Why bonds protected in 2008 but failed in 2022

Government bonds cushioned the 2008 crisis but deepened losses in the 2022 rate shock. In the FTSE global government-bond index hedged into euros, the maximum month-end loss was 2.1% within September 2007–September 2010 and 17.3% within December 2021–December 2024. Over those two windows, €10,000 grew to €12,020 in the first and shrank to €8,760 in the second, while the same amount in US equities did the reverse.

Terminal wealth over matching dates is the growth comparison here. A shallower drawdown alone is not evidence of better compounding, and these selected windows do not establish future expected geometric growth.

The asset-class crash study provides the underlying observations. Here we calculate what those results meant for terminal wealth and separate the bond-pricing mechanism from the historical comparison.

Key takeaways

1. Over the same three-year windows, government bonds beat US equities by 37.1 percentage points in 2007–2010 and trailed them by 53.4 points in 2021–2024. €10,000 in EUR-hedged global government bonds became €12,020 and €8,760, against €8,310 and €14,100 in the S&P 500 in euros.
2. The reversal is consistent with how fixed payments are priced when yields move. A hypothetical ten-year zero-coupon bond loses 17.8% when its yield rises from 1% to 3% and gains 21.7% on the reverse move.
3. The often-quoted 17.3%, 20.2% and 21.2% bond losses are all correct, because they use different start dates or indices. Moving a window boundary resets the running peak, and the 21.2% belongs to euro sovereign bonds.
4. A smaller drawdown in one crisis does not establish higher compounded wealth across a complete investment period. Euro government bonds, for example, gained 20.7% in September 2007–September 2010 but lost 17.4% in January 2020–January 2023.

Compare the same exposure over the same dates

The bond exposure below is FTSE World Government Bond Index Developed, hedged into euros. The equity comparator is S&P 500 total return, converted into euros without currency hedging. Both include reinvested income, are nominal and exclude investor fees and taxes. They are different exposures, not a controlled experiment isolating interest rates.

These derived comparisons use matching endpoints. In the first window, bond wealth finished 44.6% above equity wealth: 12,020 / 8,310 − 1. In the second it finished 37.9% below: 8,760 / 14,100 − 1.

Those are relative-wealth differences. The corresponding return gaps are +37.1 and −53.4 percentage points. Neither calculation estimates how much a mixed portfolio fell at its worst moment: each asset's maximum drawdown can occur on different dates.

The same fixed payments respond differently to falling and rising yields

When a bond's future payments are fixed, a lower market discount rate raises their present value. A higher rate lowers it. Duration summarises sensitivity to a small yield move; convexity describes the curvature that makes larger moves asymmetric.

Policy moved in opposite directions in these episodes. On 4 December 2008, the ECB announced a 75-basis-point reduction in its main refinancing rate. The ECB's historical rate table records rate increases during 2022. These decisions provide policy context. A central-bank rate is not the yield on every government bond, and markets can price a decision before it happens.

Consider a hypothetical zero-coupon bond paying €100 in ten years. At a 1% annual yield, its price is 100 / 1.01^10 = €90.53. At 3%, it is 100 / 1.03^10 = €74.41: an instantaneous 17.8% loss. Reversing the move produces a 21.7% gain, because the starting price is lower.

The repricing calculation holds maturity constant and ignores default, fees, tax and elapsed time. It illustrates a mechanism; it does not explain a measured percentage of either index's historical return. Coupon income, changing holdings, maturity mix and yield-curve changes also matter. The duration guide examines those sensitivities separately.

Why three different bond-loss numbers can all be correct

The global government-bond loss of 17.3% starts at December 2021 because that is the fixed crisis-window boundary. The whole-history table starts at the earlier July 2020 peak and records 20.2% down to October 2023. Moving the boundary resets the running high.

The 21.2% loss often quoted alongside them belongs to euro government bonds, a different index, from December 2020 to December 2022. That series ends in January 2023. It cannot establish the euro sovereign return through December 2024.

Similarly, the window labelled “Covid” runs from January 2020 to January 2023. Its global government-bond drawdown of 18.1% runs from July 2020 to December 2022. Calling that a March 2020 bond loss would misdate the observation.

Correlation describes the change; it does not explain it

The source reports global government bonds' monthly correlation with world equities excluding dividends at −0.39 in the 2008 window and +0.68 in the 2022 window. That comparator differs from the S&P 500 total-return column above. The sign change documents co-movement over these samples, without identifying causality or promising a future hedge.

The practical distinction is between issuer safety, resale-price sensitivity and the date cash is needed. High-yield credit introduces an additional credit exposure; the withdrawal analysis shows why even a temporary price loss can matter to a spending plan.

What you can do now

These are checks on your own holdings; none of them depends on a view about where yields go next.

1. Read the duration on each bond fund's factsheet. Duration multiplied by a one-point rise in yields gives a rough percentage price fall for a small move, before income.
2. Check whether each bond fund is hedged into euros. Hedged and unhedged share classes of the same global index returned −15.3% and −13.5% in 2022 and +1.4% and −5.4% in 2025, according to justETF's pages for the hedged and unhedged Xtrackers share classes.
3. Compare bond and equity holdings over identical dates. Apply each one's return between the same two statement dates, as in the table above, rather than comparing worst moments that happened at different times.
4. Write down the date of any spending you plan to fund from bonds. A bond fund has no maturity date, so its price on that date is what you receive.
5. See how much of your bond allocation is government debt. The Copio app preview shows allocation by asset class across accounts, which separates government bonds from corporate and high-yield bonds.