It’s not unreasonable to assume that a desirable retirement equates to having the financial freedom to meet one's lifestyle and personal goals. The more efficient a person is with their assets, the higher the likelihood of this, which is why sensible retirement income planning is so necessary. Today’s guest is Wade Pfau and he is arguably one of the main thinkers in the retirement income space at present – a more readable Moshe Milevsky if you will. This podcast is usually devoted to high-level discussions about portfolio investment so it was an honour to have Wade join us and have a similar kind of conversation but rather about retirement income planning. Retirees face some unique risks when it comes to strategies for asset management, insurance, and investments, which means they require tailored strategies, and today Wade weighs in on some of the different approaches we see out there. The topic of probability versus safety-first approaches, and the potential wisdom in amalgamating the two as a means of preparing for retirement, crops up a lot in this discussion. Wade talks about the four L’s of the safety-first strategy, how it recommends building up a base of savings that act as an income to reach higher legacy in the long term. He suggests that people need to account for longevity risk more and argues for the efficiency of assuming that you will live until the average oldest age. That way you don’t end up throttling your lifestyle by saving unnecessarily during retirement. In our discussion, Wade also shares valuable insight into low interest rates versus expected returns, the ineffectiveness of the 4% rule, annuities and deferred annuities concerning mortality credits, and different types of buffer assets. Tune in for all this and much more on the topic of retirement planning from one of the greats in the field today!

Key Points From This Episode:

• Notes on Wade Pfau, a leader in retirement income planning research. [0:00:43.0]

• Unique risks faced by retirees: longevity risk, sequence of returns risk, etc. [0:04:03.0]

• Retirement now vs 20 years ago: low interest rates and retirement length growth. [0:05:16]

• Safety-first retirement: build a floor and then spend more over the years. [0:06:31]

• Contractual protections (annuities) and probability vs safety-first approaches. [0:08:25]

• The four Ls of the safety-first method: longevity, lifestyle, legacy goals, liquidity. [0:12:18.0]

• Why to go for stocks/equities rather than stocks/bonds. [0:12:18.0]

• What the low interest rate environment means for expected returns. [0:16:57.0]

• The ineffectiveness of the 4% rule when applied internationally. [0:18:47]

• How people don’t properly account for longevity risk in retirement planning. [0:23:27]

• A way of covering basic needs so that higher legacy can be gained later on. [0:25:05.0]

• Strategies for buying annuities and deferred annuities at retirement. [0:28:57]

• How mortality credits from an annuity allow you to spend more in early retirement. [0:30:43]

• Mortality credits in relation to immediate and deferred annuities. [0:33:25]

• Better net incomes at the end of retirement through reverse mortgages. [0:34:36]

• Buffer assets such as reverse mortgages and permanent life insurance. [0:37:12]

• Safe savings rates in relation to historical data, bull markets, and mean reversion. [0:44:39]

• Asset accumulation conceptualised separately from the retirement plan. [0:44:39]

• Wade’s idea of a successful retirement: meeting safety-first goals. [0:48:39]

Read the Transcript:

So you've got a blog called Retirement Researcher and you've really made a name for yourself and I've read lots of your research and I've bought and read not all of your books, but some of them. You've got the specialization in retirement income research, which is something that I don't think has had as much of a scientific treatment as maybe portfolio management has over the last sort of couple of decades. Can you explain, when it comes to retirement income planning, what are the unique specific risks that retirees face that maybe other people don't face?

Mm-hmm (affirmative). Right. Yeah. So it's really been in the last couple of decades that there's been this realization that risk is different in retirement, that compared to the pre-retirement period, and it's essentially retirees have to fund spending over an unknown retirement period. So they have to meet their needs, to fund their lifestyle in the face of this longevity risk, this idea of not knowing just how long that money needs to last. And then also in the face of when you're taking distributions, the way the math works out, market volatility has a bigger impact. It's known as sequence of returns risk, where if you get a market downturn in the early retirement years, even if the markets recover and the overall kind of market environment during that retirement was fine, somebody can really be derailed by getting bad market returns early on. And so that amplifies the impacts of investment volatility. And then beyond that as well, just the idea that there's all these surprises that can happen, especially things like long-term care and so forth, where there just may be some unanticipated large expenses that you also have to be prepared. In some manner too, to be able to fund.

Is retiring today, Wade, in 2020 different from retiring say 20 years ago?

Yeah, I would say so for a few reasons. The first is the longevity side where this concept of retirement in some sense really was sort of a late 20th century idea where people might get to their mid sixties and then have a 10 or 15 year retirement and that was sustainable. But today people in their mid sixties might be looking at 30 or 40 years of retirement. And it's a lot harder to save over a 30 or 40 year period to fund another 30 or 40 years of retirement. So that longevity issue is important. And then the interest rate environment's the other big key to all of this, with interest rates being at such low levels that really puts a damper on how much a person can spend in retirement, no matter what sort of approach they're going to take, whether it's bonds or whether it's equities or whether it's annuities, everything's hit by the lower interest rate environment. And it just means looking at spending less and having to stretch that money out over longer than we ever had to in the past.

