Sequence of Returns Risk Scenario

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
Morse Code
Posts: 192
Joined: Wed Mar 14, 2007 12:02 pm
Location: Grand Rapids, MI

Sequence of Returns Risk Scenario

Post by Morse Code »

For the purpose of my question, I'd like to make a couple assumptions:

1) I can retire when I choose, which will be when I hit 25x expenses.
2) I will likely hit 25x during a bull market when expected future equity returns will be below average and consequently, sequence of returns risk will be above average.

If these assumptions are true, wouldn't the best strategy be to remain 100% equity until the very day I retire and then trade 50% of my equity for fixed income to arrive at my desired retirement allocation of 50/50, rather than use a glide path to 50/50? Wouldn't I get to 25x faster and then protect myself somewhat from sequence of returns risk with my 50% fixed income purchase on retirement day one?
Last edited by Morse Code on Thu Apr 15, 2021 3:53 pm, edited 4 times in total.
Livin' the dream
Exchme
Posts: 389
Joined: Sun Sep 06, 2020 3:00 pm

Re: Sequence of Returns Risk Scenario

Post by Exchme »

Karsten at ERN has written extensively on SORR, SWR, glidepaths, etc. His latest article on that subject suggests doing approximately what you suggest, depending on your risk tolerance:
https://earlyretirementnow.com/2021/03/ ... s-part-43/
Prudence
Posts: 619
Joined: Fri Mar 09, 2012 4:55 pm

Re: Sequence of Returns Risk Scenario

Post by Prudence »

If you hit 25x at age 50 and live to 90 you could run out of money at 75.
Topic Author
Morse Code
Posts: 192
Joined: Wed Mar 14, 2007 12:02 pm
Location: Grand Rapids, MI

Re: Sequence of Returns Risk Scenario

Post by Morse Code »

Exchme wrote: Thu Apr 15, 2021 3:46 pm Karsten at ERN has written extensively on SORR, SWR, glidepaths, etc. His latest article on that subject suggests doing approximately what you suggest, depending on your risk tolerance:
https://earlyretirementnow.com/2021/03/ ... s-part-43/
Wow! Thank you. That article speaks to exactly what was on my mind. It's a little heavy on the math for me, but the conclusion validates that my instincts are not completely crazy.
Livin' the dream
User avatar
dogagility
Posts: 1529
Joined: Fri Feb 24, 2017 6:41 am

Re: Sequence of Returns Risk Scenario

Post by dogagility »

I was planning to do something to mitigate SORR. After reading Estrada's article, I decided the cost of insuring against SORR (by reducing equity exposure) wasn't worth it. Two reasons for this decision: being affected by SORR is unlikely and the cost to my portfolio if SORR didn't show up was significant (~15% of portfolio according to my plan).

https://papers.ssrn.com/sol3/papers.cfm ... id=3685653
All children spill milk. Learn to smile and wipe it up. -- A Farmer's Wife
grok87
Posts: 9481
Joined: Tue Feb 27, 2007 9:00 pm

Re: Sequence of Returns Risk Scenario

Post by grok87 »

Morse Code wrote: Thu Apr 15, 2021 3:40 pm For the purpose of my question, I'd like to make a couple assumptions:

1) I can retire when I choose, which will be when I hit 25x expenses.
2) I will likely hit 25x during a bull market when expected future equity returns will be below average and consequently, sequence of returns risk will be above average.

If these assumptions are true, wouldn't the best strategy be to remain 100% equity until the very day I retire and then trade 50% of my equity for fixed income to arrive at my desired retirement allocation of 50/50, rather than use a glide path to 50/50? Wouldn't I get to 25x faster and then protect myself somewhat from sequence of returns risk with my 50% fixed income purchase on retirement day one?
Here is a framework for thinking about the transition to retirement
viewtopic.php?f=10&t=311560

if one is comfortable with a 100% equities risk portfolio for a time horizon of 10+ years, then arguably on retirement one should be 82% equities 18% cash.

cheers,
grok
RIP Mr. Bogle.
User avatar
JoeRetire
Posts: 8065
Joined: Tue Jan 16, 2018 2:44 pm

Re: Sequence of Returns Risk Scenario

Post by JoeRetire »

Morse Code wrote: Thu Apr 15, 2021 3:40 pm For the purpose of my question, I'd like to make a couple assumptions:

1) I can retire when I choose, which will be when I hit 25x expenses.
2) I will likely hit 25x during a bull market when expected future equity returns will be below average and consequently, sequence of returns risk will be above average.

If these assumptions are true, wouldn't the best strategy be to remain 100% equity until the very day I retire


As long as you don't care when you retire, it might make sense. You have no assurances that you will "hit 25x" in the timeframe you desire. And being at 100% equity implies a higher level of volatility.

Your assumption about the bull market and what happens to expected future equity returns doesn't make any sense. Remember, people retire (with 25x) every day. This assumption simply cannot hold true for everyone.
It's the end of the world as we know it. | It's the end of the world as we know it. | It's the end of the world as we know it. | And I feel fine.
User avatar
Brianmcg321
Posts: 1394
Joined: Mon Jul 15, 2019 8:23 am

Re: Sequence of Returns Risk Scenario

Post by Brianmcg321 »

What if the day before you retire the market drops 30% and you no longer have 25X expenses. Will you still retire or just work another year or two until you get your money back?
Rules to investing: | 1. Don't lose money. | 2. Don't forget rule number 1.
aristotelian
Posts: 9133
Joined: Wed Jan 11, 2017 8:05 pm

Re: Sequence of Returns Risk Scenario

Post by aristotelian »

The problem is most people aren't robots programmed to retire as fast as possible. Once you get to a certain level of FI, you might want to start preserving your wealth and protecting your ability to retire on a minimal budget even if it takes a little longer to become fully FI. Or you might want to retire by a certain age rather than be completely dependent on the stock market. Or you may just not feel comfortable with losing half your life savings.
Topic Author
Morse Code
Posts: 192
Joined: Wed Mar 14, 2007 12:02 pm
Location: Grand Rapids, MI

