“Don’t put all your eggs in one basket.”
Everyone knows this. Most investors still do it. They hold ENI, Intesa, Generali, Ferrari, and Stellantis and think they’re diversified because it’s “five different companies.” It isn’t. Today’s lesson is what diversification actually means, mathematically and practically, and how to achieve it.
The two kinds of risk
Stock risk splits into two categories:
Idiosyncratic risk (company-specific)
Risk specific to one company: bad management, product failure, fraud, CEO scandal, industry-specific disaster.
- Parmalat went bankrupt in 2003: idiosyncratic.
- Enron fraud in 2001: idiosyncratic.
- One specific mining company hits a dry well: idiosyncratic.
This risk can be diversified away by holding many companies. Hold 100 companies; one scandal barely moves your portfolio.
Systematic risk (market-wide)
Risk that affects the whole market: recessions, inflation surges, pandemics, wars, rate hikes.
- 2008 financial crisis: systematic.
- 2022 rate shock: systematic.
- 2020 COVID crash: systematic.
This risk cannot be diversified away by holding more stocks. Even 3,000 stocks fall when the market falls.
The diversification graph
Empirical finding: adding stocks to a portfolio reduces idiosyncratic risk, but only up to a point.
Risk
|\
| \___ (~30% diversifiable risk eliminated)
| \___________
| \_____________ (remaining ~70% is systematic)
|
|______________________________________ Number of stocks
0 10 20 30 100 500- 1 stock: 100% of its own risk (both idio and systematic).
- 10 stocks: about 60% of one-stock risk. Major idio diversification.
- 30 stocks: about 75-80% diversified.
- 100 stocks: about 85-90% diversified.
- Beyond 500 stocks: marginal improvement.
Conclusion: you need at least 30+ stocks across sectors to be meaningfully diversified. 5 Italian stocks aren’t it.
Why “five different companies” fails
Consider Sofia’s imaginary “diversified” portfolio: ENI, Intesa, Generali, Ferrari, Stellantis. Five Italian stocks.
Problems:
- Sector concentration. Two financial services, one oil, one luxury, one auto. All heavily exposed to European economic cycles.
- Geographic concentration. All Italian-HQ, all primarily European markets. Zero US, Asian, emerging market exposure.
- Small-cap concentration. All large Italian caps — a tiny slice of global market cap.
- Currency concentration. All EUR denominated (with some international operations, but EUR home base).
When Italy or the Eurozone has a bad year, all five are likely to decline together. Correlation in 2022: ~0.7 between most of these. When the Italian macro environment deteriorates, they all suffer.
What real diversification looks like
Across companies (hundreds)
Hold a global index ETF with 1,500-3,800 companies. Single-company risk essentially vanishes — any one stock is 0.05-0.5% of your portfolio.
Across sectors
Global ETFs are balanced across: tech, financial, healthcare, industrial, consumer, energy, utilities, real estate, materials, telecom.
No sector is more than ~25% of a global index. If one sector has a bad decade (energy 2014-2020; banks 2008-2015), others compensate.
Across countries
Global ETF covers US (~63%), Europe (~15%), Japan (~5%), emerging markets (~10%), others. No single country over 70% (US dominance is the most concentrated).
Countries have different economic cycles. Italy recession ≠ US recession always.
Across currencies
Global equity ETFs have implicit exposure to USD (~60%), EUR (~15%), other major currencies. Helps against EUR weakness.
Across size (cap)
Large-cap dominant, but mid-cap and some small-cap exposure. Some factor strategies tilt smaller.
Across factors
Small-cap value, momentum, quality, profitability. Some research suggests factor exposures improve long-term return. Not required for basic diversification.
Across asset classes
The biggest diversification comes from holding different asset classes entirely:
- Stocks (all the above).
- Bonds (lesson 20).
- Real estate (primary residence counts).
- Cash (emergency fund).
- Commodities (optional).
The math of correlation
Two assets’ contribution to portfolio volatility depends on their correlation, not their individual volatilities.
Simplified example: two assets, each with σ = 20%.
