“It ain’t what you don’t know that gets you into trouble. It’s what you know for sure that just ain’t so.” —Mark Twain
Do you remember the rule of thumb that says that the proportion of stock in your portfolio should equal 100 minus your age? Well, it turns out that this is very likely to be exactly the wrong advice if you want to maximize your retirement income security.
The 100-minus-your-age rule suggests that, although it’s okay own a large percentage of stock in your portfolio when you’re young – 75% at age 25, for example – you’ll want to own less and less of this asset class as you age, e.g., only 20% of your portfolio by the time you’re 80 years old. The reason for this assumption is that stocks are ‘risky’ (i.e., volatile) and that volatility is a bad thing as you begin drawing on your portfolio for income in retirement.
If volatility were the only element in the equation, this would be straightforward. Stocks, however, aren’t simply the more volatile asset class in your portfolio; they are the part of your portfolio that is going togrow over time. And since virtually all retirement portfolio strategies rely on investment returns that include growth from stock, research shows that systematically reducing a portfolio’s stock allocation throughout retirement also systematically reduces its expected return. Unfortunately, that in turn increases the likelihood that the portfolio will fail.
Let me repeat that last part: Retirement portfolios that decrease their allocations to stock over time arestatistically more likely to run out of money than portfolios with stable or increasing stock allocations.
In order, therefore, to conservatively reduce the possibility of running out of money altogether, we need to find a realistic way not only to limit the potential damage that stocks’ volatility can do to a portfolio, but also more effectively capture their long-term potential for growth.
Risk and Return in the Real World
Anybody who has ever worked in retirement planning knows that the culprit we’re all trying to tame in order to manage retirement income is “sequence of return risk.” What this means is that, although a multi-year return might average 7%, actual year-to-year results are likely to be highly volatile, and especially so if the portfolio holds a lot of stock. Over a five-year period, for example, a portfolio might be up 12% one year, down 3% the following year, down another 7%, then back up 15%, and finally down 2%. That’s an ‘average’ of 7% – but the volatility of this sequence of returns creates special planning challenges.
In fact, most retirement income studies focus on how to cope with scenarios in which there is a particularly long and bad sequence of returns during the beginning of a retirement period. This pattern is particularly problematic because portfolio withdrawals at the beginning of retirement that also happen to coincide with a bear market – at a point when the portfolio still needs to last a very long time – risk such significant capital depletion as to put the entire portfolio as risk. Then, when the market finally does recover some years later, there isn’t enough capital to take advantage of the available growth. Moreover, if the portfolio has become increasingly more fixed-income heavy as an emotional response to the stock market’s bad near-term performance, this impact is even more pronounced.
Financial planning expert Michael Kitces has written extensively about this phenomenon. The table below is from Kitces’ 2013 paper on retirement portfolio construction. It shows success rates for portfolios with varying starting and ending allocations to stock over a 30-year retirement period (the vertical and horizontal axes respectively).
In the upper left-hand corner is a 0% – 0% portfolio, i.e., an all bond portfolio. This seemingly conservative portfolio is actually the most likely portfolio of all to fail, since in a Monte Carlo analysis it survives a 4% withdrawal rate only 74.6% of the time. At the other extreme, a 100% – 100% – or all stock – portfolio, performs better, but not nearly as well as a portfolio that starts out the retirement period with a fairly conservative (30%) stock allocation and increases that allocation to 70% by the end of the retirement period.
Why, logically, does this work? Because, as we have seen, the early years of retirement income withdrawals are critical for determining a portfolio’s long-term success. If there is a disastrous sequence of market returns in early retirement, the new retiree with a 30/70 portfolio will be able to withdraw funds first from fixed-income investments, which are less likely to have suffered precipitous declines in value. At the same time, a steadily increasing allocation to stocks – yes, even during a prolonged bear market – means that, when the market finally does recover, the portfolio will be well-positioned to capture an increasing share of the pent-up growth.
Different Techniques, Similar Fundamentals
One of the most interesting things about these findings is that the general success of increasing stock allocations over time seems to underlie other traditional approaches to minimizing volatility and/or maximizing retirement income. A classic “bucketing” strategy, which creates short-term, medium-term, and long-term investment ‘buckets’, for example, is essentially a way to spend down cash and fixed income allocations first, while allowing stocks to ride in the portfolio for as long as possible. The result of this practice is precisely to increase the portfolio’s allocation to stock over time.
Similarly, various annuitization strategies, appropriately implemented, enable a new retiree to dramatically decrease immediate withdrawals from the stock portion of her portfolio. As before, this reduces the need to liquidate stock early in retirement, allowing those investments as long as possible to weather down markets and capture the growth that will inevitably re-appear.
Interestingly, Kitces’ research suggests that a plain vanilla, increasing-stock-allocation portfolio is mathematically superior to these other techniques. In the real world, however, I would say, and Kitces seems to agree, that at times they can be useful, alternative ways to implement a rising-stock-allocation strategy that may be more psychologically appealing to certain investors.
“Bonds Don’t Grow”
Despite the data, it may be difficult for some individuals to embrace these findings. I, however, found myself smiling broadly and nodding my head in agreement when I first read Kitces’ paper. The reason is that I had had a first-hand example of how successful this strategy can be. I took over the management of my dad’s portfolio in 2001, when he was about 85 years old. Dad had been investing for nearly 60 years, through more ups and downs than most of us can imagine. He was one of the smartest investors I ever knew, having turned a modest lifetime income into a portfolio that was well into seven figures. When my dad was 85, his portfolio consisted of about $50,000 in an income (bond) fund, and the remainder in a well-diversified portfolio of stocks. (For the record, he also had fixed income in the form of Social Security and a small pension.) But given the 100-minus-age rule, I had to ask him, “Hey Dad, why don’t you have more bonds in your portfolio?”
His response, which I will never forget, was short and simple: “Bonds don’t grow.”