Re: Sequence of Returns Risk Scenario

Post by Morse Code »

Brianmcg321 wrote: Thu Apr 15, 2021 5:38 pm What if the day before you retire the market drops 30% and you no longer have 25X expenses. Will you still retire or just work another year or two until you get your money back?
I would delay retirement and be grateful that it happened the day before I retired rather than the day after. :)
Livin' the dream
User avatar
1210sda
Posts: 1835
Joined: Wed Feb 28, 2007 8:31 am

Re: Sequence of Returns Risk Scenario

Post by 1210sda »

dogagility wrote: Thu Apr 15, 2021 4:16 pm I was planning to do something to mitigate SORR. After reading Estrada's article, I decided the cost of insuring against SORR (by reducing equity exposure) wasn't worth it. Two reasons for this decision: being affected by SORR is unlikely and the cost to my portfolio if SORR didn't show up was significant (~15% of portfolio according to my plan).

https://papers.ssrn.com/sol3/papers.cfm ... id=3685653
Dogagility, thanks for the link to Estrada's article.

If the cost of SORR to the portfolio does turn out to be 15%, does that mean you could reduce your SWR from 4% to 3.4% (4% times 85%) and be ok?
Topic Author
Morse Code
Posts: 192
Joined: Wed Mar 14, 2007 12:02 pm
Location: Grand Rapids, MI

Re: Sequence of Returns Risk Scenario

Post by Morse Code »

JoeRetire wrote: Thu Apr 15, 2021 5:36 pmYour assumption about the bull market and what happens to expected future equity returns doesn't make any sense. Remember, people retire (with 25x) every day. This assumption simply cannot hold true for everyone.
I think it's pretty safe to say many, if not most, people retire after a long run of good returns with high market valuations, unless they're strictly retiring when they hit a certain age like 65 when they can start collecting a pension and SS, regardless of market conditions. It certainly makes sense if you're talking about the Boglehead-type that is looking at earlier than traditional retirement. I don't know why you think this doesn't make sense.
Livin' the dream
whereskyle
Posts: 1610
Joined: Wed Jan 29, 2020 10:29 am

Re: Sequence of Returns Risk Scenario

Post by whereskyle »

Morse Code wrote: Thu Apr 15, 2021 3:40 pm For the purpose of my question, I'd like to make a couple assumptions:

1) I can retire when I choose, which will be when I hit 25x expenses.
2) I will likely hit 25x during a bull market when expected future equity returns will be below average and consequently, sequence of returns risk will be above average.

If these assumptions are true, wouldn't the best strategy be to remain 100% equity until the very day I retire and then trade 50% of my equity for fixed income to arrive at my desired retirement allocation of 50/50, rather than use a glide path to 50/50? Wouldn't I get to 25x faster and then protect myself somewhat from sequence of returns risk with my 50% fixed income purchase on retirement day one?
I think so. No glide path for me. 10-20% in long term bonds (mostly extended-duration treasuries), 20% mid/small equity tilt, the rest in total-market equities until I hit my number. Still haven't decided on FI portfolio.
"I am better off than he is – for he knows nothing and thinks that he knows. I neither know nor think that I know." - Socrates. "Nobody knows nothing." - Jack Bogle
luckyducky99
Posts: 227
Joined: Sun Dec 15, 2019 7:47 pm

Re: Sequence of Returns Risk Scenario

Post by luckyducky99 »

Morse Code wrote: Thu Apr 15, 2021 3:40 pm For the purpose of my question, I'd like to make a couple assumptions:

1) I can retire when I choose, which will be when I hit 25x expenses.
2) I will likely hit 25x during a bull market when expected future equity returns will be below average and consequently, sequence of returns risk will be above average.

If these assumptions are true, wouldn't the best strategy be to remain 100% equity until the very day I retire and then trade 50% of my equity for fixed income to arrive at my desired retirement allocation of 50/50, rather than use a glide path to 50/50? Wouldn't I get to 25x faster and then protect myself somewhat from sequence of returns risk with my 50% fixed income purchase on retirement day one?
Yes. There’s some subtlety though. You will get to 25x faster on average. But the distribution of answers for “how long until I get to 25x” is much more dispersed at 100/0. So the likelihood that you’ll get to 25x very much later than if you had a glide path is also much higher.

You’re basically swinging for the fences to try to retire earlier and like you said the odds are in your favor. But if the odds move against you, there’s a greater probability that you’ll actually get to retire later in the bad scenarios (compared to gliding). I think most people are eager to retire and conservative about that so glide to mitigate the bad-case scenarios. It’s a personal preference though.
User avatar
Watty
Posts: 22383
Joined: Wed Oct 10, 2007 3:55 pm

Re: Sequence of Returns Risk Scenario

Post by Watty »

JoeRetire wrote: Thu Apr 15, 2021 5:36 pm As long as you don't care when you retire, it might make sense. You have no assurances that you will "hit 25x" in the timeframe you desire. And being at 100% equity implies a higher level of volatility.
Another risk is that you may not be able to control when you retire if something happens like you have health problems or get laid off.

If that happens when you are 100% in equities in the middle of a bear market you could be lot of trouble.
User avatar
dogagility
Posts: 1529
Joined: Fri Feb 24, 2017 6:41 am

Re: Sequence of Returns Risk Scenario

Post by dogagility »

1210sda wrote: Thu Apr 15, 2021 6:36 pm
dogagility wrote: Thu Apr 15, 2021 4:16 pm I was planning to do something to mitigate SORR. After reading Estrada's article, I decided the cost of insuring against SORR (by reducing equity exposure) wasn't worth it. Two reasons for this decision: being affected by SORR is unlikely and the cost to my portfolio if SORR didn't show up was significant (~15% of portfolio according to my plan).

https://papers.ssrn.com/sol3/papers.cfm ... id=3685653
Dogagility, thanks for the link to Estrada's article.