So what we know, and I know I already mentioned this, but there's been this development of financial economics as a real scientific discipline and a lot of that's been focused on managing portfolios and asset pricing and things like that. Can you talk a little bit about the scientific approach to retirement income planning?

Sure. In terms of the finance theory behind it, it's lifecycle finance. It's this idea of people choosing how to allocate between spending and saving over their working years to try to effectively smooth their lifetime spending power so that they can maintain the same sort of standard of living throughout their lifetimes. And that means saving during the working years to spend during the retirement years. A lot of that sort of approach, we've known about these concepts going back at least to the 1920s or so, and then as well kind of the implications that leads to for retirement income. When you think about this idea of smoothing spending, there's this utility model, this idea that as people can spend more in a given year that makes them happier, but at a decreasing rate. And so you really want to try to smooth that and you really want to avoid having too big of drops in spending power.

And so the academic, the optimal way to spend for retirement is to have some sort of floor, some sort of contractual protection on making sure the basics can be covered no matter what happens. And then with the rest, the potential to invest more aggressively, to seek more upside from the market, but to use some sort of variable spending strategy. And so required minimum distribution rules, at least in the United States where there's a percentage you're supposed to spend each year to pay or at least pay taxes and that percentage increases as you get older. That's really close to an academically optimal approach that you build a floor to cover basic expenses. And then you spend an increasing percentage of what's left as you age to fulfill some of those more discretionary expenses.

So the floor you talk about, that's a big part of your safety first school of thought for retirement income planning, I believe. So can you compare that framework with the more traditional probabilistic framework?

Yeah, so the probabilistic framework in some sense is really newer because it's not really based on any sort of academic model. But the probabilistic approach, it's more of a total returns investment strategy where you are just spending from your investment portfolio and generally you're seeking market upside so that you can try to spend more than bonds would otherwise support, hoping that stocks will outperform bonds. And so you're just spending more than otherwise possible with fixed income. And that can be contrasted with the safety first approach, which really is more like that academic approach in terms of no, use some sort of contractual protection and bring in the power of risk pooling to help manage that longevity and market risk for the basic expenses. But it's not only thinking about risk pooling and insurance, it's just using that for core retirement expenses and then using the total returns investment portfolio for the more discretionary types of expenses so that you take the stock market risks for expenses that are, there's some more flexibility for that. But you otherwise are focused on contractual protections for the basics.

Now, when you're talking about contractual protection, just so to make sure that our listeners are clear, are we talking about annuities?

Yeah, you could do it with individual bonds but then there's no longevity component. So annuities are really what I mean by that. Whether it's the simple income annuity that pools that assets and annuitizes those assets immediately to provide an income for life or now increasingly we're seeing deferred annuities where you don't annuitize the assets, but when you add a living benefit to that, such as a guaranteed lifetime withdrawal benefit, you still get that contractually protected lifetime income. So yeah, either of those types of approaches is what I have in mind.

Okay. I find the annuities fascinating. We've talked about them on this podcast quite a bit, but I also find that interesting that people tend to be pretty averse to actually buying them.

Yeah. Especially income annuities. They're the simplest, they're the most straightforward, easiest to understand, cheapest way to approach it. But there's a lot of discomfort with the idea of doing that because you lose those assets from the portfolio statement. And so people, it's hard for them to conceptualize well yes, my portfolio is smaller now, but I now have this lifetime income that I don't see on any sort of portfolio statement. That makes it tough. And then also this idea that people feel like they're losing the upside potential for those assets and in some way they're giving up if they buy an annuity. And that's where these newer types of deferred annuities have come into play as a way to help overcome those types of behavioral constraints that people have. So you can still see the contract value and you can still seek some upside for those assets, but you also add a protection to it so that you can also draw lifetime income as well.

So if we've got the floor that the guaranteed contractual income from something like an annuity or a guaranteed minimum withdrawal benefit, do you think most investors should be combining that type of safety first thinking with a probabilistic approach to retirement income?

That's really what I've come to conclude, but in some sense it's safety first. Really it is a hybrid because it's not just in a simple sense, probability-based means investments, safety first means insurance. But safety first really goes beyond just insurance. It's about using insurance as a more efficient way to meet a basic retirement budget goal and then using investments for other goals. So in some sense, it really is that sort of hybrid that combines both investments and insurance, whereas probability based thinking as much more an anti annuity and therefore much more you can do this fine just with investments.

Wade, can you describe for our listeners the different levels that are part of your safety first framework?

