What worst case scenario should bonds cover?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

What worst case scenario should bonds cover?

Post by Progman » Sat May 04, 2019 5:57 pm

I think the investment approach should be a based on risk. The whole thing about AA and how much % stocks/bonds in general is based on risk - usually via a monte carlo simulation or age. I agree that approach is OK but one can have the same risk and better long term portfolio performance - if you believe there is a 100% chance that equities will eventually recover. History shows this to be true of course - and I do not worry particularly that society will fall apart catastrophically worse than we have seen over the 100 years. There is even a decent argument that we will not suffer another 1930's great depression (based on our recent history).

Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression, I have come to realize that drawing from purely bonds during depression is the efficient approach. The question is how much should one have in bonds to cover a worst case scenario. After some historical review, I am thinking that having 10 years of expenses (beyond annuities) in bonds might be a solid approach. Interested in others perspective.

Thesaints
Posts: 2828
Joined: Tue Jun 20, 2017 12:25 am

Re: What worst case scenario should bonds cover?

Post by Thesaints » Sat May 04, 2019 7:35 pm

A very cautious investor may have all of his future expenses in treasuries and invest the rest in stocks.
He would be too cautious though, since such an allocation would cover him under any imaginable circumstance, even those circumstances that are imaginable, but completely unrealistic, such as "stocks lose all their value and never recover".

So, why did you pick 10 years and not 5, or 15 ?
The AA is chosen precisely with a "worst case scenario" in mind. Those who pick 80/20 imagine worst cases less worse than those who pick 20/80.
That does not mean "only spending from bonds during downturns". First, because we don't know when downturns start and when they finish until much later. Second, because it is more efficient to maintain one's AA constant by rebalancing and always spend drawing in a proportional fashion.

Dandy
Posts: 5880
Joined: Sun Apr 25, 2010 7:42 pm

Re: What worst case scenario should bonds cover?

Post by Dandy » Sat May 04, 2019 9:34 pm

If you have enough you could follow Dr. Bernstein's idea of putting 20 or more years worth of drawdown needs in "safe' fixed income and invest the rest anyway you want. Does that leave potential money on the table? Most likely. But, if you have enough, the need/value of asset preservation often is more valuable than additional growth.

User avatar
arcticpineapplecorp.
Posts: 4269
Joined: Tue Mar 06, 2012 9:22 pm

Re: What worst case scenario should bonds cover?

Post by arcticpineapplecorp. » Sat May 04, 2019 9:44 pm

typically retirees may have anywhere from 30%-50% of their money in stocks. That's how the target date funds are set up. At the target date you're around 50% in stocks and you gradually move down to 30% but no lower. This is consistent with SWR. Concurrently the 4% withdrawal rate along with maintaining 30%-50% in stocks has allowed a high degree of success over a 30 year time period (see tables):

https://www.bogleheads.org/wiki/Safe_withdrawal_rates

What does this mean? If you have half your money in stocks and half your money in bonds and a 4% withdrawal rate sustained 30 years with a high likelihood then doesn't that mean you've got 15 years in bonds (and 15 in stocks)?

Would 15 years be enough to pull from bonds to wait for stocks to recover? Likely.
"May you live as long as you want and never want as long as you live" -- Irish Blessing | "Invest we must" -- Jack Bogle

User avatar
willthrill81
Posts: 11984
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What worst case scenario should bonds cover?

Post by willthrill81 » Sat May 04, 2019 9:45 pm

Progman wrote:
Sat May 04, 2019 5:57 pm
I think the investment approach should be a based on risk. The whole thing about AA and how much % stocks/bonds in general is based on risk - usually via a monte carlo simulation or age. I agree that approach is OK but one can have the same risk and better long term portfolio performance - if you believe there is a 100% chance that equities will eventually recover. History shows this to be true of course - and I do not worry particularly that society will fall apart catastrophically worse than we have seen over the 100 years. There is even a decent argument that we will not suffer another 1930's great depression (based on our recent history).

Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression, I have come to realize that drawing from purely bonds during depression is the efficient approach. The question is how much should one have in bonds to cover a worst case scenario. After some historical review, I am thinking that having 10 years of expenses (beyond annuities) in bonds might be a solid approach. Interested in others perspective.
Russia's and China's stock markets were both nationalized and went to zero with no recovery at all. No matter how remote you may believe it to be for a given country, it is a possibility.

And even if stocks do eventually recover, as a retiree, you can only wait so long before your assets are depleted. For instance, the Japanese stock market has not yet recovered to the highs it sent back in 1989. It may set new highs in the future, but if you only owned Japanese stock, could you have endured 30 years of cumulative losses?

Bonds are no sure thing either. In the period from 1978-1981, long-term U.S. treasuries had an inflation-adjusted maximum drawdown of 46.5%. Intermediate-term Treasuries were down by a third in real dollars. The 1940s were also very hard on U.S. bonds. Very few of the 'X years in bonds' plans that retirees craft account for the possibility that their bonds may lose real and significant value over time.

As such, all roads carry risk, and there's no way to eliminate this. The best that you can do is to balance out the various risks in play in a way that is logical and you can stick with.

Your strategy isn't unreasonable, but if I was going down this path, I'd strongly consider putting half of my bonds into a TIPS ladder using individual TIPS bought directly from the Treasury. That would help to reduce the risk of unexpected inflation, and owning individual TIPS means that if real yields spike or plummet, you can adjust your TIPS allocation accordingly.
“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

TomCat96
Posts: 692
Joined: Sun Oct 18, 2015 12:18 pm

Re: What worst case scenario should bonds cover?