If the cost of SORR to the portfolio does turn out to be 15%, does that mean you could reduce your SWR from 4% to 3.4% (4% times 85%) and be ok?
By portfolio cost, I mean how much smaller my portfolio would be after implementing a bond tent vs having not done so, should there not be any significant and sustained market decline. This is basically the insurance cost of the bond tent.

I calculated reducing equity allocation from 70 to 30% at retirement and then increasing equity exposure over 5-7 years to again reach 70%. If a sustained market decline didn't materialize during this time window, the portfolio size was reduced by about 15% at the end of the bond tent period.

That's a significant cost (e.g. $150K on a $1M starting portfolio). I didn't think the cost justified the risk... especially after reading Estrada's paper.

Instead of implementing some sort of bond tent, my plan is to be flexible in by portfolio withdrawals should a market decline materialize... something that is fundamental to withdrawal strategies like the VPW method.
All children spill milk. Learn to smile and wipe it up. -- A Farmer's Wife
User avatar
JoeRetire
Posts: 8065
Joined: Tue Jan 16, 2018 2:44 pm

Re: Sequence of Returns Risk Scenario

Post by JoeRetire »

Morse Code wrote: Thu Apr 15, 2021 6:37 pm
JoeRetire wrote: Thu Apr 15, 2021 5:36 pmYour assumption about the bull market and what happens to expected future equity returns doesn't make any sense. Remember, people retire (with 25x) every day. This assumption simply cannot hold true for everyone.
I think it's pretty safe to say many, if not most, people retire after a long run of good returns with high market valuations, unless they're strictly retiring when they hit a certain age like 65 when they can start collecting a pension and SS, regardless of market conditions. It certainly makes sense if you're talking about the Boglehead-type that is looking at earlier than traditional retirement. I don't know why you think this doesn't make sense.
People retire every single year. Not every year is after "a long run of good returns with high market valuations".
It's the end of the world as we know it. | It's the end of the world as we know it. | It's the end of the world as we know it. | And I feel fine.
grok87
Posts: 9481
Joined: Tue Feb 27, 2007 9:00 pm

Re: Sequence of Returns Risk Scenario

Post by grok87 »

Watty wrote: Thu Apr 15, 2021 7:32 pm
JoeRetire wrote: Thu Apr 15, 2021 5:36 pm As long as you don't care when you retire, it might make sense. You have no assurances that you will "hit 25x" in the timeframe you desire. And being at 100% equity implies a higher level of volatility.
Another risk is that you may not be able to control when you retire if something happens like you have health problems or get laid off.

If that happens when you are 100% in equities in the middle of a bear market you could be lot of trouble.
yep
RIP Mr. Bogle.
Topic Author
Morse Code
Posts: 192
Joined: Wed Mar 14, 2007 12:02 pm
Location: Grand Rapids, MI

Re: Sequence of Returns Risk Scenario

Post by Morse Code »

grok87 wrote: Fri Apr 16, 2021 7:33 am
Watty wrote: Thu Apr 15, 2021 7:32 pm
JoeRetire wrote: Thu Apr 15, 2021 5:36 pm As long as you don't care when you retire, it might make sense. You have no assurances that you will "hit 25x" in the timeframe you desire. And being at 100% equity implies a higher level of volatility.
Another risk is that you may not be able to control when you retire if something happens like you have health problems or get laid off.

If that happens when you are 100% in equities in the middle of a bear market you could be lot of trouble.
yep
That's why I started with assumption #1 in the original post, that I can retire when I choose.
Livin' the dream
qwertyjazz
Posts: 1934
Joined: Tue Feb 23, 2016 4:24 am

Re: Sequence of Returns Risk Scenario

Post by qwertyjazz »

There are two timing risks - either of which you can optimize if you can ignore the other
Stock return sequence
Employability as you age

I have yet to see a good argument looking at how you optimize the joint issue of both. There are plenty of quick thoughts in them and probable trends. I just wonder if adding the complexity of not knowing when you may be forced to retire and stock market swings can even be combined. If not, then is just optimizing on one a suboptimal strategy to a vaguer choice? If I pick the optimal AA for one, will that be worse than just somewhere in between?
G.E. Box "All models are wrong, but some are useful."
carolinaman
Posts: 4579
Joined: Wed Dec 28, 2011 9:56 am
Location: North Carolina

Re: Sequence of Returns Risk Scenario

Post by carolinaman »

If 25X expenses is your goal for retirement, I would suggest building a buffer (for example maybe 27X) so you are better protected from SORR, unexpected events or bad timing.
Random Walker
Posts: 4860
Joined: Fri Feb 23, 2007 8:21 pm

Re: Sequence of Returns Risk Scenario

Post by Random Walker »

Morse Code wrote: Thu Apr 15, 2021 3:40 pm For the purpose of my question, I'd like to make a couple assumptions:

1) I can retire when I choose, which will be when I hit 25x expenses.
2) I will likely hit 25x during a bull market when expected future equity returns will be below average and consequently, sequence of returns risk will be above average.

If these assumptions are true, wouldn't the best strategy be to remain 100% equity until the very day I retire and then trade 50% of my equity for fixed income to arrive at my desired retirement allocation of 50/50, rather than use a glide path to 50/50? Wouldn't I get to 25x faster and then protect myself somewhat from sequence of returns risk with my 50% fixed income purchase on retirement day one?
I agree a lot with what your thinking. This post I think directly addresses your issue.
viewtopic.php?t=236967
And more importantly, read the end of William Bernstein’s short ebook on Lifecycle Investing!

https://www.amazon.com/Ages-Investor-Cr ... B008CM2T2A




Dave
Random Walker
Posts: 4860
Joined: Fri Feb 23, 2007 8:21 pm

Re: Sequence of Returns Risk Scenario

Post by Random Walker »