So with safety first kind of it breaks down, there's these four goals, four financial goals in retirement that I call the four L's, the lifestyle and longevity are this retirement budget and the distinction is longevity or more like these core expenses, lifestyle's more discretionary. And then you have legacy goals and then liquidity and liquidity for unexpected expenses, to be clear in the context of retirement income, it means assets not earmarked for another goal. So once you've said okay, these assets are to fund my spending, fund my legacy, and then whatever's left over after that becomes a source of true liquidity for the financial plan. And so probability based is focused much more on just funding all that with investments. And safety first is thinking more holistically and it's not just insurance, not just investments, but really all the household assets, whether it's home equity, whether it's even things like support from the family or the community. Any sort of available resources to the household, how to put those together into a more efficient plan to match assets to the expenses, the goals of retirement, and just be more efficient in how we think about that rather than drawing everything from an investment portfolio, which really then requires us spending less to make sure we don't outlive our money due to some combination of a long life or poor market returns.

Exactly on the first floor is all about having guaranteed locked down income for your basic cost of living, correct?

So that longevity goal, that's where you want to use reliable income sources, whether it's the pension, traditional company pensions, defined benefit pensions or government based pensions, or even individual bonds to some extent and then that's where, if you still are looking for more reliable income beyond that, where an annuity can come into play as well.

I've seen research on using annuities and it's like, yes, we can talk about safety first and we can talk about guaranteeing income and giving up upside, but isn't it true also that you can have a better overall expected outcome by maybe using some annuities and then potentially increasing the equity allocation in the portfolio?

I do think so. And that's really the research I've been doing in this area. It really was highlighted for me with an article I wrote in 2013 about the efficient frontier for retirement income, which really pointed out that just a stock bond or an investment portfolio, really was the least efficient way to try to meet retirement spending goals. And the basic punchline was trade your bonds for annuities. So instead of having a stock bond allocation, you have a stock annuity allocation. Now there can be some behavioral issues here with the fact that we just mentioned, people don't necessarily like the simple lifetime income annuity. And also they may not be comfortable with everything else, not in the annuity as in the stock market, but at least mathematically that's pointing us in the direction that annuities are really supercharged bonds. They give you an underlying bond like return, but they overlay this risk pooling component that is mortality credits. The subsidies from those who don't live as long and therefore don't need as much for their retirement helping to support the spending for those who end up living longer and therefore needing more.

And by doing that risk pooling allows everyone to spend more because then everyone can spend like they're going to live to their life expectancy rather than being worried about well, what if I live to 95 or what if I live to 100? And therefore without that risk pooling again, it just means spending less or hoping that the stock market will outperform to support a higher spending level. But not everyone's comfortable with spending more today in the hope that the stock market's going to make that possible by giving you the higher returns in the future as well. So that's really the breakdown there.

That's fascinating. I have not read that 2013 article, but that it suggests that statistically you'd be better off putting all of your fixed income allocation into annuities. Did I understand that correctly?

Yeah. I mean the simple punchline is put into the annuity, what you need to meet your retirement budget, leave the rest in the stock market. And that can better support both meeting your spending goals when markets aren't cooperating, as well as on average, having a higher legacy at the end of retirement as well.

Yeah, no kidding. Annuities are supercharged bonds, what a great quote. So in a probabilistic school of thought, how do you think investors should think about expected returns?

The key to a lot of that is to just recognize that when interest rates are lower, you shouldn't just base things on historical averages. And unfortunately it's still what a lot of the simple planning software does, where they, based on the market returns going back to 1920s, they may say that okay, your bonds, assets are volatile, but on average bonds might give you a five or 6% return and then stocks with some sort of equity premium above that. Well, when you start from such a low interest rate environment, it's really mathematically impossible to get those kinds of higher returns. And so you need to make adjustments so that your average returns are going to be more reflective of the lower interest rate environment. And it has a big impact when you just plug in historical average type numbers to run your simulations.

You can see that something like the 4% rule of thumb works 95% of the time. But you update that for more realistic market return assumptions in a lower interest rate environment that 95% success could drop to somewhere between 60 or 70%. So it does have a big impact and I think that's something that a lot of people are still overlooking. They just say, "Well, we can't predict the future so the best we can do is base it on what happened historically." But with interest rates you can predict the future. When interest rates are low, bond returns are going to be lower in the future. And the only way they can average those higher returns is if interest rates continue to decline. But at some point there's a floor on just how much lower interest rates can go.

When I read your book, How Much Can I Spend In Retirement, your research on the 4% rule that you wrote about in that book was, to me at the time, totally mind blowing. I know you just mentioned it briefly, but can you expand a little bit on what you found about the 4% rule and how it probably isn't the best thing to be using?

Sure. Yeah, I mean I got my start in all this. I had a different dataset. The 4% rule is based on us historical market returns. And I had global market returns for 20 different countries and looked at did the 4% real work in other countries? Just found that it worked in the US and Canada, but not in the other 18 countries in the dataset. And if you put all the international data together with a 50-50 allocation, which is kind of a baseline for this research, the 4% rule worked about 68% of the time around the world. And if you wanted a withdrawal rate that works 90% of the time around the world, you had to drop it down to 2.8%. And so I thought that was pretty compelling that we shouldn't just base things on the US 20th century market returns when the US had such a great century as well as Canada.

