If you’ve read this blog for any length of time it’ll come as no surprise to hear that I’m rather disillusioned with the prevailing wisdom around asset allocation. It goes something like this:
Adjust a 60/40 split based on your age and risk tolerance and close your eyes for 40 years or so and the world will probably be a better place when you finally open them again. Here are some charts showing “long term” equity returns to make you feel better about enduring 50% drawdowns.
Here are my key issues with the prevailing wisdom:
- Prevailing wisdom is biased heavily toward equities based on historical experience. This is the market Skinner box in action. There are no physical laws requiring future equity returns to look like past equity returns over any particular length of time.
- Equity risk drives outcomes within most portfolios, despite these portfolios appearing more diverse when visualized in a pie chart.
- The notion of “the long term” is at best squishy. “Long term,” we’re all dead. No one’s investment time horizon is infinite. We ignore sequence risk at our peril.
- Prevailing wisdom is robust to neither inflationary nor deflationary busts. Which are really the conditions that ought to keep us up at night. Particularly inflationary busts. Because other than a few cranks no one is prepared to invest in a highly inflationary environment these days. (What? You think the inflation of the early 1970s or 1980s can’t happen again? LOL. Just Google MMT )
So this is going to be a post about the permanent portfolio, which is where I’ve landed as an alternative to the prevailing wisdom. This post should absolutely not be taken as investment advice. There are opportunity costs involved here and they may be significant. Particularly if your inclination would otherwise be to allocate 60/40 to US stocks/bonds and US equity continues its run of strong returns for an extended period of time. See my disclaimer for more on why making investment decisions based on random blog posts is an incredibly stupid thing to do.
Fundamental Assumptions & Principles
There are some key assumptions underlying my views on all this. I want to lay them out explicitly up front, because many of these can be debated endlessly. I’m not trying to argue all of this is capital-t Truth. This is simply the framework I’m operating within.* So for example, if you’re a guy or gal who wants to own five stocks forever, I’m not trying to convince you to do it differently. And you’re probably not going to agree with any of this. That’s fine.
- We should build portfolios as regret minimizers and not utility maximizers. Note that regret minimization is subjective. We can regret both realized losses AND foregone gains. The exact “regret function” will vary with each individual. The first sentence of this bullet is in bold because it’s the foundation for everything else. I’d hypothesize that human beings in general tend more toward regret minimization than utility maximization. But I can’t prove that.
- The starting point for any portfolio should be wealth allocation. A wealth allocation consists of at most three buckets. It is possible and sometimes even desirable to have fewer. Wealth allocation is consistent with regret minimization.
- Bucket #1: Protect Lifestyle (cash, annuities, etc.)
- Bucket #2: Maintain Lifestyle/Purchasing Power (traditional MPT portfolio)
- Bucket #3: Enhance Lifestyle (business ownership, concentrated single stock positions, etc.)
- This post is primarily concerned with Bucket #2.
- Equity ownership is absolutely essential for preserving and growing purchasing power over long time periods. However, equities can go through substantial and lengthy drawdowns. Major drawdowns are problematic in a number of ways:
- They create sequence of returns risk for the portfolio (e.g. massive drawdown immediately prior to retirement)
- They may encourage poor investor behavior (buying high and selling low)
- The portfolio will have the least liquidity and buying power when expected returns are highest (e.g. at the trough of a major drawdown), preventing opportunistic purchases of assets subject to forced selling, etc.
- Reliably forecasting economic cycles for the purpose of tactical asset allocation is impossible.
- Traditional methods of hedging tail risk are frequently expensive and can be a significant drag on returns if utilized in meaningful size. They can also be extremely challenging, if not impossible, for individual investors to implement.
- Ideally what we want is a core allocation capable of delivering approximately 5% real returns while minimizing drawdowns across different market regimes. The regimes that are of particular concern are:
- Inflationary Booms
- Deflationary Busts
- Inflationary Busts
- The goal of this exercise is not to build an Armageddon-proof portfolio. In the case of extreme tail events (nuclear war, zombie apocalypse, socialist revolution) your portfolio is going to be the last thing you’re worried about. And anyway, what you’ll really need in those situations are food, medicine and bullets.
The Permanent Portfolio
I think the permanent portfolio offers a solution. The investment analyst Harry Browne devised it specifically for robust performance across a range of different economic conditions. In its original form the permanent portfolio consisted of:
- 25% US Stocks
- 25% Long-Term Treasuries
- 25% Cash
- 25% Gold
The underlying intuition is a model of parsimony. This is a combination of assets where “something should always be working,” regardless of the macroeconomic environment. Long-term Treasuries and gold are less correlated and often negatively correlated with equities. Long-term Treasuries do well in deflationary busts. Gold does well in periods of high inflation.
I do have some quibbles with the permanent portfolio in its original form:
- It holds too much cash.
- It is under-allocated to equities.
- It is strongly biased toward the US.
But perhaps we can address these issues through portfolio construction.
Analyzing The Permanent Portfolio
I used Portfolio Visualizer to run some analysis using historical data. I compared two different permanent portfolio implementations with a 60/40 allocation to US Stocks/US Treasuries. I set the portfolios to rebalance any time an asset class reached +/- 10% of its target weight.
