The Golden Butterfly Portfolio has become increasingly popular in the United States over recent years, yet remains largely unknown outside English-speaking investment communities. Created by Tyler of portfoliocharts.com (also an active member of the Bogleheads community under the username Tyler9000), this defensive portfolio promises to combine the stability of the Permanent Portfolio with superior returns through optimised allocation.
But does it actually deliver? In this article, I present the first CHF backtests of the Golden Butterfly, comparing it to the 60/40, the classic Permanent Portfolio, my PP 2.x, my PFD, and the S&P 500. The results are instructive.
What is the Golden Butterfly Portfolio?
The Golden Butterfly is an evolution of Harry Browne's Permanent Portfolio. Tyler designed it to preserve the risk parity philosophy (balancing economic risks) while slightly increasing exposure to periods of prosperity — historically the most frequent economic condition.
Composition: 5 assets, 20% each
Unlike the Permanent Portfolio which uses 4 assets at 25% each, the Golden Butterfly uses 5, allocated as follows:
| Asset | Allocation | Role |
|---|---|---|
| Large Cap Blend Stocks | 20% | Growth (prosperity) |
| Small Cap Value Stocks | 20% | Growth + value premium (prosperity) |
| Long-Term Treasury Bonds | 20% | Deflation protection |
| Short-Term Treasury Bonds | 20% | Liquidity + recession protection |
| Gold | 20% | Inflation + crisis protection |
Technical note: Large Cap Blend vs Total Stock Market
On his site, Tyler specifies Large Cap Blend Stocks for the first line, but most sites referencing his portfolio use VTI (Total Stock Market) instead. Technically, VTI is not quite correct since it already contains small caps, creating partial overlap with the second line. In practice, however, the impact of this difference is negligible due to the capitalisation bias common to almost all ETFs (see my detailed analysis). SPY and VTI are therefore practically interchangeable. I use SPY in my backtests to remain faithful to Tyler's original design.
Philosophy: risk parity with a prosperity tilt
Like the Permanent Portfolio, the Golden Butterfly aims to cover the four possible economic conditions (prosperity, recession, inflation, deflation). The key difference: Tyler adds a second equity allocation (Small Cap Value), increasing equity exposure from 25% to 40%. This prosperity tilt makes statistical sense: historically, economies grow more often than they decline. Tyler also adds exposure to the value factor (small cap value premium), documented by the work of Fama and French (discussed here).
Golden Butterfly vs Permanent Portfolio: what's the difference?
Harry Browne's classic Permanent Portfolio uses 4 assets in equal weight (25% each): stocks, long-term bonds, gold, and cash (or short-term bonds). The Golden Butterfly modifies this allocation:
- 40% Stocks (20% Large Cap + 20% Small Cap Value) → +15 pp
- 20% Long-term bonds → -5 pp
- 20% Gold → -5 pp
- 20% Short-term bonds → -5 pp
The result: the Golden Butterfly is slightly more aggressive (40% stocks vs 25%) and adds value factor exposure. In return, it marginally reduces deflationary (long-term bonds) and inflationary (gold) protections. My Permanent Portfolio 2.x takes this optimisation logic even further, as the backtests below will demonstrate.
Golden Butterfly backtests: the real numbers
I conducted several backtests across different time horizons, in USD and CHF. Here are the complete results.
Backtest 1978-2025 (USD): Golden Butterfly vs S&P 500 — 47 years of data
| Metric | Golden Butterfly | S&P 500 |
|---|---|---|
| CAGR | 9.81% | 12.02% |
| Volatility | 8.32% | 15.08% |
| Max Drawdown | -17.24% | -50.97% |
| Sharpe Ratio | 0.66 | 0.55 |
| Sortino Ratio | 1.01 | 0.81 |
Verdict: Over 47 years, the S&P 500 retains its advantage in raw return (+2.21% annualised). But the Golden Butterfly divides the maximum drawdown by three (-17.24% vs -50.97%), reduces volatility by 45%, and significantly beats in Sharpe and Sortino ratios. For a defensive investor or someone approaching retirement, sacrificing 2.2% of annual return to divide drawdowns by three is an excellent trade-off.
Backtest 1987-2025 (USD): Golden Butterfly vs 60/40 vs S&P 500
| Metric | Golden Butterfly | 60/40 Boglehead | S&P 500 |
|---|---|---|---|
| CAGR | 8.36% | 7.63% | 11.14% |
| Volatility | 7.46% | 9.46% | 15.13% |
| Max Drawdown | -17.24% | -33.59% | -50.97% |
| Sharpe Ratio | 0.71 | 0.50 | 0.58 |
| Sortino Ratio | 1.07 | 0.73 | 0.84 |
Verdict: The Golden Butterfly beats the 60/40 Boglehead on every single metric. Higher return (+0.73% annualised), lower volatility (-21%), maximum drawdown halved (-17.24% vs -33.59%), and significantly superior Sharpe and Sortino ratios. Compared to the S&P 500, the Golden Butterfly sacrifices ~2.8% of annual return but divides maximum drawdown by three and cuts volatility in half.