But if you look at a more typical international experience, the 4% rule didn't work as well. I would get pushed back on this though, that people just especially like in the US context wealth, we live and invest in the US, who cares about other country market returns. But then that just led to all these other issues that the low interest rate issue that we just discussed, it's going to lower returns, lower sustainable spending. The idea that the 4% rule is calibrated to 30 years, but if you're planning for longer than 30 years of retirement, you need to look at a lower withdrawal rate. The idea that the 4% rule assumes investors earn the indexed market returns so there's no investment expenses, there's no other misbehavior. The 4% rule calls for a 50 to 75% stock allocation, which is on the aggressive side and people have to really stick to that and not panic and never misbehave.

They have to follow this perfectly rational investor type of logic. So if you take a haircut off of the returns for any of these types of issues, as well as I should say, as taxes, the 4% rule ignores taxes. So it's fine if you have some sort of tax deferred account where you're paying taxes of the 4%, but in any sort of taxable account where you have to pay taxes on an ongoing basis on interest and dividends and so forth, there's no 4% rule of that either. So that really just led to this idea that we have to look beyond the simple rule of thumb, the 4% rule to think about what is a sustainable spending strategy for retirees.

There's a great love of dividend paying stocks for many people as they plan their retirement. What do you think about using dividend paying stocks for retirement income?

Yeah. So there's not a lot of research about that, but it is definitely very popular in practice. At the end of the day there's four ways to manage this kind of, it's the longevity risk mixed with when you're taking distributions market risk, having a bigger impact. You can spend less, you can be more flexible with spending, you can reduce your volatility at key points and you can use some sort of buffer asset from outside the portfolio. And so the logic of that dividend paying stock approach is, really sequence risk is triggered by having to sell assets at a loss. And so the argument would be well, if you're just able to live off of the dividends and you have high enough dividends to cover spending, you're sort of immune from market volatility because you're never really having to sell shares. You're just spending from the dividends.

And there's just not been a lot of research because we don't have a lot of historical data to really test things like the 4% rule with a dividend paying stock strategy. The concern would just be that though dividends may be higher, the overall total returns might be less if you're getting less price returns from the more concentrated parts of the market and potentially risky or parts of the market that are paying a higher dividend. So it's a feasible strategy, it's a way to try to manage sequence risk by not selling assets at a loss, but at the end of the day, there's not a lot of research backing up pro or con in terms of how well it works in practice.

So there's a huge behavioral component to it as well. I think many dividend investors really can hold on longer, perhaps because they love their stock. That's been my experience at least.

Mm-hmm (affirmative). Yeah. And so that right, the 4% rule assumes people are these perfectly rational investors. And to the extent that they're not, it can be possible to be a more rational investor just behaviorally when you have something like a dividend paying stock strategy versus a total return investment portfolio strategy. So yeah, there can definitely be behavioral benefits for it as well.

You've mentioned longevity risk a couple of times and how risk pooling with annuities helps to address that. Do you think in general and in the 4% rule discussion, I think, speaks to this too, but do you think in general, people are properly accounting for longevity risk and the retirement planning?

Probably not. I mean, some people certainly are, but whenever you're hearing somebody thinking they might only live to 75 or 80 they're surely really underestimating their potential longevity. A lot of times people base it on their parents, which is too small of a sample size, of course. But over time people are living longer and longevity is improving by about a year per decade for the average person from age 65. So that if your parents are 30 years older than you, kind of on average people are living three years longer than their parents. And I don't think that sort of logic always comes into play as well as you're not supposed to just plan to your life expectancy because there's still a 50% chance you'll live longer than that. And so, especially with a total return strategy, you're supposed to plan to some age that you're unlikely to outlive. There may only be a five or 10% chance to live longer. And that's where the discussion starts turning towards things like well, should we plan to 95 or 100? And that's going to be, especially with an investment based strategy, a much safer way to approach the longevity question for a financial plan.

In your safety first model, the top of the pyramid is the legacy part. And the bottom is the guaranteed income to cover your basic needs part. But the more assets you put into covering your basic needs, I assume that leaves less for a legacy. Do you have suggestions for people how they think about that trade off?

In the short run, you're going to be looking at less legacy if you use more of a safety first approach, because you're not going to be getting as high of return. With safety first you're paying for insurance on your, at least your core retirement expenses so that you can definitely meet those expenses over your lifetime. So if you end up not living very long, you paid for insurance that you didn't end up needing to utilize, and you'd have a smaller legacy than otherwise. But related to that, if somebody was planning on a long retirement and doesn't live very long, even though the safety first approach might reduce legacy relative to a probability-based approach. Either way, the errors are still going to get a lot of assets because you ended up not using them, but really the focus should be more on is well, what if you live longer than average?

And in that case, now you're getting all this spending power through these mortality credits that investments are unable to provide. And if you are allocating from bonds into the annuities, so that you're not impacting the amount of stocks you're holding all that much, that is really laying the foundation where in the short run, less legacy, but in the long run more legacy with a safety first strategy. And that's where it counts more because this is scenarios where the investment strategy might even deplete. And that, to the extent that you still have money left with a safety first strategy, it's the difference between leaving some sort of positive legacy versus being dependent on your children because you outlive your assets. That's where the trade-off is, it's the safety , less legacy in the short term, but more legacy in the longterm. And therefore if you're planning for the long-term, it can be a way to spend more and have more legacy in those situations where you do end up living to 90, 95 or 100. Something like that.

