US Videos

Key Factors for Successful Withdrawal Strategies

Christine Benz
John Rekenthaler

Christine Benz: Hi. I'm Christine Benz from Is the 4% rule for portfolio withdrawals during retirement broken? Joining me to discuss that topic is John Rekenthaler. He is vice president of research for Morningstar.

John, thank you so much for being here.

John Rekenthaler: Good day, Christine.

Benz: John, there is a lot of confusion about what the 4% rule actually means. I often talk to people who think it means that you can take 4% of your portfolio--in perpetuity--that you just stick with that fixed rate. Let's talk about actually what the 4% rule posited.

Rekenthaler: Yes, the 4% rule or guideline has been around about 20 years. It's important to remember it's the output of a study. A study is different than a life. In the study, the financial planner--Bill Bengen, who did this--looked through past history up to the early 1990s from the market data--he had the '20s through the 1990s--and said, If I just used two asset classes, two asset classes alone, stocks and fixed-income bonds, and I want to run a test over various 30-year time periods to see if I'd run out of money, what figure is the highest figure I can do--what will never happen in any time period, ever--will I always be able to last for 30 years. And he came up with 4%. It's worth noting that he added another asset class a little bit later of small company stocks and he got to 4.5%, but nobody remembers that. Otherwise it could be the 4.5% rule, if it just were to come out.

What I'm trying to convey is there was work behind this, but it's arbitrary. It's arbitrary to decide on the 30-year time period. It's arbitrary to decide that the withdrawal rate adjusted for inflation is the same year after year after year--it never changes. It's arbitrary that there were only two asset classes chosen, and it’s arbitrary that in any time period ever in the study that it demanded 100% success rate as opposed to 99%, for example.

Benz: Right. It's also important to understand that the 4% rule implies that you take 4% on day one of retirement. Then, you just inflation-adjust that dollar amount as the years go by. So it's not that you're taking a fixed percentage.

Rekenthaler: That's right.

Benz: Recently there have been some arguments that that 4% rule deserves modification. People have been talking about a 3% rule. Let's talk about why people are calling for a lower withdrawal rate--why they view that as perhaps a safer rate than the old 4% rule.

Rekenthaler: Right. Well, within the context of the original study, again, it's a study with these conservative assumptions, it's reasonable to say that going forward there might not be 100% success using a 4% withdrawal rate. Bond yields are relatively low right now and stock valuation is relatively high in the period from 1926 to 1993. You don't see both of those conditions happening. Sometimes bond yields were low, but stocks were quite cheap on price/earnings ratio, or stocks were fairly expensive, but bond yields were higher than today.

Asset prices do look more expensive now than at that time period. I would guess within the context of that study over the next 40-, 50-year period, there probably will be times when a 4% withdrawal rate will not work. David Blanchett of Morningstar, among others, has done work like this in running Monte Carlo simulations and testing, looking at current asset prices: Will 4% always be successful? And it won't.

Now the point I want to make, and I think everybody here at Morningstar--I've talked with David about that, I've talked with you about this, I've talked with Michael Kitces, who writes on our site and is a planner who is quite expert in this--we all agree that the thing to convey is within the context of the study that used to be 4%, it probably is 3% now, but that study never meant that's the withdrawal rate that you had to take. It's a very conservative sort of starting point, I would say, for a withdrawal rate.

Read Full Transcript

Benz: There are aspects of the 3% rule that you disagree with, mainly because it's a worst-case scenario and also that it is asking for 100% success rate. Is that ...

Rekenthaler: It's asking for 100% success rate. It's a 30-year time horizon, which may not be appropriate, depends when you retire and health condition and so forth. I think the most important is it assumes inflexibility. It assumes absolutely no possibility of change, no adjustment to the fact. The key issue in sustaining a withdrawal rate during retirement is what happens in the first few years of the retirement. If there's pretty much any time there's a failure when you run a test situation, it's because a big bear market happens very early in retirement.

The first few years will tell. You can either start off with a little higher rate and adjust it down if you happen to hit that bear market. And you won't have to adjust it down if you don't. Or you could start off being a little conservative with a lower rate and then notch it up. But either way, if you're not hit with that bear market early, things become a lot more open. It's actually quite unlikely that you would have to stick--that anybody would have to stick--with a lower figure.

