*This Risk Parity-esque portfolio compares favorably to the Classic 60/40 in terms of both return (slightly higher expected returns) and in terms of volatility (29% reduction in standard deviation). A great portfolio for those who prize stability above all.*

This is a modern classic in the Risk Parity-influenced portfolio tracking community (it’s not a niche… it’s huge, I swear!): The Golden Butterfly created by Tyler of the oft-cited Portfolio Charts website. The portfolio itself is a bit of a tweak on a typical 50/50 portfolio, mixed with some of the insights of another portfolio, the Permanent Portfolio by Harry Browne (25% each in stocks, bonds, gold, and cash). As Tyler writes, the Golden Butterfly is set up to be prepared for the four possible economic states of prosperity, recession, inflation, and deflation, but is then slightly tilted towards prosperity with its extra allocation to small cap value stocks. The allocation to gold, which at 20% seems to be above the consensus opinion on the asset (which tends to run from 0% to about 10% at most), is the trick up its sleeve as it provides the portfolio with an non-correlated asset that stabilizes the entire portfolio. Here is an overview of the Golden Butterfly and then here is a thoughtful piece on the theory behind it, both written by the portfolio’s author.

Whereas the 60/40 is just a benchmark for research purposes, the Golden Butterfly is one I could see as the centerpiece in my decumulation strategy. Like many investors, my first reaction to the 20% gold allocation was that that was high, but then again, those are the assumptions I want to test. Maybe we’ll hit a patch of years where gold shines (yes, that pun appears every time hacks like me write about investing in gold!) and we can see what that level of investment in what is usually a fairly useless rock does. I also want to see how the Golden Butterfly does with an apples-to-apples comparison with other portfolios given the same rules for withdrawals and rebalancing. Frank Vasquez tracks the Golden Butterfly on the Risk Parity Radio site, and of course, you find lots about it on Portfolio Charts. I add to the mix by investigating it in relation to other portfolios given my common conditions, as well as by applying the same rules for withdrawals as the other portfolios.

I anticipate that the Golden Butterfly portfolio will be a stable engine of growth. It stacks up very well compared to the other nine. It is the safest portfolio on the whole, scoring best in five measures of risk and volatility. It has the highest Sharpe ratio, which measures the amount of expected return per unit of risk. Meanwhile, it ranks second for safe withdrawal rate and third for perpetual withdrawal rate, highlighted by its Ulcer Index of just 2.6, or one-quarter that of the 60/40.

In the tests from here on out, I would expect its return to be in the middle of the pack, but with a lot fewer peaks and valleys than some of the others. The gold in the portfolio is a bit like the lake effect in micro-climates: it will make the hottest periods a little cooler, and the cold periods a little warmer. Like the Classic 60/40 portfolio, this one is easy to track, and manage. Since all five assets should be equal, you can avoid the spreadsheet tricks and formulas that are relied on for tracking the others. Another great feature of this portfolio is the choice to include Small Cap Value as its own asset class. This is a clear nod to the insights of factor investing, that small cap stocks and value stocks both represent independent risk factors in investing, which in turn leads to higher rates of return to compensate investors for that risk (see this helpful video for more on factor investing).

On the flip side, I think growth in the portfolio will be fairly restrained in most economic environments. Short-term Treasuries are no great shakes as far as returns go, and then there’s gold, which attracts many doubters. This, though, is a perfect example of the need to think about the whole portfolio rather than get caught up on the individual ingredients. Baking powder doesn’t taste great on its own, but performs a role when baking a cake, and I’d expect gold to have a similar function here.

Here is the **Correlation Matrix** (*Data from 2004; substitutions made to lengthen available data; credit to **Portfolio Visualizer**):*

And, the **Backtest Analysis*** **(Data from 1970; credit to **Portfolio Charts**):*