In last month’s post, I drew a distinction between the financial services industry’s definition of risk, and a definition that might be more useful to you as an individual investor. The financial pros tend to equate risk with volatility. I suggested that in addition to volatility, the possibility of outliving one’s money ought to be included as a key component of any investor’s personal definition of risk. This possibility of outliving your money becomes particularly significant when you retire, and begin to draw down on your savings to fund your living expenses. Once you understand your personal risk not merely in the narrow, finance-based sense as volatility but in the broader view as the possibility of bankruptcy, you will be ready to find the best solutions to the problem of risk.
Now, some might think that the distinction I’m drawing here is of no real importance. After all, what investment-related risk would cause you to outlive your money other than stock market volatility—particularly the negative volatility that occurs during bear markets? If not the only danger to your nest egg in retirement, bear markets have to be by far the biggest investment-related danger, right?
Well, if this were true, then we would expect that a retiree’s chances of going bust reached their highest level during the Great Depression. No other time period can match the Depression in terms of stock market volatility. From 1929 to 1932, the Dow Jones index lost an astounding 89 percent of its value! This was by far the worst bear market in history. But the terrifying roller coaster ride didn’t end in 1932. Another major stock market downturn began in 1937, and didn’t come to an end until 1942.
Surely, then, the years stretching from 1929 to 1942 were the most dangerous of times for a retiree. Right?
Let’s do a little experiment and see. Consider a retiree—we’ll call him Paul—who’s saved $100,000, and will need to withdraw $4,500 from this nest egg to fund his living expenses in his first year of retirement. This will give him an initial withdrawal rate of 4.5 percent ($4,500 divided by $100,000 is 4.5 percent). Now, you may be familiar with the 4 percent safe withdrawal rate rule. According to this rule of thumb, a retiree can withdraw 4 percent from his nest egg in the first year of retirement, and then adjust the withdrawals upwards by the rate of inflation, without any risk of going bankrupt over the course of a 30-year retirement. The 4-percent rule is based on an analysis of historical data, which shows that a retiree starting with a 4-percent withdrawal rate could have retired in January of any year, from 1927 through 1986, and his nest egg would have lasted him for the following 30 years. We are giving Paul an initial withdrawal rate slightly above the safe 4 percent rate for a reason. At Paul’s 4.5 percent withdrawal rate, Paul will outlive his nest egg for at least a few of the 30-year retirements stretching from 1927 through 1986. We’re expecting the years that will cause his nest egg to go bust to cluster in the 1930s, when stock market volatility was at its highest level. We will see if that expectation turns out to be true.
Okay, getting back to our experiment, we will assume that Paul invests all of his money in a portfolio of U.S. large cap stocks (as represented by the S&P 500 index, including reinvested dividends). I wouldn’t recommend such a portfolio to a retiree, because without any allocation to bonds it would be too volatile for most people to stomach. But for the purposes of our experiment, we want to maximize Paul’s exposure to stock market volatility. Using historical return data for large cap stocks, along with data on the consumer price index (CPI), I’ve tested whether or not Paul’s $100,000 will survive 30-year retirements beginning in January of 1927, 1928, 1929, and so on, all the way through 1986 (for a 30-year retirement that lasts until December 2015). If the portfolio still has money in it at the end of a particular 30-year stretch, then Paul’s nest egg has survived. If his savings drop to zero before the end of the 30 years, then his portfolio failed.
Here are the results of the experiment. There are a total of sixty rolling 30-year retirements with start years ranging from 1927 through 1986. Paul’s nest egg survived 51 of these 60 retirements, for a success rate of 85 percent. His nest egg went bust in the remaining nine of the sixty retirements. Only two of these nine failures occurred during retirements that stretched over the Depression, when stock market volatility was at its highest. Instead of clustering around the 1930s, the portfolio failures instead clustered around the mid-to-late 1960s and early 1970s. In fact, Paul’s nest egg went bust in each 30-year retirement that began in 1964, 1965, 1966, 1967, 1968 and 1969. His portfolio also failed for the retirement that began in 1973 and ended in 2002.