Post by TomCat96 » Sat May 04, 2019 10:06 pm

willthrill81 wrote:
Sat May 04, 2019 9:45 pm
Progman wrote:
Sat May 04, 2019 5:57 pm
I think the investment approach should be a based on risk. The whole thing about AA and how much % stocks/bonds in general is based on risk - usually via a monte carlo simulation or age. I agree that approach is OK but one can have the same risk and better long term portfolio performance - if you believe there is a 100% chance that equities will eventually recover. History shows this to be true of course - and I do not worry particularly that society will fall apart catastrophically worse than we have seen over the 100 years. There is even a decent argument that we will not suffer another 1930's great depression (based on our recent history).

Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression, I have come to realize that drawing from purely bonds during depression is the efficient approach. The question is how much should one have in bonds to cover a worst case scenario. After some historical review, I am thinking that having 10 years of expenses (beyond annuities) in bonds might be a solid approach. Interested in others perspective.
Russia's and China's stock markets were both nationalized and went to zero with no recovery at all. No matter how remote you may believe it to be for a given country, it is a possibility.

And even if stocks do eventually recover, as a retiree, you can only wait so long before your assets are depleted. For instance, the Japanese stock market has not yet recovered to the highs it sent back in 1989. It may set new highs in the future, but if you only owned Japanese stock, could you have endured 30 years of cumulative losses?

Bonds are no sure thing either. In the period from 1978-1981, long-term U.S. treasuries had an inflation-adjusted maximum drawdown of 46.5%. Intermediate-term Treasuries were down by a third in real dollars. The 1940s were also very hard on U.S. bonds. Very few of the 'X years in bonds' plans that retirees craft account for the possibility that their bonds may lose real and significant value over time.

As such, all roads carry risk, and there's no way to eliminate this. The best that you can do is to balance out the various risks in play in a way that is logical and you can stick with.

Your strategy isn't unreasonable, but if I was going down this path, I'd strongly consider putting half of my bonds into a TIPS ladder using individual TIPS bought directly from the Treasury. That would help to reduce the risk of unexpected inflation, and owning individual TIPS means that if real yields spike or plummet, you can adjust your TIPS allocation accordingly.
+1

What bogleheads tend to assume is the concept of risk within a narrow band of possibilities, not from bad to good, but from mild to good.

When you have grown up in relative wealth, when your parents have, when your grandparents have, the concept of really bad becomes an abstraction you read about. Bad in the mind of these individuals comes along the lines of Stagflation, the Dot-Com burst, the 2009 crisis. Compare this with the Cultural revolution, or Venezuela, or with refugees of war escaping the Khmer Rouge and see how they process the question:

"What worst case scenario should bonds cover?"

I hope bogleheads realize that is a question that can speak to two completely different experiences. Your answers can be widely polarized, because the emotional conception of "worst case scenario" can be worlds apart.

Without delving further into that, let me just say that bonds should cover the worst case scenarios of :

1) Stock market crashes
2) Personal liquidity needs (loss of job, emergency fund, sickness, etc)

while assuming a functioning society, financial system, and governance.

Bonds should not cover situations of prolonged warfare, martial law, breakdown of society, hyperinflation.
Last edited by TomCat96 on Sat May 04, 2019 11:27 pm, edited 1 time in total.

User avatar
willthrill81
Posts: 11984
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What worst case scenario should bonds cover?

Post by willthrill81 » Sat May 04, 2019 10:13 pm

TomCat96 wrote:
Sat May 04, 2019 10:06 pm
while assuming a functioning society, financial system, and governance.

Bonds should not cover situations of prolonged warfare, martial law, breakdown of society, hyperinflation.
In the latter situations, the best investments may be in things like water, food, medical supplies, gold, brass, and lead. Some might look at that with a smirk and think "yeah, right," but I personally know of at least one deca-millionaire who's got all of that and more, and he's not at all a 'survivalist'. He can easily afford to use a tiny percentage of his net worth to at least partially hedge against those kinds of risk, and history has shown that those risks are very real and not as tiny as many of us think.
“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
Dialectical Investor
Posts: 527
Joined: Mon Dec 03, 2018 11:41 pm

Re: What worst case scenario should bonds cover?

Post by Dialectical Investor » Sat May 04, 2019 10:17 pm

TomCat96 wrote:
Sat May 04, 2019 10:06 pm

Without delving further into that, let me just say that bonds should cover the worst case scenarios of :

1) Stock market crashes
2) Personal liquidity needs (loss of job, emergency fund, sickness, etc)

while assuming a functioning society, financial system, and governance.

Bonds should not cover situations of prolonged warfare, martial law, breakdown of society, hyperinflation.
It's possible to have a generally functioning society while also having prolonged warfare, martial law, and inflation that might not be hyperinflation but is nevertheless much higher than what we've experienced lately. The US has had all three of those. Perhaps the degree you had in mind is much worse, but that just goes back to the original point you were supporting that some investors think only in extreme terms.

