The Permanent Portfolio

Harry Browne's Permanent Portfolio allocates 25% to stocks, 25% to long-term bonds, 25% to gold, and 25% to cash. It is designed to deliver positive returns in any economic regime, but at the cost of dampened upside and complexity.

Key takeaways

• The Permanent Portfolio is optimized for non-correlation, not return maximization
• It has delivered 4.5–5% annualized returns with 6–7% volatility since 1975
• This is exceptionally low volatility, but half the return of a typical stock-heavy portfolio
• Gold at 25% is unusual and hard to justify except as tail-risk insurance
• Equal weighting (25/25/25/25) may be suboptimal; 40/30/15/15 performs better historically

The philosophy: non-correlation as the goal

Harry Browne published "Fail-Safe Investing" in 1992, introducing the Permanent Portfolio concept. His insight was deceptively simple: most investors optimize for maximum return, which requires taking on the same risks. When a crisis comes that affects those risks, everyone suffers together. Instead, optimize for non-correlation: hold assets that do well in different conditions, so something is always performing.

His allocation:

• 25% stocks: Do well in prosperity and mild inflation
• 25% long-term bonds: Do well in deflation and recessions when rates fall
• 25% gold: Do well in inflation and geopolitical crises
• 25% cash: Do well in deflation and recessions when opportunity arises

In this allocation, no asset class dominates. If stocks crash and fall to 10% of the portfolio, you rebalance by selling cash and buying stocks at depressed prices. If inflation spikes and gold and stocks rally, you sell them and buy bonds. The portfolio mechanically forces you to be a contrarian.

Browne's argument was psychological and practical. Most investors panic-sell at market bottoms. The Permanent Portfolio, with its low volatility and built-in rebalancing discipline, makes panic-selling less likely. An investor who has lived through a 15% drawdown is more likely to stay the course than one experiencing a 40% drawdown.

Historical performance

From 1975 through 2024, the Permanent Portfolio delivered:

• Annualized return: 4.8%
• Annual volatility (standard deviation): 6.8%
• Worst single year: Down 6.3% (1981)
• Best single year: Up 21% (1980)
• Sharpe ratio: 0.41 (return per unit of risk)

These numbers are exceptional from a risk perspective. A typical 60/40 stock-bond portfolio has returned 7–8% annually with 10% volatility (Sharpe ratio ~0.65). The Permanent Portfolio cut volatility in half while only reducing returns by 3% annually.

The trade-off is harsh over long periods. From 1975–2024, a $100,000 investment in the Permanent Portfolio grew to roughly $4.2 million. The same investment in a 70/20/10 portfolio grew to roughly $13 million. The difference is nearly $9 million due to return differential compounding over 50 years.

For a 30-year working career, the impact is similar. A 35-year-old investing $50,000 annually for 30 years in a Permanent Portfolio would have roughly $4.5 million at retirement. The same person in a 70/20/10 portfolio would have roughly $8 million.

Performance in different economic regimes

The Permanent Portfolio was designed to survive four economic regimes:

1. Prosperity with low inflation: Stocks and bonds do well. Gold and cash lag.
2. Prosperity with inflation: Stocks and gold do well. Bonds lag (rates rise).
3. Recession with deflation: Bonds and cash do well (rates fall, safety). Stocks and gold lag.
4. Stagflation: Gold does well. Everything else lags.

Testing this hypothesis:

• 1975–1980 (inflation boom): Permanent Portfolio up 48%. Stocks up 68%. Bonds down. Gold up. The allocation was rescued by gold and stock gains.
• 1982–2000 (disinflationary boom): Permanent Portfolio up 250%. A 70/20/10 portfolio up 400%. Bonds did better than gold expected, and stocks dominated, so the 25% stock allocation was suboptimal.
• 2008 (financial crisis): Permanent Portfolio down 3%. A 70/20/10 portfolio down 24%. The large bond and cash allocations protected.
• 2022 (bear market with inflation): Permanent Portfolio down 9%. A 70/20/10 portfolio down 15%. Gold and cash helped, but stocks' underperformance hurt.

The evidence shows the Permanent Portfolio truly does reduce downside risk. But it also clearly underperforms in the best environments (long equities bulls).

The gold problem

A 25% gold allocation is controversial. Gold has delivered roughly 2–3% annualized returns since 1975, comparable to inflation but below stocks or bonds. Holding 25% of your portfolio in an asset with 2% expected returns is economically suboptimal if you believe the regime is "normal" prosperity.

