The single most effective way to calibrate your expectations about investing is to look at actual multi-decade data. Not this year. Not this decade. A century’s worth.
Today’s lesson is the data. The return, volatility, drawdowns, and recovery times of major equity markets across decades. With sources so you can verify.
Real long-term returns by market
Real = inflation-adjusted. In purchasing-power terms.
From Credit Suisse / UBS Global Investment Returns Yearbook (2024 edition, covering 1900-2023, 124 years):
US equities, 1900-2023
- Real return per year: 6.5% annualized.
- Nominal return: 9.5%.
- Inflation: 3.0%.
- Volatility (σ): 19.7%.
- Best year: +54% (1933).
- Worst year: −38% (1931).
UK equities, 1900-2023
- Real return: 5.2%.
- Volatility: 19.9%.
Italian equities, 1900-2023
- Real return: 2.0%.
- Volatility: 28.4%.
Italy’s long-term return is notably lower than other developed markets. Partly due to two World Wars, the post-WWII rebuild, ongoing high government-debt issues, periodic political instability, and the fact that Italian equity market is small and concentrated in low-growth sectors.
World equities (market-cap weighted), 1900-2023
- Real return: 5.2%.
- Volatility: 17.5%.
A globally-diversified equity portfolio has delivered roughly 5% real per year over 120+ years. That’s the number to anchor expectations on.
What 5% real return means in practice
€10,000 invested at 5% real for various durations:
| Years | Multiple | Final (in today’s EUR) |
|---|---|---|
| 10 | 1.63× | €16,300 |
| 20 | 2.65× | €26,500 |
| 30 | 4.32× | €43,200 |
| 40 | 7.04× | €70,400 |
| 50 | 11.47× | €114,700 |
Compound interest over decades is the magic. At 5% real, your money more than doubles every 14-15 years in real terms.
The US century, 1928-2023
Robert Shiller’s data, widely cited:
- S&P 500 with dividends reinvested, 1928-2023 (real terms).
- Mean return: ~6.5% per year.
- Volatility: ~17%.
- Positive years: ~73%.
- Negative years: ~27%.
Extended drawdowns:
- 1929-1932 Great Depression: −85% real drawdown. Full recovery: ~25 years.
- 1973-1974 oil shock: −50% drawdown. Recovery: ~7 years.
- 2000-2002 dot-com: −49% drawdown. Recovery: ~7 years.
- 2007-2009 GFC: −55% drawdown. Recovery: ~5 years.
- 2020 COVID: −34% drawdown (fastest on record). Recovery: ~5 months.
- 2022: −25% drawdown. Recovery: ~2 years.
Five major drawdowns of 40%+ in a century. Roughly once every 20 years.
MSCI World since 1970
Most relevant benchmark for modern global investors. Data 1970-2023:
- Annualized return: ~10% nominal, ~7% real.
- Volatility: ~15-17% (lower than single-country because of diversification).
- Best year: +45% (1975).
- Worst year: −40% (2008).
- Biggest drawdown: 54% (2007-2009).
FTSE MIB since 1998
The Italian benchmark. Shorter history (1998 start) but revealing:
- Total return 1998-2023: modest. Italian equities have dramatically underperformed global peers.
- Real annualized return: ~1.5% per year.
- Volatility: ~24%.
- Biggest drawdown: ~65% (2008-2009).
Italian stocks have been a terrible investment over this specific window. Home bias destroys wealth when home is Italy.
The lesson isn’t “Italy is permanently bad” — it’s that single-country exposure, even to your own country, is unnecessarily risky. A global index makes this irrelevant; you get the average.
Bond history
Harder to summarize because of varying yields across decades.
Rough annualized real return by decade (US long-term Treasuries):
- 1950s: 0%.
- 1960s: −0.5%.
- 1970s: −2.5% (high inflation).
- 1980s: +7% (falling rates, high yields).
- 1990s: +5%.
- 2000s: +3%.
- 2010s: +1%.
- 2020-2023: −2% (2022 was brutal).
Italian BTP historical real return since 1990s: roughly 2-3% per year, highly dependent on starting yield and inflation.
