Skip to Content
US Videos

What's the Most Efficient Withdrawal Strategy?

Morningstar research found that the 4% withdrawal rate is not as practical for retirees as many think, while an optimal approach combines mortality rates with portfolio performance.

Christine Benz: Hi, I am Christine Benz for Morningstar. One of the linchpins of a successful a retirement plan is settling on a sustainable withdrawal rate. Joining me to share some recent research into this area is David Blanchett. He is head of retirement research for Morningstar Investment Management.

David, thank you so much for being here.

David Blanchett: Thank you for having me.

Benz: David, you looked at various strategies for setting withdrawal rates, and you've calculated what you call the efficiency rate of each of these strategies. Let's look at what specifically you looked at when gauging the efficiency of withdrawal-rate strategies.

Blanchett: The idea behind this paper is if you have a pool of assets, you can withdraw money for income during retirement. How much can you take from the portfolio for your lifetime? And there are two key unknowns you have when you first retire. The first is how long you're going to live--so you do have to plan for a 10-year retirement period or a 30-year retirement period--and the [second is] market returns. The better the market returns, the more you can from a portfolio.

And so the idea for the research was to figure out how efficient a strategy is. If you had known how long you are going to live and the returns you are going to receive, you can know exactly how much you could take out every year during retirement. What we did is we compared the income you would use from different strategies against this efficient amount. That creates the efficiency ratio, which is, how much you actually received or what you could have received if you had perfect information.

Benz: So you tested various scenarios using different time periods, as well as different portfolio performance, to see which of the strategies was in fact the most efficient.

Blanchett: Yes. The time period is based upon life expectancy. We kind of randomized life expectancy. For someone who is 65 years old, they are going to live, say 20 years. Well, sometimes they will live two years; sometimes they live 40 years. The idea is to kind of make that random based upon mortality tables and then also randomize the returns based upon our forward-looking estimates, and say, "Given all these different simulations, what could we expect the portfolio to do if you'd known that information beforehand?"

Benz: Let's look at one of the most commonly used methods, sort of the baseline strategy that you often hear about in the context of withdrawal rates, that's the 4% rule. So there, you take 4% of your initial balance upon retirement, and then you just inflation-adjust that dollar amount. How did that do from an efficiency standpoint?

Blanchett: Not very well. I think the 4% rule is the research notion of retirement. It doesn't mesh well with actual retirement because the 4% rule is a rule that you make at one point in time. So you would decide at age 65, I'm going to take out 4% of my portfolio. If I have $1 million, [I'd take out] $40,000 a year and increase that every year by inflation. You totally ignore other things that happen, your portfolio performance, your ongoing life expectancy. What you see with that is it's very inefficient because it doesn't reflect what happened in your life. It doesn't reflect that if your portfolio goes way up in value, you can take more money out. It's a very static decision that's kind of a one-off versus a more dynamic approach.


Benz: One of the things I know people like about the 4% rule, though, is that it does let them take sort of a stable stream of income, that their income is knowable which I think a lot of people like.

Blanchett: Yes.

Benz: But it isn't very efficient?

Blanchett: Well, for example, people don't buy cars and never have them looked at. You don't drive your car until it breaks down. Every now and then, every 3,000 or 5,000 miles, you'll take it in to get the oil changed. You'll have it looked at if the brakes start working funny. So it's the same concept with income during retirement. You'd like to have kind of the same amount every year potentially as long as you're alive, increased by inflation, but it's not practical given how life changes over a 10-, 15-, 20-, 30-year time period.

Benz: So you see as kind of a best practice, at the very least, you periodically check in and see how is the market doing and maybe adjust withdrawals up or down.

Blanchett: Once a year is a good kind ofbenchmark for that.

Benz: Let's look at another strategy commonly used among a lot of our users I know. You call it the endowment approach, and that's where you take fixed percentage rate per year and you are more or less swung around by whatever your portfolio balance is doing, you can adjust up or down your portfolio withdrawals. Let's talk about how that strategy did from an efficiency standpoint.

Blanchett: If you think about a spectrum, the least kind of dynamic approach is the constant withdrawal. We just take that same dollar amount every year and increase it by inflation. Endowment is kind of one step in the right direction, because what you are doing is, every year you are taking out some percentage of your portfolio. And the good part about that is you can never be in ruin because if you are taking out, let's say 5% a year, even if your portfolio gets smaller and smaller and smaller, you still have a portfolio to withdraw assets from. So that is a better approach, because your income every year is based upon your portfolio performance.

Benz: Now, let's finally talk about the strategy that you ultimately settled on as the most efficient. Let's talk about what that strategy entails and how investors might incorporate it into their own retirement plan?

Blanchett: It has a horrible name. It's called dynamically updating mortality. But the idea is that every year that you live, the odds are you are going to live longer. So, someone who makes it to age 80 could live another 10 years. What you do is you think about the two variables as you move through time. The first is the portfolio performance, how the portfolio actually did. And the second is how long you are expected to live. And so every year, you kind of say, "I lived one more year, my portfolio is up 10%, and I'm going to live another year longer." And then you base that amount on your life expectancy to determine how much you can pull from the portfolio. Just revisit the amount you can pull out every year based upon your life expectancy and your portfolio value.

Benz: So that mortality updating piece is a component of required minimum distributions that people need to take from IRAs and 401(k)s and so forth.

Blanchett: It is. So like there's the RMD method we also test, as well, which is if you are over the age of 70 1/2, you need to take out a certain percentage from your IRAs and 401(k)s once a year. That is one over your life expectancy, and that's also an efficient way, as well, because, if for example you have 20 years left, you could take one over 20, or 5% per year. And as your life expectancy decreases, let's say you have 10 years left, you take out 10% per year. That's also an efficient way to kind of figure out how much you can take from a portfolio very simply.

Benz: So, your optimal approach combines that RMD method which takes into account mortality, but also looks at portfolio performance and melds the two.

Blanchett: Correct.

Benz: Well, David, thank you so much for sharing this research; it's very valuable. I know this is a hot topic among our users. I think they'll really appreciate hearing from you.

Blanchett: Thank you.

Benz: Thanks for watching.I am Christine Benz for