Really good way to think about it. I think we see that a lot where particularly when we're talking to people about when to take Canada Pension Plan, which is our social security equivalent here in Canada. People often tend to take a much more short-sighted view of well, I'm not going to live to be whatever age so I want to take it earlier than later. I guess the same kind of discussion there.

Yeah. We have those same issues in the US where people want to get their money out of it and don't recognize insurance aspect that it provides where you can have much more lifetime spending power through your pension if you delay when you begin those benefits.

Right. Do you think, and this kind of ties into that exact discussion we were just having about social security or Canada Pension Plan, is there, based on the research, an optimal age that people should start thinking about buying annuities?

So I think a good time to really start thinking about that seriously would be in the five or 10 years before retirement. You could start thinking about deferred annuities to help lock in some of your gains and make sure in those key pre-retirement years a big market downturn is not going to disrupt your whole retirement plan. And then right around retirement as well is another, if you haven't done it yet, a good chance to start looking at more immediate types of annuities that will begin payments at the start our retirement. Now there is a notion and actually Moshe Milevsky through York University created the Implied Longevity Yield, which shows how as people age it becomes increasingly difficult for an investment portfolio to beat an annuity. And that sort of research suggests that by your mid 70s, that's a really great time to buy an annuity. But I think it can be justified to purchase them even sooner to just lock in your retirement, have that covered and not have to worry about that anymore. Not have to worry about the markets too, to fund your goals at that point for later in retirement.

Interesting. So you mentioned deferred annuities. Is it possible to say when is a better time to buy a deferred versus an immediate annuity?

With deferred annuities there's two general ways they can be used. And what I was kind of referencing there was the prepay for retirement. So if you're thinking to retire at 65, sometime between 55, around 60. The years leading up to retirement, you could start buying deferred income annuities or other deferred annuities that would start lifetime payments around your retirement date. The other way deferred annuities get discussed would be more as a longevity insurance approach, where around the retirement date you purchased a deeply deferred annuity that, so for example you're 65, you purchase one that will start payments that around age 85. And that's a way to really supercharge the mortality credits. So you get a much higher payout rate because there's 20 years for the money to grow and the insurance company general account. And then also if you're accounting for not everyone's going to live to age 85 to start receiving those payments.

So that can be a way to really help manage that tail longevity risk of covering with a smaller asset base expenses for late retirement. So those are the two ways they get discussed. I think both can work fine as well as immediate annuities around the retirement date and even laddering annuity purchases over time so that you purchase some at retirement, purchase some more and your mid 70s, if you feel like the spending power is not keeping up with inflation and so forth. So I think any of those sorts of approaches can be made to work and it's really at the end what people are most comfortable with.

Can you expand a little bit on mortality credits? Just so that our listeners can get a better grasp of why it might make sense to use the deferred annuity as opposed to immediate?

Mm-hmm (affirmative). They both offer the mortality credits but... So the idea is with a simple sort of annuitized approach you have this risk pool and you pay the premium for the annuity, it goes to the insurance company's general account, they invest it primarily in bonds. And so they will give you a payout rate that's based on three factors, the return of your premium, interest that they're able to generate off of your premium assets that they are able to invest and then mortality credits, which are the same idea that the insurance company only has to pay people when they're alive. And so those who don't live as long they didn't spend all their premium. And so some of that premium then becomes available to those who do end up living longer. And that's where the annuity can support a higher spending rate than bonds alone can support because they're offering this third source of spending power to the longer lived than investments just can't provide.

And obviously, so if you die early, you would have been better off without the annuity to some extent because you didn't live as long. It's those who live longer that really benefit. But given that people can't predict which group they're going to fall into an advance, what it really does is it increases your spending power so that you can enjoy more in retirement. Because if you didn't have the annuity, even if you end up not living very long, you had to worry about well, what if I do live to 100? And therefore you have to spend less to stretch your assets out longer. And the mortality credits from the annuity allow you to spend more in your early retirement as well, because you can have this comfort that if you do end up living a long time, you'll have these contractually protected subsidies from the risk pool that investments can't provide and that becomes a source of spending so that you don't have to worry about outliving that asset base.

And so it raises the standard of living for everyone in the risk pool because those mortality credits become an additional source of spending that just investments are unable to provide. And that's then up against the stock market. You could fund retirement with bonds, you could fund with an annuity that lets you pool that risk and spend more or the probability based strategy says you're using aggressive asset allocation and then you get the risk premium, the idea that stocks will outperform bonds. And that's the way you're going to be able to spend more as well.

Just on the deferred versus immediate annuities, again, I know I keep coming back to the same question, but do you have the opportunity to take more advantage of the mortality credits by buying a deferred annuity that will start paying out later? As opposed to buying an immediate annuity later?