Benz: Do you think that people who had been operating with the 4% rule in mind should maybe think about ratcheting it down? If they're embarking upon retirement, should they think about starting conservatively, maybe with more like a 3% rate, take it up?

Rekenthaler: This is probably where I turn to your expertise, Christine. You look at these matters. Where I come in is just trying to step back, look at the bigger picture, and point out that this number comes from--it's a useful number and it's a useful exercise, but it's an exercise that doesn't represent how people live. Nobody is--unlike in the study--nobody is going to go 30 years and pretend they don't know what's going on or care what's going on in the market and do the same draw every single year and never change their habits. In the real world, as you loosen these assumptions, you're not trapped into that figure. There is no reason to be. In terms of the best way to execute it, I don't know, you tell me.

Benz: Well, I think a big component is what's actually in the portfolio. The Bengen research, I think, looks at a 60% equity, 40% bond portfolio. Certainly someone who has a more conservative portfolio with a heavier tilt toward bonds ought to be conservative, given what likely lies ahead.

John, I would like to cycle through what you see as sort of best practices in the realm of portfolio withdrawal rates. You touched on one, which is the value of being flexible, being willing to take your withdrawal rate down if you encounter a really bad market environment, and then maybe you can bring it up and take more if you've just come through a period like the one we've just experienced.

Rekenthaler: It sounds a little bad in a way to say, "Well, you might have to cut back," but certainly all of us during our accumulation phases--just in going through life--we have periods where we have relatively more income or more success or more or maybe just fewer costs, and then kids come or something else comes and you cut back, and we're all used to doing this. Assuming that there is a reasonable level of income so that the cuts don't really cut into absolute essentials, that seems something that we all should be comfortable with doing. Again, these are precautions that may not need to be exercised.

Benz: Right. You also note that you don't have to stick with a 100% success percentage. What is a reasonable level of success rate that someone could stick with if they wanted to ratchet that down a little bit? Where does it start to get scary? Obviously, if you're in the 50% range--50% success, 50% failure--that's not good.

Rekenthaler: That doesn't sound good, does it?

Benz: No.

Rekenthaler: I don't even know how you would do this scientifically. I have thought of 90% or 95% as numbers. When I talk to David, who's our director of retirement planning, David feels that 80% is a more appropriate number because, again, you can always adjust. You start out with a number that you project via some sort of simulation is 80%. If the first year in the markets are disappointing, you can cut back a little bit and adjust. If the first year or two are strong, you're probably already going to be within that 80%, very likely. Again, what happens early in retirement has such a big effect upon the withdrawal rate and one's withdrawal patterns 20, 25, 30 years out.

Benz: Obviously another big lever that people have is that they might be able to adjust the length of their retirement. There's been a lot of enthusiasm for this idea of working longer if people are able to do so. You note that that can actually be something that people can do to improve their overall success rate. If they're willing to shorten the time period when they expect to be retired that that improves their probabilities of success.

Rekenthaler: We also know that people tend to spend more money earlier in their retirements, as well, at least to be more active, too. A possibility is to take a little bit of a chance. But say, I know now I'm in good health and if I spend this money--there are things that I want to do with this money that are important to me--I don't know where I'll be 25 years from now. I'm not saying forget about where you're going to be 25 years from now, but I think one can overemphasize the possibilities or those issues. You don't want to forgo too much now because of what may or may not occur. This does depend on legacy issues. For those who are fortunate enough to have enough money to be thinking about a legacy, then that's an additional factor. It may be that you want to postpone some spending that you could do in order to leave more behind.

Benz: Longevity factors are also in the mix here. Part of making a good assessment here is that you've got to do some forecasting about your own longevity. Is it realistic for everybody to plan for a life that goes until 95?

Rekenthaler: Yes, my mother, she's not here anymore. She died at age 80, and if you would've looked at any sort of genetics or her health, it would have seemed when she was 75, she would have been here much longer. So, one doesn't know.

Benz: OK. A lot of uncertainties here, but it's a really important topic. We really appreciate you taking the time out of your schedule to be here with us today.

Rekenthaler: Thanks, Christine.

Benz: Thanks for watching. I'm Christine Benz for