfolio Selection in 1952. Seven decades later, his central insight remains the most reliable principle in investing: combining assets that don't move in lockstep reduces risk without sacrificing expected return.
Each crisis has punished concentration and rewarded breadth. The current market presents a familiar challenge in new clothes.
But the investor's task isn't prediction. It's preparation.
Portfolio diversification ensures no single company, sector or theme can devastate your financial future. It trades the possibility of spectacular gains for survivable outcomes — a trade-off that looks unfavourable during bubbles and invaluable after they burst.
You cannot control market returns. You cannot know which stocks will crash. What you can control is whether your portfolio's fate depends on a handful of correlated bets or the collective productivity of thousands of companies worldwide.
Resources
Abdelmalek, W. (2023). Cryptocurrencies and portfolio diversification before and during COVID-19. EuroMed Journal of Business, 19(4), 1084–1120.
Colombo, J., Cruz, F., Paese, L., & Cortes, R. (2021). The diversification benefits of cryptocurrencies in multi-asset portfolios: Cross-country evidence. SSRN Electronic Journal.
Markowitz, H. (1952). Portfolio selection. The Journal of Finance, 7(1), 77–91.
Tiwari, A. K., Aikins Abakah, E. J., Gabauer, D., & Dwumfour, R. A. (2022). Dynamic spillover effects among green bond, renewable energy stocks and carbon markets during COVID-19 pandemic. Global Finance Journal, 51, 100692.
Yousaf, I., Suleman, T., & Demirer, R. (2021). Green investments: A luxury good or a financial necessity? SSRN Electronic Journal.
Have you watched this video?
Our colleagues at Regis Media recently produced this video about Portfolio Selection, the 1952 paper which eventually earned Harry Markowitz the Nobel Prize in Economic Sciences. It's the first in a new series of videos for Index Fund Advisors, called The Papers That Changed Investing. You can follow the whole series by subscribing to the IFA YouTube channel.