Browne's argument was that gold protects against inflation and geopolitical crisis. If war, political instability, or hyperinflation strikes, gold's value persists while paper assets collapse. This is a valid tail-risk concern. But it requires believing that tail risks are more likely than simple probability would suggest.

Academic research on gold as a portfolio component shows:

• Inflation hedge: Gold loosely correlates with inflation, but is not a perfect hedge. In some high-inflation periods (1980s), bonds outperformed gold.
• Diversification: Gold correlations to stocks and bonds are low in normal times, but correlations rise during crises (it becomes another risk asset).
• Return: Gold's return above inflation has been minimal. A 2% nominal return during 2% inflation is a 0% real return.

For most investors, a 5% gold allocation (insurance) makes more sense than 25%. This gives you tail-risk protection without dragging returns as much.

Rebalancing discipline in the Permanent Portfolio

The Permanent Portfolio's greatest strength is forced rebalancing discipline. When 25% portfolio components are in fixed positions, rebalancing is very explicit.

Here's what happens in a market crash:

• Before crash: 25/25/25/25 stocks/bonds/gold/cash
• After crash: Stocks fall 40%, bonds rally 10%, gold flat, cash flat. New allocation: 15/30/25/30
• Rebalance action: Sell $200 of bonds and $300 of cash. Buy $500 of stocks at crash prices.

This is exactly the opposite of investor psychology. In crashes, you want to buy stocks, but you have no conviction. The Permanent Portfolio forces you to do it systematically.

For investors prone to panic-selling, this discipline is invaluable. The cost is borne in normal times when disciplined rebalancing forces you to sell winners and buy laggards.

Modified Permanent Portfolios

Some investors modify the 25/25/25/25 allocation to suit their goals and beliefs. Common modifications:

• 40/30/15/15 (stocks/bonds/gold/cash): Reduces drag from gold, keeps bonds as primary risk reduction.
• 40/30/20/10: Adds 5% more stocks and bonds, removes cash, reduces rebalancing frequency.
• 50/30/15/5 (stocks/bonds/commodities/cash): Replaces gold with broader commodities, reduces gold allocation.

These modifications trade some of Browne's philosophy (equal weighting across regimes) for potentially better returns. A 40/30/15/15 portfolio has delivered roughly 6.5% annualized returns with 7.5% volatility—better risk-adjusted returns than 25/25/25/25, but still conservative.

When the Permanent Portfolio makes sense

The Permanent Portfolio is appropriate for:

• Wealthy retirees drawing from savings: The low volatility is valuable when you are living off portfolio returns.
• Risk-averse investors who panic-sell: The forced rebalancing discipline is invaluable.
• Investors who believe tail risks are elevated: If you believe hyperinflation or geopolitical catastrophe is likely, the 25% gold allocation is insurance.
• Investors who want to avoid market-timing: The allocation removes the need to predict which asset class will outperform.

It is not appropriate for:

• Young investors with 40+ years to save: The return drag is severe over long periods.
• Investors with moderate-to-high risk tolerance: A 70/20/10 or 80/15/5 allocation is more suitable.
• Investors who understand and enjoy rebalancing: If you are happy rebalancing a 70/20/10 portfolio, the Permanent Portfolio's additional complexity and constraints add no value.

Permanent Portfolio implementation

A Permanent Portfolio can be implemented with:

1. 25% stocks: VTI or SPLG
2. 25% bonds: BND or SCHZ
3. 25% gold: GLD or IAU (gold ETFs)
4. 25% cash: VMFXX or Treasury Direct

Rebalancing is done annually or when allocations drift more than 5%. Because the four components have low correlation, drifts happen gradually, making rebalancing less frequent than a stock-heavy portfolio.

The philosophical choice

Choosing between a 70/20/10 portfolio and a Permanent Portfolio is ultimately philosophical. The 70/20/10 assumes normal economic conditions continue and you will be disciplined through crises. The Permanent Portfolio assumes crises are common enough that you need to hedge them constantly.

History suggests the 70/20/10 is correct for long-term investors (the average working career has 3–4 major crashes, not monthly tail risks). But psychology suggests some investors need the Permanent Portfolio's forced discipline to survive those crashes without panic-selling.

If you need forced rebalancing discipline, the Permanent Portfolio makes sense. If you can maintain discipline with a higher-return allocation, the Permanent Portfolio costs you too much.

Portfolio construction decision tree

Related concepts

• ./01-what-is-asset-allocation.md
• ./19-the-3-fund-portfolio-as-default.md

Next

Next we'll examine the All-Weather Portfolio, Bridgewater's risk-parity construction that weights assets by volatility instead of dollars.