Kristopher O’Sullivan · October 2026
Global equities: my biggest long-term bet
If you’re willing to pay the cost of admission, history suggests the odds are excellent
Global equities beat cash in 71% of 1‑year periods and 100% of 30‑year periods1900–2026
£10,000 in a passive global equity fund against £10,000 in a bank account earning interest (approximated), for every start year
Real total returns, in today’s pounds. Global equities are a historic reconstruction to 2008 with a 0.15% fund fee, and an index fund from 2009. Cash is a wholesale interest rate approximating a bank account. Sources and method are in the appendix.
Past performance is not a guide to future returns.
By global equities I mean a low-cost investment that owns a slice of thousands of listed companies across the world. If history since 1900 is anything to go by, they are one of the best long-term bets you can make with your money. I’ve bet on global equities for the last 10 years, and plan to for decades to come – they are currently around 65% of my investments.
In every rolling 30-year period since 1900, global equities finished ahead of cash (including interest), shown here. But the cost of admission is steep: falls of around 30–50% along the way, and a portfolio that can sit in the red for years. There are also other risks this data doesn’t show – these matter too.
This is an opinion piece based on historic data, not financial advice. The value of investments can fall as well as rise and you could get back less than you put in.
History gives global equities great odds
Let’s see how global equities and cash (remember, including interest) performed through different periods of history. This data is all real, meaning that inflation is taken into account.
£10,000 into global equities and £10,000 into cash starting from 1990.
What £10,000 became
The difference (global equities minus cash)
Real total returns, in today’s pounds. Global equities are a historic reconstruction to 2008 with a 0.15% fund fee, and an index fund from 2009. Cash is a wholesale interest rate approximating a bank account. Sources and method are in the appendix.
Past performance is not a guide to future returns.
Start years to try
- : the dot-com crash, starting from the highest price-to-earnings ratio on record – cash was ahead after 10 years
- : starting the year after the financial crisis – one of the strongest 10-year runs in the data
- : the only start year where cash was still ahead after 20 years, through the First World War, the post-war slump and the 1929 crash. Global equities win from 21 years
- : a long bull market gives the biggest global equities wins
All start years at once
We just looked at each start year separately, now let’s see them all together. The dots are exactly the same ones from the difference chart above, but now they’re all together. Each dot is one start year, placed by how far £10,000 in global equities ended ahead of, or behind, £10,000 in cash after that many years. Every start year from 1900 is here. Hover over a dot to see which year it is.
Global equities minus Cash1900–2026
% of start years each was aheadReal total returns, in today’s pounds. Global equities are a historic reconstruction to 2008 with a 0.15% fund fee, and an index fund from 2009. Cash is a wholesale interest rate approximating a bank account. Sources and method are in the appendix.
Past performance is not a guide to future returns.
Given more years invested, the dots move convincingly upwards. Global equities finished ahead of cash in 92% of 10-year periods, 99% of 20-year periods and 100% of 30-year periods.
How often is only part of the story. The size of the win is important too: the median difference over cash was about £18,200 at 20 years. This means that in half of the 20-year periods, global equities beat cash by at least +£18,200 (from the £10,000 starting investment in each).
The one exception at 20 years was 1912. Global equities from 1912 experienced the First World War, the post-war slump and the 1929 crash, and by 1932 was behind cash. Although losing to cash, global equities did still beat inflation: the £10,000 grew to £11,719 after accounting for inflation.
So, history gives a bet on global passive equities excellent odds if you are investing for the long term. But, it cannot give it a guarantee – one century is great evidence but not a promise, and we may experience even more extremes than we’ve seen in the past. The next sections show the short-term picture which is much more choppy and discuss other risks.
The cost of admission
Historically you’ve typically been very well rewarded for holding global equities over long time horizons. However, there have also been sharp and sudden falls in value over shorter time periods.
The pain of your investments being significantly in the red is real, and can last for years. This is the cost of admission.
Global equities: falls and recoveries1900–2026
- 1921−53%
- Fall lasted 13 years
- 1932−42%
- Fall lasted 7 years
- 2003−39%
- Fall lasted 7 years
Real total returns, in today’s pounds. Each fall is measured at year end from the highest year-end value reached before it. Each labelled year means the start of the year, so a fall during 2022 appears at 2023. Global equities are a historic reconstruction to 2008 with a 0.15% fund fee, and an index fund from 2009. Cash is a wholesale interest rate approximating a bank account. Sources and method are in the appendix.
