Is the 60/40 Portfolio Really Dead? A Century of Evidence

Is the 60/40 Portfolio Really Dead? A Century of Evidence

In 2022, a 60/40 portfolio of US stocks and bonds lost 17.5% in nominal terms; its worst calendar year since 2008, and the third-worst bond allocation in a century of data. Stocks fell as the Federal Reserve raised rates at the fastest pace in four decades. Bonds, which are supposed to cushion equity drawdowns, fell alongside them. The Bloomberg US Aggregate Bond Index posted its worst annual return in history. Commentators declared the 60/40 portfolio dead.

But was 2022 a structural break, or a predictable consequence of a specific macroeconomic regime? This article presents Quant Decoded's original backtest of five portfolio strategies across 97 years of US market data (1928-2025), segmented by inflation regime. The central finding: 60/40 is not structurally broken. It has a specific, identifiable failure mode; stagflation. In low-inflation environments, which constitute the vast majority of history, 60/40 remains one of the most efficient simple allocations available. When inflation exceeds 5% and growth stalls simultaneously, the stock-bond correlation flips positive and the diversification benefit disappears. Understanding this distinction is the difference between abandoning a sound strategy and adapting it to the regime.

A Century of 60/40: The Full Picture

Before dissecting regime-specific behavior, it is worth examining the unconditional, full-sample performance of 60/40 and its alternatives. The table below reports results for five portfolio strategies over the full 1928-2025 period, using S&P 500 total returns for equities and 10-year US Treasury total returns for bonds.

The 60/40 portfolio delivered a compound annual return of 8.8% with a Sharpe ratio of 0.54; respectable by any standard. It posted positive nominal returns in 75 of 97 calendar years (77%). But the averages mask enormous variation. The worst rolling decade (1929-1938) returned just 0.8% annualized, barely keeping pace with inflation. The maximum drawdown of -32.4% during the Great Depression required 54 months to recover.

Risk parity, which weights assets by inverse volatility rather than by capital, delivered a lower absolute return (8.1%) but a higher Sharpe ratio (0.58) and a shallower maximum drawdown. The all-weather portfolio, with its 15% commodity and 15% gold allocations, produced the lowest drawdown of any multi-asset strategy (-18.7%) and the fastest recovery time (28 months). The trend overlay improved the Sharpe ratio to 0.61 by reducing exposure during sustained drawdowns.

These results are consistent with the long-run asset return estimates of Dimson, Marsh, and Staunton (2002), whose analysis of over a century of global returns documented similar risk-return tradeoffs across balanced portfolios.

The full-sample numbers suggest that 60/40 works. But full-sample averages can be misleading when the underlying return distribution is regime-dependent. As Ilmanen (2011) argues, inflation is the dominant risk factor for balanced portfolios, and its impact is anything but uniform.

The Inflation Regime Lens

To understand when 60/40 works and when it fails, we segment the 97-year sample into three inflation regimes based on the annual change in CPI-U:

Low/no inflation (CPI at or below 5%): approximately 71 of 97 years. This includes most of the post-war period, the Great Moderation (1984-2007), and the post-GFC era (2009-2020).

High inflation (CPI above 5%): approximately 18 years. This includes the wartime inflation of the 1940s, the oil-shock era of 1973-1981, and the post-COVID inflation spike of 2021-2022.

Stagflation (CPI above 5% and real GDP growth below 1%): approximately 8 years. This is the subset of high inflation where economic growth stalls simultaneously. The canonical examples are 1973-1974, 1980-1982, and 2022.

The table below reports real (inflation-adjusted) returns by regime for each strategy.

The regime segmentation reveals a striking pattern. In low-inflation environments, which account for roughly three-quarters of the sample, 60/40 delivered 7.5% real annually, the highest of any multi-asset strategy. It outperformed risk parity by 0.7 percentage points and all-weather by 1.6 points. The simplest allocation was the best allocation.

In high-inflation environments, the picture reverses. 60/40 managed only 1.2% real; barely positive. The all-weather portfolio, with its commodity and gold ballast, delivered 3.5% real, nearly three times the 60/40 return. Risk parity (2.8%) and the trend overlay (2.9%) also outperformed substantially.

In stagflation, 60/40 was devastating. The portfolio averaged -4.5% real annually across approximately eight years of stagflationary conditions. The all-weather portfolio limited the damage to -0.5%; a 4 percentage point improvement per year. 100% equities was even worse than 60/40, averaging -7.8% real, as earnings compression from rising costs and slowing demand compounded the bond losses.

This is the central finding: 60/40 is not dead as a strategy. It has a specific failure mode that appears under identifiable macroeconomic conditions. The "death of 60/40" is really the death of 60/40 in stagflation.

Why Stagflation Breaks 60/40

The mechanism is straightforward once articulated. The 60/40 portfolio depends on a negative stock-bond correlation for its diversification benefit. When stocks fall, bonds are supposed to rally as investors seek safety, cushioning the portfolio. This relationship holds reliably in most environments, but breaks down under specific conditions.

During stagflation, two forces operate simultaneously. Rising inflation drives bond yields higher, pushing bond prices down. At the same time, rising input costs and slowing demand compress corporate earnings, pushing equity prices down. Both legs of the portfolio bleed simultaneously.

The stock-bond correlation captures this dynamic quantitatively. The table below reports the average rolling 12-month correlation between S&P 500 and 10-year Treasury monthly returns, segmented by inflation regime.

