The Permanent Portfolio is an asset allocation proposal developed by Harry Browne in the 1980s. It emerged as a direct response to the fundamental problem of economic uncertainty: if the future is impossible to predict, then the optimal (and humble) strategy is not to anticipate economic cycles, but rather to build a portfolio that performs reasonably well in any environment.
The thesis is simple yet powerful: a truly robust portfolio must include assets that both benefit and protect under different possible macroeconomic conditions. Such a portfolio allocates as follows: 25% equities, 25% long-term bonds, 25% gold, and 25% cash or equivalents.
In practical terms, its appeal lies in:
As of 2025, macroeconomic uncertainty prevails, and thus the Permanent Portfolio thesis has regained relevance. This series aims to critically examine it, measure it rigorously, and deepen the analysis with modern portfolio management methods.
The analysis is built using total weekly returns from December 2004 onward, based on representative ETFs for each asset class. This approach aims to more realistically reflect the experience of the average investor, including reinvestment of cash flows, liquidity, and implicit costs.
For this first article, rebalancing is done annually. Rebalancing refers to the periodic adjustment of asset weights back to their target allocation, counteracting distortions caused by natural market movements. The series will later explore a more sophisticated framework of dynamic rebalancing bands, designed to adapt to different investors’ risk profiles without altering the architecture of the base portfolio; for now, however, we maintain the pragmatism of annual rebalancing.
The Permanent Portfolio was designed to provide robust coverage against the four major macroeconomic environments and their combinations: growth, recession, inflation, and deflation. The underlying logic is simple yet powerful: maintain diversified exposure to assets that thrive in different regimes so that the portfolio as a whole remains resilient, without the need to anticipate the economic cycle. The philosophy behind it can be summarized as humility in the face of the inability to predict.
Accordingly, each component of the portfolio serves a specific function:
Although this design provides a structure that is neutral to the economic cycle and conceptually aligned with macroeconomic theory, its relationship with macroeconomic risks will be studied in detail in later installments of the series.
To rigorously evaluate the Permanent Portfolio, five reference portfolios were defined using the same asset universe.
The efficient frontier was constructed under the classic Mean-Variance methodology, with no short positions allowed. From this, three portfolios with different risk levels were selected: conservative (≈5% ex-ante volatility), moderate (≈9% ex-ante), and aggressive (≈13% ex-ante). Two additional cases were included: a 100% equities portfolio (VTI) and another efficient portfolio with the same volatility as the Permanent Portfolio.
The following table summarizes the metrics from a backtest using the full sample, with annual rebalancing and the optimal weights for the six portfolios:
Although the Permanent Portfolio is not efficient in mean-variance terms, this is only one dimension of analysis. In the next section, its historical performance will be evaluated against these benchmarks, incorporating annual rebalancing and more robust risk metrics.
Once the Permanent Portfolio and the comparable portfolios were defined, we proceeded with a backtest. The objective of this section is to analyze the historical performance of the Permanent Portfolio using technical and modern portfolio management metrics. All portfolios are rebalanced annually, starting 12 months after the first available observation.
The metrics are presented in the following table:
Although the Permanent Portfolio is not efficient in mean-variance terms, this is only one dimension of analysis. In the next section, its historical performance will be evaluated against these benchmarks, incorporating annual rebalancing and more robust risk metrics.
Key details:
Despite not lying on the traditional efficient frontier, the Permanent Portfolio demonstrates quantitative robustness that is hard to overlook. Its Sharpe ratio is comparable to that of optimized portfolios and exceeds that of a pure equity portfolio. It also offers a clearly defensive profile in risk metrics, with evident capital preservation: low maximum and average drawdowns, moderate tail risk, and a return distribution less prone to extreme events.
These characteristics consolidate it as a balanced, stable, and resilient strategy—ideal for those who value capital preservation.
The uniform allocation of the Permanent Portfolio does not translate into an even distribution of risk—understood as volatility— in the conventional sense used in the portfolio management industry. Each asset has a different volatility and a specific correlation structure with the rest of the portfolio, which causes certain assets to dominate the total risk.
To rigorously analyze this, the portfolio’s total volatility is decomposed into each asset’s contribution to risk. In addition, to contrast this imbalance, a Risk Parity portfolio is constructed, in which each asset is calibrated to contribute equally to total volatility.
