Investiční slovník
Correlation: Why Not to Put Everything in One Basket
Key takeaways
- Correlation tells you how much two assets move in the same or opposite direction.
- A coefficient of +1 means perfect agreement, −1 perfect opposition, 0 no relationship.
- Low or negative correlation between portfolio components reduces overall risk.
- Diversification works precisely because not everything falls and rises at the same time.
- In a crisis, correlations between equities tend to rise — so diversify in advance.
Correlation is a measure of how closely the prices of two assets move together. If they always move in the same direction, the correlation is positive; if they move in opposite directions, it is negative; if there is no relationship, it approaches zero.
The correlation coefficient in numbers
Statisticians express correlation as a number from −1 to +1. A value of +1 says that two assets always go up and down at the same time. A value of −1 means a perfect inverse relationship. In real portfolios you most often encounter values between 0.3 and 0.8 — meaning partial linkage, not absolute.
Real-world examples: stocks in the same sector (say, two car manufacturers) typically have a correlation of around 0.6–0.8. Stocks and bonds tend to be closer to 0 or even slightly negative, which is why this pair is the classic foundation of a balanced portfolio.
Why correlation matters
Imagine you hold two stocks, each individually volatile by 20% per year. If they correlate perfectly (+1), the portfolio still fluctuates by 20%. But if the correlation drops to 0, the portfolio's overall fluctuation falls to roughly 14%. Negative correlation reduces it even further.
This is the mathematical foundation of diversification: uncorrelated assets may each remain individually risky, but together they smooth out the portfolio's ride. Less volatility means the investor is less likely to panic-sell during drawdowns and more likely to stick to the plan.
When correlation deceives
The biggest trap: in times of market stress, correlations across equities rise. During the 2008 crash, almost everything fell at the same time — including assets that had otherwise moved independently. Correlation is therefore not fixed; keep in mind that at the worst moment, diversification will help you less than calm-period statistics would suggest.
- Stocks from different sectors — low correlation in calm times, higher during crises.
- Gold — historically low or negative correlation with equities, useful as a buffer.
- Real estate ETFs (REITs) — moderately high correlation with equities; less buffering than bonds.
What this means for the ordinary investor
It is not enough to hold "lots of different stocks" — if they all correlate strongly, diversification is only an illusion. That is why global ETF funds covering hundreds of companies from different countries and sectors provide better diversification than concentrating in a single market. If you want to go deeper, learn more about risk and how to measure it.
FAQ
What is correlation in simple terms?
Correlation tells you how much two assets move in the same direction. The closer the coefficient is to +1, the more they move together. The closer to −1, the more they move in opposite directions. Zero means no visible relationship.
Why is negative correlation good for a portfolio?
If one asset falls while another rises, they cushion each other's losses. This reduces the portfolio's overall fluctuation (volatility) and keeps its value more stable during turbulent times.
Is diversification the same as low correlation?
Diversification is the strategy; correlation is the tool we use to measure it. Real diversification happens when you add assets with low correlation to existing ones — sheer number of positions is not enough.
How does correlation behave during a stock market crash?
During a sharp market decline, correlations between equities typically rise — assets that otherwise moved independently start falling together. That is why it is important to hold assets from other classes, such as bonds or gold, in your portfolio.