What was it about the 1960s and early 1970s that made it such a dangerous time to retire? Beginning in 1966, the stock market entered a secular (long-term) bear market that lasted sixteen years. At the same time, inflation rose steadily until it reached double-digit levels by the late 1970s. While annualized large cap returns remained positive over this sixteen-year period, stocks did not keep pace with inflation. As a result, Paul’s living expenses grew faster than his nest egg. For a decade and longer, Paul was withdrawing money from his portfolio faster than he was replenishing it. In the long run, this is not a sustainable situation.
Of course, the 1960s and 1970s saw volatility too. How do I know that it was inflation-lagging returns, and not volatility, that were the primary cause of Paul’s portfolio failures in the 1960s and early 1970s? To eliminate volatility as an issue, I reran the experiment by replacing actual stock returns in each year from 1966 through 1981 with the annualized average of returns over these 16 years. In other words, I replaced the actual S&P returns in each year stretching from 1966 through 1981 with the average S&P returns over 1966-81. I thereby completely eliminated volatility as a risk factor. This second experiment produced results very similar to the first. There was a small reduction in the total number of portfolio failures, from nine to eight. But once again, only two of these failures occurred in the Depression years, while the remaining six failures all occurred during retirements that began from 1965 through 1970. Since there was no volatility in returns from the years 1966 through 1981 in this second version of the experiment, the only remaining explanation for these six portfolio failures is the poor, inflation-lagging stock returns.
While less important than inflation, volatility was not an inconsequential risk to past retirees—as shown by the fact that Paul’s nest egg also went bust in the two retirements beginning in 1929 and 1930. But his nest egg survived all of the retirements that began during the subsequent years of the Depression. His portfolio even survived the retirement that began in January 1931, despite the fact that that retirement took place with 1 ½ years still to go before the end of the 1929-32 bear market—and with the 1937-42 bear market still looming just six years further down the road. Why did the Depression years, and the stock market volatility they brought with them, prove less dangerous to a retiree than we might have expected? Part of the reason is that inflation was low throughout most of the 1930s—and in fact it was negative (deflation) in the early years of the Depression (up until 1933). As a result Paul was able to reduce his withdrawals and still meet his expenses up until 1933, and thereafter he could meet his expenses with only limited increases in his withdrawals to account for inflation.
But the other, crucial reason Paul’s nest egg survived most of the retirements that began in the 1930s has to do with the nature of volatility. Volatility is a risk only when it causes the market to go down. But when volatility increases, the market is characterized by sharp swings upward as well as downward. Yes, the 1930s saw two of the worst bear markets in history: the 1929 and 1937 crashes. But in between these two crashes, from 1932 to 1937, U.S stocks experienced the greatest bull market in history. Thanks to the market’s strong tendency to mean revert (that is, return to its long-term price average), bear markets have always been followed by bull markets. And, typically, the worse the bear, the better the subsequent bull. This is what we saw in the 1930s.
So here is the bottom line for our little experiment. Over the past, the biggest risk to the survival of a retiree’s portfolio was not volatility. Instead it was inflation—specifically, a long (16-year) period of high inflation, during which stock returns failed to keep up with rising consumer prices. Of the nine 30-year retirements that saw Paul’s portfolio go bust, only two of them occurred during the extremely volatile 1930s. (And remember, I stacked this experiment in favor of volatility by giving Paul a 100 percent allocation to stocks.) This is why volatility is too-narrow a definition of risk for retirees. The key risk retiree’s face is the risk of going broke—of outliving one’s savings. Volatility is a component of this broader risk definition—but so is inflation when it exceeds long-run stock returns.
So, why does all this matter? Einstein once said “If I had one hour to save the world, I would spend fifty-five minutes defining the problem, and only five minutes finding the solution.” In investing, and especially for retirees who want to be confident that their savings will last them their lifetimes, risk is the problem to be solved. As Einstein taught us, it is crucially important to understand the problem of risk in all its dimensions before you try to solve it. The financial services industry’s definition, which equates risk to volatility, would not have helped Paul navigate the dangers to his retirement posed by the secular bear market of 1966-81. The book I’m currently writing is focused exclusively on the dangers posed by long, secular bear markets like 1966-81—and on how you can reduce these dangers with the help of certain classes of stocks (namely small caps and value stocks) that have tended to outperform under the conditions characterizing secular bear markets. Stay tuned to this blog for more about secular bear markets, small caps and value stocks—the subjects of my new book. My next post will be on Monday, April 4.