Morse Code wrote: Thu Apr 15, 2021 6:37 pm
JoeRetire wrote: Thu Apr 15, 2021 5:36 pmYour assumption about the bull market and what happens to expected future equity returns doesn't make any sense. Remember, people retire (with 25x) every day. This assumption simply cannot hold true for everyone.
I think it's pretty safe to say many, if not most, people retire after a long run of good returns with high market valuations, unless they're strictly retiring when they hit a certain age like 65 when they can start collecting a pension and SS, regardless of market conditions. It certainly makes sense if you're talking about the Boglehead-type that is looking at earlier than traditional retirement. I don't know why you think this doesn't make sense.
Murphy’s Law Of Retirement! Big bull market, portfolios swell, people retire when valuations high, future expected returns lower, and whole potential future dispersion of returns shifts left.
viewtopic.php?t=220275

Dave
Random Walker
Posts: 4860
Joined: Fri Feb 23, 2007 8:21 pm

Re: Sequence of Returns Risk Scenario

Post by Random Walker »

And SORR is a strong reason to diversify across unique and independent sources of risk/return: market beta, term, size, value, Int, momentum, and maybe Alts.

Dave
User avatar
vanbogle59
Posts: 150
Joined: Wed Mar 10, 2021 8:30 pm

Re: Sequence of Returns Risk Scenario

Post by vanbogle59 »

Morse Code wrote: Thu Apr 15, 2021 3:40 pm For the purpose of my question, I'd like to make a couple assumptions:

1) I can retire when I choose, which will be when I hit 25x expenses.
2) I will likely hit 25x during a bull market when expected future equity returns will be below average and consequently, sequence of returns risk will be above average.

If these assumptions are true, wouldn't the best strategy be to remain 100% equity until the very day I retire and then trade 50% of my equity for fixed income to arrive at my desired retirement allocation of 50/50, rather than use a glide path to 50/50? Wouldn't I get to 25x faster and then protect myself somewhat from sequence of returns risk with my 50% fixed income purchase on retirement day one?
Are you trying to hit 25X as quickly as possible?
Or are you trying to have the highest odds of hitting 25X?

I am at 20X. If I were to go to 100/0, I would increase the odds that I would hit 25X in 2 years.
I would also increase the odds that I wouldn't hit it for another 10.
User avatar
JoeRetire
Posts: 8065
Joined: Tue Jan 16, 2018 2:44 pm

Re: Sequence of Returns Risk Scenario

Post by JoeRetire »

Morse Code wrote: Fri Apr 16, 2021 7:44 am
grok87 wrote: Fri Apr 16, 2021 7:33 am
Watty wrote: Thu Apr 15, 2021 7:32 pm
JoeRetire wrote: Thu Apr 15, 2021 5:36 pm As long as you don't care when you retire, it might make sense. You have no assurances that you will "hit 25x" in the timeframe you desire. And being at 100% equity implies a higher level of volatility.
Another risk is that you may not be able to control when you retire if something happens like you have health problems or get laid off.

If that happens when you are 100% in equities in the middle of a bear market you could be lot of trouble.
yep
That's why I started with assumption #1 in the original post, that I can retire when I choose.
Not exactly. You wrote "I can retire when I choose, which will be when I hit 25x expenses."

But "when I hit 25x expenses" is not under your control. You can choose the triggering event, but you cannot choose the timeframe in which it occurs (if it occurs at all).
It's the end of the world as we know it. | It's the end of the world as we know it. | It's the end of the world as we know it. | And I feel fine.
aristotelian
Posts: 9133
Joined: Wed Jan 11, 2017 8:05 pm

Re: Sequence of Returns Risk Scenario

Post by aristotelian »

luckyducky99 wrote: Thu Apr 15, 2021 7:31 pm Yes. There’s some subtlety though. You will get to 25x faster on average. But the distribution of answers for “how long until I get to 25x” is much more dispersed at 100/0. So the likelihood that you’ll get to 25x very much later than if you had a glide path is also much higher.

You’re basically swinging for the fences to try to retire earlier and like you said the odds are in your favor. But if the odds move against you, there’s a greater probability that you’ll actually get to retire later in the bad scenarios (compared to gliding). I think most people are eager to retire and conservative about that so glide to mitigate the bad-case scenarios. It’s a personal preference though.
I'm not sure that's true. If you are say 90% of the way to FI number, a stock crash would set you back significantly whether you are 60% stock or 90% stock. The only way you could still hit your number quickly without the market bouncing back would be if you are like 0-10% stock. I wonder how many real scenarios there have been where a 60/40 investor actually retires faster than 90/10 investor after a stock crash. It would be the rare scenario where the market stays down long enough that new contributions make up for the difference.

IMO the better argument for bond allocation is to protect against downside risk, not to allow you to retire faster.
Topic Author
Morse Code
Posts: 192
Joined: Wed Mar 14, 2007 12:02 pm
Location: Grand Rapids, MI

Re: Sequence of Returns Risk Scenario

Post by Morse Code »

JoeRetire wrote: Fri Apr 16, 2021 10:14 am
Morse Code wrote: Fri Apr 16, 2021 7:44 am
grok87 wrote: Fri Apr 16, 2021 7:33 am
Watty wrote: Thu Apr 15, 2021 7:32 pm
JoeRetire wrote: Thu Apr 15, 2021 5:36 pm As long as you don't care when you retire, it might make sense. You have no assurances that you will "hit 25x" in the timeframe you desire. And being at 100% equity implies a higher level of volatility.
Another risk is that you may not be able to control when you retire if something happens like you have health problems or get laid off.

If that happens when you are 100% in equities in the middle of a bear market you could be lot of trouble.
yep
That's why I started with assumption #1 in the original post, that I can retire when I choose.
Not exactly. You wrote "I can retire when I choose, which will be when I hit 25x expenses."

