Gold vs the S&P 500 in Market Crashes

In the three most recent major market crashes, gold fell less than the S&P 500 and usually recovered faster, but it still fell. In the 2007-2009 financial crisis, gold dropped 29% at its worst while the S&P 500 dropped 55%; in the 2020 Covid crash it was 12% against 34%; and in the 2022 rate-hike bear market 21% against 24%. Gold has been a partial cushion in crashes, not a place where money was safe from losses.

Below we look at each crash in turn, explain why gold behaves the way it does in a panic, and cover what this means for anyone mixing gold and stock strategies.

The numbers side by side

The figures come from the historical price data behind AlgolaneBots's backtests, measuring a simple buy-and-hold of each asset. They describe what happened to prices in the past; future crashes will play out differently.

"Recovery" means the time it took to get back to the value before the fall. Three patterns stand out.

1. Gold fell less in every case.
2. Gold recovered faster in every case.
3. The gap varied a lot. In 2008 gold's fall was about half the S&P 500's. In 2022 the two were close.

Crash by crash

2007-2009: the financial crisis

The global financial crisis was the deepest stock market fall in the data. The S&P 500 lost more than half its value and took 53 months, well over four years, to get back to its old high.

Gold is often described as a crisis asset, so it surprises many people that it also fell 29%. The reason is that in a severe panic, investors and funds often sell whatever they can to raise cash or cover losses elsewhere, and gold is easy to sell. Gold's fall was sharp but its recovery was much quicker than stocks', at 14 months.

The lesson is that gold's "safe haven" reputation is about relative performance across a crisis, not about avoiding a drop on the worst days.

2020: the Covid crash

The Covid crash was extremely fast. The S&P 500 fell 34% in a matter of weeks, then recovered within 4 months as governments and central banks stepped in with large support programmes.

Gold dipped 12% during the same scramble for cash and was back within 2 months. Both assets recovered quickly, and gold's smaller fall meant less ground to make up.

This crash shows a limitation of looking only at depth. A 34% fall that recovers in 4 months is a very different experience from a 55% fall that takes 53 months, even though both appear as "big crashes" in a summary.

2022: the rate-hike bear market

2022 was different in kind. The main driver was rising interest rates as central banks fought inflation. Higher interest rates hurt stocks, but they also hurt gold, because gold pays no interest and becomes less attractive when cash and bonds start paying more. The US dollar was also strong, which tends to weigh on the gold price.

So both fell by similar amounts: 24% for the S&P 500 and 21% for gold. Gold still recovered sooner, in 11 months against 23. But anyone who held gold expecting it to rise while stocks fell was disappointed. For more on how different strategies handled this period, see what rule-based strategies do in bear markets.

Why gold sometimes protects and sometimes doesn't

Gold's behaviour in a crash depends on what caused the crash. A rough guide, based on well-established relationships rather than precise rules:

Gold also has long bad spells of its own that have nothing to do with stock crashes. Over the full data period from 2004 to 2026, gold's worst fall was 45%, deeper than the 34% worst fall the S&P 500 had over 2011 to 2026. That decline came after gold's 2011 peak and took years to recover. Gold is volatile in its own right.

How to read crash numbers without being misled

A single "fall" figure hides a lot, and comparisons across assets are easy to get wrong. A few points help when reading tables like the one above.

• Falls are measured peak to trough. The figure is the drop from the highest value before the crash to the lowest value during it. Someone who bought halfway down had a different experience.
• Each asset has its own peak and trough. Gold and the S&P 500 did not top out or bottom out on the same day. The comparison is between each asset's worst stretch within the crash, not a same-day snapshot.
• Depth and duration are separate risks. A shallow fall that drags on for years can be harder to live with than a deep one that recovers in months.
• Three crashes is a small sample. Each was driven by different forces, and the next one will have its own causes. Patterns across three events are a guide, not a rule.
• Currency matters. These figures are in US dollars. For someone measuring wealth in another currency, both gold and US stocks would have looked different.

For more on reading drawdown figures, see max drawdown explained and how strategies held up in 2008, 2020 and 2022.

Gold over the long run

Over the period from June 2004 to November 2026, holding gold returned +11.0% a year, turning $10,000 into $103,201 in the historical data, with that worst fall of 45%. The S&P 500 data covers a different period (from 2011), so the yearly figures are not directly comparable. The point is that gold has not just been a crash hedge; it has produced meaningful long-run gains with large swings of its own.

Where a rule-based gold strategy fits

AlgolaneBots's Gold Breakout trades gold rather than holding it, and the fair comparison for it is holding gold, not the stock market. The figures below are from a historical backtest: a computer test of its rules on past prices. Nobody earned these results, and future results will differ. Its caveat is that doubling trading costs cuts its yearly result roughly in half.

At 2% risk per trade, from June 2004 to November 2026, the backtest returned +15.1% a year with a worst fall of 41%, compared with +11.0% a year and a 45% worst fall for gold held. In the crashes:

• 2007-2009: the strategy fell 13% vs 29% for gold held, but took 21 months to recover vs 14.
• 2020: 7% vs 12%, recovering in 3 months vs 2.
• 2022: 18% vs 21%, recovering in 9 months vs 11.

Its deepest fall, 41% from May 2018 to October 2019, was not in any stock crash at all, and took 41 months after the bottom to recover. Gold strategies can have their worst periods at completely different times from stocks. More detail is in holding gold vs trading gold.

What this means for a mixed portfolio

The main appeal of gold next to stocks is that the two often do not fall at the same time or by the same amount. That can smooth a combined portfolio. The effect is real but not reliable: 2022 showed both falling together.

We look at a specific example from the backtests, combining Gold Breakout with each stock strategy, in mixing gold and stock strategies. The idea of combining strategies more broadly is covered in why run more than one trading strategy. None of this is a recommendation about how much gold anyone should hold.

Key takeaways

• In the 2008, 2020 and 2022 crashes, gold fell less than the S&P 500 and recovered faster each time.
• Gold still fell in every crash, by 12% to 29%, so it is a partial cushion rather than a safe place.
• Rising interest rates, as in 2022, can push gold and stocks down together.
• Gold has deep falls of its own, including a 45% decline in the data that had nothing to do with a stock crash.
• A gold trading strategy should be compared with holding gold, and its worst periods may not line up with stock crashes.

This is general education, not personal advice. See our risk disclosure.

Test gold and stocks side by side

You can run gold and stock strategies together on paper and see how they move against each other in today's market. Start a free 90-day trial with a $10,000 paper account covering all five strategies. No card is needed.