What Stress Tests Say About the Right Way to Adjust Spending in Retirement

all right okay well good morning everyone thank you so much for joining our retirement income

0:09

Intel webinar this month my name is malcoli I will be your MC joined by our

0:14

two speakers Justin and Derek uh we are excited for another great session um for our returning users welcome back

0:21

to our new visitors welcome to retirement income Intel by income lab uh

0:27

this is a great time where we put together each month um a space for

0:32

um Justin and Derek to really talk about specific Topics in retirement income planning

0:38

um it's another great time to really hear uh your questions as well so we first have a presentation and then we

0:44

will open it up for Q a after the presentation you will see the Q a section in the zoom webinar here so

0:50

please go ahead and drop uh those questions there you can also upvote other people's questions and so

0:56

afterwards we will work through um uh the most voted questions first also after this webinar you will see a

1:04

survey to fill out your responses this is a CE approved webinar so you will also be able to give us your cfpce ID so

1:12

we can make sure you get the credit as well and then you will also be able to uh note if you would like to have one of

1:18

our team members reach out to you if you'd like to hear more about income lab specifically all right guys that is uh my Spiel

1:25

Justin Derek I will turn it over to you guys and I will be back for the Q a all right thank you madly thanks Derek

Overview

1:32

thanks everybody for for coming um yeah so this is we have two two webinars each month

1:39

um a more of a q a and a focus on practice in our um lab talk Tuesdays

1:44

this is more of our research oriented you know topics in retirement income planning uh webinar and as mcley said

1:51

there's CE here so definitely stick around to the end and fill out the um the questionnaire to get your CE credit

1:57

um so today we're really diving into some things that that Derek and I have

2:03

um have found in applying uh income Labs new stress test to some other ways to do

2:10

adjustment based planning um and specifically withdrawal rate guard rails so I I think uh this will be

2:18

really good for people who have um you know toyed with guardrails in the past uh or no people who have to kind of take

2:25

those Concepts out of Academia and see how they apply or don't as the case may

2:31

be um in real life and um you know what we can do better

2:38

okay so here's the outline um and actually Derek uh presented something

Outline

2:45

um you know adjacent to this uh on kisses.com maybe last month so also if you're interested in more of this type

2:52

of content definitely check that out if your members over there um so the outline I'm going to start

2:58

kind of high level with you know why you would want to have adjustments and guard rails in retirement I think that's a key

3:04

to just step back before we get into the Weeds on you know the specifics of how you might

3:10

um guide a client through retirement and advise them on changes in spending you know what's the point of all this

3:16

um then we'll dive into withdrawal rate guardrails what they are and why they don't work in practice and then we'll

3:22

talk about risk-based guard rails which is a more holistic approach and does does work better in practice

Why adjustments

3:30

okay so why adjustments um I think the the key behind this idea

3:36

that retirement income planning is really about an ongoing management and guidance process is that if you set

3:43

someone off on their retirement Journey spending a certain amount or following a certain spending plan

3:49

um then you get this incredible array of possible outcomes

3:54

um many of which I think you could reasonably call failure so what we're used to talking about as failure is

4:01

running out of money and if you've seen a Monte Carlo analysis chart like this with uh you know portfolio balance this

4:09

will be familiar to you this concept that hey if you just followed this plan blindly

4:15

um you you might run out of money some you know at some some percentage chance or some amount of the time uh but many

4:22

have also noticed that actually often these analyzes also show that in a huge

4:28

amount of the time you end up with way more money at the end of the plan than you had hoped for um so that is also failure now it's

4:35

probably not as bad a type of failure as running out of money but it's sort of a overly Frugal too much

4:42

um you know you've restricted your lifestyle more than you had to maybe you've foregone certain spending that

4:49

that you know really might have improved your quality of life so the main reason you might be interested in adjustments

4:56

in retirement is that this picture if it really is a representation of the

5:02

possible outcomes of retirement portfolios just shows that this would be a terrible way to run retirement the the

5:09

outcomes are just way too broad we can do better we can provide a range of outcomes with a much with a much

5:15

narrower um range simply by adjusting along the

5:21

way this is any kind of complex situation where we

5:27

don't know the future we don't know exactly what's going to happen kind of needs to have steering it needs to have

5:33

adjustments as a possibility so what you're seeing here is treating retirement as if it's like shooting a

5:39

you know a bottle rocket in the air well it's just going to land wherever it lands we'll see versus you know a a

5:46

space shuttle or something with an actual steering wheel or you know it's pinewood derby cars versus an actual car

5:52

with a steering wheel and a brake pedal and an accelerator so retirement is one of those complicated situations where we

5:59

learn what world we're living through as time goes on and so not to adjust would

6:05

be kind of crazy and actually a lot of the research on retirement income

6:11

uh even rules of thumb like the four percent rule in that research for

6:17

example Bill bengan actually talks about adjusting at the end of that article so he he is not saying this is what you

6:22

should do and you should do it without adjusting so the idea that people should have a plan for adjustments a method of

6:29

doing adjustments in retirement is very old and I think well founded

6:34

um because adjustments help us avoid two things

6:40

um running out of money or overspending so if we discover that we're in a world that you know it's a little tougher than

6:46

we had hoped adjustments help us tap the brakes on spending or make other adjustments to kind of keep things in

6:53

line and they also help us avoid leaving too much money behind or under spending so if we discover that we're actually in

6:59

a world where that's supporting a higher spending level maybe that's because returns have been particularly good or

7:06

because inflation's been particularly low or any number of other things we need to know that additional spending

7:14

is available and and have the opportunity to take advantage of that as

7:19

as retirees and clients so what our goal is to kind of optimize

Goal

7:26

income along the way and and take the best path possible given what we know

7:32

um given our resources and kind of land where we're hoping to land so this picture where the the outcomes are

