Figure 4.1 Life Expectancy—Keeps Getting Longer
Source: Internal Revenue Service, Publication 1457 (Rev. 5-2009), Table 2000CM.
In selecting an appropriate benchmark (the core of your retirement investing plan and driving all later investment decisions), your first stop is time horizon.
One of the bigger mistakes investors can make is underestimating time horizon—or having a fuzzy or ill-defined one. It can lead to major errors—ones that may not be evident immediately and maybe not for many years. Those can be the most insidious kinds of errors.
And most of the time, the errors stem not from mistakenly having too long a time horizon, but too short. By not correctly assessing time horizon, investors might, for example, underestimate how much growth is appropriate to reach other goals, which they might not realize for 10 or 20 years. At that point, it could be too late to make a material enough shift to rectify it. Or your time horizon 20 years later might indeed be much shorter, which means a different benchmark is appropriate for you. Either way, the damage may already be done.
In this chapter we’ll cover:
In the investing world, professionals, pundits, investors—most everyone—often talk about risk when they mean volatility. The two are often interchangeable—and should be! But not always. Make no mistake, volatility is a key risk every investor faces and must consider, unless you’re doing true, long-term capital preservation, which likely isn’t the case for most readers of this book.
But it isn’t the only risk! Thinking only or primarily about volatility can result in decisions that decrease the likelihood of reaching your long-term goals. For example, there’s interest rate risk—the risk interest rates fall, and when a bond matures, comparable bonds have lower (maybe much lower) coupon rates. Hence, you reinvest into a lower rate. There’s inflation risk (covered in Chapter 2). There’s political risk, exchange rate risk, liquidity risk—myriad kinds of risk.
In 1997, I wrote a paper on risk with my friend and sometimes research collaborator Meir Statman (the Glenn Klimek Professor of Finance at Santa Clara University’s Leavey School of Business) titled “The Mean-Variance-Optimization Puzzle: Security Portfolios and Food Portfolios,” which was published in the Financial Analysts Journal.
In it, we discuss how people think about food often parallels how investors think about investing. Diners don’t just want nutrition—they want it to look good and taste good. And they want to eat the food at the right time of day—people feel weird eating hamburgers in the morning, and eating waffles at night is seen as quirky. And they want prestige! They don’t want to look like a weirdo when their companion orders waffles at night in a fancy restaurant. And what people want shifts—often fast! What they feel as risk is what they want at a point in time that they think (or fear) they’re not getting. They don’t think about the things they are getting.
How does that relate to investing? You might hear investors say something like, “I don’t want any downside volatility!” They’re feeling volatility and want protection from it. Then, if stocks go on a long, sustained tear, they might feel like they’re missing out—and missing out is felt as another kind of risk. That’s the idea that what someone isn’t getting at that point (downside protection, upside return, etc.) is what they’re most concerned about—never mind if their other objectives are being met.
And a big risk investors often forget about—one that is particularly key when thinking about time horizon—is opportunity cost : the risk of giving up returns you would have otherwise gotten. If you assume a too-short time horizon and select a benchmark that may not be appropriate for you, you may not have enough exposure to equities. And over your actual (and longer) time horizon, you may find the return you get is insufficient to allow you to get the cash flow you need, suitably outpace inflation and/or grow your assets enough to accomplish all of your goals.
Again, volatility risk is a key risk and one that folks tend to feel most keenly, at least in the near term. It can be heart-stopping to watch your equity allocation—whether it’s 100% of your portfolio or just 10%—lose 20% in value fast, as can happen in a correction. And even more grinding to watch it fall 30%, 40% or more in a big bear market. Ultimately, equity investors put up with volatility because finance theory says (and history has supported), long term, you should get rewarded for that volatility—more so than in other, less volatile asset classes.