Yeah, sorry. So yes, when you buy a deferred annuity you're getting more and mortality credits. And you can see the relationship. If I buy a deferred income annuity at 55, that would start at 65 versus if I buy an immediate annuity at 65, the deferred income annuity is going to have a higher payout for two factors. And it's the fact that for 10 years, the insurance company can invest that money. And it's going to account for the people, if you're buying life only and other abilities to get cash refunds and so forth. But with life only not everyone who purchased at 55 would end up living to 65. So you get an extra boost of mortality credits for those who had died between ages 55 and 65. Now at younger ages, not as many people are dying and so it's a smaller boost but yeah, definitely a deferred income annuity would give you more mortality credits than an immediate annuity purchased at a later date.

Okay. Got it. That's what I was looking for. Perfect.

Interesting. How do you think people should think about the equity they have in their homes as they plan out the retirement income?

It should be part of the thinking. So you want to position all the household assets in a way to generate the most spending power. And I mean in the context of this discussion, I mostly just know about the US case where there's the home equity conversion mortgage, which is a type of reverse mortgage. And the conventional wisdom is don't even consider something like that until everything else has failed. And so as a last resort, if you've depleted all your investments and so forth, then you might consider a reverse mortgage so that you can create liquidity for your home equity to be able to spend that as well. And the research in the US context just generally shows that that last resort approach is the worst way to consider using your home equity asset in your retirement income plan. Part of that might not apply internationally because one of the interesting quirks of the US reverse mortgage is it creates a line of credit that actually grows over time.

And that's a big piece of why it becomes valuable to open it sooner rather than later. I don't know if that aspect might apply with the Canadian system, but at least the aspect of where most of the research about home equity in the US was it's using it as a buffer asset. I mentioned that term earlier with one of the four ways to manage the sequence of returns, risk and retirement. If you have this asset outside of your investment portfolio, that's not correlated with the portfolio. It can be a resource to draw from temporarily when markets are down to avoid selling portfolio assets at a loss and to help preserve the portfolio and give it a chance to recover so that you later resumed taking distributions from it. And research in the US context shows that using home equity as a source of spending during market downturns does really help to effectively preserve the portfolio and create more at the end. So there's a cost to a reverse mortgage, but that the benefits created for the investment portfolio exceed the costs of using the reverse mortgage to create a better net outcome at the end of retirement, in terms of meeting the spending goal and having more legacy at the end as well.

We talked about reverse mortgages on the podcast a while ago, and we thought through it and ended up deciding that it was kind of like rebalancing where if stocks drop in value, drawing from the reverse mortgage is kind of like going short bonds, which gets you back to your target asset allocation. But you're rebalancing from this uncorrelated asset, which makes it pretty compelling.

Yeah, it's this buffer asset it's outside of the portfolio and it's not losing value. So it's a way to help preserve your stocks and to not have to sell shares at a loss in a dig that hole that the whole problem of sequence of returns risk, where you can end up experiencing a much worse outcome that wouldn't be implied by the market returns over your retirement. Because you get hit hard in the early retirement years. You dig a hole for your portfolio that you're never able to dig back out of. Well, the reverse mortgage gives you a tool to not have to dig that hole that you're unable to dig back out of later because you swore spending to this buffer asset temporarily and try to give your portfolio more shot at recovering before you have to start selling from it again.

One of the other buffer assets that you've written about is permanent insurance. Can you describe how that fits into a retirement income plan?

Yeah. Permanent life insurance? And at the end of the day, I only know of three buffer assets and originally it was a big pile of cash sitting on the sidelines, but then since cash doesn't yield much, we look more at home equity or permanent life insurance, particularly whole life insurance. The cash value has that same property. It's got this stable value aspect of it. When markets are down the cash value doesn't suffer capital losses, it's protected from declining in value. And so in this, I don't know the Canadian system here, but in the US at least you can structure policy loans from the cash value that don't count as taxable income and can again provide this source of temporary spending power to spend from the cash value and preserve the investment portfolio. And then it's the same logic. The reason this works is by helping to preserve the portfolio, you'll leave it in a much better position so that even net of the costs of life insurance, you still have a better outcome when you use the life insurance strategically as a buffer asset to help support retirement spending.

It's really interesting. We do have policy loans in Canada as well. In the States I'm not sure if you have the same distinction between participating and non-participating whole life. Do you?

Yeah, we do. So that, of course, with participating you can get more upside if there are surprises in the positive direction, but at least you'd have a guaranteed return and you never have any sort of loss to cash value. It's just participating can give you more. Well, you don't know what, dividends could be less than you expected but they're not going to be negative. And so that distinction doesn't play a role in whether to use the life insurance as a cash buffer, but it is something worth looking at. Do you want something that's more fixed and guaranteed, or do you want to take, it's not really risk, how variable upside rather than fixed upside would be kind of the distinction we would make between participating and non-participating life.

Right. And I guess it gets tricky too, because you're probably paying a bit more for the participating. I guess that's the risk of the upside.