7:39

really zeroing in on in this case maybe it's a desired Legacy goal um is the is

7:44

the the ideal situation now of course we don't live in an Ideal World to do this you would have to know the future you

7:51

would have to know exactly when someone was going to die um so that's unlikely however it is

7:56

possible to at least kind of shrink the range of outcomes and get close to

8:02

optimizing for retirement spending along the way just by paying attention

8:08

and giving people advice on when they can adjust their income up and when they should adjust their income down or make

8:14

other changes to remove risk from the plan so that's at a very high level why you

8:19

would want to use adjustments in retirement it's because if you don't

8:25

um you are you're kind of just just making a decision at the beginning of time of the retirement and hoping things

8:32

work out um and and you're also ignoring a bunch of information that you get along the

8:37

way and and that you where you can update your your um your plan

Guard rails

8:44

okay so second question is what are guard rails given that we we might want to adjust along the way there's also

8:50

been a lot of work um particularly in the practitioner world on um methods for doing this analysis

8:59

methods um methods were actually delivering adjustment advice to clients and I'm

9:05

going to focus on the guardrail concept here because it is probably the most

9:10

um the most discussed and the most used approach to adjustments

9:15

um in the retirement planning space and the concept is just this

9:21

um people would prefer not to make more changes than they have to

9:27

so um guard rails really say okay there is a range at which I'll just keep spending

9:33

what I'm spending or I'll just keep following the plan that I have because there's a cost to making adjustments

9:39

there's it's just it can be kind of annoying it's administratively burdensome and frankly you can you can

9:45

overdo it and be whipsawed on making adjustments and so there is kind of this range in the Middle where you can see

9:51

this kind of light gray no change it's the the no change range um where you just follow the plan

9:58

however if risk gets quote unquote too low meaning you're getting information

10:03

that hey actually things are going pretty well the world we're living through is better than we had expected

10:08

or better than we'd hoped and there's a chance to increase spending if we want to so that's the this this green

10:16

guardrail um on the other hand there's this red guard rail which is risk is getting too high

10:23

so we need to decrease spending so things have gone worse than expected maybe inflation has been higher than

10:28

expected or returns have been worse than expected or maybe even just somebody has spent more than than they had originally

10:34

planned to and so that has made future spending have to go down um so the idea is to set these guard

10:40

rails in a way that um they trigger adjustments at a point where it really makes sense to where it's a big enough

10:47

deal that it'll impact the future risk of the plan and it's administratively

10:52

worthwhile um so you're kind of um trying to balance those those two

10:58

sides of things any adjustment will affect the risk of the plan but you don't want to do them so often that that

11:03

it's um whipsawing people all over the place making them change Lifestyles

11:09

um and so that's why you can think of these as guard rails because they're not going to be you know kind of hitting the

11:14

car all the time right we can we can sort of let the car sort of drift you know back and forth but hey if we get

11:20

too far out let's let's uh you know course correct so that's the concept of a guard rail at

11:26

again very high level this is just conceptual um let's get a little bit more uh

11:33

specific so specifically when would risk be too too hot too low rather with a

11:40

green guardrail well it's if the income that you're you're paying yourself is

11:47

too low given your situation one particularly clear way that's possible is if your balance is quote unquote too

11:54

high so you have some goal of a maybe a legacy goal could be zero could be a million dollars you have some plan

12:00

length you have some expectations and it sure looks like you got plenty of money given your current income level and so

12:06

you could afford to spend more you could afford to pay yourself extra on the other hand if income is too high given

12:12

your resources and your plan and your goals the balance is too low given your your plan of resources and goals then

12:21

um it's time to decrease the amount that you're paying yourself yeah okay one step further what a withdrawal

Withdrawal rate

12:28

rate guard rails that's the main um the main sort of guard rail that people have really focused on

12:34

um in this research have been on what is the amount you're taking from your

12:40

portfolio your retirement portfolio um as a as a rate so if I'm taking fifty

12:45

thousand dollars a year from a million dollar portfolio that's a five percent withdrawal rate so measuring what's too

12:51

high or too low based on that rate and there are a lot have been a lot of approaches to this and variations on

12:57

that theme um and and this is where okay if your

13:03

withdrawal rate is too low so maybe you you know you have a I don't know a 30-year plan and the withdrawal rate is

13:09

two percent you know that's probably too low so let's increase withdrawals um that can happen over time if the

13:16

balance is getting up there right if you keep taking fifty thousand dollars a year and your eight hundred thousand dollar portfolio goes to uh a million

13:24

and then 1.2 million your withdrawal rate is going down down down and it's it's it's good news for you you can

13:30

spend more withdrawal rate guard rails on the bottom are the opposite so if your

13:35

withdrawal rate is too high it's because the the withdrawals you're taking from that portfolio

13:41

um it's it's do I have a percentage so if that balance went from you know a million to 800 to 500 and you're still

13:46

taking fifty thousand dollars now it's a ten percent withdrawal that feels too high so you can see intuitively

13:54

um why this would make sense to people because one of the key risks that we know of in retirement is sequence of

14:00

return risk which is specifically related to portfolio withdrawals so we kind of want a a an alarm Bell to sound

14:07

if it looks like we're taking you know too much from our portfolio stressing it too much and having higher sequence of

14:15

return risk on the other hand we also want an alarm Bell to sound if if if there's good news and we actually could

14:21

be spending more so it's really intuitive how these withdrawal rates connect to this idea that risk could be

14:27

too high or too low okay there is a problem though with

Withdrawal rate guardrails

14:33

withdrawal rate guardrails there's a couple of them and Derek and I have written about this before in other contexts

14:39

um we're going to focus on some other problems today primarily but um those other I'll mention those other problems

14:45

um one is you can actually only apply withdrawal rate guard rails at least the swords that we'll be going over today

14:52

um to plans that only have withdrawals in them and typically only plans that

14:59

plan to have flat income or flat spending meaning just inflation-adjusted spending over time nothing fancy nothing

