Never Ever Rebalance Bonds into Stocks?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Blue456
Posts: 1018
Joined: Tue Jun 04, 2019 5:46 am

Re: Never Ever Rebalance Bonds into Stocks?

Post by Blue456 »

aj76er wrote: Thu Sep 24, 2020 9:37 am
Blue456 wrote: Tue Sep 22, 2020 3:51 pm
Leesbro63 wrote: Tue Sep 22, 2020 2:19 pm During 2008-9, I came to the realization that in a Japan-type scenario, rebalancing a mature portfolio (I was age 49 then; I'm 60 now) from bonds into stocks might be a "Pascal's Wager". If you're wrong and stocks keeping going down for a long time, you effectively "flush" good money into bad money that might never recover. Thus negating the reason for "safe" money in the first place. And I amended my investment plan accordingly so that I would never again rebalance from bonds to stocks.

Now, as I get even closer to retirement age, I am wondering what happens if stocks crash and stay down for a long period of time. It might be a "you're dead already" scenario if you DO NOT rebalance into stocks after a crash. In other words, what happens when stocks crash just before or after retirement and we have a 1966-1981 period?

How does one reconcile the "Pascal's Wager" problem with the "you're dead already" problem? n
1) Have two portfolios:
Portfolio one. I-bonds, TIPS, CDs and other safe income that covers your very basic expenses for 5, 10, 15, 20 years. You draw the line how much makes you comfortable.
Portfolio two. 100/0, 90/10, 80/20, 70/30, 60/40.... etc.
Don’t rebalance from portfolio 2 to portfolio 1. Do rebalance within Portfolio two.
2) Diversify risk by investing internationally?
+1

As I approach the time to start de-risking my portfolio, this is the method I’m leaning towards using. And I believe this is how Bernstein and Swenson think about risking as well.

Total Portfolio = Risk-on Assets + Risk-off Assets

Never rebalance from Risk-off to Risk-on

Risk-on: Equities, Bonds, Real-estate, Precious metals, Commodities

Risk-off: Cash, T-bills, CDs, I-bonds, EE-bonds, TIPS held to duration, Stable value funds, Money market funds

Risk-on portfolio should be equity oriented (e.g. 100/0, 90/10, 80/20, 70/30, 60/40), and should be rebalanced regularly.

Risk-off should be X number of RLE, which is traditionally calculated as your non-discretionary budget subtracting SSN and pensions. I would go further and subtract a worst-case withdrawal rate as taken from the Risk-on portfolio. For example, 2% to 2.5% of initial portfolio balance.

Example:

Risk-on portfolio of $1,000,000

Assume a worst-case WR of 2% ($20,000 per year).

Say your non-discretionary budget is $50,000 and you expect to receive $20,000 per year in SSN benefits (CPI adjusted).

RLE = $50,000 - $20,000 - $20,000 = $10,000

Say you want 20X years of RLE as Risk-off portfolio size, this would be:

Risk-off portfolio = $10,000 x 20 = $200,000
Or instead of having two buckets: Risk off and Risk on, you could have 3 for more optimal risk stratification:

Risk off bucket: I-bonds and individual TIPS
So so risk bucket: 60/40 all in one found or conservative TDF with good international diversification
Risk on bucket: 100% small cap or US stock market or whatever you think will make more money over 20-30 years

And instead of using asset allocation, use distribution allocation.
Topic Author
Leesbro63
Posts: 6787
Joined: Mon Nov 08, 2010 4:36 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Leesbro63 »

Blue456 wrote: Thu Sep 24, 2020 10:38 am
aj76er wrote: Thu Sep 24, 2020 9:37 am
Blue456 wrote: Tue Sep 22, 2020 3:51 pm
Leesbro63 wrote: Tue Sep 22, 2020 2:19 pm During 2008-9, I came to the realization that in a Japan-type scenario, rebalancing a mature portfolio (I was age 49 then; I'm 60 now) from bonds into stocks might be a "Pascal's Wager". If you're wrong and stocks keeping going down for a long time, you effectively "flush" good money into bad money that might never recover. Thus negating the reason for "safe" money in the first place. And I amended my investment plan accordingly so that I would never again rebalance from bonds to stocks.

Now, as I get even closer to retirement age, I am wondering what happens if stocks crash and stay down for a long period of time. It might be a "you're dead already" scenario if you DO NOT rebalance into stocks after a crash. In other words, what happens when stocks crash just before or after retirement and we have a 1966-1981 period?

How does one reconcile the "Pascal's Wager" problem with the "you're dead already" problem? n
1) Have two portfolios:
Portfolio one. I-bonds, TIPS, CDs and other safe income that covers your very basic expenses for 5, 10, 15, 20 years. You draw the line how much makes you comfortable.
Portfolio two. 100/0, 90/10, 80/20, 70/30, 60/40.... etc.
Don’t rebalance from portfolio 2 to portfolio 1. Do rebalance within Portfolio two.
2) Diversify risk by investing internationally?
+1

As I approach the time to start de-risking my portfolio, this is the method I’m leaning towards using. And I believe this is how Bernstein and Swenson think about risking as well.

Total Portfolio = Risk-on Assets + Risk-off Assets

Never rebalance from Risk-off to Risk-on

Risk-on: Equities, Bonds, Real-estate, Precious metals, Commodities

Risk-off: Cash, T-bills, CDs, I-bonds, EE-bonds, TIPS held to duration, Stable value funds, Money market funds

Risk-on portfolio should be equity oriented (e.g. 100/0, 90/10, 80/20, 70/30, 60/40), and should be rebalanced regularly.

Risk-off should be X number of RLE, which is traditionally calculated as your non-discretionary budget subtracting SSN and pensions. I would go further and subtract a worst-case withdrawal rate as taken from the Risk-on portfolio. For example, 2% to 2.5% of initial portfolio balance.

Example:

Risk-on portfolio of $1,000,000

Assume a worst-case WR of 2% ($20,000 per year).

Say your non-discretionary budget is $50,000 and you expect to receive $20,000 per year in SSN benefits (CPI adjusted).

RLE = $50,000 - $20,000 - $20,000 = $10,000

Say you want 20X years of RLE as Risk-off portfolio size, this would be:

Risk-off portfolio = $10,000 x 20 = $200,000
Or instead of having two buckets: Risk off and Risk on, you could have 3 for more optimal risk stratification:

Risk off bucket: I-bonds and individual TIPS
So so risk bucket: 60/40 all in one found or conservative TDF with good international diversification
Risk on bucket: 100% small cap or US stock market or whatever you think will make more money over 20-30 years

And instead of using asset allocation, use distribution allocation.
This is just a ridiculously complicated single portfolio.
Blue456
Posts: 1018
Joined: Tue Jun 04, 2019 5:46 am

Re: Never Ever Rebalance Bonds into Stocks?

Post by Blue456 »

Leesbro63 wrote: Thu Sep 24, 2020 10:14 am
Blue456 wrote: Tue Sep 22, 2020 3:51 pm
Leesbro63 wrote: Tue Sep 22, 2020 2:19 pm During 2008-9, I came to the realization that in a Japan-type scenario, rebalancing a mature portfolio (I was age 49 then; I'm 60 now) from bonds into stocks might be a "Pascal's Wager". If you're wrong and stocks keeping going down for a long time, you effectively "flush" good money into bad money that might never recover. Thus negating the reason for "safe" money in the first place. And I amended my investment plan accordingly so that I would never again rebalance from bonds to stocks.

Now, as I get even closer to retirement age, I am wondering what happens if stocks crash and stay down for a long period of time. It might be a "you're dead already" scenario if you DO NOT rebalance into stocks after a crash. In other words, what happens when stocks crash just before or after retirement and we have a 1966-1981 period?

How does one reconcile the "Pascal's Wager" problem with the "you're dead already" problem? n
1) Have two portfolios:
Portfolio one. I-bonds, TIPS, CDs and other safe income that covers your very basic expenses for 5, 10, 15, 20 years. You draw the line how much makes you comfortable.
Portfolio two. 100/0, 90/10, 80/20, 70/30, 60/40.... etc.
Don’t rebalance from portfolio 2 to portfolio 1. Do rebalance within Portfolio two.
2) Diversify risk by investing internationally?
What you're talking about is basically "buckets". This has been discussed here many times. The bottom line is that it's just mental accounting for a lower overall equity allocation.

Say your overall nest egg has portfolio #1 that's 30% I-Bonds, TIPS, CDs and other safe income that covers for up to 20 years. Portfolio 2 is 60/40. You're just fooling yourself because what you really have is one portfolio that's 42% stock and 58% Equity. Call it 40/60. It's just a head game. Why play a game like that? Just be honest with yourself and use the less aggressive 40/60 portfolio.
Except that 60/40 will fail in the following scenario while bucket approach will not.
Blue456 wrote: Wed Sep 23, 2020 2:31 pm
Doc wrote: Wed Sep 23, 2020 12:55 pm I don't understand this whole thread. I had always thought that we rebalanced to control risk not to influence return.
Imagine being 60 years old and having portfolio of 60/40 of $1,000,000. So that’s $600,000 in equity and $400,000 in bonds.

Now the following scenario occurs:
2030: Equity drops 50% —> you rebalance and now have total portfolio of $700,000 with $420,000 in equity and $280,000 in bonds
2031: Equity drops 80% —> you rebalance and now have total portfolio of $364,000
2032-2050: stock market returns are zero.