But "when I hit 25x expenses" is not under your control. You can choose the triggering event, but you cannot choose the timeframe in which it occurs (if it occurs at all).
I think you're making a distinction without a difference. Retiring when I choose means I can retire at 20x or 25x or 50x. Of course, I can't control when I hit those milestones, but I can still choose when to retire. Perhaps you would prefer, "I will not be forced to retire by loss of job or disability." :)
Livin' the dream
User avatar
JoeRetire
Posts: 8065
Joined: Tue Jan 16, 2018 2:44 pm

Re: Sequence of Returns Risk Scenario

Post by JoeRetire »

Morse Code wrote: Fri Apr 16, 2021 12:33 pm
JoeRetire wrote: Fri Apr 16, 2021 10:14 am
Morse Code wrote: Fri Apr 16, 2021 7:44 am
grok87 wrote: Fri Apr 16, 2021 7:33 am
Watty wrote: Thu Apr 15, 2021 7:32 pm

Another risk is that you may not be able to control when you retire if something happens like you have health problems or get laid off.

If that happens when you are 100% in equities in the middle of a bear market you could be lot of trouble.
yep
That's why I started with assumption #1 in the original post, that I can retire when I choose.
Not exactly. You wrote "I can retire when I choose, which will be when I hit 25x expenses."

But "when I hit 25x expenses" is not under your control. You can choose the triggering event, but you cannot choose the timeframe in which it occurs (if it occurs at all).
I think you're making a distinction without a difference. Retiring when I choose means I can retire at 20x or 25x or 50x. Of course, I can't control when I hit those milestones, but I can still choose when to retire. Perhaps you would prefer, "I will not be forced to retire by loss of job or disability." :)
Perhaps. But "I can retire at 20x or 25x or 50x" seems to conflict with "when I hit 25x expenses". You could choose to retire today, or you could choose never to retire. And of course some have life events after which they aren't physically or mentally able to work, and some call that "retiring".

I think "I have chosen to retire as soon as I hit 25x expenses" would more clearly express your intent? Or do you mean something else?

The calendar-when part of that just happens when it happens. Maybe never. That part isn't within your control.
It's the end of the world as we know it. | It's the end of the world as we know it. | It's the end of the world as we know it. | And I feel fine.
tibbitts
Posts: 13585
Joined: Tue Feb 27, 2007 6:50 pm

Re: Sequence of Returns Risk Scenario

Post by tibbitts »

Brianmcg321 wrote: Thu Apr 15, 2021 5:38 pm What if the day before you retire the market drops 30% and you no longer have 25X expenses. Will you still retire or just work another year or two until you get your money back?
Or another twenty years, as the case may be.
tibbitts
Posts: 13585
Joined: Tue Feb 27, 2007 6:50 pm

Re: Sequence of Returns Risk Scenario

Post by tibbitts »

Morse Code wrote: Thu Apr 15, 2021 3:40 pm If these assumptions are true, wouldn't the best strategy be to remain 100% equity until the very day I retire and then trade 50% of my equity for fixed income to arrive at my desired retirement allocation of 50/50, rather than use a glide path to 50/50? Wouldn't I get to 25x faster and then protect myself somewhat from sequence of returns risk with my 50% fixed income purchase on retirement day one?
It depends partly on what your fixed income choices are. I wouldn't go 100% equity with a portion of my fixed income at a 7% guaranteed for example. I think this really depends on real fixed rates at the time, and certainly they're effectively negative now in most cases. Still you have to weigh the risk of almost never getting to your number at 100% equities against the probability of getting there sooner.
luckyducky99
Posts: 227
Joined: Sun Dec 15, 2019 7:47 pm

Re: Sequence of Returns Risk Scenario

Post by luckyducky99 »

aristotelian wrote: Fri Apr 16, 2021 10:16 am
luckyducky99 wrote: Thu Apr 15, 2021 7:31 pm Yes. There’s some subtlety though. You will get to 25x faster on average. But the distribution of answers for “how long until I get to 25x” is much more dispersed at 100/0. So the likelihood that you’ll get to 25x very much later than if you had a glide path is also much higher.

You’re basically swinging for the fences to try to retire earlier and like you said the odds are in your favor. But if the odds move against you, there’s a greater probability that you’ll actually get to retire later in the bad scenarios (compared to gliding). I think most people are eager to retire and conservative about that so glide to mitigate the bad-case scenarios. It’s a personal preference though.
I'm not sure that's true. If you are say 90% of the way to FI number, a stock crash would set you back significantly whether you are 60% stock or 90% stock. The only way you could still hit your number quickly without the market bouncing back would be if you are like 0-10% stock. I wonder how many real scenarios there have been where a 60/40 investor actually retires faster than 90/10 investor after a stock crash. It would be the rare scenario where the market stays down long enough that new contributions make up for the difference.

IMO the better argument for bond allocation is to protect against downside risk, not to allow you to retire faster.
Maybe you're right. Probably the math is more subtle than I realize. It is hard I think because stocks go down faster in crashes than they go up in good times. Maybe there is some kind of equilibrium allocation where the more drastic decline in an equity-heavy portfolio is exactly of offset by its greater expected return during recovery, and moving the allocation either way from there changes the outcome for the slower or faster. At one extreme is your example 0-10% stocks portfolio. I wonder where the other extreme is. I'm not smart or knowledgeable enough to figure out.

But if not that, and if any downside is made up for by the better return of an equity heavy portfolio, what do you mean by "protect against downside risk"?
User avatar
vineviz
Posts: 9946
Joined: Tue May 15, 2018 1:55 pm

Re: Sequence of Returns Risk Scenario

Post by vineviz »

Morse Code wrote: Thu Apr 15, 2021 4:08 pm
Exchme wrote: Thu Apr 15, 2021 3:46 pm Karsten at ERN has written extensively on SORR, SWR, glidepaths, etc. His latest article on that subject suggests doing approximately what you suggest, depending on your risk tolerance:
https://earlyretirementnow.com/2021/03/ ... s-part-43/
Wow! Thank you. That article speaks to exactly what was on my mind. It's a little heavy on the math for me, but the conclusion validates that my instincts are not completely crazy.
If that’s the case, either the article is wrong or something got lost in the interpretation.