Past performance is not a guide to future returns.
Every 20-year winning period contains a fall along the way, and the investor who collected the win at the end is the one who held on through the falls.
Selling in a fall locks in the loss. The long-term investments I make are the ones I can hold through the falls, not just the ones that look good over the long term.
Starting price. A high price per pound of company earnings (a high price-to-earnings ratio, or P/E) at the start has historically been an indicator of lower returns over the following decade or two. The two most expensive starts in the data, 1929 and 2000, each took over a decade to pull ahead of cash. Expensive starts have gone on to beat cash over the long run – but it’s worth taking into account where the price-to-earnings ratio is when you start.
Other risks the data doesn’t show
Market access. The data assumes you could always buy and sell in all markets and countries. In reality there are periods you could lose access to particular countries, or indeed your entire holdings in those countries. A British investor’s American shares were compulsorily acquired by the Treasury in both world wars, and a German investor’s foreign holdings were seized outright as enemy property. Exchange controls restricted and penalised British investors in most overseas markets from 1939 to 1979, and many foreign holders of Russian assets lost theirs in 2022. Capital controls in a market as large as the US would hit a global holder hard, whatever the companies went on to earn. In the scenarios severe enough to break this pattern, cash has rarely been a safe place to be either – so the comparison is between equity and cash risks, not between risk and safety.
Fund and platform. The fund that tracks the index and the platform that holds it can both fail, like banks can. This is uncommon, and a UK-regulated firm must keep client assets separate from its own, with the Financial Services Compensation Scheme covering some losses if a firm fails. Holding more than one fund on more than one platform also reduces this risk.
Private companies. The case for global equities rests on owning the profits of the global economy. A growing share of economic activity is happening in private companies that a listed index cannot hold. If that share keeps growing, a global fund will own a smaller piece of the economy than it did over the century in the data.
The limitations of the data itself, from overlapping periods to taxes, are in the appendix.
Why I take the bet
Two things persuade me: 126 years of compelling data, and believing the pattern will continue. And I’m prepared for the cost of admission – holding through potentially prolonged and deep falls.
The data. A century covering two world wars, a depression, double-digit inflation and several crashes, and in nearly every start year the long-term global equities investor beat cash, by a margin that grew with time. It isn’t proof, but it is significant evidence.
The theory. Most of the world’s economic activity runs through public companies that make a profit. A global equity investment is a claim on those profits, so owning it means owning a piece of the global economy. That’s why I expect the pattern to continue. Two things this article doesn’t cover, which I’ll come back to: why equities should beat cash in theory, not just in the record, and why global rather than UK or US alone.
I’m happy to pay the cost of admission. I’m willing to sit through prolonged and deep downturns for years and manage my overall risks:
- This is money I won’t need for a decade or more, so the odds get the horizon they need.
- I hold cash separately, so I’m protected from being a forced seller in a fall.
- I’ve already held through the falls of late 2018 and 2022 (2019 and 2023 on the chart), and not only kept my position but bought more. I’m emotionally prepared for future falls.
A century of evidence, a reason to expect it to continue, and a position I can hold through the bad years. That’s why global equities are the biggest long-term bet I make with my money.
This is an opinion piece based on historic data, not financial advice. The value of investments can fall as well as rise and you could get back less than you put in.
Appendix – data, methodology and limitations
Every dot and line above comes from two yearly series, one for global equities and one for cash, both in pounds and both after inflation. The figures were last refreshed from the model on 15 October 2026.
Global equities
No global index fund existed for most of the century, so the series is a reconstruction up to the end of 2008 and a real fund from 2009. The reconstruction uses the Jordà-Schularick-Taylor Macrohistory Database (release 6), which holds yearly equity total returns, with dividends reinvested, for 18 countries. Those with a complete record (16 of them) from 1900 go into the reconstruction. Each country’s return is converted into dollars and the countries are weighted by their share of the group’s GDP, reset at the start of each decade, because market sizes do not exist for every country back to 1900. The dollar composite is then converted into pounds. Japan has no equity return for 1946 and 1947, while that market was being rebuilt, so those two years are set to 0%; the few missing GDP and exchange-rate values are interpolated between neighbouring years.