Right. Ultimately the insurance company charges a higher premium to collect reserves and to protect against kind of worst case scenarios from their perspective. And then when things don't look as bad as they were worried about, they return that through dividends and that's the whole source of it. So, yeah, I think that point's right.

On the podcast a while ago, Cameron and I talked about the safe savings rate, which I discovered from your research article that you'd written and we were really excited about it. We thought it was really cool. And then after we talked about in the podcast, we took it away and started thinking about should we be using this in practice? And the challenge we ran into was it's kind of a bet on mean reversion, and we don't really have that much data to build those types of expectations. Are we thinking about that the right way?

No, I mean, I think that's fair. The safe savings rate approach, it's a probability based approach that uses the same logic as the 4% rule style safe withdrawal rate, which is you're basing what you're going to do on historical data. And in historical data, you do see mean reversion. The issue though, there's two sort of scenarios that the safe savings rates, the idea is you don't always have to use a conservative 4% rule for retirement. The 4% rule tends to come up after bull markets. That's when you should be thinking about using a lower withdrawal rate. But conversely, if you retired in like 1982 or 1921, you may not be as familiar with 1921, but it would have been a lot harder to save because those years leading up to your retirement markets were pretty pitiful and you just don't have very much.

And so it would be hard to meet your spending goal with the 4% rule. But that's where I think you're concerned with mean reversion would kick in because safe savings rates would say, it's okay, you could use an eight or 9% withdrawal rate because you're retiring at the end of a bad bear market. Probably things will get better in the future. So I understand that can definitely be scary, but the issue we're in more today revolves around actually 4% might be too high because there have been these bull markets. And to the extent that they might not last forever, you might actually need to look at a withdrawal rate lower than 4%. That I don't think... The main reversion is telling you to be careful rather than telling you to spend more like in some of the historical scenarios. But it is based on historical data so it is a probability based approach. And to the extent that anyone's not comfortable basing their strategy on historical outcomes, that's definitely the same. It's going to have the same mark on it as the 4% rule would have.

Yeah, it's interesting. We took it away and said okay, well, how do we get more data? Let's build a Bootstrap Model or Monte Carlo model, but then right away you kill any mean reversion. So that's where we kind of stopped thinking about trying to put it into practice.

Right. Although it can work without mean reversion because if market returns are at least independent over time, and you want to, say, have a 90% success rate that your plan would work over your lifetime, you can end up... The traditional retirement approach assumes you have to be extra conservative pre retirement and post retirement. Because you want to say have a 90% chance to save enough to have the 4% rule work. And then you're kind of picking the 4% rule because you want to have a withdrawal rate that might work say 90% of the time. So it really, if you just wanted to have a strategy that would work 90% of the time, you could be a little more aggressive than having a strategy that works 90% of the time pre-retirement to meet a goal that will then work 90% of the time post retirement. It's really you're doubling up your 90% success rates by not thinking about this as a whole pre and post retirement period together. And so that same concept comes into play. The safe savings rate concept does show a difference, even if you assume market returns are completely independent over time and there's no mean reversion.

Interesting. Right now market prices are high and we've had this long extended bull market. Do you think people who are saving now should be thinking about saving more than someone who was saving whatever 20 years ago?

The idea of the safe savings rates is really you stick to that strategy. And so it's not necessarily that people should be saving more, but it's that they shouldn't be getting too overconfident or retiring early or thinking that they're done because they've already met these wealth accumulation goals on account of a big market run up. The safe savings rates idea would suggest no, you keep saving like the strategy told to you, because even though you met your goal early, due to mean reversion things might turn around at some point. And so you shouldn't just stop saving because it looks like you're already meeting your goals. You stick to that saving strategy, no matter whether it looks like you're ahead of schedule, which is where you should be today or in the opposite direction. Even if you look like you're way behind, don't give up because you just stick to that strategy.

Right. I guess it comes back to expected returns too where if you built a financial plan based on some level of expected returns and then realized returns are way better for 10 years or something and you get to your asset accumulation goal sooner, when you're there, you probably should revise downward your expected return based on assuming prices are relatively high at that time.

Yeah. That would be another way to get at the same sort of concept. And in the safety first context as well, that would speak to maybe using a deferred income annuity to lock in those gains and then not having to worry about it anymore. That's where annuity could fit in as well.

That's really interesting to think about separating the asset accumulation goal from the retirement plan. It's like if you hit that asset level that you expect it to be at sooner than you expected to be there, that doesn't necessarily mean you're ready to retire.

Right. Especially if you're going to stick just with investments, that would be the probability based version of it. The safety first version of it would be well, maybe you should think seriously about starting to lock in some of those gains now with an annuity.

That makes the whole conversation even more interesting. If market returns are higher than expected and your assets grow fast and you expect them to, that's maybe time to take the chips off the table.

