Serious capital losses can reduce our appetite for risk, just as surely as a night clutching the toilet bowl will put you off eating raw oysters for life.
But our psychological hard wiring presents us with a dilemma.
Foul, nausea-inducing returns now and then come with the territory in financial markets.
And we know these gut-wrenching episodes are liable to impact our future decision-making, because they trigger our impulse to avoid similar unpleasantness in the future.
In other words we’re prone to negativity bias.
But common wisdom among many investing masochists veterans is that outsized profits are made after a market meltdown.
“Buy when there’s blood on the streets!” and all that charming imagery.
And if that’s true then our natural response to shy away from whatever just hurt us could do us more harm than good.
UK equities: ten worst annual returns 1871-2022
So which view is correct?
Do awful returns fire the starting gun for massive bargains? Do you just need the testicular fortitude to scoop them up?
Or do market swan dives just as often signal that there’s more pain ahead, as feared by our savannah-ready emotional engineering?
The table below – which features real1 [1] returns – shows how UK equities bounce back – or belly-flop – after their ten most negative single years since 1871.
| Bad year | Return (%) | +1 year (%) | +3 years (%) | +5 years (%) | +10 years (%) | 10yr annualised (%) |
| 1916 | -17.4 | -12.5 | -13.8 | -36.2 | 50.1 | 4.1 |
| 1920 | -31.8 | 8.6 | 71.1 | 137.7 | 181.2 | 10.9 |
| 1931 | -16.5 | 37.8 | 99.7 | 167.6 | 78.5 | 6 |
| 1937 | -15.9 | -11.2 | -28.4 | -12.3 | 11.1 | 1.1 |
| 1940 | -18.6 | 10.8 | 31.5 | 49.2 | 35.9 | 3.1 |
| 1969 | -16.2 | -9.4 | 31.8 | -62.7 | -12.1 | -1.3 |
| 1973 | -34.2 | -57 | -22.5 | 1.2 | 75.9 | 5.8 |
| 1974 | -57 | 103.4 | 132.6 | 135.4 | 415.7 | 17.8 |
| 2002 | -23.2 | 18.6 | 57 | 83.5 | 74.9 | 5.8 |
| 2008 | -32.2 | 26.2 | 29 | 65.3 | 87.2 | 6.5 |
Real2 [2] total returns from JST Macrohistory [3]3 [4]. February 2023.
One thing jumps out from this table – the severity of the first year’s losses tells us little about what’s coming next.
The very worst year (1974) led directly to the best year in UK stock market history – the 103% doozy of 1975.
Yet the second-worst year (1973) bled straight into the 1974 nightmare. (Indeed the two years fused into the UK’s worst stock market crash [5] since the South Sea Bubble.)
Meanwhile, the third, fourth, and fifth bleakest years in our chart (2008, 1920, and 2002) were all followed by large rallies.
On the other hand, three of the five least worst-drops kept tunnelling down in year two.
More often than not, equities bounce back fast
On balance the table provides tentative evidence supporting the theory that a severe shock for shares can abate quite quickly.
This is conjecture, but perhaps in the best cases the bolder investors quickly see the panic has been overdone and pile in. Their forays restore confidence among the rest of the herd, leading to further gains.
Milder hits may not flush quite enough negativity out of the system within just a year, however. Hence there’s a fairly strong chance that escalating disquiet blows up into a deeper decline in year two.
Or maybe it’s all to do with the credit cycle or a dozen other theories…
The recovery position
Whatever the driver, a recovery is usually under way three years after the initial slump.
Seven out of ten aftermaths feature high single- to double-digit average growth. By the third-year mark, the ranges4 [6] rove from 9% annualised (after the Financial Crisis) to 32% annualised (post-1974).
Those return rates are chunky compared to the historical average return of around 5% for equities.
Less happily: we can see three events were in contrast still poisoning the water supply five years out. And one was still pishing in the pond after a decade.
Two of these periods were hamstrung by the World Wars. The other (1969) slid into the 1972-74 crash and the worst outbreak of inflation in UK history.
Yet even these observations don’t enable us to formulate a simple heuristic such as: ‘bail out for the duration of a major war or stagflationary malaise’.
For one, the ten-year returns beyond 1916 are perfectly acceptable, if nothing to brag about.