15:06

where you spend more early in the plan less later in the plan nothing where you're paying off a mortgage nothing where there's anything lumpy in the plan

15:13

so and and this is because these approaches were developed basically as as academic

15:21

Concepts and when you do that when you write articles research articles you typically simplify things so that the

15:27

the topic you're focused on can come through and not it won't be distracting that there are other issues uh in the

15:35

plan however that's not actually how most plans work it's it's pretty rare to

15:40

have someone only have portfolio withdrawals but even for those plans that do for example someone who has

15:45

already started social security has a portfolio and that's it um there's a second problem

15:52

which is that when withdrawal rate guardrail systems are tested through

15:58

tough times the at least the most commonly ones uh used ones the ones that we've tested

16:04

perform really poorly um so even if you have a plan that in

16:10

principle you could use withdraw regardless for because it has a simple shape of portfolio withdrawals uh the

16:19

problem is when it's tested you know kind of like I don't know what you'd call it FDA level one you know does it

16:25

cause harm it sure looks like it does so it looks like it doesn't work

16:30

um so just to make that first Point again uh one problem with withdrawal red guardrails is they only make sense for

16:36

withdrawal only plans with flat income but if you have a plan like this where

16:41

you've got some part-time work Social Security being delayed and so on the withdrawals themselves could look like

16:48

this dark blue um which Derek and I have called the retirement distribution Hatchet so you

16:54

can see kind of that the blade of the hatchet and then the handle and when when your plan is to take withdrawals in

17:01

a lumpy way first hire you know then lower then lower again finally changing

17:07

over time it's going to be almost impossible to apply withdrawal rate guard rails

17:13

okay however we're going to test withdraw rate guard rails using a withdrawal only

17:19

plan again if you had Social Security layered underneath this that would also be fine so that would be a a plan that

17:26

that you could in principle apply withdrawal rate guardrails to and it

17:31

wouldn't be nonsensical it would be fairly nonsensical to try to apply it to a plan like this it would not yield good

17:37

results it wouldn't even yield results that you could evaluate it would just be kind of silly

17:43

okay so let's let's take a look at this um

17:49

in order to stress test withdrawal regardless we have to choose a specific version of withdrawal regard rails and

17:57

the kind of the most um adopted approach to this is probably

18:02

the the Guyton Klinger withdrawal right guard rails again these were proposed as

18:09

part of a research project around adjustments and guardrails in retirement

18:14

um and so they were produced in a really you know a kind of a in the lab basically where everything you know is

18:20

distilled water and uh you know perfect temperature right um and and so there is something I

18:26

appreciate about this work that was done in the I think 2004 2006 2008 which is that it really it it drew the industry's

18:33

attention toward adjustments and actually coming up with systems that

18:39

could be applied in principle in a practice to help actual retirement

18:44

clients so I really appreciate that about this work um and that's one reason that I've

18:51

chosen it is that it was kind of high profile and um and has been adopted in some cases

18:56

so the approach uh of guidance Klinger we have we have here there are several articles so I've kind of taken the most

19:03

recent ones and and and you know gotten rid of some of the things from the early ones that that Guyton and Klinger also

19:09

uh kind of jettisoned along the way um but the idea is you have an initial withdrawal rate it could be anything

19:17

um and then you have a couple of ways to adjust that income

19:22

so say for example you started out your portfolio balance is a million dollars you have a five percent withdrawal rate

19:28

that's fifty thousand dollars as time goes on you keep asking okay how

19:33

much am I withdrawing and then in this case there's a um a there are a couple of variables

19:40

um that Guyton and Klinger propose um one is okay how much higher or lower

19:45

does your withdrawal rate have to get for you to make an adjustment and in this case I've gone with 20 which

19:52

is their what they do in their their articles as well and then if you hit a guardrail what do you do to that dollar

19:58

withdrawal and in this case I've gone with ten percent which is again from from their articles

20:04

um they do have a few other pieces that I we have used in this test

20:09

um one is if the trailing 12-month return is negative so skip the inflation adjustment so these are again they're

20:16

flat plans for inflation-adjusted income will stay flat but it's saying hey we'll

20:23

actually essentially will decrease our real spending um if I ever have a a trailing 12-month

20:30

period That's negative I'll just skip the inflation adjustment which has the effect of reducing my real spending and the last thing is that there are no

20:38

decreases triggered by guard rails in the final 15 years of the plan okay so

20:43

those are the four pieces of the guidance Klinger system pretty simple

20:49

um and you know because of that has been adopted in a lot of cases so let's do an example

Example

20:55

we've got let's assume it's a five percent initial withdrawal got a million dollars we're taking 50 000 a year

21:02

over time that balance is going up and down and finally it hits 1.3 million at

21:07

some point well at that point fifty thousand dollars is a about 3.8 3.9 percent withdrawal

21:17

uh that withdrawal rate is now below four percent so I'm in going to increase my withdrawal by 10 to 55 000.

21:24

time goes on further balance goes up and down eventually the balance hits 850 000

21:30

my withdrawal rates six and a half percent that's 55 000 divided by 850. and so I'm going to reduce my spend Now

21:37

by 10 you could have multiple increases you could have multiple decreases this is just the kind of the way things work

21:43

along along the way okay so what we did is um we took income

21:52

lab's new stress test feature um and this is not available uh in our

21:57

in our kind of uh normal environment so this was something that we kind of did offline

22:03

um to to test this particular guide and cleaner approach

22:08

um through some of these difficult and iconic periods in time

22:14

to see how they would have reacted how income would have changed how withdrawals would have changed how

22:19

spending would have changed over time if someone had actually followed um this this approach to guard rest and

22:26

again this is a plan that only has withdrawals so it's it's not subject to that problem that we saw before of if

22:31

you have lumpy withdrawals it just becomes nonsensical this really is a place where you could have used this