An immediate and sudden shift from 100% stocks to 50% stocks is absolutely a suboptimal strategy. You might get lucky, but it reduces the odds of accomplishing a successful retirement relative to pursuing a more gradual glide path.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
aristotelian
Posts: 9133
Joined: Wed Jan 11, 2017 8:05 pm

Re: Sequence of Returns Risk Scenario

Post by aristotelian »

luckyducky99 wrote: Fri Apr 16, 2021 8:45 pm But if not that, and if any downside is made up for by the better return of an equity heavy portfolio, what do you mean by "protect against downside risk"?
Nothing deep. Retiring as quickly as possible at 100% of FI number isn't the only goal. You may want to keep X years in expenses in case you lose your job, or you may be 90% to FI and want to preserve the ability to partially retire any time with at least 70% of your budget. Depending on the downside risks you want to protect against, a substantial bond allocation would make sense even if it means you will likely take a little longer to hit your 100% FI number.
User avatar
vanbogle59
Posts: 150
Joined: Wed Mar 10, 2021 8:30 pm

Re: Sequence of Returns Risk Scenario

Post by vanbogle59 »

luckyducky99 wrote: Fri Apr 16, 2021 8:45 pm ...

what do you mean by "protect against downside risk"?

OK. I know I am old. but no one should be allowed to be that young.
User avatar
Beensabu
Posts: 751
Joined: Sun Aug 14, 2016 3:22 pm

Re: Sequence of Returns Risk Scenario

Post by Beensabu »

vanbogle59 wrote: Fri Apr 16, 2021 10:55 pm
luckyducky99 wrote: Fri Apr 16, 2021 8:45 pm ...

what do you mean by "protect against downside risk"?

OK. I know I am old. but no one should be allowed to be that young.
This is the best thing from today. Thank you :)

You need more of the quote for full impact:
luckyducky99 wrote:...if any downside is made up for by the better return of an equity heavy portfolio, what do you mean by "protect against downside risk"?
From what I can tell in general (not OP-specific), it's less age-related and more a "state of mind" kind of thing.
"The only thing that makes life possible is permanent, intolerable uncertainty; not knowing what comes next."
Random Walker
Posts: 4860
Joined: Fri Feb 23, 2007 8:21 pm

Re: Sequence of Returns Risk Scenario

Post by Random Walker »

luckyducky99 wrote: Fri Apr 16, 2021 8:45 pm
aristotelian wrote: Fri Apr 16, 2021 10:16 am
luckyducky99 wrote: Thu Apr 15, 2021 7:31 pm Yes. There’s some subtlety though. You will get to 25x faster on average. But the distribution of answers for “how long until I get to 25x” is much more dispersed at 100/0. So the likelihood that you’ll get to 25x very much later than if you had a glide path is also much higher.

You’re basically swinging for the fences to try to retire earlier and like you said the odds are in your favor. But if the odds move against you, there’s a greater probability that you’ll actually get to retire later in the bad scenarios (compared to gliding). I think most people are eager to retire and conservative about that so glide to mitigate the bad-case scenarios. It’s a personal preference though.
I'm not sure that's true. If you are say 90% of the way to FI number, a stock crash would set you back significantly whether you are 60% stock or 90% stock. The only way you could still hit your number quickly without the market bouncing back would be if you are like 0-10% stock. I wonder how many real scenarios there have been where a 60/40 investor actually retires faster than 90/10 investor after a stock crash. It would be the rare scenario where the market stays down long enough that new contributions make up for the difference.

IMO the better argument for bond allocation is to protect against downside risk, not to allow you to retire faster.
Maybe you're right. Probably the math is more subtle than I realize. It is hard I think because stocks go down faster in crashes than they go up in good times. Maybe there is some kind of equilibrium allocation where the more drastic decline in an equity-heavy portfolio is exactly of offset by its greater expected return during recovery, and moving the allocation either way from there changes the outcome for the slower or faster. At one extreme is your example 0-10% stocks portfolio. I wonder where the other extreme is. I'm not smart or knowledgeable enough to figure out.

But if not that, and if any downside is made up for by the better return of an equity heavy portfolio, what do you mean by "protect against downside risk"?
I think you are alluding to the concept of portfolio efficiency. If a portfolio loses 50%, it needs to gain 100% to break even. If a portfolio loses 50% year one and gains 100% year 2, its simple average return is 25% buts its compounded return (what we eat) is 0%. If a portfolio loses 33%, it needs to make 50% to break even. If a portfolio loses 25%, it needs to make 33% to break even. This adverse effect of volatility, the difference between simple average return and geometric (compounded) return is called variance drain or volatility drag. The reason it occurs is that after a big loss the portfolio has less money present to participate in the subsequent gain. There is more money present for reinvestment. This is separate from any potential rebalancing bonus, which some people believe exists and some don’t. If two portfolios have the same simple mean average return over time, but different volatilities, the less volatile portfolio will accumulate more money.
So one goal in portfolio construction is to increase efficiency, bring the geometric return closer to the simple average return. This is achieved by combining less than perfectly correlated assets with hopefully high expected returns. Starting with an all equity portfolio, the best and cheapest diversifier is simple bonds. From there marginal improvements come at increasing marginal cost and decreasing marginal benefit. It’s up to the individual investor to decide where to draw the line. The path towards increased diversification, starting with TSM/TBM portfolio, includes international equity, tilting to size and value, adding other factors such as momentum and profitability, possibly adding styles like carry and defensive, increasing safe bond exposure in presence of higher risk equities, and possibly alternative investments. Marginal improvements in portfolio efficiency come at increasing marginal costs. Costs are certain and improvements are only potential. Over the last 10-15 years simpler and cheaper has blown away anything more complex. Who knows what the future holds. But there is big benefit to minimizing big losses.

