How to Hedge Your Stock Portfolio Against a Crash
Every common way to protect a stock portfolio, from cash and bonds to puts, collars and trend rules: how each works, what it costs, and how it actually behaved in 2008, 2020 and 2022.
Exent tested 12 popular diversifiers against the S&P 500: bonds, gold, T‑bills, commodities and “defensive” stock funds. Here is how each behaved in 2008, the COVID crash and 2022, and in every bad month since 2016.
What is your own portfolio exposed to? Check it with Exent →
Each hedge worked in some crises and failed in others, because each crisis had a different cause: a credit collapse in 2008, a sudden shock in 2020, and an inflation and rate shock in 2022.
| Fund | 2008 crisis: deepest fall | COVID: Feb–Mar 2020 | 2022 calendar year |
|---|---|---|---|
| S&P 500 (SPY) | −56.5% | −19.4% | −18.2% |
| T-bills (BIL) | −1.1% | +0.3% | +1.4% |
| 20+ year Treasuries (TLT) | −16.7% | +13.4% | −31.2% |
| 7–10 year Treasuries (IEF) | −7.1% | +6.8% | −15.2% |
| US aggregate bonds (AGG) | −14.6% | +1.0% | −13.0% |
| Gold (GLD) | −29.4% | −0.9% | −0.8% |
| Commodities (DBC) | −60.9% | −22.9% | +19.3% |
| High-dividend stocks (SCHD) | n/a | −20.1% | −3.2% |
| Minimum volatility (USMV) | n/a | −19.1% | −9.4% |
| Dividend growth (VIG) | −48.2% | −17.1% | −9.8% |
| Equal-weight S&P 500 (RSP) | −60.0% | −25.3% | −11.6% |
| Small caps (IWM) | −59.2% | −28.4% | −20.5% |
| Nasdaq-100 (QQQ) | −53.6% | −12.9% | −32.6% |
Exent’s data as of Sep 30, 2026, total returns. The 2008 column is the deepest peak‑to‑trough fall inside the October 2007 to March 2009 window, a harsher measure than a period return; the 2020 column is month‑end January to month‑end March 2020. Each fund has its own data page, for example TLT, GLD and DBC.
Three crises are a small sample, so Exent also looked at every bad month. From mid‑2016 to Sep 30, 2026, the S&P 500 fell more than 4% in 16 months and averaged −6.7% in them.
| Fund | Average in the 16 bad months | Correlation with S&P 500, all months | Correlation, calm months only |
|---|---|---|---|
| T-bills (BIL) | +0.2% | 0.00 | 0.05 |
| Gold (GLD) | −0.4% | 0.14 | 0.10 |
| 7–10 year Treasuries (IEF) | −0.9% | 0.18 | 0.16 |
| 20+ year Treasuries (TLT) | −1.9% | 0.19 | 0.18 |
| Commodities (DBC) | −1.2% | 0.34 | 0.28 |
| US aggregate bonds (AGG) | −1.2% | 0.40 | 0.32 |
| High-dividend stocks (SCHD) | −5.5% | 0.82 | 0.71 |
| Small caps (IWM) | −8.4% | 0.86 | 0.76 |
| Minimum volatility (USMV) | −4.9% | 0.87 | 0.77 |
| Nasdaq-100 (QQQ) | −7.6% | 0.91 | 0.85 |
| Equal-weight S&P 500 (RSP) | −7.0% | 0.93 | 0.88 |
| Dividend growth (VIG) | −5.8% | 0.95 | 0.90 |
Monthly total returns, same data. A correlation near 0 means a fund’s moves had little to do with the stock market’s; near 1 means they moved together (see correlation coefficient). None of the 12 funds had a negative correlation with stocks over the period, which is why truly negatively correlated assets are rare.
In this data, yes. For 11 of the 12 funds, the correlation with the S&P 500 was higher across all months than in calm months alone, which means the bad months pulled them closer to stocks. The only exception was T‑bills. For the stock funds the jump was about 0.05 to 0.11; for aggregate bonds it was 0.32 to 0.40.
That is the practical problem with diversification: a mix that looks well spread in normal times can behave like one position in a selloff. Why it happens, and how to measure it, is covered in correlations in crashes.
They reduce the size of losses a little, but they do not diversify. Minimum‑volatility, dividend and equal‑weight funds are still stock funds, with correlations of 0.82 to 0.95 with the S&P 500. In the 16 bad months they averaged −4.9% to −7.0%, against −6.7% for the index.
Their protection also depended on the crisis. In the COVID crash, the minimum‑volatility fund fell 19.1%, almost exactly the S&P 500’s 19.4%, and equal weight fell more (25.3%). In 2022, high dividend (−3.2%) and minimum volatility (−9.4%) did much better than the index. Growth‑heavy QQQ did the reverse: better in 2020 (−12.9%), much worse in 2022 (−32.6%).
Combining assets that fail in different crises smoothed results, at a cost in long‑run return. Hypothetical backtests on the same data, mid‑2016 to Sep 30, 2026:
| Backtest | Holdings | Feb–Mar 2020 | 2022 | Annual return |
|---|---|---|---|---|
| S&P 500 | SPY | −19.4% | −18.2% | 15.2% |
| 60/40 | SPY 60 / AGG 40 | −11.5% | −15.8% | 9.5% |
| All-weather-style | SPY 30 / TLT 40 / IEF 15 / GLD 7.5 / DBC 7.5 | −1.8% | −19.3% | 5.1% |
| Risk parity (from Jul 2017) | SPY / TLT / GLD by inverse volatility | −3.8% | −18.2% | 8.7% |
| Trend filter | SPY, to T-bills below its 10-month average | −7.7% | −20.1% | 9.4% |
Hypothetical backtests, past results, not a recommendation. The bond‑heavy mixes nearly sidestepped 2020 and then lost as much as stocks in 2022, because their main hedge failed that year. Details on 60/40, all‑weather‑style and risk parity.
For the full menu of approaches, from options to cash, start at the guide to hedging a portfolio, and see what each costs in the cost of hedging.
Of the funds Exent tested, T-bills held up best, with a deepest fall of 1.1% inside the 2007–09 window. Intermediate Treasuries fell as much as 7.1% and long Treasuries 16.7%, against 56.5% for the S&P 500.
Commodities gained 19.3% and T-bills 1.4%, and gold returned −0.8%. Bonds fell with stocks: long Treasuries lost 31.2% and aggregate bonds 13.0%.
Long Treasuries gained 13.4% in February and March 2020 while the S&P 500 lost 19.4%. Commodities lost 22.9% over the same months.
Not a true hedge. Dividend ETFs had correlations of 0.82 and 0.95 with the S&P 500 and lost money in the same bad months, though usually a little less.
In Exent’s data since mid-2016, T-bills had a correlation of 0.00 with the S&P 500 and gold 0.14. Both are close to zero rather than negative.
Upload a screenshot of your holdings and get concentration, overlap, crash exposure and a risk-matched benchmark in about a minute.
Educational content, not investment advice. Examples are illustrative; past performance does not guarantee future results.