22:37

if I started those withdrawals in November of 07 just before the global

22:42

financial crisis I end up at the worst Point 28 below my

22:48

plan I have three income decreases uh before I start recovering again you'll

22:55

notice because when the plan is 15 years or shorter we're not going to see those red lines

23:00

because we don't do the guardrail-based adjustments at that point um so that's why typically toward the

23:06

end of the plan you'll only see those green lines if you're going to see them at all

23:11

so not great before the.com bubble if I started in

23:17

January of 99 I get something a bit worse I spend some time

23:22

um you know in this case multiple years uh at 36 below my originally planned

23:28

level

23:34

if I started before the great sorry after the Great Depression or after kind

23:40

of the the core part of the 2020 or 1929 Great Depression so 1936 we're still

23:46

kind of in the Great Depression I guess technically it may have ended by this point but the recovery was a while after that

23:51

um we end up 45 below plan at a kind of in the middle of the plan

24:00

and if I start in 1965 which is about the worst time you possibly could have

24:05

started um I end up at worse 54 Below plan by following this now in

24:13

all of these cases we do recover later on because we have we've paused the red lines right and we're allowing the green

24:19

lines to still happen uh the only thing that is still applying is the skipping of inflation adjustments so that is

24:25

those are continued throughout the plan but everything else is uh is uh is only

24:31

green through the end of a plan so this that the fact that when you

24:37

stress test this approach in practice with real return real you know actual historical returns actual historical

24:43

inflation that this approach doesn't work all that well it's this is not

24:49

actually new uh information so um I want to give credit to you know Wade fowe who did a lot of

24:56

testing on adjustment approaches very comprehensive survey and he had a

25:02

particular way of kind of kind of trying to measure how adjustment approaches um perform and he found that you know in

25:08

a Monte Carlo test there was at least a 10 chance following a Guyton Klinger rule of reducing withdrawals by as much

25:16

as 84 percent um throughout a 30-year plan um now that's only a 10 chance of that

25:22

but but that's really a striking amount of reduction uh I certainly wouldn't call it anything near success or what

25:29

most people um could could handle um uh yeski also noted in he has a whole

25:38

series online if you're interested um just that you know probably shouldn't expect withdrawal

25:44

regard rails to to be some kind of Miracle um so guidance cleaner does not afford you to miraculously increase your

25:51

withdrawal amount without any drawback um the higher the initial withdrawal amount the higher the risk of massive spending

25:57

cuts in the future okay so this is not a free lunch um and and that's basically what Jessica's saying here

26:04

um part of the issue with withdrawal rate guardrails is if people focus on the

26:11

withdrawal rate itself as opposed to the dollar amount being withdrawn and spent they can it can look

26:18

better than it is so you can show that someone is able to spend you know five

26:24

percent or more over the lifetime of their plan as an average withdrawal rate

26:30

um but that's sort of irrelevant what we actually want to know is what what's my income what's my spending right I can't

26:36

go to the grocery store and give them a percentage I have to give them a dollar amount and so I think that's one reason

26:41

that the weakness of these approaches hasn't been so clear as people have often focused on withdrawal rate as the

26:48

core way to measure whether a plan is good or not whereas in fact it's it's standard of living which is which is

26:55

dollar based um in particular inflation adjusted dollar based so again this has been

27:01

known for you know eight years possibly plus um and this is you know our work here is

27:08

just extending that and hopefully um you know helping people who haven't heard that

27:13

um to to see what's going on here so why does this withdrawal right guardrail

Why does this withdrawal right guardrail approach

27:20

approach you know this particular one getting cleaner not work why is it performing so poorly

27:26

in in these tough periods in history um one reason is that the guardrails

27:33

themselves are not refigured or updated as a plan goes on and the plan gets shorter so we're using the same

27:40

guardrail level throughout the plan now we're not applying the lower guard rail at the you know in the latter part of

27:46

the plan but all of the guardrail they stay the same so if I started at five percent uh and my you know my lower

27:55

guard rail the point at which I would decrease my income was full was six it's six throughout the plan

28:02

if my upper guard rail the one that allow me to spend more was four percent it's four percent throughout the plan so

28:08

nothing is being updated to reflect the fact that as people age their risks change a six percent

28:15

withdrawal is not the same for an 80 year old as it is for a 60 year old

28:21

uh these guardrails also don't incorporate the possibility of reversion to the mean so there's nothing about you

28:26

know things have gotten worse is it now equally likely that things continue to get worse or that they revert

28:33

but probably most important is that these the values for the guardrails themselves

28:40

and the adjustments um although they sound reasonable

28:46

they are in the final analysis arbitrary so that they weren't produced by really looking

28:53

at the particular plan with its plan length its resources its risks instead

28:59

they were kind of rules of thumb again for a for a research paper that's that's reasonable but we shouldn't expect that

29:07

we could take those kind of rules of thumb General hey these feel right type numbers and apply them to a real plan

29:12

and get good results so that 20 percent buffer between the guard rails or the 10

29:18

adjustment um again they don't sound unreasonable but they are in the end arbitrary

29:24

they're not linked to the particular plan and its risks um similarly skipping an inflation

29:30

adjustment it's it's easy to implement but it's arbitrary right you're you the amount

29:36

that you're reducing spending in real terms is is just based on whatever inflation

29:42

happened to be over the last 12 months so if that could have been two percent it could have been eight percent um so the adjustment itself is now

29:48

dependent on something else uh similarly skipping the decreases in the final 15 years of plan although that

29:55

may seem reasonable if that number is arbitrary should it be 15 is it that there is no risk now at in in a plan

30:02

that's shorter than 15 years our adjustments no longer important probably not it instead it's kind of an arbitrary

30:09

um piece and that I think is the core to why withdraw rate guard rails and Guyton Klinger in particular performs um so

