Reader N.T. asked me the following thoughtful question (lightly edited for brevity and privacy):
I was reading your article Calculating My Retirement Glidepath, and I am still a little confused on your drawdown strategy. I think I understand the broad concept but I am confused on the details on how to execute.
I was hoping you can comment on what I plan on doing with my parents’ retirement drawdown strategy. They are ETF index investors like yourself with a 60% stock/40% short-term bonds split. My dad will be 73 and my mom will be 60 when they retire. I plan on withdrawing 4-4.5% from their investment portfolio. Based on the safe withdrawal of 4% study and some of the recent research done from another great Michael, Michael Kitces, I think the success rate of my parents not running out of money is like 98-99%.
I can’t tell you how you should handle your parents’ retirement spending, N.T., but I can explain how I would handle it for myself. There are three areas I’ll discuss: what to sell to generate spending money, how much to spend, success rate, and parameter selection for the plan in my paper.
What to Sell
It’s unlikely that your parents’ portfolio will generate enough dividends and interest to reach their target spending level in retirement, so they’ll have to sell assets periodically. Regardless of how often you choose to sell assets, when you have a target asset allocation, such as 60% stocks/40% bonds, the choice of what to sell is fairly easy: sell some of the ETF that is overweight to get back to the target allocation.
A more difficult decision is whether to change the 60/40 mix over time. One of my personal rules is that I don’t try to make asset allocation adjustments based on my intuition about future returns. My plans are purely mechanical. You might choose to think about what asset allocation will make sense in 10 years, 20 years, and 30 years. This will guide you to how to adjust the allocation as your parents age.
How Much to Spend
Your choice to have your parents’ annual spending 4-4.5% of their portfolio initially is in the range I’d choose. However, you need to think about how this will change over time. I’m not a fan of the inflexible plan to adjust the dollar amount by inflation every year and ignore portfolio performance.
It was fine for William Bengen to run experiments on this inflexible plan to try to minimize the odds of having to make painful spending cuts, but I plan to be more flexible than this in my own spending. I have a plan for how the percentage of my portfolio that I spend will increase slightly each year (similar to the increasing required RRIF withdrawals based on age). If my portfolio’s returns disappoint, then I’ll have to spend less, and if it outperforms, then I can spend more.
Many people have a plan that is somewhere between Bengen’s rigid spending and my flexible spending plan. They take the previous year’s spending, adjust it for inflation, and then adjust it up or down a little depending on how their portfolio performed. The more flexible you are, the safer it is to go for the upper end of your 4-4.5% starting range. However, don’t overestimate your parents’ ability to tighten their belts after a stock market crash.
Success Rate
The success rate of any plan depends greatly on several assumptions about returns. If today’s high stock market P/E ratio drops to more “normal” levels in the coming decade or two, this will have a significant effect on the probability of running out of money.
Success rate also depends on spending flexibility. As I’ve described my plan, it’s impossible for me to run out of money, because I have a planned percentage withdrawal every year based on the then current portfolio size (but I could end up spending very little if stock markets perform very poorly). At the other extreme, Bengen’s rigid plan will have a certain failure rate (for given return assumptions). This failure rate is lower when we introduce some spending flexibility.
Parameter Selection
If you want to follow the exact plan I laid out in the paper referenced in the Calculating My Retirement Glidepath article, you need to select the parameters you want to use. I don't recommend this unless you're mathematically inclined and want to automate all this in a spreadsheet or other software.
Suppose you decide that the portfolio needs to last n=35 more years, that bonds will return about 1% per year (b=0.01), and you want the starting spending level to be in the middle of the range you identified (s=0.0425). Using equation (2) in the paper, you'll find that y=9.88 years of bonds will give you the starting 40% allocation to bonds. Using equation (3), you'll find that a stock return assumption of r=0.0416 makes a 4.25% starting withdrawal rate sustainable.
So, N.T., I hope this helps fill in some of the gaps for you. I’m happy to answer further questions if I can.
Hi. A little off topic, but I was playing with your Portfolio Withdrawal Rates spd/sht and it wasn't clear to me if the calculated Retirement Magic Number includes or excludes the amount in the HISA. Thanks for all you do to try and educate us. It's greatly appreciated.
ReplyDeleteHi J.K. Reid,
DeleteYes, the Retirement Magic Number value the spreadsheet calculates does include the amount in the HISA.
Best laid plans. Not to be a downer here but some things you just can't plan for. I manage my parents retirement funds because my father suffers from dementia. He's 81. My mother is basically incapacitated from falls caused by brittle bones. All of a sudden they require a lot of expensive care so the 4% rule has gone out the window. Thankfully, they were quite frugal before becoming ill so they can afford the help for now. I'm not sure how likely it is that both parents would be unable to care for themselves, but it's worth considering in your planning.
ReplyDeleteHi Larry C.,
DeleteI'm sorry to hear about your parents' serious health problems. I agree that these things are difficult to plan for. Fortunately, the 4% rule and variants with a similar percentage turn out to be conservative in most cases, so extra money often comes available later in life.