It is easy to pick on conventional wisdom or poke fun at “rules of dumb,” but sometimes doing so is a worthy exercise. Who better than Wade Pfau and Michael Kitces to take a closer look at how wise mainstream thought might be.
Since Bill Bengen’s 4 percent safe withdrawal rate paper, and particularly over the last decade, financial planners have been flooded with studies exploring the subject of withdrawal rates and portfolio sustainability. The conventional wisdom, which lines up with our intuition, is that as we enter and advance through the retirement phase of life, we should reduce the risk level in a portfolio and own less equities than we would during the accumulation phase of life. As we get older, it seems natural that we would be more risk averse.
Monte Carlo simulations can give us some idea of the likelihood that a certain portfolio structure and spending pattern will be sustainable. One thing virtually all withdrawal rate studies show clearly, with or without MCS, is that avoiding stocks entirely through retirement is not likely to produce an acceptable spending level for a long period of time. Therefore, most retirees will own some stocks in their portfolios.
The disaster that most new retirees and their advisors seem to fear is that the stock market tanks shortly after retiring, but a meandering market can be even worse. Starting retirement in 1929, suffering a colossal collapse in equity markets, and slogging through years of economic depression isn’t actually the worst-case scenario. The 1966 retiree faired slightly worse due to rapidly increasing costs and a market that did not advance during the early years of retirement.
Simulations show bad markets early in retirement are the leading cause of failed trials. The sequence of returns matters a great deal for portfolios experiencing regular withdrawals.
To mitigate the risk of adverse markets early in retirement, most choose to find a balance between stocks and more stable holdings such as bonds and cash. Having a significant allocation to bonds and cash not only reduces the downside of the portfolio as a whole, it provides a source of cash that prevents retirees from being forced to liquidate equities in order to meet cash-flow needs for many years.
This can help avoid selling stocks during market declines, provides time for the market to recover, and it can be a useful behavioral management tool in that it provides clients a reason not to panic during a downturn. If they don’t feel they have to sell, perhaps they won’t. Historically, bad markets are followed by good markets.
Many versions of the so-called bucket approach also benefit from this behavioral element. In a market downturn, if the client can see that the short-term and intermediate-term buckets are stable sources of cash, they’re more likely to allow long-term assets to stay invested for the long term.
Kitces and Pfau noticed that one effect of spending the shorter-term buckets first was that the equity portion of the portfolio tended to rise as the fixed portion was spent down. The allocation to stocks was going up during retirement, not down. They wondered what might happen if a rising equity allocation was put in place by design. Their newest research paper examines that question. A write up will appear early in 2014 in the Journal of Financial Planning, but you can read Kitces’ description on his blog, Nerd’s Eye View; see Pfau’s take at Wade Pfau’s Retirement Researcher Blog, or the actual draft here. Table references are from the actual draft.
Using historical returns of various asset allocations over a 30-year retirement and a constant 4 percent real withdrawal rate approach, they found that the probability of success is higher than a steady allocation to stocks if the portfolio is established conservatively and adjusted annually to increase the allocation to stocks.
For instance, keeping a portfolio with a 60-40 allocation between stocks and bonds resulted in a 93 percent success rate. However starting the portfolio at 30 percent equities and increasing the allocation to stocks 1 percent each year of those 30 years resulted in a 95 percent success rate. Similar patterns resulted from most balanced portfolio mixes. (Table 6)