Diversification

What it means

Holding several stocks can reduce reliance on one company, but the number of ticker symbols alone does not tell you how diversified a portfolio is. Position sizes matter: one stock can dominate a portfolio containing many small holdings. Similar holdings can also move together because they share a sector, country, currency or broader market exposure.

Diversification can reduce company-specific, or idiosyncratic, risk. It cannot eliminate losses or the risk that many stocks decline together. A diversified stock portfolio remains exposed to stock-market risk.

How it is calculated

There is no single diversification percentage. Two useful measures answer different questions:

• Weight concentration: HHI = Σ(wᵢ²). HHI is the Herfindahl-Hirschman index, with no units when weights are decimals; wᵢ is stock i’s share of the portfolio’s current market value, and Σ means add across all N holdings. On this scale, a lower HHI generally means less concentration by position size. 1 / HHI is the effective number of equally weighted holdings, not the actual number of holdings.
• Return co-movement: σₚ² = Σᵢ Σⱼ (wᵢ × wⱼ × σᵢ × σⱼ × ρᵢⱼ). σₚ² is portfolio return variance; σᵢ and σⱼ are return standard deviations for holdings i and j; ρᵢⱼ is their return correlation, between −1 and +1; and both sums cover all holdings, including pairs where i = j. Weights and correlation have no units. If standard deviations are expressed as decimal returns per day, variance has units of squared daily return; σₚ = √σₚ² is daily return standard deviation. Use the same return period for every input.

HHI shows how capital is distributed, not whether holdings tend to rise or fall together. The variance formula incorporates that co-movement, but its result depends on the estimated inputs.

Worked example

These numbers are hypothetical and ignore costs. Suppose a US$10,000 stock portfolio contains two holdings. First compare a US$5,000 position in each stock with a US$9,000 and US$1,000 split:

1. Equal weights are 0.50 and 0.50, so HHI = 0.50² + 0.50² = 0.50. The effective number is 1 / 0.50 = 2.
2. Unequal weights are 0.90 and 0.10, so HHI = 0.90² + 0.10² = 0.82. The effective number is approximately 1.22. Both portfolios own two stocks, but the second depends much more on one position.
3. To see why equal weights are not the whole story, assume each stock has a hypothetical daily return standard deviation of 2% and their correlation is 0. For the equal-weight portfolio, daily variance is 0.50² × 0.02² + 0.50² × 0.02² = 0.0002; daily standard deviation is √0.0002 ≈ 1.41%.
4. If the same stocks instead have correlation +1, their returns move together in this simplified calculation. The equal-weight portfolio’s daily standard deviation is 2%, despite its unchanged HHI of 0.50.

Why it matters before you trade

A proposed trade changes portfolio weights. Before placing it, a concentration check can show whether it increases the largest position or adds exposure to a sector or country already heavily represented. A correlation check asks a different question: whether a new holding might behave much like existing ones. Historical correlations can change, particularly during market stress, so neither check guarantees protection.

Position size and execution conditions also matter. A portfolio can look diversified by market value while containing a large order in a less liquid stock. Diversification describes the portfolio’s exposures; it does not make a difficult trade inexpensive to execute.

How PreTrAIde uses it

In PreTrAIde, a user builds a stock basket with a side, quantity and execution conditions for each line. Its pre-trade analysis provides basket- and line-level GO, NO-GO or MODIFY verdicts as decision support, not recommendations.

PreTrAIde calculates portfolio breakdowns by country, sector, region, currency and market capitalization, along with a Herfindahl-Hirschman concentration index. Those measures help describe the basket’s exposure and position concentration. Its stated product facts do not specify a portfolio-correlation calculation or a single diversification score, so an HHI result should not be read as either.

Common mistakes

• Counting tickers instead of weights. Ten holdings are not evenly diversified if one accounts for most of the portfolio.
• Treating different company names as independent risks. Holdings may share sector, country, currency or market exposure.
• Reading HHI as a complete risk measure. It captures position-weight concentration, not correlation or the size of potential losses.
• Assuming diversification prevents losses. Several holdings can decline at once, and past co-movement may not persist.