You are left to retire on 37% of your original portfolio for the next 20 years. Finally when you are 80 the market takes off and by the time you are 90 you doubled or tripled your money. Yet it doesn’t matter at that point.

The OP is suggesting that you never rebalance into equity. In the above scenario you end up having $600,000 in bonds and $40,000 equity. You are left with 64% of your original portfolio.
flyingaway
Posts: 2973
Joined: Fri Jan 17, 2014 10:19 am

Re: Never Ever Rebalance Bonds into Stocks?

Post by flyingaway »

willthrill81 wrote: Tue Sep 22, 2020 2:34 pm Another possibility is that you could use a liability matching portfolio (LMP) approach using I bonds and TIPS to cover your essential spending in retirement and everything else in stocks.
Looks like a bucket strategy in a different disguise.
Topic Author
Leesbro63
Posts: 6787
Joined: Mon Nov 08, 2010 4:36 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Leesbro63 »

flyingaway wrote: Thu Sep 24, 2020 11:01 am
willthrill81 wrote: Tue Sep 22, 2020 2:34 pm Another possibility is that you could use a liability matching portfolio (LMP) approach using I bonds and TIPS to cover your essential spending in retirement and everything else in stocks.
Looks like a bucket strategy in a different disguise.
That was my conclusion as well.
User avatar
Uncorrelated
Posts: 1057
Joined: Sun Oct 13, 2019 3:16 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Uncorrelated »

Blue456 wrote: Thu Sep 24, 2020 10:56 am
Except that 60/40 will fail in the following scenario while bucket approach will not.
That is just a different from of cherry picking. You might as well claim that 60/40 fails when both stock and bond markets crash spectacularly but a strategy that invests everything in the correct lottery ticker succeeds. Investors aren't concerned about a single possible, an investor wants to pick the strategy that results in the best collection of outcomes over all the probability distribution of future events.

As I mentioned earlier, it is easy to prove that bucket or one-way rebalance approaches are always suboptimal. You only have to understand bellman's principle of optimality to see why.
User avatar
HomerJ
Posts: 15194
Joined: Fri Jun 06, 2008 12:50 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by HomerJ »

Uncorrelated wrote: Thu Sep 24, 2020 11:28 amAs I mentioned earlier, it is easy to prove that bucket or one-way rebalance approaches are always suboptimal. You only have to understand bellman's principle of optimality to see why.
Many of us doing it this way do not care about finding the "optimal" approach.

If you've made the decision to retire, you've said "I have enough".

It is NOT necessary at that point to rebalance from bonds to stocks to have "more".

We already have "enough". By definition, "more" is not very important if one already has "enough".

I'm all about preservation in retirement. Don't care about optimization, or increasing my odds of having a larger portfolio.

If stocks go down, I'll live off the bonds until stocks come back. In the long-run, it's very likely stocks will still give me a good return. I don't need to rebalance into stocks to get an even better return.

Because that puts my safe money and my retirement at risk if the Great Depression II happens and stocks DON'T recover for a decade or two. It's happened before. It can happen again.

I don't need to put my retirement at risk, even if the odds are very high that I will get "more" with that strategy. I already have "enough". I don't need "more".
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
flyingaway
Posts: 2973
Joined: Fri Jan 17, 2014 10:19 am

Re: Never Ever Rebalance Bonds into Stocks?

Post by flyingaway »

HomerJ wrote: Thu Sep 24, 2020 11:46 am
Uncorrelated wrote: Thu Sep 24, 2020 11:28 amAs I mentioned earlier, it is easy to prove that bucket or one-way rebalance approaches are always suboptimal. You only have to understand bellman's principle of optimality to see why.
Many of us doing it this way do not care about finding the "optimal" approach.

If you've made the decision to retire, you've said "I have enough".

It is NOT necessary at that point to rebalance from bonds to stocks to have "more".

We already have "enough". By definition, "more" is not very important if one already has "enough".

I'm all about preservation in retirement. Don't care about optimization, or increasing my odds of having a larger portfolio.

If stocks go down, I'll live off the bonds until stocks come back. In the long-run, it's very likely stocks will still give me a good return. I don't need to rebalance into stocks to get an even better return.

Because that puts my safe money and my retirement at risk if the Great Depression II happens and stocks DON'T recover for a decade or two. It's happened before. It can happen again.

I don't need to put my retirement at risk, even if the odds are very high that I will get "more" with that strategy. I already have "enough". I don't need "more".
I like the concept of "enough", although I don't know what is really THE "enough" (for me).
User avatar
Uncorrelated
Posts: 1057
Joined: Sun Oct 13, 2019 3:16 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Uncorrelated »

HomerJ wrote: Thu Sep 24, 2020 11:46 am
Uncorrelated wrote: Thu Sep 24, 2020 11:28 amAs I mentioned earlier, it is easy to prove that bucket or one-way rebalance approaches are always suboptimal. You only have to understand bellman's principle of optimality to see why.

I'm all about preservation in retirement. Don't care about optimization, or increasing my odds of having a larger portfolio.
That is a contradiction. You do care about optimization, you want to pick the portfolio that has the best preservation properties.

It follows directly from Bellman's principle of optimality that the portfolio with the best preservation properties does not involve one-way rebalancing. The portfolio allocation methods you describe do not minimize the chance of a failed retirement.
User avatar
Doc
Posts: 9807
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Never Ever Rebalance Bonds into Stocks?

Post by Doc »

KlangFool wrote: Wed Sep 23, 2020 4:38 pm
Doc wrote: Wed Sep 23, 2020 4:32 pm
KlangFool wrote: Wed Sep 23, 2020 4:15 pm But, I am only comfortable if my fixed income portion is at least 5 years of annual expense.
If your interest and dividends were greater than your annual expenses would you still want another five years of expenses in FI?

None of us has the same circumstances as everyone else. The problem I have with the OP is the "Never Ever". If that is supposed to apply to every single one of us the answer is not only NO it's HECK NO.

If he said "IF XYZ don't rebalnce into stocks" at least we could talk about what XYZ means to each of us.

Doc,

<<If your interest and dividends were greater than your annual expenses would you still want another five years of expenses in FI?>>


The numbers do not add up. Intererest/Dividend is around 2%.

...

KlangFool
My numbers do add up. But my numbers are not the same as your numbers. I do not have to take money from my investment accounts to meet routine expenses. Our income from SS and pensions meet our normal expenses and also allow us to make a monthly contribution to our "car" fund. (A few years ago I spent several $k from the car fund a new roof. Eh, so what.

The whole idea of this thread could be discarded if everyone just had their own "car" fund. (Some people like more expensive "cars". That's fine. They just have a bigger car fund.)
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.
KlangFool
Posts: 17767
Joined: Sat Oct 11, 2008 12:35 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by KlangFool »

Doc wrote: Thu Sep 24, 2020 12:43 pm
KlangFool wrote: Wed Sep 23, 2020 4:38 pm
Doc wrote: Wed Sep 23, 2020 4:32 pm
KlangFool wrote: Wed Sep 23, 2020 4:15 pm But, I am only comfortable if my fixed income portion is at least 5 years of annual expense.
If your interest and dividends were greater than your annual expenses would you still want another five years of expenses in FI?

None of us has the same circumstances as everyone else. The problem I have with the OP is the "Never Ever". If that is supposed to apply to every single one of us the answer is not only NO it's HECK NO.

If he said "IF XYZ don't rebalnce into stocks" at least we could talk about what XYZ means to each of us.

Doc,

<<If your interest and dividends were greater than your annual expenses would you still want another five years of expenses in FI?>>


The numbers do not add up. Intererest/Dividend is around 2%.

...

KlangFool
My numbers do add up. But my numbers are not the same as your numbers. I do not have to take money from my investment accounts to meet routine expenses. Our income from SS and pensions meet our normal expenses and also allow us to make a monthly contribution to our "car" fund. (A few years ago I spent several $k from the car fund a new roof. Eh, so what.

The whole idea of this thread could be discarded if everyone just had their own "car" fund. (Some people like more expensive "cars". That's fine. They just have a bigger car fund.)
Doc,

<<I do not have to take money from my investment accounts to meet routine expenses. Our income from SS and pensions meet our normal expenses and also allow us to make a monthly contribution to our "car" fund.>>

In summary, you do not need any money from your portfolio to meet your normal expense. That does not apply in my case.


KlangFool
User avatar
HomerJ
Posts: 15194
Joined: Fri Jun 06, 2008 12:50 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by HomerJ »

Uncorrelated wrote: Thu Sep 24, 2020 12:03 pm
HomerJ wrote: Thu Sep 24, 2020 11:46 am
Uncorrelated wrote: Thu Sep 24, 2020 11:28 amAs I mentioned earlier, it is easy to prove that bucket or one-way rebalance approaches are always suboptimal. You only have to understand bellman's principle of optimality to see why.

I'm all about preservation in retirement. Don't care about optimization, or increasing my odds of having a larger portfolio.
That is a contradiction. You do care about optimization, you want to pick the portfolio that has the best preservation properties.

It follows directly from Bellman's principle of optimality that the portfolio with the best preservation properties does not involve one-way rebalancing. The portfolio allocation methods you describe do not minimize the chance of a failed retirement.
That statement is incorrect.
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
User avatar
vineviz
Posts: 7842
Joined: Tue May 15, 2018 1:55 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by vineviz »

HomerJ wrote: Thu Sep 24, 2020 1:30 pm
Uncorrelated wrote: Thu Sep 24, 2020 12:03 pm
HomerJ wrote: Thu Sep 24, 2020 11:46 am
Uncorrelated wrote: Thu Sep 24, 2020 11:28 amAs I mentioned earlier, it is easy to prove that bucket or one-way rebalance approaches are always suboptimal. You only have to understand bellman's principle of optimality to see why.