Why not updating withdrawal rate guardrail

30:16

poorly just to go over the first one um

30:22

why not updating withdrawal rate guardrails is such a big problem

30:27

um the risk of a given withdrawal rate again in a plan with only withdrawals in

30:32

it and flat inflation-adjusted withdrawals um is not the same as you age so uh

30:39

whatever your success rate that you'd like to maintain whether it's you know 50 or 100

30:45

um you'll see that you can withdraw more and more as you age because of course your plan is getting shorter and shorter

30:51

now it's probably never zero years right we always have some life expectancy ahead of us but we can see here if we

30:58

had a six percent um reduce income uh guard rail or a four

31:03

percent increase income guardrail um as time goes on these curves are getting up into the six

31:10

percent and so I'm going to be reducing my income even if my risk hasn't changed

31:16

so maybe I want to start out at the you know the green level here the 75 success

31:21

well by the time I'm about 75 76 years old um a six percent withdrawal is is has

31:28

the same risk as my original five percent withdrawal um and so I wouldn't want to adjust at

31:33

that point because my risk hasn't actually changed um however the withdraw rate guardrails

31:38

would have me adjust um so this is a case where withdrawal regardless can actually um if they were followed

31:45

um they could cause you to make adjustments that are not necessary or even in the

31:51

opposite direction of what really would be would be right if what you're focusing on is keeping people's risk in

31:57

line so what's an alternative the alternative is to take a broader

Retirement income risk

32:04

view of risk um we call this income risk or retirement income risk it is simply what

32:11

are all of the things that could make your current spending level your current

32:18

plan for future spending unsustainable and um suggest that you might want to adjust

32:25

your spending downward that's what retirement income risk is and it can vary by plan because plans

32:32

have different lengths because people have different longevity expectations people have different resources that

32:37

they bring to retirement you might have a a pension that's not adjusted for inflation that can add inflation risk

32:43

you might have pensions that adjust if one person dies that can add mortality risk you might have delayed Social

32:51

Security which increases sequence of return risk early in the plan but decreases it later in the plan there are

32:58

so many things that can be unique to a plan that really we need a really ample

33:03

definition of income risk that takes all of this into account so it's it's really

33:08

just the chances that the current plan won't be sustainable without adjustment

33:14

um so what does that look like if we test it in the same

Riskbased guard rails

33:20

in the same periods so this we call risk-based guard rails and it is the the

33:27

approach to guard rails and the income lab software um and so although the guiding Klinger

33:32

test is not available in our software the the risk-based tests are so if we start this same plan which is a you know

33:39

accrued plan with withdrawals only nothing fancy um

33:45

it does see a reduction in the global financial crisis but at most it's three

33:50

percent below plan versus 28 below plan

Thecom bubble

33:58

for the.com bubble um with risk-based guardrails this is something we've we've seen quite a bit that really the the.com

34:05

bubble for reasonably conservative plans um rarely has an adjustment

34:12

um so although you know at the time this felt particularly bad and and certainly there were segments of the of the um of

34:19

the economy that were you know really really hurt uh in this period and it was a relatively extended period

34:26

um if you had a balanced portfolio if you were not taking excess risk if you weren't trying to push your portfolio

34:31

too hard um you you actually didn't have any adjustments uh down

34:37

and we see that here which compares

34:43

um really well to the guide and cleaner approach which which had you decreasing income through the.com bubble and

34:49

actually ending up 36 36 below plan Never Getting Back above plan

The Great Depression

34:58

uh what about this this particularly bad period in the Great Depression so starting in 1936

35:04

um one reason this is uh a little bit worse than 29 um and this is actually banging also

35:10

mentions this in his article in 94 was that there was a lot of deflation um in the in the early Great Depression

35:17

which helped a lot of plans so this skipped some of it but we still get some of the bad uh returns and so I end up

35:24

again decreasing quite a bit I'm 45 below plan with Gaiden Klinger uh I am

35:30

eight percent below plan at worst with risk-based guardrails it's worth looking at the the time

35:36

period as well now it's a relatively long period that's eight percent below plan for the risk-based guard rails but

35:42

no it's it's almost the entire plan that you're below plan for for a cut and cleaner

Stagflation

35:48

the worst performing period for both risk-based guardrails for this plan and for um Guyton Klinger is this

35:55

um stagflation ERA this is the source of the four percent rule um I purposely pushed income high enough

36:02

that um that we would see some some uh some turmoil if I kept it to four

36:08

percent obviously this would have performed a time because that's where the four percent rule came from

36:13

um and so I spend almost all the plan below plan with getting Klinger at worst 11 years

36:19

at about 54 Below less than half of my original plan spending

36:24

um it's pretty bad for risk-based guardrails too in this case three years spent at 32 below plan

36:31

um so probably in this case uh it would have been you may have been looking at some some other options

36:37

um this one is actually worth pulling up the

36:42

the details out because it's it's sort of interesting this is a little bit of a a side note but

36:48

um this period of high inflation

36:54

um is really instructive it's instructive to look at

36:59

what um what the nominal amounts

37:07

would have been uh over time because you had a lot of inflation in this period so we were

37:14

yeah this is the one um we had so much inflation

37:20

that just spending and inflation adjusted you know fixed real dollar

37:26

amount that the light blue that you see here I mean look at this from from 65 where we start with you know 35 90 until

37:35

96 would have been um it would have been 17 600. uh so and

37:43

that is the same dollar amount from from one period to the end in purchasing power terms and so uh maybe this this

37:51

gives like a slightly better view of what the actual experience would have been like for people

37:56

um so it's essentially flattening the nominal dollar amount for both for for

38:02

guide and cleaner from 69 you know roughly through 85 and then for

38:07

risk-based guard rails it's it's from about 79 to 85 that we're just sort of

38:13

going sideways with our nominal um with our nominal withdrawals so that

38:19

might give you a better feel for what that would have felt like it would have felt like your your your spendable uh withdrawals were just

38:26

stagnating um as as things got more and more expensive so that's that's what belt tightening would have looked like and I

