The 4% rule and other fallacies of retirement planning

There’s a reason it’s called a retirement plan.

When you go to a financial planner or try out a quickie retirement calculator, it’ll often come back with a result that reads something like “Sticking to this allocation and savings plan, you’ll have a 90% chance of meeting your retirement needs.” A good financial planner will translate that for you, but a retirement calculator will just leave you hanging.

Is a 90% chance of meeting the target high? Is it low? Is a 10% chance of not meeting your target particularly high or low? Who knows?

The invention of the 4% rule

The “4% rule” is a common rule-of-thumb used to guide retirees when they’re getting ready to tap their nest egg. Let’s say you retired with $1 million. The rule says that in the first year, you can withdraw $40,000 (4% of $1 million). In year two, you can withdraw $40,000 plus a slight adjustment up for inflation. And so on.

You might not be as familiar with how that rule started. In 1998, a trio of professors at Trinity University penned a study looking at historical returns, withdrawal rates, and life expectancies. A table in the study broke down probabilities of success based on a 2% withdrawal rate, 3%, 4%, 5%, and so on. Four percent was simply the highest withdrawal rate where, most of the time, the probability to success was near 100%. So, all the media folks and financial planners turned it into mantra.

But was that what the professors intended? Not at all. In fact the study concludes: “For stock-dominated portfolios, withdrawal rates of 3% and 4% represent exceedingly conservative behavior.” Why did everybody gravitate to 4%? They were mesmerized by that 100% success rate.

How retirement really works

I wrote something recently for Consumerism Commentary that basically argued a 7% chance of not meeting your target is pretty high. I mean, it doesn’t seem high until you’re part of the 7%. Then you’ll wish you had invested as if you had a 99% or 100% success rate.

I still agree with what I wrote a little while ago, but I think I’m guilty of committing a wonkish mistake. You see, if you deal with these kinds of calculators and statistics and historical returns and probabilities long enough, you start to forget how people actually act. It’s like getting so caught up in batting averages and ERAs that you forget what a game looks like when it’s played.

I can set up a retirement plan for myself that the calculator says will give me a 100% chance of meeting my retirement income goal, but in the end, that ends up being a pretty useless exercise. Retirement calculators necessarily make you plan out 40 or 50 years of your life, but they don’t take into account humans’ abilities to adapt.

If someone saw their account balance drop dramatically in 2008, they probably ratcheted back their spending. On the other hand, if someone hits age 85 and still has millions in the bank, they probably decide to splurge a bit. We don’t set a plan at age 65 and blindly follow it as if nothing has changed for 30 years. That’s just stupid. (And that’s what a retirement calculator assumes.)

Your plan is more flexible than you think.

Where does that leave you and your retirement calculator? You could set up a plan that’s effective 100% of the time. Your calculator will tell you that you have to save a ton of money or work well beyond age 65. But in exchange, you’ll sacrifice a lot — maybe too much — of today’s standard of living to get those extra percentage points of certainty.

So next time I use a retirement calculator, I might not panic if it says I only have an 85% chance of meeting my retirement goal. Even when I enter retirement, I might try a 5% withdrawal rate, which according to the Trinity study, would give me an 80% chance of not outliving my money. As time goes on, I’ll adjust up or down depending on what life and the market throws at me.

After all, in 2020, I know I won’t look at the plan I wrote in 2010 and keep following it blindly. So why should I plan as if I will?

Tagged as: Personal Finance, retirement