Why Did Both Stocks and Bonds Fall in 2022?

Key takeaways

• UK gilts lost 24.00% in 2022 while the FTSE All Share returned 0.70% and global developed shares lost 7.83% in sterling, so almost all the damage sat on the bond side.
• A 60/40 of global shares and gilts lost 14.30% in 2022 before inflation. With UK CPI up 10.5% in the year to December 2022, that is a real loss of 22.44%.
• Across 154 years of US data, 2022 was the worst calendar year for bonds after inflation, at -20.16%. Equities lost 20.14% in the same twelve months.
• US equities fell 39.29% in real terms in 2008, almost twice the 2022 fall, and yet a 60/40 lost 15.86% that year against 20.15% in 2022.
• The real US bond series still has not regained its December 2021 level. In October 2026, 58 months on, it sat 27.36% below it.

What made stocks and bonds fall in 2022 at the same time

You hold bonds so that one half of the portfolio keeps its footing when the other half doesn't. In 2022 neither half did. The reason stocks and bonds fall in 2022 together is that the shock came from the price of money rather than from company earnings, and the price of money sets the value of both.

Here's the mechanism in a paragraph. Inflation forced central banks to raise rates quickly. A bond already in issue pays a fixed coupon, so the only way it can offer a higher yield is for its price to fall. The same higher discount rate lowers what a share is worth today, because a company's future profits get discounted back at that rate. When a shock arrives from the growth side, as in 2008, money moves into government bonds and the bond side rises. When it arrives from the rate side, there's nowhere inside a stock and bond portfolio for that money to go.

That's why 2022 reads so differently from every crash most investors had lived through. It wasn't a crisis of earnings, and it wasn't a crisis of solvency. It was a repricing of the discount rate, applied to everything at once.

For a UK investor it was barely a stock market crash at all

The search phrasing most people use is about shares, and for a sterling investor that's the wrong half of the story. The FTSE All Share, measured net of tax on the index BlackRock's tracker follows, returned 0.70% in 2022. Global developed shares, on the MSCI World net total return index in sterling, lost 7.83%. UK gilts, on the FTSE Actuaries UK Conventional Gilts All Stocks index, lost 24.00%.

Deflating those by the December 2022 UK CPI rate of 10.5% gives the numbers that actually describe the year. CPI averaged 9.1% across 2022 and peaked at 11.1% in October.

In money, a £100,000 portfolio split 60/40 between global shares and gilts ended 2022 at about £85,700. In December 2021 money that was worth roughly £77,600. The gilt sleeve did the work: it fell 31.22% in real terms, against 8.87% for UK shares.

So the usual framing has it backwards for a UK portfolio. The supposedly safe asset produced the loss, and the supposedly risky one produced almost none of it. That inversion ran right through the retail fund range, where the ladder a crisis proof portfolio is built on turned upside down. Over the 12 months to 30 September 2022 Vanguard's LifeStrategy 20% equity fund fell 15.40% and its 100% equity fund fell 4.47%, a gap of 10.93 percentage points the wrong way round.

2022 was the worst year for bonds in 154 years of American data

One market's bad year proves little on its own. The question is whether the way stocks and bonds fall in 2022 has precedent, or whether the year sits outside the record. Robert Shiller's monthly series lets you test that properly, because it carries a real total return index for US equities and another for a constant-maturity 10-year Treasury, both deflated by CPI, running from 1871 to October 2026.

Line up every complete calendar year from 1872 to 2025 and you get 154 observations. In 2022, real equity returns came in at -20.14% and real bond returns at -20.16%. That bond figure is the lowest of all 154. Not the lowest since the war, not the lowest in living memory: the lowest in the series.

Both assets losing real money in the same year is not rare, though. It happened in 18 of those years, among them 1887, 1947, 1978, 1994 and 2018. Overlapping 12-month windows give a similar answer: 249 of 1858 windows, or 13.4%, saw both series lose real money. Roughly one stretch in seven, then, the ballast doesn't work.

What made 2022 exceptional wasn't that both fell. It was how far the bond side fell. The median real bond loss across those 18 years was 8.66%, and in 12 of them it came in under 10%, which an equity recovery absorbs without much trouble. A loss of 20.16% on the half of the portfolio held for stability is a different kind of event.

2008 is the control case, and it settles the argument

Here's the cleanest test of whether 2022 was really about shares. In 2008 US equities lost 39.29% in real terms, almost twice the 2022 fall. A 60/40 portfolio lost 15.86%. In 2022 equities lost roughly half as much, and the same 60/40 lost 20.15%. The bond side is what flipped the ranking: it returned 19.28% in 2008 and -20.16% in 2022.

The chart above plots the real bond return in each of the six calendar years when US equities fell hardest, plus 2002 for the dot-com comparison and 2022 itself. Bonds made money in five of those eight years. They lost money in 1917 and 1974. Then there's 2022.

Read down the first column and 2022 is the mildest equity year on the list. Read down the last and it's the third worst year for a 60/40 portfolio in 154, behind only 1917 and 1974. US CPI rose 18.10% in 1917 and 12.34% in 1974, against 6.45% in 2022, so inflation is the thread connecting all three. The stock bond correlation turns positive when inflation is doing the driving, and that's the regime in which a 60/40 portfolio drawdown stops being cushioned by anything.

The long end of the gilt curve moved further than Bank Rate did

The size of the gilt loss follows from arithmetic rather than panic. The Bank of England raised Bank Rate from 0.25% on 16 December 2021 to 3.5% on 15 December 2022, a rise of 3.25 percentage points in eight steps, passing 2.25% in September.

