Diversification: correlation matters more than the number of holdings
Owning 20 assets that move together is still one bet. In 2022 both stocks and bonds fell, and a 60/40 blend lost 18.4% in under six months — the cushion failed because the correlation flipped.
Twenty holdings, one direction
Most people count diversification in holdings: ten stocks, five funds, a bond or two, and the job feels done. Investopedia opens its guide to correlation with a warning against exactly that habit: you can own 20 different assets and still have a portfolio that moves in lockstep without realizing it. True diversification depends on more than spreading money across different asset types — it requires spreading money across assets with low or negative correlation.
Correlation is the measure of how two investments move in relation to each other, on a scale from -1.0 to +1.0. A correlation of +1.0 means the two move in lockstep, up and down together. A correlation of 0.0 means there is no consistent relationship between how they move. A correlation of -1.0 means they move in exact opposition — when one rises, the other falls. The point of the number is practical: combine assets that do not move together and the portfolio's ride gets smoother, because a fall in one is not automatically a fall in all of them.
2022: the year the cushion failed
The textbook case of a correlation breaking arrived in 2022. Stocks and bonds tend to have a negative correlation — when equities sell off, bonds often rise and cushion the fall — but in 2022 returns on both fell. Morningstar's Rekenthaler column measures the damage on a standard 60/40 blend of 60% Morningstar US Market Index and 40% Morningstar 5-10 Year Treasury Bond Index: it lost 18.4% between January 2 and June 17, 2022, and 23.3% after considering the effect of inflation — well above the 20% downturn that customarily defines a bear market.
Vanguard's research, cited by Investopedia, says positive stock/bond correlations occur during supply-side economic shocks or when inflation exceeds central bank targets — exactly the conditions of 2022, when the Federal Reserve raised rates fast to fight inflation and both legs of the classic mix wobbled at once. Morningstar notes the 60/40 portfolio floundered in 2022 just as it periodically floundered during the 1970s and early 1980s, and that in July and August 2022 balanced funds endured their steepest redemptions since the global financial crisis of 2008. The cushion did not fail because diversification is useless; it failed because the two holdings were answering to the same shock.
- 60/40 blend, Jan 2 - Jun 17, 2022-18.4%
- After inflation-23.3%
- Bear-market line-20%
- Balanced-fund redemptions, Jul-Aug 2022Steepest since 2008
- Tactical funds vs balanced, Jan-Jun4 basis points below
What actually moved differently
If the lesson of 2022 is that counting holdings is not enough, the next question is what genuinely behaved differently. Investopedia's correlation table for ETFs gives concrete pairings: the S&P 500 ETF (SPY) correlates 0.29 with a 7-10 year Treasury ETF (IEF) and 0.18 with a gold ETF (GLD), while IEF correlates -0.32 with a commodity ETF (COMT) — one of the few negative pairings on the board. Gold ETFs, Investopedia notes, offer long-term diversification with return drivers largely independent of stocks and bonds, and commodity ETFs help hedge against inflation and supply shocks — the very shocks that broke the stock-bond cushion.
Morningstar adds a humbling footnote for anyone tempted to time the fix. Recognizing that bonds were precariously priced was not the same as being able to act on that suspicion: from January through June 2022, when the 60/40 portfolio plummeted, the average return of tactical asset allocation funds — the professionals paid to adjust — was 4 basis points below that of traditional balanced funds. The cure, in that stretch, was very slightly worse than the disease. Choosing holdings that move differently is a design decision made in advance, not a trade made in the middle of the fall.
Checking your own mix
So the useful question is not how many positions a portfolio holds, but what drives each one. Investopedia's checklist runs beyond stocks and bonds: commodities, real estate, managed futures — anything whose returns answer to different economic drivers, risk exposures and investor behaviour. The test is simple to state and harder to pass: if every holding falls for the same reason — the same rate shock, the same inflation surprise — the portfolio is one bet wearing several costumes.
Two cautions belong with that test. First, correlation scores are only backward-looking: they describe how assets moved together, not how they will move next, and the 2022 flip is the proof. Second, the old relationship has not vanished: over the long run the negative stock-bond correlation still holds, and most equity downturns are still accompanied by positive bond returns. Diversification still works — it just works on behaviour, not on headcount. Educational content, not investment advice.