Next, let’s examine the difference in an investor’s fate after 1973 compared to 1974.
What a difference a year makes
The post-1973 path took a decade to straighten itself out. In contrast, you were skipping along like it’s the Yellow Brick Road straight after 1974.
But realistically, how many investors who’d just been through the 1973 shoeing would be itching to double-down after the -72% roasting inflicted by the end of 1974?
You’d have to be a robot – or rich enough not to really care about losing money – to wade in after that two-year bloodbath.
Still, if you held your nerve you were handsomely rewarded. Returns were close to an extraordinary 18% annualised for the next decade.
The really unlucky cohort were the 1969-ers. These guys suffered a relatively mild recession at the tail-end of the ’60s, but they then ran smack into the 1972-74 W.O.A.T.5 [7], and ended up with negative returns after ten years.
Ultimately, these investors recovered to 5% annualised respectability.
But it took 16 years of keeping the faith to get there.
World equities: ten worst annual returns 1970-2021
How does the picture change if we look beyond UK equities? We have good data on the MSCI World index going back to 1970.
Let’s see how quickly (or not) global equities bounce back from the abyss:
| Year | Return (%) | +1 year (%) | +3 years (%) | +5 years (%) | +10 years (%) | 10yr annualised (%) |
| 1970 | -10.2 | 2.1 | -2.2 | -25.3 | -37.9 | -4.6 |
| 1973 | -22.6 | -38.1 | -10.5 | -27.8 | 7.2 | 0.7 |
| 1974 | -38.1 | 23.4 | 16.1 | 0.8 | 116.8 | 8 |
| 1977 | -19.7 | 0.6 | -11.5 | 15.8 | 136.2 | 9 |
| 1979 | -13.7 | 1.9 | 33.4 | 115.1 | 311.1 | 15.2 |
| 1987 | -11.8 | 22 | -2.6 | 26.5 | 112.5 | 7.8 |
| 1990 | -35.5 | 14 | 59.1 | 83 | 213 | 12.1 |
| 2001 | -15.6 | -28.7 | -11.3 | 8.8 | 4 | 0.4 |
| 2002 | -28.7 | 18.1 | 49.4 | 60.4 | 57.4 | 4.6 |
| 2008 | -20.3 | 12.4 | 14.1 | 50 | 126.8 | 8.5 |
Real total returns (GBP) from MSCI [8]. February 2023.
Quick aside: last year’s -16.6% loss slots in at no.7 on the World Annus Horribilis chart. But I’ve excluded that result because, well, we don’t know how it turns out yet.
The pattern of the worst routs leading to the best rebounds mostly holds true on the world stage, too. 1973 proves to be the exception once more.
We can also see the past 50 years has been much kinder to stocks than the first half of the 20th Century. There were no World Wars, Great Depressions, or what have you.
Nevertheless it still takes five years before a majority of the sample periods turn positive.
At the three-year mark, half the pathways are underwater.
But five years on, and only two scenarios are negative. Of the goodies, two are positive but miserable, two have average returns, and four above-average to superb.
Finally, at the ten-year mark, three of the timelines were all told a thankless slog. (Think working in the laundromat in Everything Everwhere All At Once [9].)
The others are all excellent though. Well, except for post-2002. It hovers right around average.
UK gilts: 10 worst annual returns 1871-2021
Now let’s consider UK government bonds [10].
| Year | Return (%) | +1 year (%) | +3 years (%) | +5 years (%) | +10 years (%) | 10yr annualised (%) |
| 1916 | -32.5 | -17.7 | -36.7 | -35.2 | 8.6 | 0.8 |
| 1917 | -17.7 | -7.5 | -38.3 | 6.1 | 44.8 | 3.8 |
| 1919 | -16.8 | -19.7 | 38 | 65.4 | 98.7 | 7.1 |
| 1920 | -19.7 | 27.4 | 90.5 | 106 | 188.1 | 11.2 |
| 1947 | -19.9 | -6.5 | -17.1 | -38.3 | -50 | -6.7 |
| 1951 | -17.2 | -10.2 | 2.3 | -19.3 | -31.7 | -3.7 |
| 1955 | -14.5 | -7.7 | -5.3 | -12.4 | -10.1 | -1.1 |
| 1973 | -16.6 | -27.2 | -20.5 | -8.7 | 26.8 | 2.4 |
| 1974 | -27.2 | 10.9 | 38.3 | 17.3 | 73.3 | 5.7 |
| 1994 | -12.2 | 14.5 | 38.2 | 56 | 99.3 | 7.1 |
Real total returns from JST Macrohistory. February 2023.
