Do Correlations Go to 1 in a Crash? What the Data Shows
The saying is that in a crash everything falls together. Here is what ten years of monthly data and five real crises show, and which holdings actually behaved differently.
GlossaryGlossary
What Is the Correlation Coefficient? Definition and Example
What the correlation coefficient measures, which levels count as good for diversification, and why correlations change when markets fall.
Key takeaways
- The correlation coefficient measures how closely two assets’ returns move together, from −1 (opposite) through 0 (unrelated) to +1 (in lockstep).
- For diversification, a common rule of thumb treats below 0.3 as low, 0.3–0.7 as moderate and above 0.7 as high.
- In Exent’s data (60 months to Sep 30, 2026), the Nasdaq-100 ETF (QQQ) had a 0.92 correlation with the S&P 500 ETF (SPY), gold (GLD) 0.17.
- Correlations rise in sell-offs, so use rolling and stress-period figures, not one long-run number.
What is your own portfolio exposed to? Check it with Exent →
What is the correlation coefficient?
It is a number between −1 and +1 that describes how two series move together. Correlation = covariance(A, B) ÷ (σA × σB), where σ is each asset’s standard deviation. It says nothing about size: two assets can be highly correlated even if one swings three times as much (that is beta’s job).
What correlation is good for diversification?
| Asset vs SPY | Correlation | Rule-of-thumb reading |
|---|---|---|
| Vanguard S&P 500 ETF (VOO) | 1.00 | Same exposure |
| Nasdaq-100 (QQQ) | 0.92 | High |
| Schwab U.S. Dividend Equity (SCHD) | 0.71 | High |
| Johnson & Johnson (JNJ) | 0.19 | Low |
| Gold (GLD) | 0.17 | Low |
| ExxonMobil (XOM) | 0.13 | Low |
Source: Exent’s data, monthly returns over 60 months to Sep 30, 2026. The bands are a rule of thumb, not a standard.
Why use rolling correlations?
Correlation changes over time, and it tends to rise when markets fall. The same data on calm months versus all months shows it: QQQ’s correlation with SPY was 0.85 in calm months but 0.91 overall. In the 16 months since mid-2016 when SPY fell more than 4%, SPY averaged −6.7% and QQQ −7.6%, while T-bills (BIL) averaged +0.2% and gold −0.4%. Rolling correlations, such as a 36-month window recalculated each month, show the drift. More in correlations in crashes.
Why it matters for a self-directed investor
Owning many stocks with correlations above 0.7 to each other is one bet with many tickers. Grouping holdings into correlation clusters makes that visible. See portfolio correlation and covariance.
Common mistakes
- Trusting a long-run average that hides crisis behavior.
- Reading low correlation as protection: an uncorrelated asset can still lose money in a crash.
- Mixing daily and monthly data, which give different numbers.
See correlation in your portfolio with Exent
- Upload a screenshot of your holdings, or a CSV, Excel, PDF or Google Sheet; no brokerage login.
- Exent groups holdings that move together into correlation clusters and replays the portfolio through past crashes.
Frequently asked questions
It reduces overall swings, because one asset tends to rise when the other falls. Reliably negative correlations are rare and often come with low expected returns.
The two assets tend to move in the same direction, but loosely. Under the usual rule of thumb that is moderate correlation.
Usually, yes. In monthly data to Sep 2026, most stock funds were more correlated with the S&P 500 across all months than in calm months alone.
See what your portfolio is really made of.
Upload a screenshot of your holdings and get concentration, overlap, crash exposure and a risk-matched benchmark in about a minute.
Check my portfolioLimited offer$1.95 for unlimited checks· unlimited checks
Educational content, not investment advice. Examples are illustrative; past performance does not guarantee future results.