But focusing solely or primarily on volatility risk—thinking just about the lack of volatility that you want but aren’t getting at a point in time—and ignoring opportunity cost can be very, well, costly. What’s more, making up for lost opportunity can be very difficult. Investors may simply have to swallow a lower projected cash flow, lower expected returns, etc.
And that can be hard to do, even dispiriting, if you’ve been counting on a larger income—particularly if you’re already retired or nearing retirement—more so if your spouse was also counting on that income. (Try explaining that.) Or you could work longer to expand your portfolio size—also hard on you and the spouse. These are all tough realities to deal with—and you’re likelier to avoid them if you don’t underestimate your time horizon.
Understanding what an appropriate time horizon is ties into my Chapter 1 point that people often think about their investing future in chopped up phases. Investors often believe they have one time horizon that ends when they retire or start taking cash flow or hit some other milestone, then they enter a new phase. Many investors think this way: “I’m 55. I plan to retire at 65. That means I have a 10-year time horizon.” In my view, this is harmful thinking that potentially cuts you off at the knees.
Rather, your time horizon is how long you need your assets to work toward your objectives. For most investors, this is their entire lives. If you don’t want to risk the ire of your spouse and prefer he/she remembers you fondly (and not with crass epithets), include his/her lifetime, too—particularly if your spouse is younger and/or healthier.
Figure 4.1 is a good start in thinking about time horizon. It’s basic US life expectancy, produced by IRS actuaries. The x-axis shows current age, and the y-axis shows average life expectancy. So if you’re 30, average life expectancy is another approximately 48 years (i.e., 78 years old). If you’re 70, average life expectancy is another approximately 14 years (i.e., 84).
Figure 4.1 Life Expectancy—Keeps Getting Longer
Source: Internal Revenue Service, Publication 1457 (Rev. 5-2009), Table 2000CM.
And keep in mind, this is average life expectancy. Will you be average? Or above or below? No way to know with certainty. But a good way to estimate is considering the age your grandparents and parents died—as long as they died of natural causes. A car wreck at a young age is a tragedy but likely won’t impact any genetic predisposition to longevity. If your grandparents died in their 80s and your parents are alive in their 90s, odds are good you’ll live at least that long or longer, IRS tables be damned. Then, consider your own health—perhaps high cholesterol runs in your family, but you diet and exercise and keep it under control. Your better health should, theoretically, add to your longevity.
And you want to err a bit on the longer side for a few reasons. First, you don’t want to presume you’ll live until 85, plan for your assets to last until then and no longer and end up living into your 90s (as my parents did—it’s very easy for me to envision long time horizons). Then, too, there are many comforts money can buy as you get older. (And don’t forget the grandparent effect. Spoiling your grandkids is so darn fun—something I wouldn’t have realized until I became a grandparent myself.)
But perhaps more important, average life expectancy has done nothing but increase. A white male born in 1850s America had an average life expectancy of 38.3 years. Ladies had it slightly better—their average life expectancy was 40.5. In 1900, that improved to 48.2 and 51.1.
Along came penicillin (discovered in 1928 and put into heavy rotation thereafter), and American children born in 1950 could expect to live to 66.3 and 72.0.1 Add in the polio vaccine, the MMR vaccine, better screening for cancers and heart disease, new pharmaceuticals and myriad other life-extending miracles, and US life expectancy has increased an average 1.6 years for men and 1.5 for women every decade since 1950. Life expectancies globally are extending, too—in emerging nations faster, but starting from a lower base.
And life expectancies likely only keep increasing. Loss of mobility can be a major factor in shortening life expectancy. Our hearts rely on mobility to help pump blood—if you’re not moving, your heart naturally wants to stop pumping so healthily. Just think of all the innovations in recent years—not just for older folks—but for anyone with a loss of mobility. We’ve seen huge gains in prosthetics. In the 2012 Olympics, a young man with no legs competed in a relay event! He ran on carbon fiber “blades.” As this and other technologies mature, they naturally become more readily available and prices fall (as they always do when technologies mature). And they allow folks who would, not long ago, be immobile, live much fuller, healthier, longer lives.
