Risk is sometimes the elephant in the investing room, especially for retirees. People understand stocks as ownership and bonds as debt, but risk is hard to grasp and instinctively dangerous.
Later Living has recently published four posts on risk. Risk and high returns go together, so retirees who want high returns must deal with risk. Here are the four earlier posts knit together into one risk story:
Risk is often described as the prospect of loss. For example, the Securities and Exchange Commission says risk tolerance is your willingness to lose your investments in exchange for potential gain. That sounds as if the loss is surer than the gain, offering tentative investors a reason to fear risk.
Professional investors understand risk as variability. Risky investments (stocks) fluctuate in value more than non-risky ones (bonds/cash). More importantly, however, risky investments (stocks) offer greater long-term gains. Stocks rise and fall around a generally rising trend, and that trend line is usually steeper than that of non-risky investments (bonds/cash).
When people are investing for short-term goals, however, the potential harm of falling stock values near the time of the need becomes very important. The prospect of trend-line gain gets swamped by that of potential loss should the stock market sink just when the money is needed, so people shift away from risky investments (stocks) as the goal approaches.
Retirement investment is mostly long-term investing, so portfolios weighted to stocks make sense for most of the retirement investing time period. That period may be 60 or more years, and the potential for long-term gain is great.
The Danger of Risk at Retirement Time
There is one time in the retirement scenario when investing is effectively short-term, and that is the time of retirement. The retirement date is when a portfolio switches from merely accumulating value to supporting retirees through periodic withdrawals.
If their portfolio is heavily tilted to stocks and the market caves just before or after retirement, retirees face a grim choice. Either they must cut their planned withdrawals (those based on previously higher portfolio values) or face the prospect of running out of money as they age. At retirement time, the risk of loss is poignant.
Two Vanguard mutual funds, total stock market and total bond market funds, brought these ideas to life over a 14-year period, from 2000 through 2013. That period started with three years of sharply negative stock returns. When a portfolio of 100% stocks was subjected to annual, inflation adjusted withdrawals set at 4% of the initial value, the portfolio shrank over the 14-year period to just 57% of its beginning value, which is dangerously low.
That simulation was compared to an all-stock portfolio that experiences a different sequence of returns in which the negative returns occur at the end of the 14 years instead of the beginning. The ending value of that portfolio was 95% larger than the first and 11% larger than the beginning value.
The example shows that a stock market drop at the time of retirement is much more harmful than the same drop eleven years later.
The post went on to show how diversification among asset classes and moderation of withdrawals ameliorates the harshness of negative stock returns early in retirement. These two points can not be overemphasized: retirees approaching retirement should have portfolios diversified among stocks and bonds (and perhaps other assets). That diversification was an important factor helping to sustain a portfolio. Second, being able to live without inflation adjustments on withdrawals also helped preserve the portfolio.
As retirement continues, a portfolio often slowly grows while life expectancy slowly declines. Gradually retirees can re-introduce more risk (stocks) to their portfolios. A recent article by Michael Kitces and Wade D. Pfau in the April 2014 issue of the AAII (American Association of Individual Investors) Journal helps quantify the idea that risk levels in retirement portfolios can optimally increase through retirement. (The idea is also on Mr. Kitces’s blog.)
After simulating different risk levels (stock/bond allocations), they concluded that a portfolio with 30% stocks at retirement (70% bonds) increasing to 70% stocks (30% bonds) over a 30-year period was the optimal path (produced the highest probability of NOT exhausting a portfolio subject to withdrawals starting at 4% of the initial value and increasing with inflation).
Retirees Can Wrestle Investment Risk and Win
A third blog post went on to highlight different risk characteristics between annual withdrawals from retirement portfolios and larger withdrawals for special needs or plans. Home repairs, tuition for grandchildren, motor homes, travel or long-term care can all present special spending needs that will require lumps of cash.
For example, a person in his early 80s whose health is declining may need long-term care in five or six years. The-risk situation is very similar to that of a young couple saving to buy a home in five or six years. In each case, there is a need for a predictable amount of money in a few years, and that sum should not be exposed to much risk.
Four questions help assess the risk characteristics of such special needs. They are:
- Is the goal near in time, say within four or five years, or is it more distant?
- Is the goal certain in time? Will it occur at a specific time that cannot be easily advanced or postponed?
- Will the goal require a known, specific amount of money, or can it be funded at flexible levels?
- Is the amount a large part of the portfolio?
The questions prompt a qualitative analysis of risk that is specific to the people and spending needs at issue. There isn’t a collection of correct answers.
Killer Moves to Wrestle Lumpy Retirement Spending
In the fourth post on risk, I offered a more detailed heuristic for thinking through the four questions. I used the case of long-term care.
I hypothesized the long-term care need at $300,000 and ended with an allocation for that part of the portfolio at 17% stocks and 83% bonds. That shifted the overall allocation of a $1 million portfolio to 51% stocks and 49% bonds. These results are specific to the example, and every reader thinking about a large, special spending need will need to think through the questions with their own facts in mind. The analysis will work with any of the special goals or plans listed in the post, “Retirees Can Wrestle … .”
Conclusions
What arises from the four-posts, now five, is a useful way of thinking analytically about risk. Risk is variability. Stocks, which are risky, tend to grow in value more over the long term than bonds, which are less risky. Investors who want good long-term growth will invest large amounts of their portfolios in stocks.
But when an investor is saving for a short-term, special spending goal, then risk is an enemy and bonds or cash are better.
A family saving for college tuition understands that when college starts, their savings will switch from accumulation to spending and that they will need enough to last through the spending goal. In such cases, investors should avoid risk as that switch approaches. Otherwise, a severe drop in stock values near the time of the goal jeopardizes their ability to pay for it.
At retirement time, a portfolio switches from accumulation to spending, but the spending is a smaller part of the portfolio (often 4%) than college tuition or long-term care. Nonetheless, the analytics are similar: if the portfolio’s value plummets just before or just after retirement, if the annual withdrawals are based on the earlier, higher portfolio value, then they will be too high. Unless they reduce their withdrawals, the retirees are more likely to run out of money late in life.
As retirement proceeds, however, portfolios subject to modest withdrawals tend to slowly grow while life expectancy slowly declines. The prospects of running out of money shrink, and, absent special spending needs, the risk level—or stock allocation—can gradually increase.
If lumpy spending needs are part of the retirees’ plans, however, then stock exposures should be scaled back depending on an analysis of the four questions posed above.