Here’s some handy data on historical asset class returns for the UK. The chart below shows UK asset class returns – with income reinvested – since 1870:
[1]
Data in this article from JST Macrohistory [2]1 [3], FTSE Russell [4], and JP Morgan Asset Management [5].
As you can see, over the long-term equities (shares) have done much better than gilts [6] (UK government bonds) and cash (here the return on UK treasury bills).
Gilts meanwhile beat cash. But the lead has changed hands a few times – most recently during the 1970s inflation outbreak towards the end of the UK’s biggest bond crash [7].
A few other things to note:
- The chart shows real historical asset class returns. So the returns strip out inflation [8], which gives us a more realistic understanding of capital growth in relation to purchasing power.
- The huge 2008/9 bear market [9] and others seem irrelevant on such a long-term graph [10], yet such corrections are painful at the time and can last for years.
- The lines include divots when all three asset classes trended down simultaneously. Most obviously during the catastrophe of World War One, the 1972-1974 crash [11], and just last year in 2022. You can deal with this possibility by using an even more diversified portfolio [12].
- The returns are total returns. That is, you got them so long as you reinvested dividends and interest. If you don’t reinvest income – perhaps because you’re understandably spending it to support a modestly lavish income in your old age – then those stupendous returns come down a lot.
Incidentally, if you’re wondering whether UK historical asset class returns have much relevance to global portfolios, I’d argue that they do.
Firstly, long-run returns of the main UK asset classes are correlated with their global counterparts.
Secondly, few other financial markets can offer such a rich a seam of historical data as the UK’s. Global data is particularly scant before 1970 for equities.
Finally, paying attention to UK historical returns is more pragmatic than relying on US data [13] biased towards the century that the Americans won.
A number of analysts warn that even the US market may struggle to deliver such outstanding results in the future. UK historical asset class returns still rank highly, but are perhaps more reflective of a world in which it’s impossible to pick the winners and losers in advance.
Historical asset class returns: annualised results
Few of us are going to live a century or more (though Gen Z-ers who eat their greens have a decent chance), so let’s break down those historical asset class returns into more manageable chunks:
Historical asset class returns (% annualised)
| 2022 | 10 years | 20 years | 50 years | 152 years | |
| Equities (shares) | -8.1 | 4.0 | 4.9 | 5.3 | 5.3 |
| Government bonds (Gilts) | -30.2 | -2.2 | 0.9 | 3.2 | 1.4 |
| Index-linked bonds | -15.8 | -0.8 | 1.7 | ||
| Cash (Treasury bills) | -6.4 | -1.8 | -0.8 | 1.1 | 0.9 |
Note: the longest annualised return available for index-linked bonds is 2.9% over 40 years.
Again, the table shows real total returns – the annual rate at which the asset class grows (or shrinks) over any particular period after inflation – and with income reinvested.
Equities had a poor 2022. But they typically deliver superior long-term returns as the timeline stretches beyond a decade.
However, nothing is guaranteed.
The longest period of negative annualised returns suffered by UK equities was 25 years.
A combination of World War One, Spanish Flu, overhanging war debt – and the financial and social trauma that followed – kept the stock market suppressed for quarter of a century.
Which is why every investor should be diversified, despite 2022’s harrowing fixed income returns that turned the last ten year’s returns negative for gilts and index-linked bonds.
Gilt returns across the decades seem particularly variable, given this is meant to be a relatively stable asset class.
A glance back at the orange line in the first chart shows that government bond returns seem to be subject to long-term super-cycles that correspond to eras of falling or rising bond yields [14].
For example, gilt yields peaked in 1975 and drifted down thereafter until 2021. That trend of falling yields – and hence rising prices – pepped up bond returns with capital gains adrenaline shots, until 2022’s rapid interest rate hikes ended the party.
Thus, while the historical 50-year return for equities doesn’t look unachievable in the years ahead, we should be probably be much less hopeful about equalling that 3.2% 50-year return for bonds.
Our post on expected returns [15] offers a realistic perspective on the potential of bonds right now.
Finally, cash is often thought of as a safe haven, but it’s delivered the worst long-term returns of all.
Notice that cash has a negative return for the past 20 years after inflation. The end of the low interest rate era has prompted many Monevator readers to switch out bonds for cash. But there can be significant long-term consequences if you take this too far.
Treasury bills as cash proxy – Treasury bills are ultra short-term UK government debt. Academics use the total return of bills as a stand-in for cash interest rates. One reason being that treasury bills are often a big component of cash-like holdings such as money market funds [16].
Historical asset class returns: the long and short of it
It can be misleading to look at just the last couple of years of asset class returns when you’re deciding how to invest over the long-term.
Returns from asset classes are volatile [17], so a few years of history gives you no useful information.
Shares may do very well one year and bonds do poorly. The next year the returns may be different, or it may take years before their relative performance changes.
Equally, looking at the long-run, historical asset class return averages can leave you unprepared for the wild swings in fortune regularly inflicted by equities, sometimes by bonds, and once in a blue moon by cash.
