Diversification: Why It Works — and Why It Has Limits for Real-World Investors
The classic portfolio principle explained, illustrated, and reframed through a practical lens.
I recently came across a post titled This Is the Only Free Lunch in Investing that walks through a key insight often attributed to Nobel laureate Harry Markowitz:
“Diversification is the only free lunch in investing.”
It’s one of the most widely repeated principles in investing — and for good reason. Diversification is grounded in decades of theory, data, and experience. By spreading your investments across different assets, you reduce your exposure to any one source of risk.
But like many good ideas, diversification has limits — especially for real-world investors with constraints like limited capital, no access to leverage, or short time horizons.
In this post, we’ll explore what diversification actually does, where it helps most, and where it can quietly fall short.
What Diversification Actually Does
In plain terms, diversification reduces volatility without necessarily sacrificing expected return — at least in theory. When you combine assets that don’t move perfectly together, your portfolio's ups and downs can smooth out, even if the individual assets remain volatile.
This is easiest to see with a mix of stocks and bonds. Stocks are volatile but have higher expected returns. Bonds are more stable but offer lower returns. Together, they tend to offset each other — creating a portfolio that’s smoother and potentially more efficient.
Here’s what that looks like in practice:
Cumulative Returns with Annual Rebalancing: VTI/BND Blends
For this example, we use two well-known, low-cost index funds:
VTI (Vanguard Total Stock Market ETF): A broad fund that represents the entire U.S. stock market — large, mid, and small caps. It serves as our equity building block, with high expected return and volatility.
BND (Vanguard Total Bond Market ETF): A diversified U.S. bond fund covering investment-grade government and corporate bonds. It serves as our lower-risk, lower-return anchor.
We chose these because they’re widely available, low-cost, and representative of the major asset classes most retail investors can access easily in the U.S.
Here’s what different combinations look like over time:
Total cumulative returns for portfolios rebalanced annually. Each line shows a different stock/bond allocation using VTI and BND. Sharpe ratios are annualized.1
Note for UK and EU-based investors:
If you can’t access U.S.-domiciled ETFs like VTI or BND due to local regulations, the same principle applies using UCITS-listed equivalents.2 For example:
IWDA (iShares Core MSCI World UCITS ETF) — a global developed markets ETF that includes the U.S. and serves as a diversified equity building block
AGGH (iShares Core Global Aggregate Bond UCITS ETF) or VUCP (Vanguard Global Bond UCITS ETF) — diversified bond exposures similar in purpose to BND
These UCITS ETFs offer comparable building blocks for diversified portfolios — and are widely available to EU and UK investors.
Why Sharpe Ratio Alone Can Mislead
Let’s say you build a beautifully diversified portfolio of equities and bonds. You backtest it and find that the Sharpe ratio — return per unit of volatility — is higher than just holding equities.
That’s great in theory. But here’s the catch:
Suppose your equity portfolio returns 10% annually with volatility of 15% (Sharpe ≈ 0.67). Your diversified portfolio returns 6% with volatility of 7% (Sharpe ≈ 0.86).
The Sharpe ratio is higher — but as a retail investor, you can’t easily leverage that smoother portfolio to boost returns to equity levels. The result? Lower long-term wealth, despite better risk-adjusted metrics as shown in the figure above.
When Diversification Works Best
Diversification shines when:
You want to manage drawdowns (e.g. near retirement)
You want a smoother ride you can stick with
You have uncorrelated assets with similar returns
But diversification isn’t magic — and its benefits depend on how long you’re willing to stay invested.
A Useful Rule of Thumb: Time = (1.64 / Sharpe)^2
If returns are normally distributed, this formula tells you how many years it takes for the mean return to dominate the noise with 95% confidence.3
That means:
A portfolio with Sharpe of 0.5 takes ~11 years to reliably outperform cash.
A Sharpe of 1.0 needs just ~2.7 years.
Final Thought
Diversification is about balance — between risk and return, and between short-term noise and long-term growth. But it’s not magic. You still need to think in terms of goals, time horizon, and real-world constraints.
The free lunch is there — but sometimes, you need to bring your own cutlery.
Coming Soon
Next week’s post:
Where Should You Park Your Cash?
Understanding T-Bills, ETFs, and the Real Return on Safe Money
Until then, stay invested—and stay curious.
👉 Subscribe to The Practical AInvestor to get posts and tools straight to your inbox:
Sharpe = (Mean − RiskFree) / Vol, where Mean is the annualized average return, RiskFree is the annualized risk-free rate (often assumed 0 in simplified cases), and Vol is the annualized standard deviation of portfolio returns.
UCITS stands for “Undertakings for Collective Investment in Transferable Securities” — a framework that allows funds to be sold across the EU and UK with standardized regulation.
The value 1.64 corresponds to the 95th percentile of the standard normal distribution.