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 6:28 am

Thesaints wrote:
Sat May 04, 2019 7:35 pm
That does not mean "only spending from bonds during downturns". First, because we don't know when downturns start and when they finish until much later. Second, because it is more efficient to maintain one's AA constant by rebalancing and always spend drawing in a proportional fashion.
I was thinking about most efficient use of capital - which to me means that if equities are down, pulling from bonds to feed expenses while equities are down so you don't lock in a 'loss'. Agreed one doesn't know when a real downturn starts but one could have a defined trigger. I like to be well insulated from the market by having around 2 years of expenses in a money market (that is my medicine to sleep good!), so I refill that once a year and have to decide at that point if I pull from equities or bonds. So I could just look at the last year, if equities are down pull from bonds, if equities are up pull from equities.

I get your second point about efficiency. It is totally valid. I am a geeky engineer accustomed to doing risk analysis on technical processes and wanted to explore quantifying the specific risk(s).

Ferdinand2014
Posts: 486
Joined: Mon Dec 17, 2018 6:49 pm

Re: What worst case scenario should bonds cover?

Post by Ferdinand2014 » Sun May 05, 2019 6:41 am

Progman wrote:
Sat May 04, 2019 5:57 pm
I think the investment approach should be a based on risk. The whole thing about AA and how much % stocks/bonds in general is based on risk - usually via a monte carlo simulation or age. I agree that approach is OK but one can have the same risk and better long term portfolio performance - if you believe there is a 100% chance that equities will eventually recover. History shows this to be true of course - and I do not worry particularly that society will fall apart catastrophically worse than we have seen over the 100 years. There is even a decent argument that we will not suffer another 1930's great depression (based on our recent history).

Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression, I have come to realize that drawing from purely bonds during depression is the efficient approach. The question is how much should one have in bonds to cover a worst case scenario. After some historical review, I am thinking that having 10 years of expenses (beyond annuities) in bonds might be a solid approach. Interested in others perspective.
1.) The idea that drawing from bonds during a depression is an efficient approach sounds a bit like a bucket approach, meaning only one way rebalancing. There is some data suggesting this approach may not actually be ideal despite it’s intuitive sense. In short, the article below found a simple 60/40 balanced fund over the past 115 years using annual rebalancing regardless of market conditions and using a withdrawal of 4% inflation adjusted had a failure rate of 0 vs a 2.3% failure rate using a bucket approach.

https://poseidon01.ssrn.com/delivery.ph ... 05&EXT=pdf

Abstract
“A bucket approach, which broadly consists of parking a few years of annual withdrawals safely in cash and investing the rest of the portfolio more aggressively, is a popular strategy often recommended by financial planners and typically embraced by retirees. Although this strategy is not devoid of merit, the comprehensive evidence discussed here, from 21 countries over a 115-year period, questions its effectiveness. In fact, simple static strategies, which by definition involve periodic rebalancing, clearly outperform bucket strategies, and they do so based not just on one but on four different ways of assessing performance.”
October, 2018

2.) Most every stock downturn at least in the U.S. according to Jeremy Siegel (Stocks for the long run) recovers to the previous high within 5 years. The article above however, seems to refute the idea of a number of years of safe savings as a hedge against failure rates in retirement.
“You only find out who is swimming naked when the tide goes out.“ — Warren Buffett

Ferdinand2014
Posts: 486
Joined: Mon Dec 17, 2018 6:49 pm

Re: What worst case scenario should bonds cover?

Post by Ferdinand2014 » Sun May 05, 2019 6:53 am

Dialectical Investor wrote:
Sat May 04, 2019 10:17 pm
TomCat96 wrote:
Sat May 04, 2019 10:06 pm

Without delving further into that, let me just say that bonds should cover the worst case scenarios of :

1) Stock market crashes
2) Personal liquidity needs (loss of job, emergency fund, sickness, etc)

while assuming a functioning society, financial system, and governance.

Bonds should not cover situations of prolonged warfare, martial law, breakdown of society, hyperinflation.
It's possible to have a generally functioning society while also having prolonged warfare, martial law, and inflation that might not be hyperinflation but is nevertheless much higher than what we've experienced lately. The US has had all three of those. Perhaps the degree you had in mind is much worse, but that just goes back to the original point you were supporting that some investors think only in extreme terms.
My view, although it is certainly important to consider range of risk severity and likelihood, I think there are simply some who shouldn’t invest in stocks at all. You have to be an optimist to a degree about the future of civilization in order to invest in stocks.
“You only find out who is swimming naked when the tide goes out.“ — Warren Buffett

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 6:54 am

Dandy wrote:
Sat May 04, 2019 9:34 pm
If you have enough you could follow Dr. Bernstein's idea of putting 20 or more years worth of drawdown needs in "safe' fixed income and invest the rest anyway you want. Does that leave potential money on the table? Most likely. But, if you have enough, the need/value of asset preservation often is more valuable than additional growth.
That goes along with the idea that money does not buy happiness - which I totally agree. And that is the idea I am after, to have a rock solid financial basis for the rest of my life (and wife), while maximizing the Legacy which for me might be primarily charity.

I chose 10 years without reviewing history. So if I say 'how long does it take to get back to the high value before the loss' as an example. Looking at historical S&P 500 we went 27 years for the great depression of 1930's. Then the next big one was from the early 70's to the late 80's - say 15 years. Looking at this context, 20 years might be as good as any estimate. This makes me like annuities even more....