Innovation breeds future innovation, and there’s no chance humanity wakes up tomorrow and decides, “To hell with profits! And to hell with problem solving!” Never underestimate the power of profit motive in driving life-changing and life-extending innovations. (That, combined with the desire to keep up with the grandkids for a little longer.) If you’re 50 now, for example, and your median life expectancy (according to the IRS) is another nearly 36 years, what are the odds, 30 years later, life expectancy has extended still more? Probably pretty good.
My view is folks tend to underestimate time horizons much more often than overestimate. But is it possible someone could have a shorter time horizon than their life expectancy? Of course! It can happen that an investor may have a set of circumstances making the time horizon shorter. But generally, for most investors, life expectancy can be a good jumping off point to consider time horizon.
Can you have a time horizon much longer than an individual’s or couple’s life expectancy? Certainly. Many institutional investors—like college endowments, charities, pension funds, etc.—have literally an infinite time horizon. They aren’t investing (often) for this life, but for the long life they hope the endowment (charity, pension, etc.) has, which can stretch on for decades or even centuries, in many cases.
For individuals, super-long time horizons like that are unusual. However, many investors say, “This money isn’t for me or my spouse. This is for the kids or grandkids or the Save the Piping Plover Foundation or some combination of the grandkids and plovers.” In that case, the time horizon can indeed get pretty long—longer than your life expectancy.
Some readers may have a hard time thinking about a longer time horizon. Maybe they say, “I get it. I don’t need cash flow now. But I need it in 10 years, and that’s a big change. So shouldn’t I think about the next 10 years one way and the following set of years another?”
Again, for most readers of this book, this is why all investing is retirement investing (usually). Your time horizon shouldn’t be chopped into neat (or, heck, messy) segments. You shouldn’t invest one way for this set of years, then radically change everything once you hit a milestone, just because you hit said milestone.
Whether you started investing today, 5 years ago or 37 years ago. Whether you’re 22 or 35 or 57. Your retirement investing plan should be crafted with an eye to your long-term goals, whether that means you plan on taking cash flow in 5, 10 or 50 years or never.
Now, maybe your milestone necessitates a change—a shift in benchmarks—maybe not. As discussed in Chapter 1, that change (big or small) may need to come 5 years in advance of that milestone, or 7.2 years in advance or 4.6 years after. But those decisions should all be driven not by your retirement party date or the first distribution you take from your portfolio, but by a long-term, forward-looking (always) assessment of what’s appropriate for you to increase the likelihood you achieve your goals the entirety of the period you need your money working for you. If someone says to you, “I retire in five years, and after that, I’ll need to make a big change,” your response should be, “Why are you investing without a long-term plan?”
Why the heck is time horizon so important? Very generally, the longer the time horizon, the more a bigger allocation of equities likely becomes appropriate in your benchmark.
This isn’t to say time horizon is the only deciding factor. No! That would be akin to saying age is the only factor. But it is key. It must be considered alongside and in concert with return expectations, cash flow needs and other factors. But a longer time horizon means investors have more time to grow beyond near-term equity volatility.
A fact that shocks no readers: Over short time periods, equities are generally more volatile than bonds. Figure 4.2 shows standard deviation (a common measure for volatility) and average annual rates of return over five-year rolling periods for a 100% equity allocation, a 70% equity/30% fixed income, a 50/50 and a 100% fixed-income allocation.
Figure 4.2 Five-Year Time Horizon—Volatility
Source: Global Financial Data, Inc., as of 06/22/2012. US 10-Year Government Bond Index, S&P 500 total return, average rate of return from 1926 through 12/31/11.2
* Standard deviation represents the degree of fluctuations in historical returns. This risk measure is applied to five-year annualized rolling returns in the chart.