I'm all about preservation in retirement. Don't care about optimization, or increasing my odds of having a larger portfolio.
That is a contradiction. You do care about optimization, you want to pick the portfolio that has the best preservation properties.

It follows directly from Bellman's principle of optimality that the portfolio with the best preservation properties does not involve one-way rebalancing. The portfolio allocation methods you describe do not minimize the chance of a failed retirement.
That statement is incorrect.
It’s not. I don’t always agree with Uncorrelated, but the bolded statement is inarguably true.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
HomerJ
Posts: 15194
Joined: Fri Jun 06, 2008 12:50 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by HomerJ »

Doc wrote: Thu Sep 24, 2020 12:43 pmI do not have to take money from my investment accounts to meet routine expenses. Our income from SS and pensions meet our normal expenses
Then none of this pertains to you. You have zero risk no matter what you do. Congrats.

I am in different situation.

I understand you don't like having "never ever" applied to you. (Although technically, most of us are just defending why WE do it, not telling you to do it).

But I also don't appreciate posts where you tell me I've done a poor job determining my AA and retirement plan.
If you don't rebalance when the stock market tanks you have decided to ignore all the thought and work you put into that AA in the first place. If you don't rebalance into stocks you you may not lose more money but if the market doesn't recover you will not achieve your original objective like having enough to eat during retirement. And if you do rebalnce and the market goes up again as it always does (eventually) you recover faster.

If you can't emotionally buy into a down market you have not set your original AA where it needs to be both emotionally and financially. The answer is not to ignore rebalancing but to somehow save more or reduce your spending desires.

Sith happens. Markets crash. Take advantage of that crash by buying more not less.
You're set. Of course, there's no emotional toll for you to rebalance into a down market. No risk.

If Sith does happen, and continues to happen, you won't be affected by your decision to rebalance.
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
User avatar
HomerJ
Posts: 15194
Joined: Fri Jun 06, 2008 12:50 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by HomerJ »

vineviz wrote: Thu Sep 24, 2020 1:32 pm
HomerJ wrote: Thu Sep 24, 2020 1:30 pm
Uncorrelated wrote: Thu Sep 24, 2020 12:03 pm
HomerJ wrote: Thu Sep 24, 2020 11:46 am
Uncorrelated wrote: Thu Sep 24, 2020 11:28 amAs I mentioned earlier, it is easy to prove that bucket or one-way rebalance approaches are always suboptimal. You only have to understand bellman's principle of optimality to see why.

I'm all about preservation in retirement. Don't care about optimization, or increasing my odds of having a larger portfolio.
That is a contradiction. You do care about optimization, you want to pick the portfolio that has the best preservation properties.

It follows directly from Bellman's principle of optimality that the portfolio with the best preservation properties does not involve one-way rebalancing. The portfolio allocation methods you describe do not minimize the chance of a failed retirement.
That statement is incorrect.
It’s not. I don’t always agree with Uncorrelated, but the bolded statement is inarguably true.
Nope. I'm guessing the issue is we have different definitions of "failed" retirement.

Edit: Yes, I changed the goalposts - see below - Sorry, you guys are right in general.
Last edited by HomerJ on Thu Sep 24, 2020 1:57 pm, edited 1 time in total.
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
User avatar
HomerJ
Posts: 15194
Joined: Fri Jun 06, 2008 12:50 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by HomerJ »

I think the problem is that most models think of only one spending level.

I have two different levels. I can drop down to the second level if needed and be perfectly happy. So as long as I can cover the second level, my retirement won't "fail".

Odds are good that I'll stay at higher level if I don't rebalance into stocks during a crash. The odds may indeed be higher I can stay at the higher level if I DO rebalance into stocks, but I'm more interested in protecting the second level.

My chances of running out of money over 30 years with a set amount in bonds/cash/CDs (along with SS) at the second level is nearly zero. I will spend less than I could have, but my chances of a "failed" retirement is extremely low... i.e. minimized.

Sure, other methods will give a better outcome most of the time.

But I'm looking at possible outliers. Rebalancing bonds into stocks add risk. Tiny risk, probably, but non-zero.

If stocks come back, I'll be fine either way (I will have more if I rebalanced - but then again, I don't need more)

if stocks stay down for 20-30 years, I'll be in trouble if I kept rebalancing all the way down, but fine if I didn't.


The problem is a matter of semantics and context. Don't tell me I'm wrong without trying to understand the full picture. You guys are right in general, but maybe not for me and my particular scenario.

This isn't a mathematical exercise. This is my life and my family's life. It's complicated. It's messy. Emotions ARE part of it. A legitimate part.
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
User avatar
Doc
Posts: 9807
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Never Ever Rebalance Bonds into Stocks?

Post by Doc »

The OP was addressing Never to Rebalance from Bonds to Stocks.

Several people have posted their need or desire to have some basic amount of fixed income as reserve that they don't want to drop below. There is nothing in this desire to mean that you should Never Ever rebalance from bonds to stocks.

It is possible to have a policy that does rebalance from Bonds to Stocks and still maintains that "security" position.

Simply separate the amount you want to keep in reserve no matter what and apply the more Boglehead method of percentage allocations to the rest with rebalncing by selling bonds to buy stocks or the other way around based on bands or a schedule.

The thing is to have a rational plan that works for you and stick to it. (Review and change it as your age and total investment balance dictate.) Don't abandon your plan simply because the market tanks.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.
User avatar
abuss368
Posts: 21619
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Never Ever Rebalance Bonds into Stocks?

Post by abuss368 »

Leesbro63 wrote: Tue Sep 22, 2020 2:19 pm During 2008-9, I came to the realization that in a Japan-type scenario, rebalancing a mature portfolio (I was age 49 then; I'm 60 now) from bonds into stocks might be a "Pascal's Wager". If you're wrong and stocks keeping going down for a long time, you effectively "flush" good money into bad money that might never recover. Thus negating the reason for "safe" money in the first place. And I amended my investment plan accordingly so that I would never again rebalance from bonds to stocks.

Now, as I get even closer to retirement age, I am wondering what happens if stocks crash and stay down for a long period of time. It might be a "you're dead already" scenario if you DO NOT rebalance into stocks after a crash. In other words, what happens when stocks crash just before or after retirement and we have a 1966-1981 period?

How does one reconcile the "Pascal's Wager" problem with the "you're dead already" problem? n
No one knows in advance! Japanese investors in the mid-80s had no idea what was over the horizon. We investors can focus on the things we can control such as asset allocation, keeping costs very low, stay the course, keep our asset allocation, save more than we make, look for ways to always increase savings and investing, pay down / off debt, keep taxes low, and tune out the noise.

Anything else? We are kidding ourselves that we can make a difference or be in control.
John C. Bogle: “Simplicity is the master key to financial success."
User avatar
Uncorrelated
Posts: 1057
Joined: Sun Oct 13, 2019 3:16 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Uncorrelated »

HomerJ wrote: Thu Sep 24, 2020 1:48 pm
vineviz wrote: Thu Sep 24, 2020 1:32 pm
It’s not. I don’t always agree with Uncorrelated, but the bolded statement is inarguably true.
Nope. I'm guessing the issue is we have different definitions of "failed" retirement.
That is not the issue here. Bellman's principle of optimality says your methods are suboptimal for any definition of "failed".


I understand you arrived at this asset allocation because you are in a good financial position and didn't want to spend 10000 hours to become an expert in optimization. That is totally fine. But it's probably a good idea to communicate that with the rest of our readers...
KlangFool
Posts: 17767
Joined: Sat Oct 11, 2008 12:35 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by KlangFool »

Uncorrelated wrote: Thu Sep 24, 2020 2:30 pm
HomerJ wrote: Thu Sep 24, 2020 1:48 pm
vineviz wrote: Thu Sep 24, 2020 1:32 pm
It’s not. I don’t always agree with Uncorrelated, but the bolded statement is inarguably true.
Nope. I'm guessing the issue is we have different definitions of "failed" retirement.
That is not the issue here. Bellman's principle of optimality says your methods are suboptimal for any definition of "failed".


I understand you arrived at this asset allocation because you are in a good financial position and didn't want to spend 10000 hours to become an expert in optimization. That is totally fine. But it's probably a good idea to communicate that with the rest of our readers...

Uncorrelated,

<<That is not the issue here. Bellman's principle of optimality says your methods are suboptimal for any definition of "failed".>>


If this statement is true, you just need to provide one definition of "failed" to show that you are correct. Please show us that.


KlangFool
User avatar
HomerJ
Posts: 15194
Joined: Fri Jun 06, 2008 12:50 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by HomerJ »

Doc wrote: Thu Sep 24, 2020 2:23 pm The thing is to have a rational plan that works for you and stick to it. (Review and change it as your age and total investment balance dictate.) Don't abandon your plan simply because the market tanks.
This I agree with.

My plan always included not rebalancing to stocks when stocks fell. I rebalance the other direction; when stocks rise, I sell them and buy more bonds to keep myself at 50/50.

Being doing that since 2009. Sure, I'd have more if I had rebalanced both ways all this time, but emotionally, I was able to stick with the plan doing it my way.
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
Northern Flicker
Posts: 6519
Joined: Fri Apr 10, 2015 12:29 am

Re: Never Ever Rebalance Bonds into Stocks?