Over a century, bonds returned roughly 2% real in most developed markets. Less than equities; less volatile.
Correlations
How asset classes move together matters for diversification.
Historical correlations (global data, 20-year):
| Pair | Correlation |
|---|---|
| US stocks / US bonds | 0.0 to +0.3 (varies by decade) |
| US stocks / EU stocks | +0.7 to +0.9 |
| US stocks / EM stocks | +0.6 to +0.8 |
| US stocks / gold | −0.1 to +0.2 |
| US stocks / real estate | +0.4 to +0.6 |
| US stocks / oil | +0.2 to +0.4 |
Developed equity markets are highly correlated, particularly during crises. Stocks and bonds typically have low correlation (good for diversification), though 2022 saw both fall together — a reminder that historical correlations can break down.
The power of dividends
Price charts alone understate actual return. Dividends and reinvestment matter.
S&P 500 price-only return 1928-2023: ~6% nominal per year. S&P 500 total return (with dividends reinvested): ~10% nominal per year.
That’s ~4% per year just from dividends. Over a century, dividends double or triple the cumulative return.
Accumulating ETFs capture this automatically. Distributing ETFs require you to actively reinvest.
The “lost decade” phenomenon
Sometimes a market delivers near-zero real return for a decade.
- Japan 1989-2019 (30 years!): Japanese Nikkei was flat nominally, negative in real terms. Thirty years.
- US 2000-2009: S&P 500 flat-to-negative nominally in real terms. Dot-com bust + GFC = “lost decade.”
- Italy 2000-2024: FTSE MIB up maybe 20% nominally over 24 years, deeply negative real.
Single-country concentration exposes you to lost decades. Global diversification reduces but doesn’t eliminate.
For a long-horizon investor (20+ years), even lost decades usually recover. For a short-horizon investor, timing matters a lot.
The rare wealth-destruction events
Most analysis assumes “ordinary” market risk. Some countries have experienced worse:
- Russia 1917: stock market zero after revolution.
- Germany 1945: near-total collapse.
- China 1949: communist takeover, pre-war holdings zeroed.
- Argentina repeatedly: hyperinflation and default.
For developed-world investors in 2025, these outcomes seem unlikely but not impossible over 50+ year horizons. Diversification across countries is partial protection.
Italian wealth-destruction events within living memory: WWII (of course), the 1970s inflation (real assets better than financial), 1992 currency crisis (lira devaluation).
Interpreting “past performance”
The universal disclaimer: past performance isn’t a guarantee of future results.
But it’s the best we have. Multi-decade data across many countries shows:
- Equities returned roughly 5-7% real over the very long run.
- Diversified global portfolios outperform single-country.
- Bonds returned 1-3% real.
- Cash returned ~0% real.
Future expected returns might be higher or lower than historical averages. Current valuations (e.g., Shiller PE) suggest expected returns below long-term average from 2024 onward. Planning for 4-5% real for equities over the next 20 years is reasonable, not 7%.
What to do with this lesson
Three things:
- Print a chart of your chosen index’s 50-year history. Stare at the ups and downs. Prepare yourself emotionally for the inevitable.
- Use real, not nominal, return assumptions in your planning. 5% real is reasonable for global equity over 30-year horizons.
- Plan for at least one 40% drawdown over any 20-year window. It’s not “if” — it’s “when.”
Sources
- Credit Suisse / UBS — Global Investment Returns Yearbook 2024.
https://www.ubs.com/global/en/investment-bank/in-focus/2024/global-investment-returns-yearbook.html. - Robert Shiller — Online data.
http://www.econ.yale.edu/~shiller/data.htm(US stock market 1871-present). - Dimson, Marsh, Staunton — Triumph of the Optimists, Princeton University Press, 2002.
- MSCI — Historical index data.
https://www.msci.com/end-of-day-data-search. - FTSE Russell — FTSE MIB Index.
https://www.ftserussell.com/products/indices/italia.
Next lesson: crashes, bears, and the worst cases — a closer look at specific market disasters and how long recoveries took.