In low-inflation environments, the stock-bond correlation averages -0.31. This is the negative correlation that makes 60/40 work: when equities draw down, bonds rally, buffering the portfolio. In stagflation, the correlation flips to +0.51. Stocks and bonds move in the same direction, and that direction is down.

This finding aligns with Campbell, Sunderam, and Viceira (2017), who demonstrated that the stock-bond correlation is not constant but varies systematically with the inflation regime. Their model shows that when inflation is the dominant macroeconomic shock (as in stagflation), both asset classes respond negatively to the same impulse, generating positive correlation.

Commodities and gold, by contrast, tend to benefit from rising inflation. Commodity prices rise directly with input cost pressures. Gold has historically served as an inflation hedge, rising when real yields are negative or declining. This is why the all-weather portfolio, with 30% of its weight in real assets, holds up comparatively well during stagflation: the commodity and gold legs offset some of the losses in the stock and bond legs.

The Starting Yield Dimension

Inflation regime is the primary determinant of 60/40 performance, but it is not the only one. The starting yield on bonds adds a second, complementary dimension.

The logic is intuitive. When bond yields are high at the time of investment, the coupon income from the 40% bond allocation provides a substantial cushion against price declines. When yields are low, the bond allocation contributes little income and is highly sensitive to rate increases.

To test this, we regress the subsequent 10-year real return of a 60/40 portfolio on the 10-year Treasury yield at inception, using overlapping annual observations from 1928 to 2015 (the last year for which a full 10-year forward return is available).

The starting 10-year Treasury yield explains 73% of the variance in subsequent 10-year 60/40 real returns. This is a remarkably high R-squared for a financial variable. The result is driven primarily by the bond component: starting yield explains 82% of next-decade bond returns but only 8% of equity returns. Since bonds are 40% of the portfolio, their predictability dominates the portfolio-level result.

The practical implication is captured in the table below.

When the 10-year Treasury yields less than 3% at the start of the decade, the subsequent 10-year real return on 60/40 averages just 2.1%; barely above inflation. When starting yields exceed 5%, the average return nearly triples to 6.8%.

This finding extends the foundational work of Campbell and Shiller (1988) on return predictability. Their insight that starting valuations predict future returns applies not only to equities (via the CAPE ratio) but even more powerfully to the bond component of a balanced portfolio.

The two dimensions, inflation regime and starting yield, together form a practical framework. A low-inflation environment with starting yields above 4% represents the best-case scenario for 60/40. A stagflationary environment entered from low starting yields is the worst case.

Robustness Checks

Several robustness tests confirm the stability of these findings.

Sub-period analysis: Splitting the sample into 1928-1975 and 1976-2025 reveals consistent patterns. The stagflation penalty is present in both halves, with 60/40 real returns averaging -3.8% in the first sub-period's stagflationary years and -5.1% in the second. The correlation flip from negative (low inflation) to positive (stagflation) appears in both sub-periods.

CPI threshold sensitivity: The results are not an artifact of the 5% CPI threshold. Using 4% or 6% as the cutoff shifts the year counts but preserves the relative ordering: 60/40 underperforms alternatives during high inflation regardless of the exact threshold, and the underperformance is most severe in the stagflationary subset.

Rebalancing frequency: Annual rebalancing (the baseline specification) produces nearly identical results to quarterly rebalancing for 60/40, risk parity, and all-weather. The trend overlay benefits modestly from monthly monitoring versus quarterly (Sharpe improves from 0.59 to 0.61), as expected for a strategy that depends on timely regime detection.

Real versus nominal: All regime-level results are reported in real terms. Nominal returns during high inflation are mechanically higher but economically misleading, as they do not reflect purchasing power erosion. The stagflation penalty is even more pronounced in nominal terms when taxes on nominal gains are considered, though this analysis does not model tax effects.

Practical Takeaways for Investors

The data suggests a straightforward, two-variable framework for evaluating 60/40's forward prospects.

In low-inflation regimes (CPI below 5%), which have constituted roughly 73% of the last century, 60/40 has historically been one of the most efficient simple allocations. It delivered 7.5% real with lower volatility than 100% equities and higher returns than the more complex alternatives. For most investors in most environments, the traditional allocation has been difficult to beat on a risk-adjusted basis.

When CPI rises above 5% and real GDP growth simultaneously decelerates below 1%, the data points to adding real-asset exposure. Allocating 15-20% of the portfolio to commodities and gold (reducing both the equity and bond legs proportionally) has historically cut real losses roughly in half during stagflation, improving the annualized real return from -4.5% to approximately -0.5%.

The trend overlay offers a complementary layer of protection. Switching the equity allocation to cash or short-term bonds when the S&P 500 trades below its 200-day moving average has historically improved the Sharpe ratio from 0.54 to 0.61 across all regimes. The mechanism is drawdown reduction: the filter gets investors partially out of equities during sustained bear markets, at the cost of whipsaw losses during volatile but ultimately rising markets.

Starting yield adds a secondary lens. When the 10-year Treasury yields below 3%, subsequent decade returns for 60/40 have historically averaged just 2.1% real, regardless of the inflation regime. This does not mean 60/40 should be abandoned at low yields, but it calibrates expectations and argues for a modest tilt toward alternatives (risk parity or all-weather) that are less sensitive to the bond starting point.

The two variables to monitor are CPI trajectory and real GDP growth. Together, they define the inflation regime. The starting 10-year Treasury yield provides a second input for calibrating decade-ahead expectations. This is not a timing model; it is a regime-awareness framework. The evidence suggests that 60/40's death is neither permanent nor unpredictable. It is conditional on a specific macroeconomic configuration that has appeared in roughly 8 of the last 97 years.