I’ll walk through a couple observations in this post but if you’d like to explore the analysis yourself here is the link to exactly what I ran. (aside: I can’t recommend Portfolio Visualizer enough as a free analytical tool) Below are my three portfolios.
I was able to backtest these allocations back to 1986 with the available data. Unfortunately, the periods where the permanent portfolio really shines versus 60/40–the stagflation of the 1970s and the high inflation and interest rates of the early 1980s–lie outside this time period. If you drop the international equities you can capture the late 1970s and early 1980s, however.
Starting in 1986, Portfolio 1 rebalanced 7 times. Portfolio 2 rebalanced 9 times. Below is a growth chart comparison.
This 30-year period has been truly extraordinary for US stocks and long-dated Treasuries. It comes as no surprise that the permanent portfolios have lagged a 60/40 allocation. And, of course, the permanent portfolio with ex-US equity exposure lagged even more.
But you’re still getting your 5% annualized real returns, with milder drawdowns than the 60/40 portfolio. And in the bargain, you’re better protected from an inflationary regime than you would be with a 60/40 portfolio.
Obviously, if you don’t believe in allocating to ex-US equity you will prefer Portfolio 2 over Portfolio 1. I don’t share that belief, personally. But I certainly can’t prove US equity returns won’t continue to dominate going forward.
We can also look at returns over rolling periods, which paint a similar picture.
If you’re willing to use volatility as a quantitative proxy for risk, you can see the permanent portfolios are significantly more diversified in their sources of risk and return than the 60/40. Equity risk dominates the 60/40 allocation. Imagine the extent to which it dominates in a 70/30 or 80/20 split.
The tradeoff here is simple: give up some upside for a more attractive risk profile.
But what if you could juice the returns a bit?
Because this strategy is robust across market regimes it should also be fairly amenable to leverage. In an ideal world I’d take the more diversified flavor (Portfolio 1) and lever it something like 1.25x to 1.50x. This is the intuition behind risk parity: take a well-diversified portfolio with the risk exposures you want, then lever them to reach your target return. So instead of being limited to 33%/33%/33% you would be allocated maybe 45%/45%/45% for 135% notional exposure.
Unfortunately, as an individual investor it’s not straightforward to lever a portfolio. So, there are some implementation issues to work around. The simplest solution appears to be to use mutual funds or ETFs that apply leverage via either equity or Treasury futures.** This essentially allows you to “bolt on” gold and/or other alternative strategies without having to cut back on your equity or fixed income exposure. I may do a follow-up on this analysis exploring this form of implementation in more detail.
The Permanent Portfolio In A Barbell Portfolio
There are a couple less obvious, ancillary benefits to the permanent portfolio structure I want to mention in closing.
First, because of the attractive drawdown characteristics, it may obviate the need for large cash allocations (e.g. “emergency funds” or “cash buckets” for individuals). There are significant opportunity costs associated with large cash allocations, particularly in real terms.
Second, in keeping with the above, the permanent portfolio provides an excellent stable core around which to build a satellite portfolio of opportunistic investments. For example, at the wealth allocation level you could implement a structure where 70% of the portfolio is permanent portfolio, and the remaining 30% of capital is allocated to private market investments, or high risk/high return single hedge fund investments, or concentrated single stock positions. From a wealth allocation perspective you would be looking at something like 0% Protect Lifestyle / 70% Maintain Lifestyle / 30% Enhance Lifestyle.
In the above configuration, you would also likely be able to use the permanent portfolio as a source of liquidity during major market dislocations, to fund opportunistic investments at precisely the times when expected future returns are highest.
In my view, this strikes a nice balance between staying rich, maybe getting a bit richer but without dying trying.
Essentially, what you’re doing here is building a barbell portfolio. You’re using the permanent portfolio to set a floor for the value of the overall portfolio. You’re then taking the “excess” capital and buying call options with it.
There are other ways for individuals to implement a barbell portfolio structure. You could just use cash to create the floor. Except that’s an extremely inefficient use of capital, in my view. You could also use an annuity.
My quibbles with the annuity approach:
- A fixed annuity with a modest inflation escalator leaves you vulnerable to inflationary booms and busts.
- I am deeply suspicious of variable annuities–in fact, any insurance product with bells and whistles designed to “protect” you from various risks. The pricing of the bells and whistles is usually opaque and therefore not a good deal for the buyer. It is a timeless truth of economics that opaque pricing always and everywhere obscures profitability (see: healthcare; college).
- No matter what route you go the insurance company will extract its pound of flesh.
- You give up the ability to opportunistically redeploy capital from the annuitized core of your portfolio.
That said, I think buying a simple fixed annuity with an inflation escalator is a straightforward option for individuals who want to implement a barbell portfolio, and who are unable or unwilling to go the permanent portfolio route.
*I’d encourage everyone reading this to spend some time writing out your investment framework as explicitly as you can. This is your Investing Code. Once you’ve written down your Code, compare it to your actual portfolio and see if they match. The results of this exercise may surprise you. Incidentally, my friend Rusty Guinn wrote a phenomenal series of articles on Investing Codes and portfolio construction, called Things That Matter/Don’t Matter. I can’t recommend it enough. It touches on the issues discussed in this post as well as many, many more.
** Thanks to @choffstein and @EconomPic for their help conceptualizing this via Twitter.