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 6:59 am

arcticpineapplecorp. wrote:
Sat May 04, 2019 9:44 pm
typically retirees may have anywhere from 30%-50% of their money in stocks. That's how the target date funds are set up. At the target date you're around 50% in stocks and you gradually move down to 30% but no lower. This is consistent with SWR. Concurrently the 4% withdrawal rate along with maintaining 30%-50% in stocks has allowed a high degree of success over a 30 year time period (see tables):

https://www.bogleheads.org/wiki/Safe_withdrawal_rates

What does this mean? If you have half your money in stocks and half your money in bonds and a 4% withdrawal rate sustained 30 years with a high likelihood then doesn't that mean you've got 15 years in bonds (and 15 in stocks)?

Would 15 years be enough to pull from bonds to wait for stocks to recover? Likely.
That's an interesting thought. I'll chew on that one....

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 7:10 am

willthrill81 wrote:
Sat May 04, 2019 9:45 pm

Russia's and China's stock markets were both nationalized and went to zero with no recovery at all. No matter how remote you may believe it to be for a given country, it is a possibility.
I don't see this as a possibility in my lifetime. This kind of thing is what I meant by society falling apart.This approach makes the idea of AA and monte carlo simulation not valid at all. So let's put that aside.

User avatar
tennisplyr
Posts: 2268
Joined: Tue Jan 28, 2014 1:53 pm
Location: Sarasota, FL

Re: What worst case scenario should bonds cover?

Post by tennisplyr » Sun May 05, 2019 7:11 am

Retired 8 years; I basically withdraw from the asset class that will roughly keep my AA at my target over time.
Those who move forward with a happy spirit will find that things always work out.

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 7:19 am

TomCat96 wrote:
Sat May 04, 2019 10:06 pm

I hope bogleheads realize that is a question that can speak to two completely different experiences. Your answers can be widely polarized, because the emotional conception of "worst case scenario" can be worlds apart.

Without delving further into that, let me just say that bonds should cover the worst case scenarios of :

1) Stock market crashes
2) Personal liquidity needs (loss of job, emergency fund, sickness, etc)

while assuming a functioning society, financial system, and governance.

Bonds should not cover situations of prolonged warfare, martial law, breakdown of society, hyperinflation.
Exactly - Let's steer clear of that stuff. Hey, if that stuff happens, then we start again with different income and expenses.

Sconie
Posts: 675
Joined: Sun Feb 07, 2010 10:23 am
Location: Arizona

Re: What worst case scenario should bonds cover?

Post by Sconie » Sun May 05, 2019 7:25 am

I keep approximately 10 years worth of add'l spending----that is, spending beyond pensions and social security----in my cash and fixed pools, with the remainder in equities. For me, this results in about a 60% equity/40% fixed asset allocation.

Most recent economic history (excluding data from the great Depression of the 1930s) demonstrates that in the event of a stock market crash, if you just do nothing----but at least don't sell out----you'll be even about 5 years, and in about 10 years, you will be pleased that you "hung in there."
I know you think you understand what you thought I said but I'm not sure you realize that what you heard is not what I meant. - Alan Greenspan

MnD
Posts: 4299
Joined: Mon Jan 14, 2008 12:41 pm

Re: What worst case scenario should bonds cover?

Post by MnD » Sun May 05, 2019 7:32 am

Progman wrote:
Sat May 04, 2019 5:57 pm
Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression, I have come to realize that drawing from purely bonds during depression is the efficient approach.
How did you "come to realize" this?
Although popular among some here, this approach decreases returns and increases the risk of overall portfolio failure as tested over the past 115 years across 21 countries versus maintaining a fixed asset allocation with rebalancing. A balanced portfolio contains many years of fixed income and rebalancing and/or spending from the overweight asset class naturally results in spending from fixed income during market declines and from equity during hot markets. There is zero need for the mental accounting and complexity of some bucket of cash/bonds for down markets.

https://www.marketwatch.com/story/do-bu ... 2019-02-12
https://papers.ssrn.com/sol3/papers.cfm ... id=3274499

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 8:56 am

Ferdinand2014 wrote:
Sun May 05, 2019 6:41 am
1.) The idea that drawing from bonds during a depression is an efficient approach sounds a bit like a bucket approach, meaning only one way rebalancing. There is some data suggesting this approach may not actually be ideal despite it’s intuitive sense. In short, the article below found a simple 60/40 balanced fund over the past 115 years using annual rebalancing regardless of market conditions and using a withdrawal of 4% inflation adjusted had a failure rate of 0 vs a 2.3% failure rate using a bucket approach.

Abstract
“A bucket approach, which broadly consists of parking a few years of annual withdrawals safely in cash and investing the rest of the portfolio more aggressively, is a popular strategy often recommended by financial planners and typically embraced by retirees. Although this strategy is not devoid of merit, the comprehensive evidence discussed here, from 21 countries over a 115-year period, questions its effectiveness. In fact, simple static strategies, which by definition involve periodic rebalancing, clearly outperform bucket strategies, and they do so based not just on one but on four different ways of assessing performance.”
October, 2018

2.) Most every stock downturn at least in the U.S. according to Jeremy Siegel (Stocks for the long run) recovers to the previous high within 5 years. The article above however, seems to refute the idea of a number of years of safe savings as a hedge against failure rates in retirement.
On point 1 - you are correct at the 10,000 foot level that I am describing a bucket approach. And the article does elicit a good point that the 60/40 w/annual rebalancing is simple and effective. Details are important though on the bucket approach. The author of the paper had to make specific bucket criteria which I think is doomed to fail - the maximum bond bucket was 5 years expenses - which is getting at the specific question of this thread - how big is the bond bucket?. So I think picking this short of time periods is doomed to fail - this thread seems to be in the 15-20 yr range is more appropriate. I wonder if there is an article somewhere on bigger bond buckets?