38:33

think that's a pretty interesting you know just bit of context um

38:42

okay so why are risk-based guardrails um working better

38:48

um the the primary things are it now the

38:54

guardrails themselves are are not arbitrary they're they're based on all the risks of the plan itself at each

39:00

point in time so it incorporates the whole plan um even if it's withdrawals only it's

39:06

incorporating the whole plan including uh you know the the amount of plan that remains uh the

39:14

um the you know what's happened with the with the um you know Capital Market

39:19

assumptions economic context and so on and it's updating the guardrails along the way so it's it's much more Dynamic

39:25

they're not static guard rails um and then for a broader plan you're taking more risks into account you're

39:32

taking longevity and mortality risk inflation risk and Market risk so you're able to apply these guard rails to a

39:38

broader set of of plans um

39:44

so so that's really the the the core of why these are are performing uh so much

39:50

better um with that I know we have um some questions but I want to bring

39:56

bring Derek in here to kind of talk a little bit about uh well hey if he has any uh comments on on the materials so

40:03

far but also um I know advisors

40:08

often find withdrawal regardless really attractive because they're easy to talk about they feel intuitive and so on so I

40:14

wondered if you have some thoughts on client communication you know risk-based guardrails versus withdrawal rate guard

40:22

rails and and you know are there things I should be worried about as an advisory if I've been using withdrawal rate guardrails

40:28

yeah I think for me I try to keep it really simple I I don't I think there's always the tendency that you know a

40:34

client has read not even so much about risk-based guardrails um but you know four percent rule anything like that

40:40

where they want to get you know overly fixated on a particular distribution rate and so I like to use the graphic

40:46

Inc in income lab where I can show cash flows and I can show you know what

40:52

you're actually taking from your portfolio doesn't fit this you know similar pattern all across time unless

40:59

you know it's for some people it could be if they're particularly you know uh 72 and Beyond they're taking rmds or you

41:06

know doing all that that would be a situation where maybe that's going to be a little bit more flat distribution in

41:12

that case but for most of my clients it looks like the hatchet that we see here where there is a at least a period where

41:18

they're taking at a higher distribution rate and then of course if you're modeling in the spending smile things

41:23

like that that's going to change distribution rates as well so I like to you know explain main kind of why the

41:29

four percent rule may not be relevant to them why it might be okay that they're starting out taking five six seven eight

41:34

percent from a portfolio if they're ultimately dropping down to a lower distribution rate longer term

41:41

um and then when I talk about the guard rails I really just like to talk about risk and say you know at this point we

41:48

feel the risk has gotten too high or at this point the risk is low enough if you could go ahead and spend more and I

41:54

found that that for me has really worked well because people kind of understand okay yeah we're we're just talking about

42:00

a general risk level here um and I tend to not go deeper into the weeds and talk about what's driving that risk level

42:06

exactly what thresholds are of course I could go there with a client if they wanted to but most of the time that's

42:12

not a um a question that I'm getting from clients

Withdrawal rates

42:17

yeah I think that point about withdrawal rates is a really important one

42:22

because like you said if somebody's kind of out there you know Googling things they will probably see a lot of chatter

42:29

about that and you know we even see you know I've seen a lot of headlines in the

42:35

last year or two of you know oh maybe it's uh three and a half percent now instead of four and right it's really treated in the press as if this is the

42:43

key question about retirement planning is what initial percentage of your

42:48

portfolio can you withdraw um and I think it's really important to

42:53

find a way to talk with clients about you know in fact even in the in the software I could probably find a way to

43:00

uh I mean let me find an example here um that that is not the shape of your

43:06

withdrawals and so do you actually think that we'll be able to find a um

43:13

a a rule of thumb that would tell you exactly the withdrawal rate you should have

43:20

um right now or in the future that would apply to everyone so for example here is

43:25

you know it's not the same graph but it's it's similar right here I just have one so security that's delayed

43:32

um if I take these out you know this is what my portfolio withdrawals look like that is far from a

43:38

flat line right and so what makes me think that I would have any any rule of thumb that could apply across the the

43:45

length of this plan um and that I could apply to everyone's plan

43:50

I I've sometimes toyed with the metaphor of like a swimming pool

43:56

um you know it's deeper in some spots um but you know depending on the shape of the bottom the same amount of water

44:03

might you know fill different different types of of swimming pools you got to understand that like we're trying to

44:09

sort of this is one of your resources the yellow stuff in this case and we're sort of spreading it out filling in the

44:14

gaps where we need it and so on and and that's really realistic planning um so there's a need to step back from

44:21

just to focus on withdrawals I understand why people have that Focus because they feel like that's the one

44:26

decision they have right they they actually have some others like when to take Social Security and so on but it

44:32

does feel like oh that's the lever I can pull so I just want an easy answer to it the issue is that there just isn't one

44:37

for most people yeah and I do think um you know it certainly the point to get get across to

44:45

people is not that the initial withdrawal rate doesn't matter it does matter because even at different risk levels right that you're still moving

44:51

that initial and the long-term withdrawal rate up or down but it's just the fact that for most people it's not

44:57

this smoothed out one constant withdrawal rate all the way through retirement and when that's not the case

45:03

you really need to look at total risk and not just withdrawal rays right yeah and then I guess the the

Staying at high level

45:11

other thing I don't know if you've it sounds like you're able by just staying at high level you don't have to

45:16

you know address oh okay why would this change over time but if if someone were

45:22

particularly attached to the four percent rule for example I think you could it would be relatively intuitive

45:27

to say look you know when you're 80 you should want me to let you spend more

45:33

than four percent right you probably don't still have a 30-year plan at that point um so I think that's also kind of an

45:38

intuitive um concept that yeah their risks of different Behavior are different at

45:44

different ages yeah absolutely and I think the other confusion though that can creep in there

45:50

is when people start to think the four percent rule is talking about four percent of the portfolio about balance