From 2009 the series is a real index fund that tracks the FTSE Global All Cap Index, a benchmark of large, mid and small companies across developed and emerging markets: its yearly total returns, converted from dollars into pounds. The fund is deliberately not named. This article recommends no product. The two series are joined at the end of 2008 by chaining the fund’s yearly returns onto the reconstruction.
Cash
Cash data in this article is the UK short-term interest rate, a rate at which banks lent money for short periods, also from the Jordà-Schularick-Taylor Macrohistory Database (release 6). The cash returns are compounded each year to 2018, then the Bank of England’s SONIA compounded index to 2025. This is a proxy for a bank account, not a record of one: a wholesale rate, with no tax taken and no bank’s margin.
The nature of cash changed during the period. Sterling was tied to gold until 1914, again from 1925 to 1931, and then to the dollar under Bretton Woods until 1971. Interest rates and inflation behaved differently under those regimes than they have since, with an independent central bank targeting inflation. The pre-1971 cash dots are evidence from a different monetary world.
Inflation and currency
Both series are deflated by UK consumer prices: the Macrohistory Database’s UK CPI to 2020, extended with ONS CPI from 2021. So every figure in the article is in today’s money. Dollar-to-pound conversions use the database’s exchange rates to 2020 and Yahoo Finance rates after that.
The equity returns are unhedged, converted to pounds at each year’s exchange rate, so sterling’s moves over the century are inside the dots. Over thirty-year holds, exchange rates have broadly followed the difference in inflation between countries, so in today’s money currency is not a big factor. It matters more over short periods and in the year you sell.
Costs
The reconstruction is reduced by a fund fee of 0.15% a year, deducted from every year 1900 to 2008. That is representative of current offerings: at least five passive global equity funds over £1bn in size charge less. The fund’s own returns from 2009 are already net of its fee. Platform fees, dealing charges and bid-offer spreads on purchases, sales and dividend reinvestment are not deducted; each would take a little more from the equity dots.
Limitations of the data
Overlapping periods. Neighbouring 30-year windows share 29 years, so the 97 dots at 30 years are closer to 4 independent trials than a hundred. The odds quoted in the article are a summary of one century, not a sample of many. A 100% record over 4 trials is encouraging, not a law.
Yearly readings. Both series are annual. Falls inside a year (1987, March 2020) were deeper than the chart shows, and some short recoveries are invisible.
Survivorship. The database’s 18 countries are the ones with continuous records. Russia, China and Austria-Hungary, together over a tenth of world market value in 1900, are absent, and Russian and Chinese shareholders lost everything in 1917 and 1949. A world-weighted investor in 1900 held Russian shares. When Dimson, Marsh and Staunton added those three countries to their own world index, it lowered the annualised real return by 0.14% a year. The reconstruction here likely overstates by a similar margin, and the Japan patch adds a bit more.
Not an investable index. Nobody could buy this portfolio in 1900: the first global index funds arrived in the 1990s. The reconstruction records what the markets did, not what a British investor could have collected at the time.
Taxes. The dots are before tax. Outside an ISA or pension, tax on dividends and gains takes a slice of the equity return, and tax on interest a slice of the cash return.
Adding and withdrawing. The dots are a lump sum left alone. Monthly contributions change the result, and drawing money out changes it most: a bad early decade matters far more when you are withdrawing through the fall. The 1912 line is a lump-sum result. Someone drawing an income from it would have done worse.
Market access, fund and platform risk and private companies are covered in the section on other risks the data doesn’t show.
Sources
- Òscar Jordà, Moritz Schularick and Alan M. Taylor, Jordà-Schularick-Taylor Macrohistory Database, release 6. Equity returns from Jordà, Knoll, Kuvshinov, Schularick and Taylor, “The Rate of Return on Everything, 1870–2015”, Quarterly Journal of Economics (2019).
- MeasuringWorth, UK short-term interest rates, as carried in the Macrohistory Database.
- Bank of England, SONIA compounded index.
- Office for National Statistics, CPI annual rate.
- Yearly total returns of an index fund tracking the FTSE Global All Cap Index, and GBP/USD exchange rates from Yahoo Finance.
- Elroy Dimson, Paul Marsh and Mike Staunton, Credit Suisse Global Investment Returns Yearbook 2013, for the survivorship estimate.