On point 2 - As I noted before we had a 15 year period 70's-80's for the S&P 500 to recover.

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 9:01 am

Ferdinand2014 wrote:
Sun May 05, 2019 6:53 am
Dialectical Investor wrote:
Sat May 04, 2019 10:17 pm
TomCat96 wrote:
Sat May 04, 2019 10:06 pm

Without delving further into that, let me just say that bonds should cover the worst case scenarios of :

1) Stock market crashes
2) Personal liquidity needs (loss of job, emergency fund, sickness, etc)

while assuming a functioning society, financial system, and governance.

Bonds should not cover situations of prolonged warfare, martial law, breakdown of society, hyperinflation.
It's possible to have a generally functioning society while also having prolonged warfare, martial law, and inflation that might not be hyperinflation but is nevertheless much higher than what we've experienced lately. The US has had all three of those. Perhaps the degree you had in mind is much worse, but that just goes back to the original point you were supporting that some investors think only in extreme terms.
My view, although it is certainly important to consider range of risk severity and likelihood, I think there are simply some who shouldn’t invest in stocks at all. You have to be an optimist to a degree about the future of civilization in order to invest in stocks.
Agreed - I really value high quality annuities for this reason....and having a 2 year emergency fund....

Ferdinand2014
Posts: 486
Joined: Mon Dec 17, 2018 6:49 pm

Re: What worst case scenario should bonds cover?

Post by Ferdinand2014 » Sun May 05, 2019 9:02 am

Progman wrote:
Sun May 05, 2019 8:56 am
Ferdinand2014 wrote:
Sun May 05, 2019 6:41 am
1.) The idea that drawing from bonds during a depression is an efficient approach sounds a bit like a bucket approach, meaning only one way rebalancing. There is some data suggesting this approach may not actually be ideal despite it’s intuitive sense. In short, the article below found a simple 60/40 balanced fund over the past 115 years using annual rebalancing regardless of market conditions and using a withdrawal of 4% inflation adjusted had a failure rate of 0 vs a 2.3% failure rate using a bucket approach.

Abstract
“A bucket approach, which broadly consists of parking a few years of annual withdrawals safely in cash and investing the rest of the portfolio more aggressively, is a popular strategy often recommended by financial planners and typically embraced by retirees. Although this strategy is not devoid of merit, the comprehensive evidence discussed here, from 21 countries over a 115-year period, questions its effectiveness. In fact, simple static strategies, which by definition involve periodic rebalancing, clearly outperform bucket strategies, and they do so based not just on one but on four different ways of assessing performance.”
October, 2018

2.) Most every stock downturn at least in the U.S. according to Jeremy Siegel (Stocks for the long run) recovers to the previous high within 5 years. The article above however, seems to refute the idea of a number of years of safe savings as a hedge against failure rates in retirement.
On point 1 - you are correct at the 10,000 foot level that I am describing a bucket approach. And the article does elicit a good point that the 60/40 w/annual rebalancing is simple and effective. Details are important though on the bucket approach. The author of the paper had to make specific bucket criteria which I think is doomed to fail - the maximum bond bucket was 5 years expenses - which is getting at the specific question of this thread - how big is the bond bucket?. So I think picking this short of time periods is doomed to fail - this thread seems to be in the 15-20 yr range is more appropriate. I wonder if there is an article somewhere on bigger bond buckets?

On point 2 - As I noted before we had a 15 year period 70's-80's for the S&P 500 to recover.
“Even including the recent financial crisis, which saw the worst bear market since the 1930s, the longest it has ever taken an investor to recover an original investment in the stock market (including reinvested dividends) was the five-year, eight-month period from August 2000 through April 2006.”

Jeremy Siegel “Stocks for the long run” He is using peer reviewed data going back to 1802. Lots of recovery data does not include reinvested dividends, only price.
“You only find out who is swimming naked when the tide goes out.“ — Warren Buffett

dbr
Posts: 29798
Joined: Sun Mar 04, 2007 9:50 am

Re: What worst case scenario should bonds cover?

Post by dbr » Sun May 05, 2019 9:03 am

Progman wrote:
Sat May 04, 2019 5:57 pm
Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression,
This is not the purpose of bonds. At least in the usual model of taking withdrawals from a mixed portfolio of stocks and bonds, bonds have nothing directly to do with a person's income flow. The effect of bonds on a portfolio in this scenario is to control how variable the evolution of the portfolio will be over time while taking money from the portfolio. Part of that picture of variability is how high the range of value achieved by the portfolio lies.

See the portfolio trajectories displayed in FireCalc for different choices of asset allocation to see this effect.

Understanding the effect of varying the asset allocation over time either as a plan in advance or in response to market variations is a complicated problem with only a highly variable statistical answer as an output. The conventional approach is to maintain the target allocation by rebalancing.

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 9:27 am

Sconie wrote:
Sun May 05, 2019 7:25 am
I keep approximately 10 years worth of add'l spending----that is, spending beyond pensions and social security----in my cash and fixed pools, with the remainder in equities. For me, this results in about a 60% equity/40% fixed asset allocation.