Gilt yields moved further than that at the long end, which is where the capital losses live. The 10-year nominal spot rate went from 0.9785% on 31 December 2021 to 3.7312% a year later, a rise of 2.75 percentage points. The 30-year rate started 2022 at 1.1250% and hit 4.8504% on 27 September, up 3.73 points. Duration, the rough sensitivity of a bond's price to a one point change in yield, runs at 6.96 years for the All Stocks gilt index. Multiply that by the move in yields and you have most of the 24.00% loss without needing any other explanation.

Part of the autumn move wasn't about the rate outlook. On 28 September 2022 the Bank announced temporary purchases of long-dated gilts, saying the repricing "is particularly affecting long-dated UK government debt" and warning: "Were dysfunction in this market to continue or worsen, there would be a material risk to UK financial stability." It described the operation as intended "to tackle a specific problem in the long-dated government bond market", with auctions running until 14 October. The 30-year yield peaked the day before the announcement; the 20-year reached 5.0406% on 12 October, inside the purchase window. Gilt yields in 2022 were doing two jobs at once, pricing inflation and clearing a liquidity scramble.

The recovery is where the detail sits, and it is still incomplete

Peak to trough, a real 60/40 rebalanced monthly fell 23.88% between December 2021 and its low in October 2022. The equity series fell 24.47% over the same months. The bond series kept going for another year, bottoming in October 2023 at 27.79% below its December 2021 level.

The 60/40 got its real value back in December 2024, 36 months after the peak. Equities got there earlier, in March 2024. The bond side hasn't got there at all. In October 2026 the real US bond series was still 27.36% below December 2021, having spent all 58 months in between underwater.

UK gilts tell the same story in sterling. The All Stocks index returned -5.09% in 2021, -24.00% in 2022 and -3.13% again in 2024. Over the five years to 30 September 2026 it compounded at -4.37% a year. An investor at the cautious end of the fund range lived the same thing: Vanguard's 20% equity fund returned 2.42% in total over those five years while UK CPI rose 28.20%, which is a real loss of 20.11%. The 100% equity fund returned 73.63% nominal, or 35.43% real.

That's the part a one-year table hides. The equity drawdown of 2022 has been repaired. The bond drawdown, in real terms, mostly hasn't, and at five years and counting, calling it a temporary mark-down is no longer a neutral description.

The strongest objection is that the starting yield can't repeat

There's a serious case against reading 2022 as evidence about diversification, and it deserves its best shot. A 10-year gilt yielding 0.9785% has almost no income to offset a price fall. A 30-year at 1.1250% has even less. At those levels the asset was arithmetically incapable of cushioning much, so 2022 tested a starting point rather than a principle.

That argument has force, and the current numbers support it. The All Stocks gilt index carried a yield to maturity of 5.18% at 30 September 2026. A portfolio starting from there has several years of income to absorb a yield rise of the same size, which the 2021 version plainly did not.

Two things qualify it. The first is symmetry. If the 2022 bond loss was a function of the starting yield, then the long run of years when bonds cushioned equity falls was partly a function of starting yields too, and those were falling for four decades. The second is the record. Shiller's series predates the low-yield era entirely, In 1917 bonds lost 13.85% in real terms from a December 1916 10-year yield of 4.215%, and in 1974 they lost 8.68% from a December 1973 yield of 6.74%. Neither starting point resembles the 1.47% on offer in December 2021. Inflation, rather than a near-zero starting point, is the common factor across every one of those years.

What this data cannot tell you

The 154-year series is American. A UK investor's bond holdings are gilts, and the two markets diverged sharply in the autumn of 2022 for reasons specific to the long-dated gilt market. The US figures establish the pattern. They don't measure a British portfolio.

Shiller's bond column is a constant-maturity 10-year Treasury total return, which isn't a bond fund. It holds duration steady by construction, where a real fund's duration drifts and its manager trades. The 6.96 year duration quoted above is a 2026 reading of the gilt index rather than a 2021 one, so it illustrates the mechanism instead of reproducing the exact loss.

Calendar years are an arbitrary frame. A fall that straddles a December shows up smaller in both years than it felt, and the count of 18 is sensitive to that. Overlapping 12-month windows are not independent observations either, so 13.4% describes this sample rather than a probability. And one year is one year. A single observation that cuts against a long pattern is a warning, not a refutation of it.

The fund figures are a short sample as well. Vanguard's LifeStrategy range launched in 2011, so its record contains no 2008 and no 1974. Deflating returns by headline CPI also assumes your own spending looks like the published basket, which for most households it doesn't.

What would change the conclusion

If inflation settles near target and stays there, the mechanism behind 2022 goes quiet. The evidence that bonds and shares fall together is concentrated in high-inflation years, and in the other regime a diversification failure on the 2022 scale reads as an episode rather than a pattern.

If the next equity bear market arrives from the growth side, the bond side has room to do what it did in 2002 and 2008, because 5.18% of income absorbs a great deal more than 0.9785% did. A growth shock with gilts falling anyway would be the genuinely new information, and the thing that would make 2022 look like a beginning rather than an episode.

If the real bond series is still below its December 2021 level in a seventh year, the recovery arithmetic changes character. 36 months to repair a 60/40 is a bad patch. A bond half that is 27.36% down after 58 months is not a patch at all, and the case for holding it has to rest on the income rather than on the cushion.

The number worth watching isn't the drawdown. It's whether gilts and global shares fall on the same days, because that one relationship decides whether a 60/40 holds together. LedgerTouch measures that correlation across the holdings you actually own rather than across two indices. For the shape of other crisis windows, there's the record of what held up in each episode of diversification in a crisis.