Post by Northern Flicker »

HomerJ wrote: Wed Sep 23, 2020 8:27 pm
Northern Flicker wrote: Wed Sep 23, 2020 4:04 pm Regarding rebalancing after a crash... if you were comfortable with a 55/45 allocation before a market crash, why would the same allocation become more risky after the crash? The downside risk was obviously much greater before the crash.
I was comfortable with $1 million in bonds, not 45%... If I rebalance and I only have $800,000 in bonds, I'm no longer comfortable.

It's really not that hard to understand. You don't have to agree, but it's not hard to understand.

If I get to $1.1 million in bonds during retirement, and a crash happens then, I might rebalance $100,000.
If stocks fall and stay down for 5 years, where will you take withdrawals? Isn't the purpose of having N years of expenses in bonds to have it available to meet up to N years of expenses when stocks are in the toilet? If you just withdraw from stocks to keep holding the sacred N years of expenses in bonds, this is rebalancing in the other direction and will amplify sequence of return risk.

Lastly, if you are holding N*(current year's expenses) in nominal bonds, you do not have N years of expected expenses in bonds. You have N years of expected expenses conditional on inflation being zero as an inflation outcome.
Last edited by Northern Flicker on Thu Sep 24, 2020 10:56 pm, edited 2 times in total.
Risk is not a guarantor of return.
Tattarrattat
Posts: 116
Joined: Wed Aug 19, 2020 6:05 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Tattarrattat »

Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
User avatar
Uncorrelated
Posts: 1057
Joined: Sun Oct 13, 2019 3:16 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Uncorrelated »

Tattarrattat wrote: Thu Sep 24, 2020 3:25 pm Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.

For example, suppose we have a retiree with $1000, a 50/50 asset allocation, $100 spending per period (spend stocks first), and a "don't rebalance into stocks" rule.

Sequence 1: bonds -50%, rebalance, spend $100. After this sequence of returns, retiree has $650 left, $275 in stocks and $375 in bonds.
Sequence 2: stocks -50%, rebalance, spend $100. After this sequence of returns, retiree has $650 left, $150 in stocks and $500 in bonds.

We see that both sequences result in the same net worth but different suggested asset allocations. This contradicts Bellman's principle of optimality, so the strategy cannot be optimal. Note that we were able to prove this without knowing what the actual optimization objective is.


Proving that bucketing with two-way rebalancing is also suboptimal is significantly harder. However there are at least three good arguments that the general concept of a bucket just doesn't work: Basically, there is no known mathematical model for which the optimal asset allocation (in retirement) even vaguely resembles a bucket.
User avatar
WoodSpinner
Posts: 1669
Joined: Mon Feb 27, 2017 1:15 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by WoodSpinner »

Leesbro63 wrote: Thu Sep 24, 2020 10:14 am
Blue456 wrote: Tue Sep 22, 2020 3:51 pm
Leesbro63 wrote: Tue Sep 22, 2020 2:19 pm During 2008-9, I came to the realization that in a Japan-type scenario, rebalancing a mature portfolio (I was age 49 then; I'm 60 now) from bonds into stocks might be a "Pascal's Wager". If you're wrong and stocks keeping going down for a long time, you effectively "flush" good money into bad money that might never recover. Thus negating the reason for "safe" money in the first place. And I amended my investment plan accordingly so that I would never again rebalance from bonds to stocks.

Now, as I get even closer to retirement age, I am wondering what happens if stocks crash and stay down for a long period of time. It might be a "you're dead already" scenario if you DO NOT rebalance into stocks after a crash. In other words, what happens when stocks crash just before or after retirement and we have a 1966-1981 period?

How does one reconcile the "Pascal's Wager" problem with the "you're dead already" problem? n
1) Have two portfolios:
Portfolio one. I-bonds, TIPS, CDs and other safe income that covers your very basic expenses for 5, 10, 15, 20 years. You draw the line how much makes you comfortable.
Portfolio two. 100/0, 90/10, 80/20, 70/30, 60/40.... etc.
Don’t rebalance from portfolio 2 to portfolio 1. Do rebalance within Portfolio two.
2) Diversify risk by investing internationally?
What you're talking about is basically "buckets". This has been discussed here many times. The bottom line is that it's just mental accounting for a lower overall equity allocation.

Say your overall nest egg has portfolio #1 that's 30% I-Bonds, TIPS, CDs and other safe income that covers for up to 20 years. Portfolio 2 is 60/40. You're just fooling yourself because what you really have is one portfolio that's 42% stock and 58% Equity. Call it 40/60. It's just a head game. Why play a game like that? Just be honest with yourself and use the less aggressive 40/60 portfolio.
Count me as a proud Mental Accountant!!

This essentially describes how my portfolio works now that I am retired.

One key factor you are overlooking is that my target AA will change every year as I spend down the Liability Matching Portfolio (LMP).

In my situation. I need to fund a LMP to cover Cashflow through age 70. After that SS kicks in and I shouldn’t need to touch the portfolio unless I have some very unexpected expenses. Why would I hold a constant AA during this time since my need to take a risk is changing.

A more detailed explanation is available in this post.

I don’t understand why your suggested approach would be superior or preferable.


WoodSpinner
Last edited by WoodSpinner on Thu Sep 24, 2020 9:30 pm, edited 1 time in total.
invest2bfree
Posts: 72
Joined: Sun Jan 12, 2020 9:44 am

Re: Never Ever Rebalance Bonds into Stocks?

Post by invest2bfree »

Leesbro63 wrote: Tue Sep 22, 2020 2:19 pm During 2008-9, I came to the realization that in a Japan-type scenario, rebalancing a mature portfolio (I was age 49 then; I'm 60 now) from bonds into stocks might be a "Pascal's Wager". If you're wrong and stocks keeping going down for a long time, you effectively "flush" good money into bad money that might never recover. Thus negating the reason for "safe" money in the first place. And I amended my investment plan accordingly so that I would never again rebalance from bonds to stocks.

Now, as I get even closer to retirement age, I am wondering what happens if stocks crash and stay down for a long period of time. It might be a "you're dead already" scenario if you DO NOT rebalance into stocks after a crash. In other words, what happens when stocks crash just before or after retirement and we have a 1966-1981 period?

How does one reconcile the "Pascal's Wager" problem with the "you're dead already" problem? n


This is why you buy VT, if us becomes japan then there will be someone else who can hold the growth mantle. Japanese investors lost because they were 100% in Japanese large caps like many people who are totally in VOO.
User avatar
aj76er
Posts: 823
Joined: Tue Dec 01, 2015 11:34 pm
Location: Austin, TX

Re: Never Ever Rebalance Bonds into Stocks?

Post by aj76er »

Blue456 wrote: Thu Sep 24, 2020 10:38 am
aj76er wrote: Thu Sep 24, 2020 9:37 am
Blue456 wrote: Tue Sep 22, 2020 3:51 pm
Leesbro63 wrote: Tue Sep 22, 2020 2:19 pm During 2008-9, I came to the realization that in a Japan-type scenario, rebalancing a mature portfolio (I was age 49 then; I'm 60 now) from bonds into stocks might be a "Pascal's Wager". If you're wrong and stocks keeping going down for a long time, you effectively "flush" good money into bad money that might never recover. Thus negating the reason for "safe" money in the first place. And I amended my investment plan accordingly so that I would never again rebalance from bonds to stocks.

Now, as I get even closer to retirement age, I am wondering what happens if stocks crash and stay down for a long period of time. It might be a "you're dead already" scenario if you DO NOT rebalance into stocks after a crash. In other words, what happens when stocks crash just before or after retirement and we have a 1966-1981 period?

How does one reconcile the "Pascal's Wager" problem with the "you're dead already" problem? n
1) Have two portfolios:
Portfolio one. I-bonds, TIPS, CDs and other safe income that covers your very basic expenses for 5, 10, 15, 20 years. You draw the line how much makes you comfortable.
Portfolio two. 100/0, 90/10, 80/20, 70/30, 60/40.... etc.
Don’t rebalance from portfolio 2 to portfolio 1. Do rebalance within Portfolio two.
2) Diversify risk by investing internationally?
+1

As I approach the time to start de-risking my portfolio, this is the method I’m leaning towards using. And I believe this is how Bernstein and Swenson think about risking as well.

Total Portfolio = Risk-on Assets + Risk-off Assets

Never rebalance from Risk-off to Risk-on

Risk-on: Equities, Bonds, Real-estate, Precious metals, Commodities

Risk-off: Cash, T-bills, CDs, I-bonds, EE-bonds, TIPS held to duration, Stable value funds, Money market funds

Risk-on portfolio should be equity oriented (e.g. 100/0, 90/10, 80/20, 70/30, 60/40), and should be rebalanced regularly.

Risk-off should be X number of RLE, which is traditionally calculated as your non-discretionary budget subtracting SSN and pensions. I would go further and subtract a worst-case withdrawal rate as taken from the Risk-on portfolio. For example, 2% to 2.5% of initial portfolio balance.

Example:

Risk-on portfolio of $1,000,000

Assume a worst-case WR of 2% ($20,000 per year).

Say your non-discretionary budget is $50,000 and you expect to receive $20,000 per year in SSN benefits (CPI adjusted).