45:55

every single year not at all with the four percent rule saying and saying you started out

46:01

initial four percent and then down the road that could be 10 15 particularly in

46:06

the uh the scenarios where you were depleting the portfolio right right you know very very good point

How does asset allocation affect guard rails

46:13

okay I think we have a bunch of questions I don't know if you've had a chance to

46:18

pick some out yes I have um and for our users please uh feel free

46:25

to check out the Q a upvote any of the existing questions and drop in um any of your questions as well

46:31

um the one that we have the most votes on right now uh is a question um asking how does the asset allocation

46:37

one sets up affect the guard rails and how do I explain that in plain English

46:42

to clients great question Derek do you have thoughts on that first

46:49

um I mean I think you know just generally speaking it's going to be driving the um depending on the type of analysis

46:55

you're doing it's going to be driving the market returns that you're seeing in the plan and what it's projecting out or the different risk levels

47:02

um you know how to explain that to a client I I can I like to keep things simple and

47:09

so I like to really try and set up a plan that you know matches clients portfolio or is a reasonable proxy for

47:15

what we're doing for a client and just say you know this this really reflects a portfolio like yours or a portfolio with

47:22

a similar risk level as yours just to keep it really high level and not get into the weeds of like here's how much

47:28

is allocated to large cap value versus large cap growth or anything like that

What are Klinger guardrails

47:35

yeah I think um to address that question um

47:41

with the withdrawal rate guard rails are getting Klinger guardrails in in mind that is another one of those those

47:48

issues with again these these these research pieces were put together you know in an idealized context where it's

47:54

a 30-year plan and it's a you know a particular type of uh portfolio and so on but that is one of the other problems

48:01

with actually following a guy or coming up with a guiding cleaner with raw Ray

48:07

guardrail plan is really they should be different um if you have a different portfolio

48:12

right so you don't expect as much volatility in a bond portfolio is in the stock portfolio and so on so really

48:18

these the the guardrail should be different um but it's again because it's this you

48:24

know I used 20 and 10 which is the what you'll find in those papers but they probably should be different uh in a

48:30

different one the problem is it's not entirely clear what they should be um in uh

48:38

risk-based guard rails what we're always doing is saying okay what would it take

48:45

for my current plan with my current spending level to have a higher risk

48:52

level and like an uncomfortably higher risk level and so that's done with the

48:58

Capital Market expectations in mind so if you have a less you know less return but also less volatility in your

49:03

portfolio you'll get a different answer than you will if it has uh you know more

49:09

a higher return and higher volatility portfolio so you can simply create two plans in an income level and you'll see

49:15

exactly the the difference there keeping all the other risk uh the same

49:21

you'll see how that changes uh it's similar to any other plan changes like

49:27

postponing Social Security typically increases what you can spend at a risk

49:32

level uh because you're getting that eight percent you know deferral bonus um but

49:38

it also typically tightens the guard rails at least a little bit which is

49:43

saying you have higher um sequence of return risk because you're you're stressing your portfolio more early on so total risk guardrails

49:50

have a great way of reflecting everything about the plan in that way

49:57

Okay so we've got a few different questions all around Guyton here as you can imagine so I will kind of go through

50:03

these next few ones um I love this one just so quick summary so in summary guide Klinger results and income being

50:10

too low and Legacy being too high compared to income lab results correct

50:16

yes Derek you might have more subtle or uh nuanced answer to that but

50:22

yeah I mean I you know in summary in terms of you know what you're trying to take away from it I would also take away from it that it's really not something I

50:29

would use with clients right that I don't think this is a framework that provides a uh a level of downside

50:37

adjustment risk that most clients are going to be comfortable with and so really having that conversation taking

50:43

clients through the stress tests with their particular plan seeing you know okay are they comfortable with that then

50:49

you know you've got a good plan for a client

When riskbased guardrails perform better than Klinger guardrails

50:54

I think this similar similar response here uh question is under tough times when risk-based guardrails perform

51:00

better than guy in Klinger with the gun cleaner ending portfolio be larger absolutely absolutely so it's it's

51:08

really it's always um in all these examples it's penalizing

51:13

current spending in order to protect the portfolio balance but it's doing so

51:18

unnecessarily now in this plan I had a a zero Legacy goal

51:24

um now you're never gonna you know hit a zero Legacy and have your last check bounce

51:29

um not in you know the real world where there's always you know some life expectancy um

51:35

but clearly as the plan is is farther and farther along we have more and more of an idea of how much we can spend and

51:42

still you know have a positive balance um at the end of our then current life

51:47

expectancy but that is exactly what is happening is it's it's really it's always vastly overstating the need to

51:55

keep portfolio balance in there and penalizing current spending

52:01

um now people have different ways of prioritizing Legacy and spending goals but

52:07

um I've mostly seen people prioritize spending over Legacy I don't have to argue of a similar experience

52:14

yeah I'd say for the most part you know sometimes people have a particular goal in mind what they want their legacy to

52:20

do but uh for most most of the clients I work with I'd say is more common to prioritize spending but the nice thing

52:26

of course with income lab is you could put in whatever those preferences are and you can identify any Legacy goal if

52:32

desired as part of your plan assumptions

When the final 15 years of a plan is a variable

52:38

um next question is how do you know when the final 15 years of a plan is

52:43

um life expectancy is a variable right

52:48

so in this we um so the retirement stress test and and the plan test in

52:54

income lab is a is a is a real simulation of how you would have followed a plan

53:00

down to the very details that every month you're updating life expectancy and plan length

53:06

so in this case we simply said hey if the plan at that point in time is 15 years

53:12

or shorter don't apply the the downward guardrail

53:18

um so it's pretty simple in income lab to do that that's exactly what's happening even when you're looking at risk-based guardrails it's really saying

53:24

if if you as an advisor followed this plan exactly as it's supposed to be followed including updating it every

