Investors are hungry for success stories, especially tales that include high returns with low risk. And the investment industry is always happy to stoke that appetite.
One of the most popular stories today is the so-called All Seasons portfolio, whose virtues are trumpeted in the massive bestseller Money: Master the Game, by motivational speaker Tony Robbins. The book has been out since last November, and I thought the hype would blow over quickly, but I'm still getting inquiries about it, so I thought I'd take a closer look.
The All Seasons portfolio was created by Ray Dalio of Bridgewater Associates, one of the largest hedge fund managers in the world. It's based on Dalio's similarly named All Weather fund, which reportedly has more than $80 billion USD in assets. The portfolio has the following asset mix:
40% Long-term bonds
15% Intermediate bonds
In a backtest covering the 30 years from 1984 through 2013, the All Seasons portfolio had an annualized return of 9.7% (net of fees) and only four years with a loss. Its worst year was a modest -4% in 2008. With a risk-return profile like that, it's no wonder so many investors have been attracted to the All Seasons portfolio. In fact, a service run by Robbins' own advisor has been swamped with requests from investors who want a piece of this seemingly miraculous strategy.
So, is the All Seasons portfolio really a recipe for stellar returns with minimal risk? Or is it just another example of investors chasing hypothetical past performance?
The reasons for the seasons
The All Seasons portfolio is based on the idea that asset prices move in response to four forces: rising economic growth, declining economic growth, inflation and deflation. In each of these economic "seasons," some asset classes thrive and others suffer. For example, when growth is strong and inflation is low, stocks are likely to perform well, whereas commodities and gold benefit from rising growth and rising inflation. Bonds do well when economic growth and inflation are both falling. By including all of these asset classes in your portfolio, you'll do well under all conditions. It's like travelling with sunscreen, an umbrella, a swimsuit and a parka.
There's nothing wrong with this general idea: most investors understand that a portfolio should include asset classes with low (or even negative) correlation. Nor is it an original thesis: it's very similar to what Harry Browne wrote in the early 1980s. Browne's Permanent Portfolio was also based on the principle that you should hold asset classes that would thrive during four economic scenarios: stocks for prosperity, cash for recessions, gold for inflation protection, and long-term bonds for deflation. (If you're interested in learning more, read Part 1 and Part 2 of my 2011 interview with Craig Rowland, co-author of The Permanent Portfolio.)
Was the performance really remarkable?
But while the premise of the All Seasons portfolio is reasonable, there's nothing astonishing about its performance during the last 30 years. Moreover, anyone expecting it to deliver 9.7% with low risk in the future is likely to be disappointed.
The returns were unremarkable. A 9.7% annualized return doesn't mean much unless you compare it to the alternatives. The truth is that all diversified portfolios performed well during the last three decades.
Despite the carnage of the dot-com bust at the turn of the millennium and the financial crisis of 2008-09, most of those 30 years were extremely kind to stocks. Late 1987 to the spring of 2000 saw the longest bull market in history, and the one we're enjoying now ranks third all-time. From 1984 through 2013, the S&P 500 returned a whopping 11.1%.
And what about bonds, which make up 55% of the All Seasons portfolio? In the US, long-term government bonds returned 9.4% during the period. In Canada, they did even better: the FTSE TMX Canada Long-Term Bond Index returned 10.3% over those 30 years.
Once you consider the context, a 9.7% annualized return since 1984 isn't remarkable at all. Anyone who stayed invested in a diversified portfolio would have seen similar results.
The risk was not "extremely low." OK, maybe the returns of the All Seasons portfolio were in line with a traditional balanced portfolio, but risk was much lower, right? In an article for Yahoo Finance, Robbins reports that the standard deviation of the portfolio during the 30-year period was 7.63%, which he declares is "extremely low risk and low volatility."
I'm not sure investors would agree with that assessment. If a portfolio has an average expected return of 9.7% and a standard deviation of 7.6%, that means in 19 years out of 20, its annual return can be expected to range between -6% and 25%. That's not "extremely low volatility": it's about the same as that of a traditional balanced portfolio. In our white paper Great Expectations, my colleague Raymond Kerzhéro and I found that a portfolio of 40% bonds and 60% global stocks had a standard deviation of about 7.8% over a similar period (1988 to 2013).
What about the fact that the All Seasons had only four negative years, all with only modest losses? Robbins and Dalio frequently compare the All Seasons portfolio to the S&P 500, which certainly saw much larger and more frequent drawdowns. But this is a totally inappropriate benchmark, as the All Seasons portfolio includes just 30% stocks. Dalio's portfolio holds 55% bonds, which are far less volatile than stocks. More important, bonds only lose value when interest rates rise, and from 1984 to 2013, the yield on 30-year Treasuries fell from over 11% to about 3.5%. Any bond-heavy portfolio would have seen rare and modest drawdowns during that period.
There were many disappointing periods. The long-term returns of almost any diversified portfolio look impressive, but unfortunately you can't buy 30 years of performance in advance: you have to earn those returns by doggedly sticking to your plan even when it disappoints. And let's be clear: the All Seasons portfolio would have tried your patience many times.
While the portfolio never suffered huge losses, it would have significantly lagged a traditional balanced portfolio during the many periods when stocks delivered double-digit returns. That's why this strategy - and the Permanent Portfolio, for that matter - had few followers during the 1980s and 1990s. A portfolio with just 30% stocks would have been met with derision during that long, giddy bull market.
Gold would have been even harder to hold. Sure, it glittered during the most recent financial crisis, but during the 21 years from 1984 through 2004, the real return on gold in Canadian dollars was -2.3% annually. Would you have had the guts to hold it through two money-losing decades?
Don't make the mistake of thinking it's easy to stick with a strategy when it underperforms during strong bull markets, as the All Seasons portfolio is almost certain to do. Bridgewater's own All Weather fund returned -3.9% in 2013, one of the best years for stocks in recent history (the MSCI World Index was up almost 34% in Canadian dollars). My guess is that Dalio's clients took little comfort in the fact that strategy performed well in historical backtesting.
Couldn't stand the weather
My goal here is not to beat up on the All Seasons portfolio specifically: on the contrary, I wanted to show that in many ways it's not fundamentally different from other balanced portfolios.
My concern is that the All Seasons portfolio is presented as a magic formula that will dramatically outperform a traditional stock-and-bond portfolio with far less risk. The very name implies that it will perform well during all market conditions. But that wasn't true over the last 30-plus years, and it's even less likely to be the case during a period of low interest rates. (No one knows where rates are headed, but it's absurd to expect 9% or 10% returns on bonds when yields are 1% to 3%.)
A well-diversified, low-cost portfolio executed with discipline offers your best chance of enjoying market returns with moderate volatility. But there's no secret recipe, no optimal asset allocation, and no reward without risk. Be skeptical of anyone who suggests otherwise.