RLE = $50,000 - $20,000 - $20,000 = $10,000

Say you want 20X years of RLE as Risk-off portfolio size, this would be:

Risk-off portfolio = $10,000 x 20 = $200,000
Or instead of having two buckets: Risk off and Risk on, you could have 3 for more optimal risk stratification:

Risk off bucket: I-bonds and individual TIPS
So so risk bucket: 60/40 all in one found or conservative TDF with good international diversification
Risk on bucket: 100% small cap or US stock market or whatever you think will make more money over 20-30 years

And instead of using asset allocation, use distribution allocation.
Technically I’m doing this, but my high Risk-on bucket is a small amount in Hedfundie’s excellent adventure :sharebeer
"Buy-and-hold, long-term, all-market-index strategies, implemented at rock-bottom cost, are the surest of all routes to the accumulation of wealth" - John C. Bogle
User avatar
aj76er
Posts: 823
Joined: Tue Dec 01, 2015 11:34 pm
Location: Austin, TX

Re: Never Ever Rebalance Bonds into Stocks?

Post by aj76er »

Leesbro63 wrote: Thu Sep 24, 2020 10:46 am
Blue456 wrote: Thu Sep 24, 2020 10:38 am
aj76er wrote: Thu Sep 24, 2020 9:37 am
Blue456 wrote: Tue Sep 22, 2020 3:51 pm
Leesbro63 wrote: Tue Sep 22, 2020 2:19 pm During 2008-9, I came to the realization that in a Japan-type scenario, rebalancing a mature portfolio (I was age 49 then; I'm 60 now) from bonds into stocks might be a "Pascal's Wager". If you're wrong and stocks keeping going down for a long time, you effectively "flush" good money into bad money that might never recover. Thus negating the reason for "safe" money in the first place. And I amended my investment plan accordingly so that I would never again rebalance from bonds to stocks.

Now, as I get even closer to retirement age, I am wondering what happens if stocks crash and stay down for a long period of time. It might be a "you're dead already" scenario if you DO NOT rebalance into stocks after a crash. In other words, what happens when stocks crash just before or after retirement and we have a 1966-1981 period?

How does one reconcile the "Pascal's Wager" problem with the "you're dead already" problem? n
1) Have two portfolios:
Portfolio one. I-bonds, TIPS, CDs and other safe income that covers your very basic expenses for 5, 10, 15, 20 years. You draw the line how much makes you comfortable.
Portfolio two. 100/0, 90/10, 80/20, 70/30, 60/40.... etc.
Don’t rebalance from portfolio 2 to portfolio 1. Do rebalance within Portfolio two.
2) Diversify risk by investing internationally?
+1

As I approach the time to start de-risking my portfolio, this is the method I’m leaning towards using. And I believe this is how Bernstein and Swenson think about risking as well.

Total Portfolio = Risk-on Assets + Risk-off Assets

Never rebalance from Risk-off to Risk-on

Risk-on: Equities, Bonds, Real-estate, Precious metals, Commodities

Risk-off: Cash, T-bills, CDs, I-bonds, EE-bonds, TIPS held to duration, Stable value funds, Money market funds

Risk-on portfolio should be equity oriented (e.g. 100/0, 90/10, 80/20, 70/30, 60/40), and should be rebalanced regularly.

Risk-off should be X number of RLE, which is traditionally calculated as your non-discretionary budget subtracting SSN and pensions. I would go further and subtract a worst-case withdrawal rate as taken from the Risk-on portfolio. For example, 2% to 2.5% of initial portfolio balance.

Example:

Risk-on portfolio of $1,000,000

Assume a worst-case WR of 2% ($20,000 per year).

Say your non-discretionary budget is $50,000 and you expect to receive $20,000 per year in SSN benefits (CPI adjusted).

RLE = $50,000 - $20,000 - $20,000 = $10,000

Say you want 20X years of RLE as Risk-off portfolio size, this would be:

Risk-off portfolio = $10,000 x 20 = $200,000
Or instead of having two buckets: Risk off and Risk on, you could have 3 for more optimal risk stratification:

Risk off bucket: I-bonds and individual TIPS
So so risk bucket: 60/40 all in one found or conservative TDF with good international diversification
Risk on bucket: 100% small cap or US stock market or whatever you think will make more money over 20-30 years

And instead of using asset allocation, use distribution allocation.
This is just a ridiculously complicated single portfolio.
Not really. Most people already do this naturally by maintaining a separate emergency fund from their portfolio (typically thought of in terms of X living expenses).

It’s really just a formalized way of keeping a minimum amount in FI (or an FI floor), which has been mentioned several times in this thread as being a reasonable approach.

As I approach my own de-risking path, I like the idea of having a giant teddy bear of cash-like assets to hug when the excrement hits the fan.
"Buy-and-hold, long-term, all-market-index strategies, implemented at rock-bottom cost, are the surest of all routes to the accumulation of wealth" - John C. Bogle
User avatar
willthrill81
Posts: 20957
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Never Ever Rebalance Bonds into Stocks?

Post by willthrill81 »

AlohaJoe wrote: Thu Sep 24, 2020 12:04 am
willthrill81 wrote: Tue Sep 22, 2020 9:17 pm His 'stop playing the game' saying makes it sound like stocks are little better than the craps table, which really irks me.
I understand where you're coming from and I feel the same frustration about overly glib statements. But keep in mind Bernstein is a financial advisor who only deals with clients of at least $30,000,000. Given that background, my impression is his comment about "stop playing the game" probably isn't based on his experience working with Bogleheads who have got to 27x expenses. He's probably got in mind his clients who have reached 200x expenses and he still has to work hard to convince them deleverage from the concentrated risk of a family business (or a single industry they think they are experts about) and to put money in munis and Treasuries.
Yes, for that audience, it makes a lot more sense. But I suspect that most of those who parrot that saying like it was etched in stone on a mountaintop don't have anything close to $30 million or 200x. No doubt if I had anywhere near that much, I would buy a truckload of TIPS.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings
User avatar
willthrill81
Posts: 20957
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Never Ever Rebalance Bonds into Stocks?

Post by willthrill81 »

Uncorrelated wrote: Thu Sep 24, 2020 4:26 pm
Tattarrattat wrote: Thu Sep 24, 2020 3:25 pm Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.
Of course, the problem is greatly complicated by the fact that no two investors have the same utility function.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings
User avatar
willthrill81
Posts: 20957
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Never Ever Rebalance Bonds into Stocks?

Post by willthrill81 »

flyingaway wrote: Thu Sep 24, 2020 11:01 am
willthrill81 wrote: Tue Sep 22, 2020 2:34 pm Another possibility is that you could use a liability matching portfolio (LMP) approach using I bonds and TIPS to cover your essential spending in retirement and everything else in stocks.
Looks like a bucket strategy in a different disguise.
Strategies that do not involve rebalancing between buckets are not mere mental accounting. With a LMP approach, a set percentage of I-bonds and/or TIPS are consumed every year. These are not sold to buy stocks, nor are stocks sold to buy more of them.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings
User avatar
abuss368
Posts: 21619
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Never Ever Rebalance Bonds into Stocks?

Post by abuss368 »

Uncorrelated wrote: Thu Sep 24, 2020 4:26 pm
Tattarrattat wrote: Thu Sep 24, 2020 3:25 pm Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.
All investors are different. They have different goals, timeframes, and tolerances for risks. That changes everything!
John C. Bogle: “Simplicity is the master key to financial success."
User avatar
Uncorrelated
Posts: 1057
Joined: Sun Oct 13, 2019 3:16 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Uncorrelated »

willthrill81 wrote: Thu Sep 24, 2020 7:11 pm
Uncorrelated wrote: Thu Sep 24, 2020 4:26 pm
Tattarrattat wrote: Thu Sep 24, 2020 3:25 pm Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.
Of course, the problem is greatly complicated by the fact that no two investors have the same utility function.
abuss368 wrote: Thu Sep 24, 2020 7:23 pm
Uncorrelated wrote: Thu Sep 24, 2020 4:26 pm
Tattarrattat wrote: Thu Sep 24, 2020 3:25 pm Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.
All investors are different. They have different goals, timeframes, and tolerances for risks. That changes everything!
The example doesn't use two different investors. It uses one investor who faces two possible futures.

Honestly, it looks as if neither of you have bothered to read the rest of my post.
User avatar
Uncorrelated
Posts: 1057
Joined: Sun Oct 13, 2019 3:16 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Uncorrelated »

willthrill81 wrote: Thu Sep 24, 2020 7:12 pm
flyingaway wrote: Thu Sep 24, 2020 11:01 am
willthrill81 wrote: Tue Sep 22, 2020 2:34 pm Another possibility is that you could use a liability matching portfolio (LMP) approach using I bonds and TIPS to cover your essential spending in retirement and everything else in stocks.
Looks like a bucket strategy in a different disguise.
Strategies that do not involve rebalancing between buckets are not mere mental accounting. With a LMP approach, a set percentage of I-bonds and/or TIPS are consumed every year. These are not sold to buy stocks, nor are stocks sold to buy more of them.
Which is still mental accounting unless you can perfectly forecast future inflation, interest rates, future spending needs and remaining lifespan.

The idea that it's possible to build an LMP portfolio without rebalancing between stocks is hubris. What are you going to do when the US government defaults on their debt? Rebalance from stocks. What are you going to do when inflation is higher than expected? Rebalance from stocks. What are you going do to if CPI inflation imperfectly forecasts your personal inflation? Rebalance from stocks. What are you going to to when your washing machine breaks down earlier or later than expected, you outlive your expected lifespan, tax law changes? The answer is clear: rebalance from stocks.
Dude2
Posts: 1087
Joined: Fri Jun 08, 2007 3:40 pm
Location: Florida, USA

Re: Never Ever Rebalance Bonds into Stocks?

