How Retirement Spending Affects Withdrawal Rates

Christine Benz: Hi, I'm Christine Benz for Morningstar.com. How does spending change over retirees' life cycles, and what are the implications for safe withdrawals rates? Joining me to discuss that topic is financial planning expert Michael Kitces.

Michael, thank you so much for being here.

Michael Kitces: Thanks, Christine. It's good to be here.

Benz: You and a colleague wrote a piece about retirees' spending patterns looking at how retirees spend over their own life cycles. So, I know there's a body of research in this area. Let's talk about what the data tell us and talk specifically about how when you look at the data you do tend to see spending trail off a little bit as people get older.

Kitces: Yeah. So, there has been a lot of interesting data that's starting to come forth over the past five and 10 years, some very good national data sets that look at spending behavior patterns and even now some major financial institutions that can like literally look internally--how do all their bank customers behave.

And we're learning some interesting stuff. The prevailing view has always been people want to live a stable standard of living in retirement. I mean, the whole essence of a pension is built around you're going to get the same check every year for life, hopefully, it's an inflation-adjusted check, like at least an inflation-adjusted standard of living is going to remain constant throughout retirement. And now, we're digging into the data and we're finding out it's not what people actually do.

So, your colleague actually here at Morningstar, David Blanchett, did a nice slice of this data a couple of years ago and found this effect where when we look at inflation-adjusted spending what we find is, as people get older, the spending trails off. It trails off by about 1% a year through our 60s, 2% a year through our 70s, and then another 1% a year through our 80s. So, if you, kind of, draw that, it goes negative, more negative and then a little less negative. So, David called it like a smile, retirement spending smile.

Benz: Healthcare expenses at the end, right?

Kitces: Right. So, early on our spending slows down a little as our lifestyle slows down. In the middle, our lifestyle slows down a lot. We get kind of the slow-go years, much less traveling, discretionary expenses fall off. And then even in our 80s the spending still trails off. So, inflation-adjusted it starts slowing down by less, but even in our 80s the data is showing that the average household spending continues to decline in our 80s even with the uptick of healthcare because the other stuff, mostly kind of our discretionary buckets, slows down even more. So, yeah, I have more healthcare spending, but we only need one car now. Those kinds of trade-offs that start shifting in the household means the spending continues to slow down even in the later years.

Benz: OK. Let's take a look at the implications then for sustainable withdrawal rates, because a lot of the research there, sort of, assumes that, oh, yes, I'm going to need 4% and then I'm going to inflation-adjust that dollar amount over time, and it assumes a fairly stable level of consumption when in fact the data don't necessarily support that.

Kitces: Yeah. The early safe withdrawals rate study--I mean, I've been guilty of this as someone that published to that literature as well--we were all essentially looking at replicating pension streams of income which were stable inflation-adjusting streams of income, and so we shot for the same thing in a lot of the research. Now we get all this data coming forth that says, well, actually people's behavior doesn't match that.

So, we actually re-ran some of the numbers and published a study a couple of weeks ago about what does it look like when you actually assume decreasing spending by something like maybe 1% a year through retirement or 10% a decade or like maybe we stay steady for the first decade but then we slowdown in the second and third. And what we found, I mean, no great surprise, your withdrawal rate gets a little higher when you assume spending decreases.

Now, the interesting effect is, it's actually not as much of a boost as maybe you would think. If you imagine spending going down by 1% or 2% a year for 30 years, like that cuts 30% to 40% off your inflation-adjusted spending by the end. So, it's a big number. But what we found is the safe withdrawal rate only goes up a little. It goes from 4% to maybe something like 4.4% to 4.8%. So, we get maybe a 10% or 20% increase in safe withdrawal rate for what could actually be a 30% or 40% decrease in cumulative spending.

Benz: And what sort of asset allocation do you assume there?

Kitces: So, we look at over a range of asset allocations, what we generally find is balanced portfolios really do still work the best where we can rely on the interest and the dividends and the capital gains and tapping the principal when we need to. The good news of diversified portfolios is there's always something that we can tap from year to year to generate the cash flows and that still matters. Too conservative, inflation is damaging; too aggressive, market volatility is damaging.

But when we get back to those balanced portfolios, generally in the 40-60 to 60-40 kind of range, we would find these effects where your initial safe withdrawal rate lifts up some, but not a huge amount, in part because even though your spending declines, it's only in the later years. So, kind of, cumulatively over a retirement, your spending does decrease quite as much. And the effect that we still have to deal with good old-fashioned sequence risk.

You might realize, hey, yeah, I'm not probably going to do as much world cruising when I'm 80, but unfortunately if you do lots when you're in your 60s and there's a bear market and you run out of money, it doesn't really matter that you're going to do less travel in your 80s because you might have no money left. And so, that need to have a little bit of conservatism in the early years to defend against sequence risk is part of--even if we assume spending goes down a bit, we don't quite get as much of an offsetting increase in the initial safe withdrawal rate, because we still have to hedge a little against that sequence risk.

Benz: OK. Fascinating research, Michael. Thank you so much for being here to discuss it with us.

Kitces: My pleasure. Thanks for having me.

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