Most recent economic history (excluding data from the great Depression of the 1930s) demonstrates that in the event of a stock market crash, if you just do nothing----but at least don't sell out----you'll be even about 5 years, and in about 10 years, you will be pleased that you "hung in there."
This is exactly the idea I am at. So how do you decide when to draw from your fixed pools? That article posted about the buckets listed 3 criteria for deciding when to draw from fixed pool. The winners were when equity return drops below long term avg return or 5 year avg return (as opposed just being positive).

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 9:36 am

Ferdinand2014 wrote:
Sun May 05, 2019 9:02 am
Progman wrote:
Sun May 05, 2019 8:56 am
Ferdinand2014 wrote:
Sun May 05, 2019 6:41 am
1.) The idea that drawing from bonds during a depression is an efficient approach sounds a bit like a bucket approach, meaning only one way rebalancing. There is some data suggesting this approach may not actually be ideal despite it’s intuitive sense. In short, the article below found a simple 60/40 balanced fund over the past 115 years using annual rebalancing regardless of market conditions and using a withdrawal of 4% inflation adjusted had a failure rate of 0 vs a 2.3% failure rate using a bucket approach.

Abstract
“A bucket approach, which broadly consists of parking a few years of annual withdrawals safely in cash and investing the rest of the portfolio more aggressively, is a popular strategy often recommended by financial planners and typically embraced by retirees. Although this strategy is not devoid of merit, the comprehensive evidence discussed here, from 21 countries over a 115-year period, questions its effectiveness. In fact, simple static strategies, which by definition involve periodic rebalancing, clearly outperform bucket strategies, and they do so based not just on one but on four different ways of assessing performance.”
October, 2018

2.) Most every stock downturn at least in the U.S. according to Jeremy Siegel (Stocks for the long run) recovers to the previous high within 5 years. The article above however, seems to refute the idea of a number of years of safe savings as a hedge against failure rates in retirement.
On point 1 - you are correct at the 10,000 foot level that I am describing a bucket approach. And the article does elicit a good point that the 60/40 w/annual rebalancing is simple and effective. Details are important though on the bucket approach. The author of the paper had to make specific bucket criteria which I think is doomed to fail - the maximum bond bucket was 5 years expenses - which is getting at the specific question of this thread - how big is the bond bucket?. So I think picking this short of time periods is doomed to fail - this thread seems to be in the 15-20 yr range is more appropriate. I wonder if there is an article somewhere on bigger bond buckets?

On point 2 - As I noted before we had a 15 year period 70's-80's for the S&P 500 to recover.
“Even including the recent financial crisis, which saw the worst bear market since the 1930s, the longest it has ever taken an investor to recover an original investment in the stock market (including reinvested dividends) was the five-year, eight-month period from August 2000 through April 2006.”

Jeremy Siegel “Stocks for the long run” He is using peer reviewed data going back to 1802. Lots of recovery data does not include reinvested dividends, only price.
Thanks for pointing that out - I was looking at price only.

Ferdinand2014
Posts: 486
Joined: Mon Dec 17, 2018 6:49 pm

Re: What worst case scenario should bonds cover?

Post by Ferdinand2014 » Sun May 05, 2019 9:45 am

Progman wrote:
Sun May 05, 2019 9:27 am
Sconie wrote:
Sun May 05, 2019 7:25 am
I keep approximately 10 years worth of add'l spending----that is, spending beyond pensions and social security----in my cash and fixed pools, with the remainder in equities. For me, this results in about a 60% equity/40% fixed asset allocation.

Most recent economic history (excluding data from the great Depression of the 1930s) demonstrates that in the event of a stock market crash, if you just do nothing----but at least don't sell out----you'll be even about 5 years, and in about 10 years, you will be pleased that you "hung in there."
This is exactly the idea I am at. So how do you decide when to draw from your fixed pools? That article posted about the buckets listed 3 criteria for deciding when to draw from fixed pool. The winners were when equity return drops below long term avg return or 5 year avg return (as opposed just being positive).
When you need money, you draw proportionally from bond and stock allocation based on your asset allocation, then at some point every year or so rebalance as needed to your preset asset allocation in a 2 step approach. The amount you draw is based on your needs and estimated SWR. This will vary based on inflation and expense needs. The second step of rebalancing back into your predetermined asset allocation, forces rebalancing in both up and down markets in a 2 way fashion. The bucket approaches inferior outcomes are presumed to be from the only replenishing of bonds occurring after markets are up. This means never buying into stocks when the market is down, so you do not reap any benefit from 2 way rebalancing. Or to say more succinctly, not buying low and selling high.
“You only find out who is swimming naked when the tide goes out.“ — Warren Buffett

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 9:47 am

dbr wrote:
Sun May 05, 2019 9:03 am
Progman wrote:
Sat May 04, 2019 5:57 pm
Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression,
This is not the purpose of bonds. At least in the usual model of taking withdrawals from a mixed portfolio of stocks and bonds, bonds have nothing directly to do with a person's income flow. The effect of bonds on a portfolio in this scenario is to control how variable the evolution of the portfolio will be over time while taking money from the portfolio. Part of that picture of variability is how high the range of value achieved by the portfolio lies.

See the portfolio trajectories displayed in FireCalc for different choices of asset allocation to see this effect.

Understanding the effect of varying the asset allocation over time either as a plan in advance or in response to market variations is a complicated problem with only a highly variable statistical answer as an output. The conventional approach is to maintain the target allocation by rebalancing.
Referring to the Bold above - I am not talking about the usual model of mixed portfolio. As you point out, bonds stabilize the return over time. I am poking at a 'more efficient' approach (and yes, more complicated). The mixed model sells equities at a 'loss' during times of protracted market downturns. Drawing from only Bonds during those periods (Bonds more nearly do not lose value like equities), keeps equities in place so they can recover value over time - giving an improvement in LT portfolio growth/durability.

staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 9:40 am

Re: What worst case scenario should bonds cover?