Post by Dude2 »

Uncorrelated wrote: Thu Sep 24, 2020 4:26 pm Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.

For example, suppose we have a retiree with $1000, a 50/50 asset allocation, $100 spending per period (spend stocks first), and a "don't rebalance into stocks" rule.

Sequence 1: bonds -50%, rebalance, spend $100. After this sequence of returns, retiree has $650 left, $275 in stocks and $375 in bonds.
Sequence 2: stocks -50%, rebalance, spend $100. After this sequence of returns, retiree has $650 left, $150 in stocks and $500 in bonds.

We see that both sequences result in the same net worth but different suggested asset allocations. This contradicts Bellman's principle of optimality, so the strategy cannot be optimal. Note that we were able to prove this without knowing what the actual optimization objective is.


Proving that bucketing with two-way rebalancing is also suboptimal is significantly harder. However there are at least three good arguments that the general concept of a bucket just doesn't work: Basically, there is no known mathematical model for which the optimal asset allocation (in retirement) even vaguely resembles a bucket.
Great post. This helps the never-ending debate concerning whether to include the emergency fund as part of the portfolio. If we can finally conclude that buckets are suboptimal, then we should not consider it such. HomerJ's thinking could be re-tooled to include a huge eFund, but the remainder of his retirement money should reflect desired AA (and a plan to rebalance to it).
Henceforth I’ll bear Affliction till it do cry out itself, “Enough, enough,” and die.
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Steve Reading »

Uncorrelated wrote: Fri Sep 25, 2020 12:53 am
willthrill81 wrote: Thu Sep 24, 2020 7:11 pm
Uncorrelated wrote: Thu Sep 24, 2020 4:26 pm
Tattarrattat wrote: Thu Sep 24, 2020 3:25 pm Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.
Of course, the problem is greatly complicated by the fact that no two investors have the same utility function.
abuss368 wrote: Thu Sep 24, 2020 7:23 pm
Uncorrelated wrote: Thu Sep 24, 2020 4:26 pm
Tattarrattat wrote: Thu Sep 24, 2020 3:25 pm Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.
All investors are different. They have different goals, timeframes, and tolerances for risks. That changes everything!
The example doesn't use two different investors. It uses one investor who faces two possible futures.

Honestly, it looks as if neither of you have bothered to read the rest of my post.
Deja Vu.
In the past, I’ve made similar argument against strategies with a “cash buffer that you don’t rebalance into”, a strategy that actually was Will who brought it up at the time. Look:
viewtopic.php?p=5395755#p5395755

If you read a bit (I wouldn’t waste my time, it doesn’t seem to go anywhere), you’ll see I get pushback for the same “well two investors aren’t the same”. It’s interesting that that’s where people get hung up.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
User avatar
willthrill81
Posts: 20957
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Never Ever Rebalance Bonds into Stocks?

Post by willthrill81 »

Uncorrelated wrote: Fri Sep 25, 2020 1:27 am
willthrill81 wrote: Thu Sep 24, 2020 7:12 pm
flyingaway wrote: Thu Sep 24, 2020 11:01 am
willthrill81 wrote: Tue Sep 22, 2020 2:34 pm Another possibility is that you could use a liability matching portfolio (LMP) approach using I bonds and TIPS to cover your essential spending in retirement and everything else in stocks.
Looks like a bucket strategy in a different disguise.
Strategies that do not involve rebalancing between buckets are not mere mental accounting. With a LMP approach, a set percentage of I-bonds and/or TIPS are consumed every year. These are not sold to buy stocks, nor are stocks sold to buy more of them.
Which is still mental accounting unless you can perfectly forecast future inflation, interest rates, future spending needs and remaining lifespan.
No, it isn't. Something being referred to as 'mere mental accounting' is generally intended to mean that the strategy could be replicated by using a traditional asset allocation approach, like 60/40. When there is no rebalancing between the 'buckets', then a bucket strategy cannot be replicated using an AA approach.
The idea that it's possible to build an LMP portfolio without rebalancing between stocks is hubris. What are you going to do when the US government defaults on their debt? Rebalance from stocks. What are you going to do when inflation is higher than expected? Rebalance from stocks. What are you going do to if CPI inflation imperfectly forecasts your personal inflation? Rebalance from stocks. What are you going to to when your washing machine breaks down earlier or later than expected, you outlive your expected lifespan, tax law changes? The answer is clear: rebalance from stocks.
Most here aren't worried about the U.S. defaulting on its debt, so that's a moot example. The point of using I bonds and TIPS in an LMP is that doing so protects you from higher than expected inflation. To your other examples, it's important to have a good grasp on what your future expenses are likely to be but also keeping in mind that they can only be estimated, not predicted with great accuracy.

It sounds like you don't like the LMP approach. I don't personally intend to use it, but I understand entirely why many others do.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings
Northern Flicker
Posts: 6519
Joined: Fri Apr 10, 2015 12:29 am

Re: Never Ever Rebalance Bonds into Stocks?

Post by Northern Flicker »

Uncorrelated wrote: Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence.
Bellman's principle applies to optimization by dynamic programming. Unless you discuss a dynamic programming algorithm for optimizing rebalancing decisions, it has no interpretation in the current context.
Risk is not a guarantor of return.
User avatar
Uncorrelated
Posts: 1057
Joined: Sun Oct 13, 2019 3:16 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Uncorrelated »

Northern Flicker wrote: Fri Sep 25, 2020 2:56 pm
Uncorrelated wrote: Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence.
Bellman's principle applies to optimization by dynamic programming. Unless you discuss a dynamic programming algorithm for optimizing rebalancing decisions, it has no interpretation in the current context.
The principle is valid for all optimization problems that can be described by the bellman equation, this certainly includes stuff such as maximizing an arbitrary utility function or minimizing the probability that some event happens. If you're aware of any relevant economic problems that cannot be described by this equation, I would like to know.
Tattarrattat
Posts: 116
Joined: Wed Aug 19, 2020 6:05 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Tattarrattat »

Appreciate the math/economics lessons I agree with those that say that that buckets are basically mental accounting, because it all adds up to a certain AA at any given time, no matter what you call it.

Setting that aside, I don't understand the Bellman business. It sounds like it is saying their is one correct most efficient path, which to me will be true retrospectively, but how does that help you make a forward looking decision? Plus how does it take into account the non-quantitative emotional and psychological vagaries of an investor's decision making? Just can't wrap my head around it.
User avatar
HomerJ
Posts: 15194
Joined: Fri Jun 06, 2008 12:50 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by HomerJ »

Tattarrattat wrote: Fri Sep 25, 2020 3:51 pm Appreciate the math/economics lessons I agree with those that say that that buckets are basically mental accounting, because it all adds up to a certain AA at any given time, no matter what you call it.

Setting that aside, I don't understand the Bellman business. It sounds like it is saying their is one correct most efficient path, which to me will be true retrospectively, but how does that help you make a forward looking decision? Plus how does it take into account the non-quantitative emotional and psychological vagaries of an investor's decision making? Just can't wrap my head around it.
It doesn't.

Never ask an economist for investing advice.

ESPECIALLY an economist who doesn't understand the human element. Economics is a social science, not a physical science.
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
User avatar
Uncorrelated
Posts: 1057
Joined: Sun Oct 13, 2019 3:16 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Uncorrelated »

Tattarrattat wrote: Fri Sep 25, 2020 3:51 pm Setting that aside, I don't understand the Bellman business. It sounds like it is saying their is one correct most efficient path, which to me will be true retrospectively, but how does that help you make a forward looking decision? Plus how does it take into account the non-quantitative emotional and psychological vagaries of an investor's decision making? Just can't wrap my head around it.
The bellman equation can be used to make decisions under uncertainty. If your goal is to maximize your expected wealth and your possible actions are "bet on black at the roulette wheel" and "do nothing", then the optimal action is to do nothing. Of course the ball might end on black, but on average doing nothing will be the best action under this utility function.

By specifying the appropriate utility function, you can get a lot of control over the outcome. if you can choose between $20k guaranteed or a 50/50 chance between $15k and $40k, then the best option depends on your personal risk preferences. If you prefer $20k guaranteed with socially responsible investments over $25k guaranteed with dirty oil investments, there is no reason why that can't be included in the utility function (it's tricky, but possible). This is generally described as risk aversion and not emotion. With emotion I'm imagining actions such as abandoning the strategy when a large drawdown happens, but abandoning the strategy is clearly a bad idea if the original risk preferences were correctly specified. If you care about drawdown, then the correct approach is to specify your utility function in such a way that the probability of a large drawdown is acceptable to you.

How does one actually find the optimal policy for retirement? Good question. This area of research is in it's infancy, there is only one resource that is worth mentioning and that is https://www.aacalc.com/about. If you take a look throughout the papers, you should get a good idea of the general shape of the optimal solution. I (independently) did some similar research myself. If you are still in the accumulation pase, you can find many resources with the search term lifecycle investing, if you are in a hurry aacalc.com has an approximation to the most general solution here.
Tattarrattat
Posts: 116
Joined: Wed Aug 19, 2020 6:05 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Tattarrattat »

Very interesting resource. I will have to chew on it a little. In terms of the decumulation phase, have you read Jim Otar's Unveiling the Retirement Myth, and if so, what what your thoughts on it?
RadAudit
Posts: 3915
Joined: Mon May 26, 2008 10:20 am
Location: Second star on the right and straight on 'til morning

Re: Never Ever Rebalance Bonds into Stocks?