Equities have a superior return over five-year periods—the 100% equity allocation averages 9.9%—but with much higher standard deviation. And as the share of equities decreases, so does the standard deviation. The 100% fixed-income allocation has the lowest average annual return at 5.5%. But it also has the lowest standard deviation. Less wiggling means lower returns. Pretty straightforward.
But if you look at longer periods, something happens. Figure 4.3 shows standard deviation and average annual returns, but over 20-year rolling periods. Standard deviation for 100% equities is much lower over 20 years than over 5 years. And over 20 years, average standard deviation for 100% equities and 100% fixed income are darn near identical—3.4% versus 3.2%.
Figure 4.3 20-Year Time Horizon—Volatility
Source: Global Financial Data, Inc., as of 06/22/2012. US 10-Year Government Bond Index, S&P 500 total return, average rate of return from 1926 through 12/31/11.3
* Standard deviation represents the degree of fluctuations in historical returns. This risk measure is applied to 20-year annualized rolling returns in the chart.
Figure 4.4 shows 30-year rolling periods. Over 30 years, the average standard deviation for 100% equities is actually lower than for 100% fixed income—1.4% for equities versus 2.8% for fixed income. Lower standard deviation with materially better long-term average annual returns.
Figure 4.4 30-Year Time Horizon—Volatility
Source: Global Financial Data, Inc., as of 06/22/2012. US 10-Year Government Bond Index, S&P 500 total return, average rate of return from 1926 through 12/31/11.4
* Standard deviation represents the degree of fluctuations in historical returns. This risk measure is applied to 30-year annualized rolling returns in the chart.
Day to day, month to month and year to year over those 20- and 30-year periods, equities swung much more wildly than fixed income (as depicted in Figure 4.2). But the data show, over longer periods, the wild wiggles resolve to more consistent upward volatility. (And yes, remember, volatility goes up, too.) Folks often think when stocks rise, that’s good, but when they fall, that’s volatility. But they’re wrong. They’re two sides to the same coin, and you don’t get big returns without volatility—up and down.
Said another way, volatility is normal. Bear markets are normal. They happen. Sometimes they’re big. But if you’re an investor with a 20-year time horizon or more, even a big bear market like 2008’s likely becomes a blip after several decades of equity-like growth. Near-term vicissitudes simply get swamped by the later, longer and stronger overall upsweep of equity markets—history shows that.
Now, if you believe “this time, it’s different,” and in the future, equities aren’t likely to continue delivering superior returns relative to similarly liquid asset classes, I direct you to my 2011 book, Markets Never Forget. Investors nearly always believe “this time, it’s different”—that phraseology pops up repeatedly through history and can be tracked in media. But “this time” isn’t ever all that different as folks fear. Yes, details differ and the scenery changes, but humans don’t evolve fast enough to make our behaviors change. And markets are a reflection of the behavior of billions of people doing trillions of transactions daily.
There have always been and always will be contractions and bear markets. There will always be outsized fear the future is doomed—particularly following recessions and bear markets (sometimes for years after). But the future won’t be doomed—economies will keep rising in fits and starts, and, very likely, equity markets will keep notching new highs, given enough time.
This doesn’t necessarily mean an equity strategy is appropriate for all investors because that will depend on a number of considerations. Rather, this highlights why determining the correct time horizon is critical for your benchmark and your retirement investing plan.
Notes
1. Sources: Department of Health and Human Services, National Center for Health Statistics; National Vital Statistics Reports, vol. 54, no. 19, June 28, 2006.
2. The S&P 500 Total Return Index is based upon GFD calculations of total returns before 1971. These are estimates by GFD to calculate the values of the S&P Composite before 1971 and are not official values. GFD used data from the Cowles Commission and from S&P itself to calculate total returns for the S&P Composite using the S&P Composite Price Index and dividend yields through 1970, official monthly numbers from 1971 to 1987 and official daily data from 1988 on.
3. Ibid.
4. See note 2.