Risk Contributions: Marginal (MRC) and Total (RC)
There are two key ways to break down a portfolio’s total risk: Marginal Risk Contribution (MRC) and Total Risk Contribution (RC). The MRC quantifies how much the portfolio’s total volatility increases with a small increase in exposure to a specific asset. RC is obtained by multiplying the MRC by the asset’s weight in the portfolio.
MRC and RC depend not only on the individual volatilities of the assets but also on how they interact with each other through the covariance matrix (since volatility is used). This means that the correlation between assets plays a central role.
In this exercise, because the Permanent Portfolio weights are equal, the RC becomes a simple scale of the MRC; therefore, when expressed as percentages of the total, both metrics turn out to be identical.
The disproportionate weight of gold and equities in total risk (<75%) is due to their higher individual volatility and positive correlations with the other assets, while short bonds, with low volatility and minimal correlation, contribute almost no risk. This highlights that an equal-weight allocation does not imply an equal distribution of volatility and that, therefore, the high-frequency risk structure of the Permanent Portfolio is dominated by the more volatile assets.
Optimal Weights of the Risk Parity Portfolio
The Risk Parity portfolio seeks to assign asset weights so that each contributes equally to the portfolio’s total risk (volatility). The result is a portfolio that symmetrically distributes volatility sources, giving more weight to more stable and less correlated assets, and less weight to the more volatile or redundant ones.
The portfolio would be as follows:
Based on these weights, a backtest was conducted with annual rebalancing, and the results were as follows:
Due to the highly defensive nature of short bonds (zero correlation with other assets and minimal volatility), the portfolio inevitably concentrates excessively in this asset. As a result, Risk Parity offers a very conservative profile: while its compound annual return (CAGR) is limited, it achieves very low volatility and a contained maximum drawdown (−10.18%), characteristics typical of a portfolio designed primarily for pure capital preservation.
A first practical takeaway is that if the Permanent Portfolio appeals to you but seems to carry more systematic risk than desired, Risk Parity can serve as a useful guide.
Cash Constraint in Risk Parity
In practice, cash (and equivalents) is often treated as an exogenous variable, determined by investment mandate restrictions, desired liquidity, or the client’s risk profile. It is not the result of an endogenous risk-optimized allocation, but rather imposed as a prior limit or preference. For this reason, cash is excluded from the optimization process and the weights of the remaining assets are proportionally adjusted.
For example, if a 20% cash allocation is required, the Risk Parity weights for the three remaining assets are multiplied by 0.8 to preserve their relative proportions. This approach maintains the Risk Parity logic for the actual risk assets while respecting the investor’s tactical or strategic liquidity structure. At the same time, by keeping cash out of the optimization, extreme results that compromise the model’s practical usefulness are avoided.
The resulting portfolio is as follows:
This design better reflects the original intention of Risk Parity—to equalize marginal risk contribution—without the extreme bias toward low-volatility instruments such as short-term bonds.
In performance terms, the portfolio achieved a CAGR of 6.97%, with an annualized volatility of 9.24% and a maximum drawdown of −24.7%, recorded in October 2022.
Compared to the traditional Permanent Portfolio, this design assumes higher volatility and higher expected return. Compared to the full Risk Parity portfolio, it better balances macroeconomic conditions. Therefore, this version is more useful as a strategic core, leaving the liquidity portion to be managed according to the portfolio’s mandate or at the discretion of risk management.
In real life, if an investor needs to hold 40% in cash and wishes to avoid constant macro regime analysis or having a single asset dominate portfolio volatility, they can allocate the remaining 60% equally to equities, long bonds, and gold (20% each). This way, they achieve balanced macro coverage, with similar risk contributions and without the need for frequent tactical management.
While the construction of the Permanent Portfolio does not aim for efficiency under the mean-variance framework, its results position it close to a “moderate–conservative” optimal portfolio, with particular advantages in operational simplicity, resilience to unpredictable adverse macroeconomic environments, and ease of understanding.
Unlike optimized portfolios, its diversification logic is based on protection against major macroeconomic regimes, without explicitly considering volatility or correlations between assets. The portfolio’s relationship with macroeconomic risks will be detailed starting in the third article of this series.
The contrast with a Risk Parity portfolio highlights that equality in weights does not imply equality in volatility. Being blind to volatility, the Permanent Portfolio ends up overexposed to certain risks when analyzed through modern risk metrics such as MRC.
The series will continue to develop these ideas, delving deeper into risk decomposition, proposing methodological adjustments, and exploring practical ways to strengthen the original strategy. The next article will take the risk contribution analysis as its starting point to move toward more sophisticated structures.