Post by RadAudit »

I'm familiar with the concept of never rebalance bonds into stocks. (I was prepared to follow it.) Unfortunately, the author did not know my wife who believes a four fund portfolio was too complicated for her to rebalance. So, I switched to a LifeStrategy fund - constant risk. By definition, that fund fund rebalances very frequently (daily?).

I once asked this forum if a constant risk portfolio could rebalance to zero. The consensus of opinion was no. (Thank you longinvest viewtopic.php?t=297459)

During the latest downturn (March?) the portfolio lost a fair chunk of change. However, it left enough to meet anticipated needs for the duration, even if the market hadn't recovered. Don't know if that down turn was an adequate stress test. Hope I don't have to find out. But I believe the never rebalance from bonds to stocks approach is one of many approaches that could result in successful outcomes for a portfolio.
FI is the best revenge. LBYM. Invest the rest. Stay the course. - PS: The cavalry isn't coming, kids. You are on your own.
000
Posts: 2805
Joined: Thu Jul 23, 2020 12:04 am

Re: Never Ever Rebalance Bonds into Stocks?

Post by 000 »

Uncorrelated wrote: Fri Sep 25, 2020 1:27 am The idea that it's possible to build an LMP portfolio without rebalancing between stocks is hubris. What are you going to do when the US government defaults on their debt? Rebalance from stocks.
No way would I rebalance from stocks or other assets into US debt if the US Government defaults.

Of course I choose not to make loans to the government or anyone else.
langlands
Posts: 572
Joined: Wed Apr 03, 2019 10:05 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by langlands »

Uncorrelated wrote: Thu Sep 24, 2020 4:26 pm
Tattarrattat wrote: Thu Sep 24, 2020 3:25 pm Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.

For example, suppose we have a retiree with $1000, a 50/50 asset allocation, $100 spending per period (spend stocks first), and a "don't rebalance into stocks" rule.

Sequence 1: bonds -50%, rebalance, spend $100. After this sequence of returns, retiree has $650 left, $275 in stocks and $375 in bonds.
Sequence 2: stocks -50%, rebalance, spend $100. After this sequence of returns, retiree has $650 left, $150 in stocks and $500 in bonds.

We see that both sequences result in the same net worth but different suggested asset allocations. This contradicts Bellman's principle of optimality, so the strategy cannot be optimal. Note that we were able to prove this without knowing what the actual optimization objective is.
Thanks for a thought provoking post. I agree that if you assume stock markets are memoryless (Markov process) and that your own utility function is memoryless, then your asset allocation should depend only on your current wealth and not how you got to that wealth. But I think that's an assumption one has to make (which assuming EMH is usually an approximation one is willing to make on the stock market side). I don't think that follows from Bellman's principle of optimality. Rather, Bellman's equation assumes a Markov environment and derives the recursive substructure of the decision making process from it.

To see the merit of one way balancing or other path dependent strategies, consider a utility function that really hates stock market drops. So if this person has $100 in the stock market and the stock market drops by 50% and returns back to the same level, this person experiences significant negative utility. Usual utility functions that only take into account current wealth would not be able to model this. Bellman's equation can still be applied in this scenario, but the "state" now needs to take into account the entire trajectory and the solution is presumably much more complicated.

I think that most people in fact have such path dependent utility functions, but the economic literature rarely considers them since they don't really make sense from the point of view of homo economicus. Perhaps the behavioral finance people have looked into this. If one is self aware enough to realize that he has such a utility function, it would be hugely beneficial for wealth accumulation to eliminate the path dependency from his utility function. I used to think this is just obvious and anyone who doesn't is simply irrational. Now I'm a little more mature and realize that for most people, this is nearly impossible. Without getting into too much armchair psycho-pop, IMO it's probably much easier for INTP/INTJ personalities to modify their own utility functions by consciously realizing which parts of their utility function are purely emotional (have no material impact) and to thereby eliminate them from consideration.
User avatar
Uncorrelated
Posts: 1057
Joined: Sun Oct 13, 2019 3:16 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by Uncorrelated »

Tattarrattat wrote: Fri Sep 25, 2020 6:30 pm Very interesting resource. I will have to chew on it a little. In terms of the decumulation phase, have you read Jim Otar's Unveiling the Retirement Myth, and if so, what what your thoughts on it?
I have not. I found a copy of the book online and skimmed it. Unfortunately I have to reject the authors' findings on both accumulation and decumulation. This work on accumulation is significantly inferior to Merton's work on the impact of human capital in asset allocation going back to 1970 (now commonly known as lifecycle investint). For decumulation, the author makes three false assumptions: success rate and median wealth are meaningful indicators of investor preference, retirees never adjust their spending even when doom is inevitable, and the asset allocation stays constant during the entire retirement period. Additionally, the author only mentions annuities in passing, but aacalc.com argues that annuities easily improve certainty-equivalent consumption by 20% or more. See his papers for more detailed calculations.

I agree with the author that most retirement research is useless, unfortunately this work appears to be one of those.

langlands wrote: Sat Sep 26, 2020 12:46 am
Uncorrelated wrote: Thu Sep 24, 2020 4:26 pm
Tattarrattat wrote: Thu Sep 24, 2020 3:25 pm Can someone explain how "Bellman's Principle of Optimality" proves anything in this scenario?
Bellman's principle of optimality says that the optimal solution to any optimization problem must have a property called path independence. Informally, this means that if two retirees follow an optimal strategy and have the same remaining lifespan and net worth, they must have the same asset allocation.

For example, suppose we have a retiree with $1000, a 50/50 asset allocation, $100 spending per period (spend stocks first), and a "don't rebalance into stocks" rule.

Sequence 1: bonds -50%, rebalance, spend $100. After this sequence of returns, retiree has $650 left, $275 in stocks and $375 in bonds.
Sequence 2: stocks -50%, rebalance, spend $100. After this sequence of returns, retiree has $650 left, $150 in stocks and $500 in bonds.

We see that both sequences result in the same net worth but different suggested asset allocations. This contradicts Bellman's principle of optimality, so the strategy cannot be optimal. Note that we were able to prove this without knowing what the actual optimization objective is.
Thanks for a thought provoking post. I agree that if you assume stock markets are memoryless (Markov process) and that your own utility function is memoryless, then your asset allocation should depend only on your current wealth and not how you got to that wealth. But I think that's an assumption one has to make (which assuming EMH is usually an approximation one is willing to make on the stock market side). I don't think that follows from Bellman's principle of optimality. Rather, Bellman's equation assumes a Markov environment and derives the recursive substructure of the decision making process from it.

To see the merit of one way balancing or other path dependent strategies, consider a utility function that really hates stock market drops. So if this person has $100 in the stock market and the stock market drops by 50% and returns back to the same level, this person experiences significant negative utility. Usual utility functions that only take into account current wealth would not be able to model this. Bellman's equation can still be applied in this scenario, but the "state" now needs to take into account the entire trajectory and the solution is presumably much more complicated.

I think that most people in fact have such path dependent utility functions, but the economic literature rarely considers them since they don't really make sense from the point of view of homo economicus. Perhaps the behavioral finance people have looked into this. If one is self aware enough to realize that he has such a utility function, it would be hugely beneficial for wealth accumulation to eliminate the path dependency from his utility function. I used to think this is just obvious and anyone who doesn't is simply irrational. Now I'm a little more mature and realize that for most people, this is nearly impossible. Without getting into too much armchair psycho-pop, IMO it's probably much easier for INTP/INTJ personalities to modify their own utility functions by consciously realizing which parts of their utility function are purely emotional (have no material impact) and to thereby eliminate them from consideration.
Drawdown based utility functions are one of those things that sound intuitive but fail in spectacular ways. Here are some examples:

Scenario 1: the price starts at $100. There is a small intraday blip to $110 and then the price decreases to $100. It stays here for five years, then it drops to $50.
Scenario 2: same as scenario 1 without the intraday blip.

Scenario 1 undoubtedly has a larger drawdown than scenario 2 and therefore should have a lower utility. But at the same time, awarding scenario 1 a lower utility because of a intraday blip five years ago is probably not intended.

Another one. The user specifies a max drawdown of 50% and picks a 50/50 position. The portfolio drop by 40%. What should the user do? He should pick a much more conservative asset allocation, since that minimizes the probability of further drawdowns. Despite the many users insisting that they care about the max drawdown, I have never seen an user implementing such a strategy. (No, a bucket approach doesn't work).

A third problem occurs when evaluating strategies that are designed to limit drawdown. This is frequently done by running backtests, this approach pretty much always results in extremely poor parameter estimation and accidental market timing.


The conventional approach to account for emotional issues is to simply calculate the optimal portfolio and then cap the max. equity allocation. For example, with lifecycle investing, infinite leverage theoretically results in the highest expected utility, but most users cap the leverage at 2 or even avoid leverage altogether. I don't think that's irrational. I would like to have access to better tools that allow users to specify their behavioral biases and see the result ("if you specify max drawdown 50%, that results in a SWR of X. Without this rule, SWR of Y"), which hopefully enables users to decide if they really want to keep those biases. Unfortunately I don't see that happening any time soon due to the difficulty of computing such asset allocations.
User avatar
HomerJ
Posts: 15194
Joined: Fri Jun 06, 2008 12:50 pm

Re: Never Ever Rebalance Bonds into Stocks?

