By James White and Victor Haghani¹
ESTIMATED READING TIME: 3 min.

A number of clients have asked us about a recent Buttonwood column in The Economist titled “The Eccentric Investment Strategy That Beats the Rest.” The strategy in question is sometimes called the “Permanent Portfolio,” and holds 25% each in equities, gold, nominal bonds and cash. Buttonwood describes some of the performance metrics that have made this portfolio look attractive recently – beating more traditional portfolios over the last few years – and also notes that researchers have found it have to have a higher Sharpe Ratio (ratio of excess returns to risk) than standard portfolios over 40-year periods.

Buttonwood elegantly explains the core objections to a portfolio like this:

…[the] investment strategy is unusual because it ignores some basic principles of asset allocation. One is that a portfolio should have big positions only in investments that generate cash-flows, and therefore have an identifiable source of expected returns. Shares do (implicitly, from their underlying earnings) and so do bonds (from their coupons). But gold does not, and historically the returns from cash have barely beaten inflation, which is why conventional portfolios tend not to hold much of either for the long run.

Then there are the weights. Orthodox theory suggests these should be determined by balancing the expected returns from each asset class against its volatility, while accounting for its correlations with the others. The 60/40 strategy is a rule of thumb that comes from such considerations. The 25/25/25/25 strategy does not: it just picks four asset classes and lumps equal parts of your portfolio into each.

We couldn’t have said it better ourselves! And yet…reasoned, abstract objections like these sometimes don’t seem so powerful when set next to a concrete track record that looks pretty good recently and over long periods too. Assuming that an “intuitive” portfolio like this has outperformed in Sharpe ratio terms over the last 40 years (though our casual research found that the improvement in Sharpe ratio is quite modest, and the return was substantially lower than more traditional stock/bond portfolios), shouldn’t that be telling us something?

The answer is: not really. The main ingredient in a metric like historical Sharpe Ratio is the historical return, and you know the saying: past performance is not necessarily indicative of future returns. Even over long periods, past returns do not reliably predict future returns, and so historical metrics which rely heavily on historical returns just can’t tell us much for how we should invest in the future.

Indeed, it will always be the case that we can find portfolios which are obviously and indisputably suboptimal in terms of prospective expected return and risk, but which have done really well historically, even over long periods. For example, it’s easy to find one-to-three stock portfolios which have handily trounced the total market index over 30 years based on any backward-looking metric you care to name, including Sharpe ratio and risk-adjusted return. This pulls against our innate desire for the world to make sense; we want portfolios that have done great to have done so because they should have outperformed.

But these dissonant outcomes are intrinsic to investing in a highly uncertain world. If you bet all your wealth on the flip of a fair coin, that’s clearly not a sensible thing to do – but the flip will still land in your favor half the time! Similarly, the theories of portfolio construction Buttonwood is drawing on guide us to making a sound decision before the coin gets flipped. Actual, experienced outcomes sometimes differing wildly from the forward-looking prescription are fundamental to financial markets being risky and uncertain.

Traditional forward-looking expected return metrics are in general going to be better, more sensible predictors of returns than history will be. To make an exaggerated example: say you buy a ten-year par bond at a 10% yield, and then over the next year, its price goes up enough that it’s now trading at a 1% yield. That 1% yield gives you a much better idea what return to expect in the future than the roughly 90% total return you just got over the past year.

So, the Permanent Portfolio may have a catchy name and an appealingly simple structure, but “permanent” is doing a lot of work in a world where the only constant is change. Financial markets are evolving systems where yesterday’s winning strategy can easily become tomorrow’s cautionary tale. If we’re going to ignore basic principles of asset allocation, we should have a better reason than “it worked for a while.”

1. This is not an offer or solicitation to invest, nor are we tax experts and nothing herein should be construed as tax advice. Past returns are not indicative of future performance.