Backtest 1994-2025 (CHF): Golden Butterfly vs S&P 500
| Metric | Golden Butterfly CHF | S&P 500 CHF |
|---|---|---|
| CAGR | 6.33% | 9.11% |
| Volatility | 10.48% | 17.17% |
| Max Drawdown | -27.93% | -65.84% |
| Sharpe Ratio | 0.44 | 0.48 |
| Sortino Ratio | 0.60 | 0.64 |
Verdict: In CHF, the S&P 500 retains its return advantage (+2.78% annualised). But the Golden Butterfly divides the maximum drawdown by 2.4 (-27.93% vs -65.84%) and reduces volatility by 39%. Sharpe/Sortino ratios are comparable, with a slight edge to the S&P 500. Note that the CHF appreciation against the USD structurally penalises USD-denominated assets over this period.
Backtest 2007-2026 (CHF): Golden Butterfly vs Permanent Portfolio
| Metric | Golden Butterfly CHF | Permanent Portfolio CHF |
|---|---|---|
| CAGR | 5.21% | 4.49% |
| Volatility | 9.49% | 8.90% |
| Max Drawdown | -27.76% | -25.92% |
| Sharpe Ratio | 0.47 | 0.41 |
| Sortino Ratio | 0.64 | 0.61 |
Verdict: The Golden Butterfly beats the Permanent Portfolio in return (+0.72% annualised), Sharpe and Sortino. The classic PP retains a slight advantage in pure stability (volatility -0.59%, MDD -1.84%), but the Golden Butterfly offers a better performance/stability trade-off.
Backtest 2006-2026 (CHF): Golden Butterfly vs PFD
| Metric | Golden Butterfly CHF | PFD |
|---|---|---|
| CAGR | 5.35% | 12.98% |
| Volatility | 9.31% | 6.78% |
| Max Drawdown | -28.02% | -14.69% |
| Sharpe Ratio | 0.48 | 1.67 |
| Sortino Ratio | 0.65 | 2.53 |
Verdict: Unsurprisingly, the PFD (quantitative value portfolio) dominates the Golden Butterfly on every metric. The PFD outperforms by +7.63% annualised, with lower volatility and a smaller maximum drawdown. Sharpe/Sortino ratios are 3-4× higher. This comparison is apples to oranges: the Golden Butterfly is a passive defensive lazy portfolio, while the PFD is an optimised active quantitative portfolio. The point is simply that a well-constructed quantitative portfolio can beat lazy portfolios in both return and stability simultaneously.
Backtest 2011-2026 (CHF): Golden Butterfly vs PP 2.x
| Metric | Golden Butterfly CHF | PP 2.x CHF |
|---|---|---|
| CAGR | 7.05% | 10.65% |
| Volatility | 8.64% | 8.69% |
| Max Drawdown | -18.95% | -21.61% |
| Sharpe Ratio | 0.78 | 1.13 |
| Sortino Ratio | 1.11 | 1.57 |
Verdict: The PP 2.x beats the Golden Butterfly by +3.6% annualised, with superior Sharpe/Sortino ratios. The Golden Butterfly retains a slight stability advantage (MDD -18.95% vs -21.61%, near-identical volatility). The PP 2.x represents the ultimate optimisation of the defensive lazy portfolio, beating the Golden Butterfly in return while maintaining excellent stability.
Implementing the Golden Butterfly in Europe and Switzerland
To replicate the Golden Butterfly in Europe with accessible ETFs, here are the proxies I recommend:
| Original asset (USA) | Recommended ETF | Ticker |
|---|---|---|
| Large Cap Blend (20%) | iShares Core S&P 500 | CSPX |
| Small Cap Value (20%) | SPDR MSCI USA Small Cap Value | ZPRV |
| Long-Term Bonds (20%) | iShares USD Treasury 20+yr | DTLA |
| Short-Term Bonds (20%) | iShares USD Treasury 1-3yr | IBTA |
| Gold (20%) | Invesco Physical Gold | SGLD |
Recommended brokers
- General (France + Switzerland): Interactive Brokers (low fees, broad ETF selection)
- Switzerland only: Saxo Bank Switzerland or Cornertrader
Tax considerations
The Golden Butterfly is a buy & hold portfolio with annual rebalancing. The tax impact is therefore minimal:
- Switzerland: No capital gains tax (except professional traders). Annual rebalancing generates no tax event.
- France: 30% PFU on capital gains realised during rebalancing. To minimise the impact, prioritise using new contributions to rebalance rather than selling existing positions.
Who is the Golden Butterfly for?
Ideal profile
- Defensive investor prioritising stability over raw return
- Retiree or pre-retiree unable to tolerate large drawdowns
- Minimum 10+ year horizon (as with all lazy portfolios)
- Seeking simplicity: 5 ETFs, annual rebalancing only
Unsuitable profile
- Young investor (<35) and/or 30+ year horizon: too defensive an allocation; 70-100% equities will outperform over that timeframe
- Seeking maximum return: S&P 500 or quantitative value portfolio beats GB in raw return
- Seeking ultimate optimisation: PP 2.x beats GB in return (+3.6%) with comparable stability
Where does the Golden Butterfly sit among lazy portfolios?