53:31

month which happens automatically in income Lab at some point your plan would have shrunk to 15 years or less

53:37

um and so at that point we're no longer applying the the red lines here

53:44

um on the dining cleaner versus income lab comparison uh you started the in this you started at the same initial

53:50

income would the income lab calculation have recommended the initial draw rate be higher or lower

How to determine the initial withdrawal rate

53:58

um so I wouldn't say getting Klinger and maybe I'm just misremembering but I

54:03

don't I don't think there was a way to determine the initial withdrawal rate

54:08

that was kind of you know systematic or something um and so actually in the test what we

54:15

did is we took the income lab plan with ritz-based guardrails and the initial

54:20

income was based on on that risk level um which was you know I just chose the

54:26

default which is right in the middle of that risk slider which is equivalent in our world to a risk of 20 out of 100 or

54:33

you could think of it as you know it would have had an 80 probability of success if you were doing typical Monte

54:38

Carlo um and we just made guidance cleaner started the same at the same level so

54:44

that's the way that we made them apples to apples comparisons um the examples I've used here were five

54:50

percent withdrawals basically because then the 20 is an easy easy you know in your head math to go up to six or down

54:56

to four um and and there may even have been that example in in those articles but I don't

55:02

think there was anything you know particularly important about starting at

55:07

five percent of those examples Derek do you recall yeah I mean I think actually in the the

55:13

guidance cleaner paper I'm pretty sure they just give a range of potential starting distribution rates they don't

55:18

even give a there's no predefined level in in my experience seeing advisors that have spreadsheets built or other tools

55:24

that they're using they tend to just they do gravitate towards some typical starting

55:29

distribution rate that I think they prefer um yeah I think the from the methodology

55:35

as Justin described really just trying to get Apples to Apples comparison yeah I think typically if it is a

Adjusting the initial withdrawal rate

55:42

30-year plan with a 60 40 portfolio or something people like to start above four percent basically because you know

55:48

there should be some payoff to having adjustments which is well you know maybe I can at least spend a bit more sometimes than than that kind of you

55:57

know basic rule of thumb and to bring this though to you know practical context of if you are playing around

56:02

with the the sliders in income lab right that's that is something that is going to be changing if you go to a more

56:09

conservative risk setting you're going to be bringing that initial distribution down it's not going to be holding that

56:15

fixed um at any particular level and so that's part of what you see is okay yes

56:21

um you know if you're trying to compare different income lab scenarios to each other is that that initial income is

56:26

dropping to help provide more of a flattened income experience yeah that's a really good point so for

56:33

example you know the worst performing one on income lab was here you know you might say if you really think yeah a

56:40

stagflation type is uh environment is really what I wanna I wanna pay

56:45

attention to here I want to build a plan that would do well in that scenario you might say well then this is this is too

56:50

risky and so you might move that slider over and say hey let's start with let's spend a little less today so that we

56:56

know we wouldn't have to make such a huge reduction in real spending in you know the worst possible periods of

57:02

stagflation so it's just a trade-off between um spending and the volatility of the

57:08

income experience over time so you can spend more um if you can stomach volatility in bad

57:15

times um or you spend less if you'd rather you know pay the price of spending Less in

57:21

order to have less volatility in bad times and quick while we're on this kind of you know finding the guardrails that

57:27

work for a client topic the other thing I find myself using a lot now is that income ceiling uh setting where I can

57:34

actually set and say okay the client's not going to spend over a certain level because that'll help flatten out

57:40

something like this you know scenario I don't know this scenario specifically but when you do have a dip that's a

57:45

little bit too much of a dip that's another way to potentially adjust that is by adjusting the ceiling taking less

57:51

upside uh potential but then smoothing out the the downside yeah that's right

57:56

especially if there if there are increases early in the plan you know for example here we go from I think there's about you know 36 3700 a month

58:05

you know and there is a reduction which is pretty late in the plans or maybe it's not not actually that big of a deal but

58:10

um you know you might have said okay this is enough let's let's not take that next increase and so you would have

58:16

smoothed things out if you said hey there's a ceiling to how much I'm going to spend at some point I'll just let it

58:22

accrue in my portfolio all right I think we have like two

58:28

minutes left um let me see there's a few

58:36

and just a reminder I asked something on this comment um after you exit the webinar you will have a quick survey

58:41

where you can fill out the information and give us the cfpce credit um Justin I'm seeing one here just an

58:48

interesting one uh any plants include the guide and cleaner plan in income lab for modeling purposes

58:55

um no plans at the moment I mean if we had people clamoring for it I suppose we

59:00

might but it would really would be there really just as a way to show like this is probably not what you want to do so

59:05

um yeah I I don't think so at the moment but we're always listening

59:17

covers the all right we can try to address some of these uh over over email for folks

59:24

but uh right somebody did have a question I can address this one really quickly if I set

59:29

the Legacy to go to zero will there be a lot more adjustments late in life um in theory there would be like if we

59:36

knew you were dying on a certain day you could just say Okay spend the rest of your money but in these plans we're

59:41

always assuming you do have a life expectancy so you never hit the point where you are tomorrow your plan ends

59:47

and so there is you you say you know is there basically a buffer uh built in yeah essentially there's a longevity

59:53

buffer which is we're never assuming we know when you'll you'll die until you'll never spend that last dollar

59:59

and that's the same in this test for risk-based or even we did it with with Guyton cleaner so

1:00:06

yeah perfect well um thank you everyone for your amazing questions here Justin Derek

1:00:13

thank you so much for putting this together um as I mentioned um please fill out the

1:00:18

survey at the end if we didn't get to your questions we will review some of them and send you some some email responses uh to the ones that we can and

1:00:25

then please uh look out for the invites for next month's laptop Tuesday as well as next month's retirement income Intel

1:00:32

and then we will look forward to see you all on the next one thanks so much everyone have a wonderful day everybody