Post by staythecourse » Sun May 05, 2019 9:49 am

I have 1 year of cash to cover emergencies and worst case scenario. I'm a physician married to one as well and I own my own practice so job stability is not an issue. I don't really care if the market tanks for another 10+ years as our monthly income covers our expenditures.

For me I get more worried about paying monthly bills (mortgage) then I do about how the market is doing. I have not seen the market move yet that has made be pay attention (including 2008) so I doubt it ever will. Now if I had a job that I could lose and NOT pay my bills from income I would be much more conservative. So I guess it all depends on what you consider anxiety provoking. For me it isn't the market.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

User avatar
HomerJ
Posts: 13136
Joined: Fri Jun 06, 2008 12:50 pm

Re: What worst case scenario should bonds cover?

Post by HomerJ » Sun May 05, 2019 9:54 am

willthrill81 wrote:
Sat May 04, 2019 9:45 pm
Bonds are no sure thing either. In the period from 1978-1981, long-term U.S. treasuries had an inflation-adjusted maximum drawdown of 46.5%. Intermediate-term Treasuries were down by a third in real dollars. The 1940s were also very hard on U.S. bonds. Very few of the 'X years in bonds' plans that retirees craft account for the possibility that their bonds may lose real and significant value over time.
There's nothing intrinsically risky with short to medium duration bond funds. They self-correct fairly quickly. Interest rates go up, sure the bond fund loses money, but it starts paying more too, and you're back to even fairly quickly.

Inflation is what hurts bonds. Inflation is a real danger.
Your strategy isn't unreasonable, but if I was going down this path, I'd strongly consider putting half of my bonds into a TIPS ladder using individual TIPS bought directly from the Treasury. That would help to reduce the risk of unexpected inflation, and owning individual TIPS means that if real yields spike or plummet, you can adjust your TIPS allocation accordingly.
I've been thinking about this as well. Why individual TIPs instead of a TIPs fund though?
The J stands for Jay

User avatar
JoMoney
Posts: 7264
Joined: Tue Jul 23, 2013 5:31 am

Re: What worst case scenario should bonds cover?

Post by JoMoney » Sun May 05, 2019 9:55 am

Individuals will have different risk preferences, personally I think 10 years of expenses in bonds is more than I would want (especially at current rates :shock: )... but I don't think it's entirely out of line for some.
I find it an interesting coincidence that if one is also considering the "4% SWR" as a guideline, that the classic 60/40 stock/bond allocation effectively has 10 years of 4% withdrawals in bonds.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

dbr
Posts: 29798
Joined: Sun Mar 04, 2007 9:50 am

Re: What worst case scenario should bonds cover?

Post by dbr » Sun May 05, 2019 9:55 am

Progman wrote:
Sun May 05, 2019 9:47 am
dbr wrote:
Sun May 05, 2019 9:03 am
Progman wrote:
Sat May 04, 2019 5:57 pm
Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression,
This is not the purpose of bonds. At least in the usual model of taking withdrawals from a mixed portfolio of stocks and bonds, bonds have nothing directly to do with a person's income flow. The effect of bonds on a portfolio in this scenario is to control how variable the evolution of the portfolio will be over time while taking money from the portfolio. Part of that picture of variability is how high the range of value achieved by the portfolio lies.

See the portfolio trajectories displayed in FireCalc for different choices of asset allocation to see this effect.

Understanding the effect of varying the asset allocation over time either as a plan in advance or in response to market variations is a complicated problem with only a highly variable statistical answer as an output. The conventional approach is to maintain the target allocation by rebalancing.
Referring to the Bold above - I am not talking about the usual model of mixed portfolio. As you point out, bonds stabilize the return over time. I am poking at a 'more efficient' approach (and yes, more complicated). The mixed model sells equities at a 'loss' during times of protracted market downturns. Drawing from only Bonds during those periods (Bonds more nearly do not lose value like equities), keeps equities in place so they can recover value over time - giving an improvement in LT portfolio growth/durability.
No, the mixed model buys equities when equities are down. How much you lose by simply leaving them alone and not rebalancing is open to discussion on the merits. It could be by accident that just selling bonds and leaving equities alone would rebalance the portfolio, but usually a serious decline in equity prices would require selling bonds for withdrawal and selling additional bonds to buy more stocks. An alternative is to rebalance out of stocks only when they are high but not to rebalance into stocks when the are low. This is probably somewhat equivalent to holding a little less in stocks and rebalancing rigorously.

dbr
Posts: 29798
Joined: Sun Mar 04, 2007 9:50 am

Re: What worst case scenario should bonds cover?

Post by dbr » Sun May 05, 2019 9:56 am

JoMoney wrote:
Sun May 05, 2019 9:55 am
Individuals will have different risk preferences, personally I think 10 years of expenses in bonds is more than I would want (especially at current rates :shock: )... but I don't think it's entirely out of line for some.
I find it an interesting coincidence that if one is also considering the "4% SWR" as a guideline, that the classic 60/40 stock/bond allocation effectively has 10 years of 4% withdrawals in bonds.
Which is why one wonders what is the thinking behind the idea that one has to have x% income in bonds.

dbr
Posts: 29798
Joined: Sun Mar 04, 2007 9:50 am

Re: What worst case scenario should bonds cover?