Quick aside part two. Last year’s -30.2% ranks at number two in the UK gilt all-time losses chart. But again 2022 is excluded due to crystal ball malfunction.
First thing to notice is that the UK’s worst one-year bond losses aren’t much more gentle than our grimmest stock market losses. (And they’d be nastier still if we threw 2022 into the mix.)
Partially that’s because the UK’s historical gilt benchmark was stuffed full of highly-volatile long bonds. Bond drops are gentler if you stick to shorter durations [11].
But much of the story hinges on inflation. In fact the only three positive years in the ‘+1 year’ column occurred because heightened inflation fears subsided, rather than escalated.
Roll the time-tape on three years, and the only middle-ground is the post-1951 nothing burger.
Every other path is either a double-digit return spectacular, or else it’s negative growth purgatory.
But it’s the five-year column that really shows how a bond bounce-back can be arduous.
Fully 50% of this sample still remains in the red at that point. Whereas we’d seen 70% of UK equities bounce back by the five-year post-crash mark.
What was that about slow and steady?
Remember, over the long-term we’re not expecting much more than 1% annualised real returns from government bonds.
Yet by the time a decade has elapsed, only one outcome from our sample of worst starting points has delivered anything like that.
Four of the following decennial returns are equity-hot. (That’s good!) Two are great, at least for bonds. But three would leave you ruing the day.
That latter trio of roads to nowhere (1947, 1951, 1955) were all caught in the middle of the UK’s biggest bond crash. Inflation kept slipping its leash and mauling the real returns from fixed income.
Hope for the best, but be ready for the worst
While none of this data is predictive of future outcomes, I think we can draw a few general lessons.
Firstly, the worst equity crashes are not predictive of more slaughter to come. The majority are a reset that auger better days ahead. Equities bounce back and usually sooner rather than later.
If you’ve just taken a heavy hit in the stock market then your best (but far from guaranteed) route back to profit is to hang in there. The market should fairly quickly pick up speed again.
Eventually any market will almost certainly right itself. That’s why equities and bonds have positive return records going back 150 years [12] and more.
But the rebound may not happen according to a timetable that suits you. The longest string of successive negative returns for UK equities was 12 years straight.
Incidentally there’s also an outlier pathway in the historical record that does nicely for 18 years, and then collides with World War One. That calamity saddled 1897 equity investors with a negative return after 25 years!
An extreme event for sure. But it helps illustrate why 100% equities is a risk. The expected returns [13] you’d planned for may not be there when you want them.
Do you have bouncebackability?
Most of us are likely to go through the investing meat grinder at some stage in our lifetimes. That’s the price of entry as an investor.
Just think of all the big crashes recently. How many investing experts managed to swerve the Global Financial Crisis? The Covid crash? Or the inflationary shock of 2022?
Predictive power is in short supply. Rather it’s staying power that we need.
We say keep your head together after a bad run and don’t chase the market. Give it time and it should turn in your favour. Sooner or later your patience will very likely be rewarded.
Take it steady,
The Accumulator
P.S. This concept was inspired / shamelessly cribbed from US asset manager and author Ben Carlson. See his post on US stock and bond rebounds [14]. But I’d just like to say in my defence that I’m a big fan of Ben’s work. And I’d do it again, so help me!
- That is, inflation-adjusted. [↩ [19]]
- Real returns subtract inflation from your investment results. In other words, they’re a more accurate portrayal of your capital growth in relation to purchasing power than standard nominal returns. [↩ [20]]
- Òscar Jordà, Katharina Knoll, Dmitry Kuvshinov, Moritz Schularick, and Alan M. Taylor. 2019. “The Rate of Return on Everything, 1870–2015.” Quarterly Journal of Economics, 134(3), 1225-1298. [↩ [21]]
- Three-year annualised return, not shown in the table. [↩ [22]]
- Worst of All-Time! [↩ [23]]