- Correlation = 1.0 (perfect): combined σ = 20%. No benefit.
- Correlation = 0.7: combined σ = 18.4%.
- Correlation = 0.3: combined σ = 14.8%.
- Correlation = 0.0: combined σ = 14.1%.
- Correlation = −0.5: combined σ = 10.0%.
Lower correlation = more diversification benefit. Most major equity markets have correlations of 0.7-0.9 with each other, so international equity diversification helps a bit but not dramatically.
Stocks and bonds typically have correlation 0.0-0.3 (much lower). That’s why holding both helps diversify more.
The “diworsification” trap
Overdoing it hurts. If you hold 20 ETFs covering similar ground (MSCI World + MSCI ACWI + MSCI Emerging + MSCI EAFE + FTSE Developed + …), you:
- Pay 20 TERs when 1-2 would do.
- Incur extra transaction costs.
- Get tiny marginal diversification because they overlap.
- Create rebalancing complexity.
Peter Lynch coined “diworsification” for this pattern. Don’t.
Simple three-fund portfolio (global equity + bond + cash) achieves 95% of the diversification of a 20-fund portfolio, with 90% less complexity.
How many stocks do you actually need
Empirical studies suggest:
- For 90% of idiosyncratic risk diversified: ~30-50 stocks across sectors.
- For 95%: ~100 stocks.
- For effectively all: ~500+ stocks.
You don’t need to hand-pick these. A single global equity ETF (Vanguard FTSE All-World, iShares Core MSCI World) holds 1,500-3,800 stocks for you, automatically maintained. One trade, total diversification.
Diversification in bonds
Similar logic. A corporate bond ETF holds 1,000+ issuers across many sectors. Single-issuer default barely affects the portfolio.
For government bonds, “diversification” is less about many issuers (you already get a few countries in a broad sovereign ETF) and more about managing duration and credit risk across the curve.
Diversification limits
Things diversification cannot fully protect against:
1. Systematic market crashes
Even a globally diversified portfolio fell 40%+ in 2008 and 34% in 2020. Diversification reduces idiosyncratic risk, not market risk.
2. Regulatory and political risk
If EU-wide regulations change to hurt a sector (e.g., massive restrictions on tech), all EU tech companies suffer together regardless of diversification.
3. Currency collapse
Extreme scenarios (total EUR collapse) would hurt everyone holding EUR-denominated assets regardless of diversification within EUR.
4. Time-correlated liquidity crises
In 2008 and 2020, correlation between asset classes briefly spiked as everyone liquidated. “Normal” correlation benefits temporarily vanished.
Specific recommendations for different portfolio sizes
Under €5,000
Single global equity ETF. Add bonds only if time horizon is short.
€5,000 - €50,000
Two or three ETFs:
- Global equity.
- EU or global bond.
- Optional: EM or factor tilt.
€50,000 - €500,000
Four or five ETFs reasonable:
- Global equity.
- EU bond.
- Optional tilts: small-cap, value, EM.
- Possibly Italian-tax-advantaged (BTP direct).
Above €500,000
Consider more complexity: private placements, individual bonds, real estate, international diversification at country level. Benefit from fee-based financial advisor.
What to do with this lesson
Three steps:
- Count your holdings properly. How many unique companies? How many sectors? How many countries? If it’s under 30 companies, your idiosyncratic risk is high.
- Consolidate overlapping ETFs. 5 ETFs covering the same equity isn’t 5× diversified — it’s 1× diversified with 5× overhead.
- If you have concentrated single-stock positions (employer stock, inherited shares), make a plan to diversify. Not overnight if tax hurts; but within a few years.
Sources
- Harry Markowitz — Portfolio Selection, Journal of Finance, 1952 (origin of modern portfolio theory).
- William Bernstein — The Intelligent Asset Allocator, 2000.
- Peter Lynch — One Up On Wall Street, 1989 (origin of “diworsification” term).
Next lesson: global vs home-country bias — Italians hold 50%+ Italian equity when Italy is 0.7% of global market. The math on what this costs.