If you've met your goal, start thinking this is where people might be getting, feeling more greedy, but how would you feel if your wealth doubled again versus how would you feel if your wealth dropped in half? And if the benefits of having your wealth double, which you're probably not going to really spend it anyway, so it might not really matter that much. If that's less than if your wealth dropped in half that might ruin your whole retirement plan. So think about those trade-offs on whether it's worth it to just get greedy and seek more and more upside once you have the chance to meet your goal.

The marginal utility piece that you mentioned there is fascinating, but I'm thinking about the idea of, if you get to a certain asset level based on good market returns, if you're taking the asset only approach or the probabilistic approach, you don't necessarily want to think about retiring at that time because future expected returns are lower. But if you get to that asset accumulation goal and then convert to guaranteed income sources, that changes the whole conversation, because you're no longer worried about what expected returns are for your income.

Right. But in the context of probability base, yeah, if you're giving it the same idea, if asset returns were higher and therefore you project lower future market returns, well then you're projecting a lower withdrawal rate as well. To attach to those lower future market returns. So that's the same sort of idea that after a bull market, expect that you might have to use a lower withdrawal rate and vice versa. If the vice versa is not where we are today.

Right.

That's kind of where the probabilistic approach meets the safety first approach and always have that in the back of your mind. So interesting. So it's a perfect lead up to my next question, which is how do you define a successful retirement planning outcome?

So I mostly focus on the financial side and of course so much of retirement is more about lifestyle and kind of realizing your goal, your personal goals, and on the financial side it means having the resources to make that retirement worthwhile and what you expected and the lifestyle you hope to have. And so financially speaking then, a successful retirement is you have the resources to meet your goals with the level of comfort or level of risk you are willing to take. So that if you're very worried about outliving your money, you're forced to spend less and you might end up having a much bigger legacy than otherwise anticipated, but having a strategy that just lets you be more efficient in terms of I'm meeting my goals, my spending goals, having the amount of legacy or the safety margin for my assets that I wanted to be comfortable and getting more spending power while still meeting that legacy goal in an efficient manner. But being able to put that all together into an overall plan that meets those four L's, the financial goals of retirement. That's really how I would focus on a successful retirement. And also just being efficient that you didn't have to underspend and then end up leaving too much legacy at the end, striking the right balance between your spending and your legacy as well.

Efficiency is a piece that we haven't touched on maybe implicitly but we haven't explicitly touched on efficiency in our conversation. But that's probably a whole other discussion but retirement spending efficiency is huge because like you said, you could end up spending way too little of your portfolio and then realizing at age 95 when you're about to be dead, that while geez, I could have spent way more on my portfolio, but you were too scared to do it while you were alive.

Right. Yeah. And by efficiency and just being kind of in the economics context that if you're comparing strategy A to strategy B, strategy A lets you spend either just as much whilst having more legacy at the end or spending more in having the same legacy or even spending more and having more legacy at the end, then it would be a more efficient strategy. With a given resource space you're getting more lifetime spending power from those assets. That would be a more efficient strategy.

So Wade, you've taken this academic rigor to the science of retirement planning. And it's like, when you look out there and maybe this is just my limited knowledge set of who's out there, but there's really Moshe Milevsky and you in terms of being leaders in this space, that's my perception anyway. How do you, other than your academic successes and all the writing that you've done, how do you define success in your life?

Yeah. I mean, thank you. And I really think for people who watch a lot of TV on South Park they had an episode about... Well, The Simpsons did everything and I feel like that's the case with Moshe Milevsky that he's really done it all and I'm really just following in his footsteps. But yeah, I mean just trying to help get that message out, but helping people. I mean, my focus is on retirement income planning, being efficient with your assets and just helping more people be able to achieve the kind of retirement that they're hoping for. It would be how I would really define that.

Makes sense. Well Wade, this has been a fascinating conversation and we really appreciate you coming on to again, bring that academic rigor that on this podcast we talk a lot about but usually in the context of portfolio management, but you're able to bring that from the perspective of retirement income planning, which brings a lot of unique insights. So we really appreciate it.

Thanks. Yeah, it's been a pleasure talking with you both.

Books From Today’s Episode:

How Much Can I Spend in Retirement? — https://amzn.to/2WbMrCa

Safety-First Retirement Planning — https://amzn.to/2Nm75di

Links From Today’s Episode:

Rational Reminder on iTunes — https://itunes.apple.com/ca/podcast/the-rational-reminder-podcast/id1426530582.
Rational Reminder Website — https://rationalreminder.ca/ 

Shop Merch — https://shop.rationalreminder.ca/

Join the Community — https://community.rationalreminder.ca/

Follow us on Twitter — https://twitter.com/RationalRemind

Follow us on Instagram — @rationalreminder

Benjamin on Twitter — https://twitter.com/benjaminwfelix

Cameron on Twitter — https://twitter.com/CameronPassmore

Wade Pfau — https://retirementresearcher.com/wade-pfau/

Wade Pfau on LinkedIn — https://www.linkedin.com/in/wpfau/

The Retirement Researcher — https://retirementresearcher.com

'An Efficient Frontier for Retirement Income' — https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2151259