Dave
Random Walker
Posts: 4860
Joined: Fri Feb 23, 2007 8:21 pm

Re: Sequence of Returns Risk Scenario

Post by Random Walker »

vineviz wrote: Fri Apr 16, 2021 9:16 pm
Morse Code wrote: Thu Apr 15, 2021 4:08 pm
Exchme wrote: Thu Apr 15, 2021 3:46 pm Karsten at ERN has written extensively on SORR, SWR, glidepaths, etc. His latest article on that subject suggests doing approximately what you suggest, depending on your risk tolerance:
https://earlyretirementnow.com/2021/03/ ... s-part-43/
Wow! Thank you. That article speaks to exactly what was on my mind. It's a little heavy on the math for me, but the conclusion validates that my instincts are not completely crazy.
If that’s the case, either the article is wrong or something got lost in the interpretation.

An immediate and sudden shift from 100% stocks to 50% stocks is absolutely a suboptimal strategy. You might get lucky, but it reduces the odds of accomplishing a successful retirement relative to pursuing a more gradual glide path.
Hi Vineviz, why do you say this? Simply because more gradual could mean longer time with greater exposure to the high expected return equity asset class? To me it makes sense to take the William Bernstein approach: when you hit your number, lock it in. When equities have an SD in the neighborhood of 15-20%, it seems to me very rational to make a definitive move when the opportunity arises. Now of course I did this in about 2017, and have way less money than I would otherwise :-). But alternative histories can play out.

Dave
User avatar
watchnerd
Posts: 8551
Joined: Sat Mar 03, 2007 11:18 am
Location: Seattle, WA, USA

Re: Sequence of Returns Risk Scenario

Post by watchnerd »

Random Walker wrote: Fri Apr 16, 2021 9:37 am

Murphy’s Law Of Retirement! Big bull market, portfolios swell, people retire when valuations high, future expected returns lower, and whole potential future dispersion of returns shifts left.
viewtopic.php?t=220275

Dave
I hope to retire in a deep recession.

I have my fingers crossed for 2025.
60% Global Market Stocks (VT,FM) | 38% Global Market Bonds | 2% Alts || LMP TIPS/STRIPS || RSU + ESPP
User avatar
vineviz
Posts: 9946
Joined: Tue May 15, 2018 1:55 pm

Re: Sequence of Returns Risk Scenario

Post by vineviz »

Random Walker wrote: Sat Apr 17, 2021 11:20 am Hi Vineviz, why do you say this? Simply because more gradual could mean longer time with greater exposure to the high expected return equity asset class? To me it makes sense to take the William Bernstein approach: when you hit your number, lock it in. When equities have an SD in the neighborhood of 15-20%, it seems to me very rational to make a definitive move when the opportunity arises. Now of course I did this in about 2017, and have way less money than I would otherwise :-). But alternative histories can play out.
Because we can never know the exact path the future will take, the best strategy must employ at least some elements of maximizing the likelihood of success.

The strategy you attribute to Bernstein is directionally correct, but it's dangerous to take the advice too literally. It's also dangerous to wait for an "opportunity to arise" for two reasons. First, "opportunity" is entirely subjective (and subject to redefinition under emotional duress). Second, the "opportunity" may not arise when you need it to.

For example, take the hypothetical case of a worker who starts saving at age 35 in 1977. The graph below shows the portfolio balance each year as "number of years of expenses", aka "the number". It's not necessarily a representative example, but it does show the dangers of making a hard shift from 100% stocks to bonds.

Image

The blue line shows the scenario in which the investor stays 100% stocks until they "hit the number", which I'm showing as 25x expenses. I'd describe this path as frustrating. They got close at age 57 in 1999, only to watch retirement slip away. They got close again at age 65 in 2007, but it slipped away again. They finally crossed the line in 2013 at age 71.

Contrast that to the grey line, which shows a scenario in which the investor listens to the economists and uses a glide path from stocks to bonds. Like the blue line investor they start at 100% stocks, but starting at age 55 they start gliding down to a 40/60 portfolio by age 61. Like the blue line investor, they miss the early retirement at age 57 in 1999. But not only do they reach the 25x finish line a full four years earlier (age 67, in 2009) than the blue line investor, if for some reason they had been FORCED to retire at any time after 1999 they would have been able to do so with much more of their life savings. About 20% more, on average.

Again, we almost never get the optimal result (except by chance) because the future is uncertain. So we are left to aim for the optimally sub-optimal result, which in this context means some sort of glide path from 100%+ in equities early in the accumulation phase to some allocation (determined by risk profile) of stocks and/or bonds at retirement. Almost never does that optimally sub-optimal path involve a sharp transition from 100% stocks to 50% stocks in a single moment.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Topic Author
Morse Code
Posts: 192
Joined: Wed Mar 14, 2007 12:02 pm
Location: Grand Rapids, MI

Re: Sequence of Returns Risk Scenario

Post by Morse Code »

I'm having trouble understanding what is sub-optimal about a strategy to convert from 100% equity to a retirement allocation (50/50 for example) in one fell swoop once a desired balance is reached.

Let's say it's April 30th, 2022 and the market has been on a tear for two years now, so I've hit my number (25x or whatever). Everyone thinks the market is over valued, so people are speculating here and everywhere that the bubble has to pop soon, but, of course, nobody knows when. It could happen tomorrow or the bull could rage for another year or more. This is SORR the way I understand it. If I make the decision to retire, I could be retiring into a bear market right at the get go. So the way to mitigate SORR is to move 50% of my retirement portfolio to bonds on April 30th and punch out of the work force.

If the bubble had popped earlier or there had been no bull market at all, but an extended bear market instead, well then I would not have reached my number on April 30th and would continue to work until I did. How is this worse than a multi-decade glide path that almost guarantees a delayed retirement versus what I am suggesting? How does a glide path mitigate SORR any better? It seems to me the SORR is the same whether I arrived at 50/50 over several decades or in one day.
Livin' the dream
User avatar
watchnerd
Posts: 8551
Joined: Sat Mar 03, 2007 11:18 am
Location: Seattle, WA, USA

Re: Sequence of Returns Risk Scenario

Post by watchnerd »

Morse Code wrote: Thu Apr 29, 2021 10:49 am I'm having trouble understanding what is sub-optimal about a strategy to convert from 100% equity to a retirement allocation (50/50 for example) in one fell swoop once a desired balance is reached.
Well, that's directionally aligned, but not tactically, with what Bernstein says.