But we can get a sense of how tempestuous each asset class is by looking at the distribution of its annual returns. That is, how widely dispersed returns are and how violently they skew towards large gains or losses.
Equities
[18]
Annual UK equity returns range widely and wildly – anywhere from -57% to 103%. The positive, overall return of equities is revealed by the right-ward bias of the columns. But low and negative returns are still a frequent occurrence. (Statistically-speaking, we can expect equity returns to be negative every one year in three on average.)
Bonds
[19]
The dispersion of gilt returns is much tighter than equities. Lower, positive returns are common, and negative returns quite frequent. But left-tail / right-tail outlier events are fewer and less extreme than in the stock market.
Treasury bills / cash
[20]
Finally, here’s why we all love cash. Those steady, if low, returns keep trickling in. Additionally, the chance of a hideous car-crash is minimal.
Different strokes
The historical asset class return distribution charts help illustrate that different assets are good for different purposes:
- Cash is king [21] for short-term requirements. Think paying bills or saving up a house deposit.
- Government bonds are most useful for planning outgoings in 5-10 years. That’s because you can know the return you’ll get in advance, so long as you hold them until they are redeemed. Bonds are also used [22] to dampen down the volatility of your portfolio, according to your risk tolerance [23]. A young and brave investor might hold no bonds. A 65-year old might be advised to have 50% of their money in bonds.
- Shares are best for long-term investing, since they deliver the highest real returns over longer periods. Adding new money regularly, holding for many years, and reinvesting the income can help you manage the volatility.
By owning a simple portfolio [24] of different assets you can benefit from diversification [25]. When one asset class has a bad year another will likely have a good one. As a result you dampen the ups and downs of your portfolio’s value.
Moreover rebalancing [26] can help smooth out the zigging and zagging of the different asset classes.
The price you pay for this reduced volatility is a potentially lower overall long-term return. That’s because your holdings of lower-risk assets like bonds and cash will typically deliver less in the way of gains than shares.
If you’re investing for the long-term into a pension, say, it may make sense when you’re young to pound-cost average [27] into shares alone. Try to ride out the volatility to maximise your returns.
But whatever you do, never take more risk [28] than you’re comfortable with. Always think about your personal risk tolerance.
And remember that a stock market crash can hit you hard [29] if it strikes as you approach retirement.
Returns and tactical asset allocation
A final – and riskiest – option in deciding how to allocate your money is to take a view on what assets look cheap and expensive at any point in time.
You then tilt your portfolio to try to capture a reversion to the mean. That is, you invest presuming that asset classes will tend towards the average historical returns we saw above.
I do this to some extent. But I wouldn’t recommend it unless you’re sure you can avoid following the crowd – and you understand your poor calls could cost you by actually reducing your returns.
Wealth warning: There’s no proven method for forecasting long-term stock market returns. Studies [30] show even the best predictive metric (the longer-term CAPE ratio [31]) only explains about 40% of future returns.
Anyone can see that different asset classes have good and bad years. It’s obvious from tables of discrete annual returns [32].
But timing when reversion to the mean will happen is very different from just predicting it will happen someday.
Historical asset class returns: a brief history
These things do tend to sort themselves out over time – even if such a reversion feels unthinkable at any given moment.
Just look at the following table from the 2010 edition of the Barclays asset class report:
1899-2009: UK real asset class returns (% per annum)
| 2009 | 10 years | 20 years | 50 years | 110 years | |
| Equities (shares) | 25.9 | -1.2 | 4.6 | 5.2 | 5.0 |
| Government bonds (Gilts) | -3.3 | 2.6 | 5.4 | 2.3 | 1.2 |
| Index-linked bonds | 3.1 | 1.9 | 3.8 | ||
| Cash (treasury bills) | -1.7 | 1.8 | 3.1 | 1.9 | 1.0 |
Source: Barclays Capital Equity Gilt Study 2010
From this table, it’s again pretty clear that different asset classes can deviate from their long-run returns for substantial periods of time.
Moreover equities were showing a very unusual negative real return over the decade to 2009.
As I wrote in the 2010 version of this very article:
UK shares have struggled to advance over the past 10 years, as the markets have been felled by the dotcom crash and the financial crisis.
- You wouldn’t normally expect shares to deliver a negative (-1.2% per year) real return over a decade, or for them to be beaten so handsomely by safe and secure Government bonds.
Over the long term such periods even themselves out, which is one reason why a strong decade for shares [33] may follow the terrible 2000-2010 period.
The FTSE did indeed go on to rally nicely for several years. You did even better with dividends.
Yet in 2009, in the midst of the greatest buying opportunity [34] for a generation – and with the table above showing how badly shares had done for a decade – buying them was not easy [35].
Take comfort from history
Many people said they had sworn off shares for good by 2009.
But you should never say never again [36] if you want to be a successful investor.
- You might also like my article on US asset class returns [13].
Remember if you’re using an investment return or compound interest calculator [37] then it’s legitimate to use long-term historical returns as a proxy for the interest rate function in the calculator.
Note: Comments below may refer to a previous version of this article, so please check their date.
- 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. [↩ [42]]