No pain, no gain.
That’s an important lesson clients have to understand when building their portfolios. To maximize their potential returns, they need to be willing to stomach some risk and experience losses, at least part of the time.
But just how frequent and how deep can those losses be, before the risks no longer warrant the rewards those investments could deliver? Analyzing mutual funds over the past 20 years, shows there may be a sweet spot for your clients.
In the “Mutual Fund Universe” chart below, I categorized every mutual fund (including every type of equity fund and every type of fixed-income fund) that survived over the past 20 years by how often it experienced a negative return in a calendar year. There were 2,482 funds overall, and this study only included the mutual fund with the oldest share class to avoid double- and triple-counting funds that have multiple share classes.
Among that group, 273 funds (11% of the total) never had a calendar-year loss over the past 20 years. As you would probably expect, this group was dominated by money-market mutual funds and short-term bond funds. Their average 20-year standard deviation of annual returns was just 1.97% — a very low level of volatility, indeed.
However, low risk tends to deliver lower rewards. After all, the average 20-year annualized return for a lump-sum investment on January 1, 1998 was a meager 1.86% per year. This means a $10,000 investment made in 1998 would total only $14,563, on average, after 20 years.
The message is quite clear: if your client wants to avoid negative calendar-year returns, he or she needs to also be OK with a bare minimum level of returns.
The next group of mutual funds experienced a negative calendar year only once in the 20 years — a 5% frequency of loss. There were 100 funds in this group, nearly all of which were bond funds. Interestingly, there were a couple of funds with equity exposure (Vanguard Life Strategy Income and T. Rowe Price Capital Appreciation). The average 20-year standard deviation for the 100 funds was 3.55% and the average size of the one negative calendar-year return was -3.43%. The average 20-year annualized lump-sum return was 3.9%. However, as might be expected because of their equity exposure, the Vanguard Life Strategy Income fund had a 20-year average annualized return of 5.17%, while the T. Rowe Price Capital Appreciation fund came in at 9.99%. Clearly, those two funds were the outliers in a group of primarily fixed-income funds.
The next group of mutual funds (153 of them) had a calendar-year loss 10% of the time (that is, two years out of the 20). As expected, the standard deviation for this group was higher at 5.38%, as was the average calendar-year loss of -4.88%. The reward for exposure to a 10% frequency of losses was an average 20-year average annualized return of 4.86%.
As you further examine the data in the table, it becomes clear that a higher frequency of losses is positively correlated with a higher standard deviation of returns and higher average size of annual losses (when they occur). For example, among the 88 funds that had calendar-year losses 40% of the time, the average standard deviation of returns was 28.06%, and the average calendar-year loss was -15.87% — both of which were considerably higher than funds which had losses 20% of the time.
-
If history is a guide, your clients can stop freaking out about having enough money during their post-work life.
July 19 -
Some may be afraid to deploy cash into stocks now because they think the market will tank sooner rather than later.
May 31 -
Booms and busts both have the potential to create unrealistic outlooks on both ends of the performance spectrum.
March 29
That said, it’s also important to notice that there was not a material performance reward provided by the funds that had losses 40% of the time versus the funds with losses 20% of the time. The 88 funds with losses 40% of the time had an average 20-year annualized return of just 6.82%, which was only 17 bps higher than the 6.65% average annualized return for the 505 funds that had calendar-year losses 20% of the time.
Clearly, there is a point at which the frequency of losses and the size of the annual losses begin to diminish the performance potential of a mutual fund or ETF.
As highlighted in the table, this transition point appears to coincide with a 25% loss frequency. In other words, the 452 mutual funds that experienced losses 25% of the time had the highest average annualized return. When the frequency of annual losses exceeded 25%, there was a consistent decrease in 20-year performance.
What does all this suggest?
First observation: To maximize potential returns, your clients have to be willing to experience losses some of the time — but not more often that 25% of the time on an annual basis. We’re not talking about daily or monthly returns here — that sort of behavior begins to look like day trading.
Second observation: Loss frequency needs to be measured over a sufficiently long period of time — in this case, 20 years — to reasonably capture the general behavior of the fund.
Third observation: Funds in the loss-frequency range of 25% to 40% have calendar-year losses that are surprisingly similar in size, on average. Only when the loss frequency escalates to 45% or 50% does the average size of the negative calendar-year return begin to spike upward. This suggests that some of the annual losses among funds that have annual losses 25% to 40% of the time are relatively small — but losses nevertheless. The takeaway from this is to not overreact to small calendar-year losses. It’s the large, frequent losses that we are really trying to avoid when building portfolios for clients.
This is a vital point. The primary job of a well-diversified portfolio is to deliver the appropriate level of return needed by the investor, but to do so in a way that minimizes both the frequency and magnitude of annual losses. Why? Because clients observe and often react badly to investment losses.
A typical measurement of risk is standard deviation of return — but that is not an intuitive statistic for most people. Very few clients are inclined to calculate the standard deviation of their portfolio, but they can probably recite from memory the last time their portfolio had a negative annual return and also remember the size of the loss.
The point is that the common measure of risk — standard deviation — is likely not how actual investors perceive and internalize risk in their portfolios.
Risk-return spectrum
Shown below in “From Less to More Diversified” are three investment approaches. The first is a 100% investment in the S&P 500 Index (via VFINX). This is an all-equity approach.
Next is Vanguard STAR, a fund-of-funds that comprises 11 other Vanguard funds to achieve a 60% equity/40% fixed-income asset allocation.
Finally, there is a broadly-diversified approach that incorporates 12 Vanguard ETFs and achieves a 65% allocation in equities and diversifies the balance in various fixed-income funds. This approach, which I developed, is known as the 7Twelve Portfolio.
Over the past 20 years, among these three approaches, the most diversified portfolio had the best performance. Interestingly, the 7Twelve model and Vanguard 500 Index both had loss frequencies of 20% (meaning four calendar-year negative returns over the past 20 years). However, the key differentiator was that the 7Twelve model’s average negative annual return was -8.6%, whereas the Vanguard 500 Index had an average calendar-year loss of -20.1%, while Vanguard STAR’s average loss was -11.7%.
Building a portfolio that does not experience annual losses more than 25% of the time is important, but it’s also crucial to keep the losses small when they do occur. That is one of the key reasons to build a broadly-diversified portfolio. It won’t make losses go away, but it will likely reduce the size of the losses when they do occur.
