Deutsche Bank's sweeping new study spanning 56 economies and 200-plus years delivers some uncomfortable arithmetic for modern portfolios — and a few surprises.
The Long Arc of Compounding
The arithmetic of long-term investing is unforgiving. Deutsche Bank's Jim Reid and his co-authors draw on one of the most comprehensive datasets ever assembled to make the case that patient, risk-tolerant investors have consistently been rewarded. Over a 200-year median series, global inflation-adjusted returns in USD show equities returning 4.9% annually, a 60/40 portfolio at 4.2%, government bonds at 2.6%, and cash effectively destroying wealth at negative 2.0% per year. Reid states that "history shows that investors have been consistently rewarded for taking risk and compounding the dividends and coupons available in equities and bonds." The evidence is not a footnote. It is the foundation.
Gold complicates the narrative. Over two centuries it managed just 0.4% annually in real terms — far behind income-generating assets. Yet since 1999, gold has returned 7.45% per year in real terms, outpacing US equities at 5.8%, German equities at 3.9%, and UK equities at 3.3%. The 21st century has been an anomaly, and the report is careful not to project it forward.
The GDP Anchor
Nominal GDP growth is the load-bearing beam of this entire analytical structure. The report establishes that global nominal GDP has grown at an average annual rate of 5.7% since 1900, and that asset prices are ultimately claims on that growth. The problem is that developed market nominal growth has deteriorated sharply. By end of 2024, 25-year rolling nominal DM equity returns were the lowest since 1877. Reid warns that "forecasters rarely incorporate future policy-driven inflation in their long-term models," noting that since 1971 no economy has averaged annual inflation below 2% as policymakers expand the monetary base in bad times without reversing course in good ones.
Demographics compound the pressure. Both DM and EM are experiencing their slowest population growth in two centuries. Of 56 economies studied, 32 are projected to see working-age populations shrink by 2050. Reid raises the open question: can artificial intelligence compensate for demographic contraction by boosting productivity? The answer remains genuinely unknown.
Valuation Is Destiny
The single most powerful predictor of long-term equity returns is starting valuation. The report constructs portfolios sorted by above- and below-median P/E ratios, rebalanced annually. The low P/E portfolio returned 20.2% per year over 70 years; the high P/E portfolio returned 11.4%. Using dividend yields back to 1819, the high-dividend portfolio returned 12.8% against 9.3% for the low-dividend basket. Reid identifies the US as "the exception, not the norm" globally and historically — a market that has continued to generate returns despite elevated CAPE and P/E ratios and historically low dividend yields driven by buybacks.
The CAPE-based analysis of the S&P 500 is instructive: the only prior period of higher CAPE ratios was the dot-com build-up of 1999-2000, and the 10-year real returns following that peak were negative.
Bonds: The Asymmetric Risk No One Talks About
For fixed income, the report documents something advisors should not ignore. The probability of underperforming inflation is roughly 25% across 5-, 10-, and 25-year horizons for government bonds, reflecting the long-cycle nature of fixed income super-cycles. Real bond drawdowns are not merely cyclical. France's bond market entered a drawdown in 1790 that lasted 234 years. Germany's began in 1910 and stretched 114 years. Italy's 1913 drawdown still registers as ongoing at 111 years.
Reid observes that "when bond investors experience substantial real losses, the damage is often near-permanent," reflecting the asymmetric nature of bond returns where upside is capped but deep drawdowns can endure for decades. Despite this, the 60/40 portfolio has delivered the lowest probability of nominal loss across all measured horizons, with just 0.1% probability of negative 25-year returns.
Five Key Takeaways for Advisors and Investors
- Starting valuations are the single strongest predictor of long-term equity returns globally. Portfolios tilted toward cheaper markets have consistently outperformed across centuries.
- Equities lose to inflation roughly one time in four over any given 5-year period. Time horizon is not an abstraction; it is a risk management tool.
- Bonds retain portfolio value not primarily through diversification but through income. Starting yield levels determine nearly everything about future bond returns.
- The 60/40 portfolio is not a compromise. It is historically the lowest-risk nominal-loss construction over 25-year horizons, outperforming equity-only on a risk-adjusted basis in almost every country studied.
- Demographics and nominal GDP trajectories suggest structurally lower future returns for developed markets, absent an AI-driven productivity surge. Return expectations should be recalibrated accordingly.
Footnote: Reid, Jim, Henry Allen, and Galina Pozdnyakova. The Ultimate Guide to Long-Term Investing. Deutsche Bank Research Institute, 27 Oct. 2025.