- GB vs Permanent Portfolio: GB wins on return (+0.7%), comparable ratios and stability
- GB vs PP 2.x: PP 2.x beats GB (+3.6% return), superior ratios, comparable stability
- GB vs 60/40 Boglehead: GB beats 60/40 on every metric — return, stability, ratios
- GB vs All Weather: GB beats All Weather on every metric (see my All Weather analysis)
- GB vs PFD: PFD (quantitative value) dominates GB, but requires active management. Different categories.
Practical implementation
Initial setup
- Open an account with Interactive Brokers (or Saxo Bank Switzerland)
- Buy the 5 ETFs in equal weight (20% each)
Rebalancing
- Frequency: Annual
- Time required: ~1 hour per year
- Method: Sell overweighted positions, buy underweighted positions to return to 20% each
- Tax-efficient alternative: Use new contributions to rebalance (avoids taxable sales)
The importance of Small Cap Value: Fama-French factors
The Golden Butterfly distinguishes itself from the classic Permanent Portfolio through the addition of Small Cap Value (20% of the portfolio). This is not an arbitrary choice: it targets two risk premia documented by Eugene Fama and Kenneth French — the size premium (small caps outperform large caps over the long term) and the value premium (undervalued companies outperform growth companies). The Small Cap Value allocation combines both, theoretically delivering superior long-term returns compared to a Total Stock Market fund.
Conclusion: should you adopt the Golden Butterfly?
The Golden Butterfly Portfolio is an excellent defensive lazy portfolio, offering a remarkable compromise between stability and return. My CHF backtests confirm that it beats the classic Permanent Portfolio and outperforms the 60/40 Boglehead (as well as the All Weather) on every metric.
However, superior alternatives exist for different investor profiles:
- If you seek ultimate optimisation: My PP 2.x beats GB by +3.6% annualised with comparable stability
- If you tolerate active management: A quantitative value portfolio like my PFD dominates all lazy portfolios (CAGR 12.98% since 2006)
- If you are young (<35): 70 to 100% equities will beat GB over a 30+ year horizon
The Golden Butterfly remains relevant for: retirees or pre-retirees seeking maximum stability, defensive investors unwilling to accept -30% drawdowns, and anyone seeking absolute simplicity (5 ETFs, 1 hour per year of management).
Frequently asked questions
VTI or SPY for the Large Cap Blend component?
Tyler recommends Large Cap Blend (SPY — S&P 500), because VTI (Total Stock Market) already contains small caps and theoretically creates overlap with the Small Cap Value line. In practice, the impact is negligible due to the capitalisation bias common to ETFs. SPY is more theoretically consistent; VTI works equally well.
Why 20% in gold in the Golden Butterfly?
Gold plays two roles: inflation protection and a hedge during major crises. Historically, gold is decorrelated from stocks and bonds, which improves the portfolio's overall risk/return ratios. The 20% allocation is slightly lower than in the classic Permanent Portfolio (25%), reflecting the Golden Butterfly's greater prosperity tilt.
Is the Golden Butterfly accessible for beginner investors?
Yes — it is one of the most accessible lazy portfolios: 5 ETFs only, trivial annual rebalancing (~1 hour per year), no market timing required. Ideal for investors seeking simplicity and stability without active management.
How does the Golden Butterfly perform in CHF for Swiss and European investors?
The CHF backtests (1994-2025) show a CAGR of 6.33% for the Golden Butterfly versus 9.11% for the S&P 500 in CHF. The gap is explained partly by CHF appreciation against the USD, which structurally penalises USD-denominated assets. The Golden Butterfly's key advantage in CHF remains its risk profile: maximum drawdown of -27.93% versus -65.84% for the S&P 500. For Swiss investors seeking stability, this difference is decisive.
Golden Butterfly vs 60/40: which to choose?
The Golden Butterfly beats the 60/40 Boglehead on every metric: superior return (+0.73% annualised over 1987-2025), lower volatility (-21%), drawdown halved (-17.24% vs -33.59%), significantly superior Sharpe/Sortino ratios. There is no rational argument for choosing the 60/40 over the Golden Butterfly for a defensive investor.
Sources and data
Backtests conducted with: PortfolioVisualizer (USD data) and Portfolio123 (CHF data)
ETFs used for CHF backtests: SPY (Large Cap), VBR (Small Cap Value), TLT (LT Bonds), SHY (ST Bonds), GLD (Gold)
- Portfolio Charts — Golden Butterfly (Tyler, creator of the portfolio)
- The Theory Behind the Golden Butterfly (Tyler)
- Les Déterminants de la Richesse — my book on quantitative value investing
Rebalancing: Annual for all backtests. Fees: Not parameterised (negligible impact on lazy portfolios). Analysis period: January 1978 — January 2026 (USD), December 1994 — January 2026 (CHF)
En savoir plus sur dividendes
Subscribe to get the latest posts sent to your email.