Post by HomerJ »

Uncorrelated wrote: Sat Sep 26, 2020 3:56 am Another one. The user specifies a max drawdown of 50% and picks a 50/50 position. The portfolio drop by 40%. What should the user do? He should pick a much more conservative asset allocation, since that minimizes the probability of further drawdowns. Despite the many users insisting that they care about the max drawdown, I have never seen an user implementing such a strategy. (No, a bucket approach doesn't work).
Why doesn't a bucket approach work? If I'm 50/50 and I specify a max drawdown of 50%, then that means stocks can drop 80%-90%, and I can still achieve my goals. If I rebalance into those stocks as they fall, and they keep falling, THAT's how it becomes possible for my drawdown to go higher than 50%

Which is precisely the point of people who don't like to rebalance into stocks near or during retirement.

We're not SELLING stocks, which is a real victory of sorts, compared to most people, but we are't buying either. We have our reserve that will pay the basic bills for years and years and years, and we don't want to risk a big chunk of that.

Sure, it's emotional. That's part of investing.
The conventional approach to account for emotional issues is to simply calculate the optimal portfolio and then cap the max. equity allocation. For example, with lifecycle investing, infinite leverage theoretically results in the highest expected utility, but most users cap the leverage at 2 or even avoid leverage altogether. I don't think that's irrational. I would like to have access to better tools that allow users to specify their behavioral biases and see the result ("if you specify max drawdown 50%, that results in a SWR of X. Without this rule, SWR of Y"), which hopefully enables users to decide if they really want to keep those biases. Unfortunately I don't see that happening any time soon due to the difficulty of computing such asset allocations.
Yes, it is indeed difficult to compute user behavioral and emotional biases.
A Goldman Sachs associate provided a variety of detailed explanations, but then offered a caveat, “If I’m being dead-### honest, though, nobody knows what’s really going on.”
Dottie57
Posts: 9198
Joined: Thu May 19, 2016 5:43 pm
Location: Earth Northern Hemisphere

Re: Never Ever Rebalance Bonds into Stocks?

Post by Dottie57 »

HomerJ wrote: Wed Sep 23, 2020 1:48 pm
Doc wrote: Wed Sep 23, 2020 12:55 pm I don't understand this whole thread. I had always thought that we rebalanced to control risk not to influence return.

Many posters have referred to matching retirement needs. Fine. We set our asset allocation to meet that objective. We then rebalance to keep the risk of that AA constant.

If you don't rebalance when the stock market tanks you have decided to ignore all the thought and work you put into that AA in the first place. If you don't rebalance into stocks you you may not lose more money but if the market doesn't recover you will not achieve your original objective like having enough to eat during retirement. And if you do rebalnce and the market goes up again as it always does (eventually) you recover faster.

If you can't emotionally buy into a down market you have not set your original AA where it needs to be both emotionally and financially. The answer is not to ignore rebalancing but to somehow save more or reduce your spending desires.

Sith happens. Markets crash. Take advantage of that crash by buying more not less.
Once you get close to retirement, controlling risk is about dollar amounts, not percentages.

If I have a million in bonds, that, plus SS, can cover all my base expenses probably for the full 30 years.

I don't want to rebalance that into stocks after a 50% drop, and watch stocks drop ANOTHER 50%, and then maybe never recover (or take 20 years).

That million in bonds would be locked in. That would be my safe money.

Sure, it's a 99% chance I'd have more/recover faster if I rebalanced into stocks during a downturn... But I'm controlling for the 1% risk that we have a Japan type or Great Depression downturn.
Someone on this board made me realize I could rebalance into stocks until I had too little left. So in retirement now, I am using FI during stock turn downs. Like HomerJ I have enough bonds to carry me through. Stocks are icing on the cake.
longinvest
Posts: 4433
Joined: Sat Aug 11, 2012 8:44 am

Re: Never Ever Rebalance Bonds into Stocks?

Post by longinvest »

RadAudit wrote: Fri Sep 25, 2020 6:32 pm (Thank you longinvest viewtopic.php?t=297459)
Thanks RadAudit for reminding me of that specific thread. On it I had a post detailing the calculations to select a sensible asset allocation for the portfolio and then staying the course by rebalancing it.

For the benefits of this thread's readers, here are the key points I made on that thread:
longinvest wrote: Sat Dec 14, 2019 7:11 pm Those who claim that not rebalancing a portfolio is "less risky" than rebalancing it are almost always guilty of comparing portfolios which had differing average asset allocations over the comparison period.

Why is this important? Because, instead of not rebalancing a portfolio, an investor should lower the allocation to stocks BEFORE the downturn to the level of risk that the investor is really willing to accept, and then the investor should rebalance the portfolio. I've shown this in the following posts: It's worth restating the conclusion:
longinvest wrote: Wed Oct 24, 2018 10:37 am There's no reason to add bonds to a portfolio if the goal isn't to manage risk. Rebalancing is part of this. Avoiding rebalancing (or not rebalancing into stocks) to reduce losses is illogical; the same loss protection can be achieved by simply choosing a higher allocation to bonds in the first place. This will, of course, reduce the potential upside of the portfolio, but it will do so consistently. A non-rebalanced portfolio reduces potential gains at the worst of times, at the bottom of a bear market, and it increases potential losses at the worst of times, at the top of a bubble; it's an illogical approach to risk management.
I strongly encourage readers to take the time to read both of these posts.

Rebalancing is part of the 10th principle of the Bogleheads investment philosophy: Stay the course.
longinvest wrote: Sun Dec 15, 2019 12:02 pm Here's an example of how to put in action the logic/mathematics contained in my previous post.

Let's take an investor with a $1,000,000 portfolio who doesn't want her portfolio to drop below $500,000. The investor considers that a potential yet low-probability catastrophic scenario for stocks is a loss of -80%.

A naive (and broken) approach would be for the investor to put $625,000 in stocks and $375,000 in bonds and not rebalance her portfolio. That's because (($625,000 - 80%) + $375,00) = $500,000. This is broken because it doesn't stand the test of logic. Portfolios aren't static. If the retiree is in retirement, she'll need to withdraw money from the portfolio. If she withdraws from bonds, the portfolio could drop lower than $500,000. If she withdraws from stocks, she'll sell them at low (and lower) prices and she might quickly run out of stocks to sell. Anyway, let's put these considerations aside and continue our example with a static portfolio from which no money is withdrawn and to which money isn't added.

The logical approach is for the investor to consider what would happen if the -80% loss of stocks spanned over 200 consecutive days, while the portfolio was rebalanced daily. The daily stock loss would be: ((1 - 80%)^(1/200) - 1) = -0.8015%. As the investor's goal is to limit the portfolio's loss to 50% over that period, the target daily portfolio loss would be: ((1 - 50%)^(1/200) - 1) = -0.3460%. As a consequence, the investor puts (-0.3460% / -0.8015%) = 43.17% of her portfolio in stocks. In other words, the investor puts $431,663 in stocks and $568,336 in bonds and won't fear regularly rebalancing her portfolio because she knows that her portfolio won't lose more than half of its value, even in a catastrophic scenario where stocks lost -80% of their value.
Let's calculate a sensible asset allocation for an investor with a $1,000,000 portfolio allocated 55/45 stocks/bonds who is unwilling to rebalance bonds into stocks for fear of letting her portfolio shrink too much in a downturn.

We'll consider that a potential yet low-probability catastrophic scenario for stocks is a loss of -80%.

Our retiree has $550,000 in stocks and $450,000 in bonds. If stocks were to lose -80% and bonds remained $450,000 (without rebalancing), the portfolio would shrink to (($550,000 - 80%) + $450,000) = $560,000 and end up with a 20/80 stock/bond allocation after the loss.

Our retiree is thus aiming to limit portfolio losses to -44% if stocks ever lost -80%. Over 200 consecutive days, a cumulative -80% loss represents a ((1 - 80%)^(1/200) - 1) = -0.8015% daily loss. and a cumulative -44% loss represents a ((1 - 44%)^(1/200) - 1) = -0.2895% daily loss. So, if our retiree allocates (-0.2895% / -0.8015%) = 35% (rounded) of her portfolio to stocks, she'll protect her portfolio against catastrophic stock losses while staying the course and rebalancing her portfolio.

The fact that our retiree is considering not to rebalance her 55/45 stock/bond portfolio, trying to maximize potential portfolio gains when stocks are more expensive, yet she's willing to accept much lower potential portfolio gains with a smaller 20/80 stock/bond portfolio when stocks are -80% cheaper, is indicative of behavioral pitfalls. I would suggest that she simply puts her entire portfolio into Vanguard's Target Retirement Income Fund (VTINX), a globally-diversified One-Fund Portfolio with a 30/70 stock/bond allocation (close enough to 35% stocks). This way, even if stocks gradually lose a cumulative -80% over 200 days, she'll end up with a bigger portfolio ($1,000,000 X ((((1 - 0.8015%) X 30%) + 70%)^200) = $617,873) than if she kept a non-rebalanced 55/45 stock/bond portfolio, and a higher exposure (30%) to stocks once they get dirt cheap. Even better, she won't need to do anything as the fund is automatically rebalanced; she'll be able to enjoy her retirement, instead.
Last edited by longinvest on Sat Sep 26, 2020 10:29 pm, edited 1 time in total.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Post Reply