Bernstein says to lock in your number, once you hit it, by putting your essential living expenses an LMP portfolio -- annuity or non-rolling TIPS/bond ladder being the usual choices.

That's not necessarily tactically the same as going 50/50 or using a rising equity glide path, but they're all attempts to de-risk equities and reduce SOR risk.
60% Global Market Stocks (VT,FM) | 38% Global Market Bonds | 2% Alts || LMP TIPS/STRIPS || RSU + ESPP
User avatar
vineviz
Posts: 9946
Joined: Tue May 15, 2018 1:55 pm

Re: Sequence of Returns Risk Scenario

Post by vineviz »

Scenario: it’s April 21st, 2022 and you’re ALMOST at your number (say 24.98 x or whatever).

Stocks drop by 50% in a week and you lose your job.

Now what?

Glide paths are the conventional wisdom because they actually solve the problem people face: accumulating enough in the face of an uncertain future.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 9946
Joined: Tue May 15, 2018 1:55 pm

Re: Sequence of Returns Risk Scenario

Post by vineviz »

Also note that for most investors the glide path actually shouldn’t be multi-decade in length. Ten years is usually plenty.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Topic Author
Morse Code
Posts: 192
Joined: Wed Mar 14, 2007 12:02 pm
Location: Grand Rapids, MI

Re: Sequence of Returns Risk Scenario

Post by Morse Code »

vineviz wrote: Thu Apr 29, 2021 11:07 am Scenario: it’s April 21st, 2022 and you’re ALMOST at your number (say 24.98 x or whatever).

Stocks drop by 50% in a week and you lose your job.

Now what?

Glide paths are the conventional wisdom because they actually solve the problem people face: accumulating enough in the face of an uncertain future.
I already addressed what happens if the bubble bursts before I hit my number. I'm specifically talking about SORR here and losing my job is a different kind of risk we face every day, not just prior to retirement. We mitigate job loss risk in other ways like un-employment insurance, emergency funds, skill/education diversification, etc. In the OP I said to assume I can choose when to retire.
Last edited by Morse Code on Thu Apr 29, 2021 11:18 am, edited 2 times in total.
Livin' the dream
lostdog
Posts: 4044
Joined: Thu Feb 04, 2016 2:15 pm

Re: Sequence of Returns Risk Scenario

Post by lostdog »

I got lucky with 100/0 for most of my investing career. Moved to 80/20 and 60/40 very quickly when we hit 22x expenses. 60/40 can get us 3x expenses with less risk.
Topic Author
Morse Code
Posts: 192
Joined: Wed Mar 14, 2007 12:02 pm
Location: Grand Rapids, MI

Re: Sequence of Returns Risk Scenario

Post by Morse Code »

vineviz wrote: Thu Apr 29, 2021 11:12 am Also note that for most investors the glide path actually shouldn’t be multi-decade in length. Ten years is usually plenty.
This still doesn't address how a glide path mitigates sequence of returns risk better than a fell swoop.
Livin' the dream
User avatar
vineviz
Posts: 9946
Joined: Tue May 15, 2018 1:55 pm

Re: Sequence of Returns Risk Scenario

Post by vineviz »

Morse Code wrote: Thu Apr 29, 2021 11:16 am
vineviz wrote: Thu Apr 29, 2021 11:07 am Scenario: it’s April 21st, 2022 and you’re ALMOST at your number (say 24.98 x or whatever).

Stocks drop by 50% in a week and you lose your job.

Now what?

Glide paths are the conventional wisdom because they actually solve the problem people face: accumulating enough in the face of an uncertain future.
I already addressed what happens if the bubble bursts before I hit my number. I'm specifically talking about SORR here and losing my job is a different kind of risk we face every day, not just prior to retirement. We mitigate job loss risk in other ways like un-employment insurance, emergency funds, skill/education diversification, etc. In the OP I said to assume I can choose when to retire.

If you start with an assumption that SORR can’t exist then you’ll have a hard time explaining strategies that reduce the impact of SORR.

If you’re willing and able to work forever, and you don’t have a preference for retiring earlier rather than later, the utility of a glide path probably isn’t very high.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vineviz
Posts: 9946
Joined: Tue May 15, 2018 1:55 pm

Re: Sequence of Returns Risk Scenario

Post by vineviz »

Morse Code wrote: Thu Apr 29, 2021 11:20 am
vineviz wrote: Thu Apr 29, 2021 11:12 am Also note that for most investors the glide path actually shouldn’t be multi-decade in length. Ten years is usually plenty.
This still doesn't address how a glide path mitigates sequence of returns risk better than a fell swoop.

It does this because SORR isn’t only a risk AFTER retirement. It’s a risk before retirement too, and most people prefer not to expose themselves to this risk more than is necessary.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
Ben Mathew
Posts: 1575
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle

Re: Sequence of Returns Risk Scenario

Post by Ben Mathew »

Morse Code wrote: Thu Apr 29, 2021 11:20 am
vineviz wrote: Thu Apr 29, 2021 11:12 am Also note that for most investors the glide path actually shouldn’t be multi-decade in length. Ten years is usually plenty.
This still doesn't address how a glide path mitigates sequence of returns risk better than a fell swoop.
The problem with sudden changes in glide path is that you are taking very different risk on two different days. This results in poor time diversification. Holding 50% stocks over two days is better than holding 100% stocks on one day and 0% stocks on the other day. You get the same expected return, but lower risk because of you're averaging across the returns of two days. It's similar to diversifying and averaging across assets. So switching from 100% stocks one day to 50% stocks the next means you took on too much risk @ 100% and too little @ 50%.

But you've described an extenuating circumstance here where you say you are very risk tolerant prior to retirement and very risk averse right after leaving the job. So that can argue for a sudden drop in AA at retirement. You would have to balance that against the time diversification benefits of a more consistent glidepath.
Post Reply