Post by dbr » Sun May 05, 2019 9:58 am

JoMoney wrote:
Sun May 05, 2019 9:55 am
Individuals will have different risk preferences, personally I think 10 years of expenses in bonds is more than I would want (especially at current rates :shock: )... but I don't think it's entirely out of line for some.
I find it an interesting coincidence that if one is also considering the "4% SWR" as a guideline, that the classic 60/40 stock/bond allocation effectively has 10 years of 4% withdrawals in bonds.
Which is why one wonders what is the thinking behind the idea that one has to have x% income in bonds.

If people are really worried about this kind of thing the answer that truly serves the financial concern is an inflation indexed SPIA or a TIPS ladder poor man's pension. The monkey wrench that is always in the works is inflation, which is an explicit part of the whole SWR study of portfolios.

User avatar
HomerJ
Posts: 13136
Joined: Fri Jun 06, 2008 12:50 pm

Re: What worst case scenario should bonds cover?

Post by HomerJ » Sun May 05, 2019 10:15 am

MnD wrote:
Sun May 05, 2019 7:32 am
Progman wrote:
Sat May 04, 2019 5:57 pm
Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression, I have come to realize that drawing from purely bonds during depression is the efficient approach.
How did you "come to realize" this?
Although popular among some here, this approach decreases returns and increases the risk of overall portfolio failure as tested over the past 115 years across 21 countries versus maintaining a fixed asset allocation with rebalancing. A balanced portfolio contains many years of fixed income and rebalancing and/or spending from the overweight asset class naturally results in spending from fixed income during market declines and from equity during hot markets. There is zero need for the mental accounting and complexity of some bucket of cash/bonds for down markets.

https://www.marketwatch.com/story/do-bu ... 2019-02-12
https://papers.ssrn.com/sol3/papers.cfm ... id=3274499
It's pretty much the same thing.

"Withdrawing from the asset class that is doing better" is the same thing as "spending bonds when stocks are down".

I plan to be 50/50 stocks/bonds with a 3 year CD ladder. Every year, I'll cash out a CD, and buy a new 3-year CD pulling money from whichever account is higher to get me back to 50/50.
The J stands for Jay

User avatar
willthrill81
Posts: 11984
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What worst case scenario should bonds cover?

Post by willthrill81 » Sun May 05, 2019 10:34 am

HomerJ wrote:
Sun May 05, 2019 9:54 am
willthrill81 wrote:
Sat May 04, 2019 9:45 pm
Bonds are no sure thing either. In the period from 1978-1981, long-term U.S. treasuries had an inflation-adjusted maximum drawdown of 46.5%. Intermediate-term Treasuries were down by a third in real dollars. The 1940s were also very hard on U.S. bonds. Very few of the 'X years in bonds' plans that retirees craft account for the possibility that their bonds may lose real and significant value over time.
There's nothing intrinsically risky with short to medium duration bond funds. They self-correct fairly quickly. Interest rates go up, sure the bond fund loses money, but it starts paying more too, and you're back to even fairly quickly.

Inflation is what hurts bonds. Inflation is a real danger.
I agree that inflation, not interest rate fluctuations, is the real danger for nominal bonds. That's a problem I have with putting all of one's fixed income in nominal bonds.
HomerJ wrote:
Sun May 05, 2019 9:54 am
willthrill81 wrote:
Sat May 04, 2019 9:45 pm
Your strategy isn't unreasonable, but if I was going down this path, I'd strongly consider putting half of my bonds into a TIPS ladder using individual TIPS bought directly from the Treasury. That would help to reduce the risk of unexpected inflation, and owning individual TIPS means that if real yields spike or plummet, you can adjust your TIPS allocation accordingly.
I've been thinking about this as well. Why individual TIPs instead of a TIPs fund though?
Individual TIPS give you greater control. Depending on your circumstances, you can shorten or lengthen the duration, and if real yields become sufficiently attractive, you can 'back up the truck'.
“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

Ferdinand2014
Posts: 486
Joined: Mon Dec 17, 2018 6:49 pm

Re: What worst case scenario should bonds cover?

Post by Ferdinand2014 » Sun May 05, 2019 10:42 am

HomerJ wrote:
Sun May 05, 2019 10:15 am
MnD wrote:
Sun May 05, 2019 7:32 am
Progman wrote:
Sat May 04, 2019 5:57 pm
Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression, I have come to realize that drawing from purely bonds during depression is the efficient approach.
How did you "come to realize" this?
Although popular among some here, this approach decreases returns and increases the risk of overall portfolio failure as tested over the past 115 years across 21 countries versus maintaining a fixed asset allocation with rebalancing. A balanced portfolio contains many years of fixed income and rebalancing and/or spending from the overweight asset class naturally results in spending from fixed income during market declines and from equity during hot markets. There is zero need for the mental accounting and complexity of some bucket of cash/bonds for down markets.

https://www.marketwatch.com/story/do-bu ... 2019-02-12
https://papers.ssrn.com/sol3/papers.cfm ... id=3274499
It's pretty much the same thing.

"Withdrawing from the asset class that is doing better" is the same thing as "spending bonds when stocks are down".

I plan to be 50/50 stocks/bonds with a 3 year CD ladder. Every year, I'll cash out a CD, and buy a new 3-year CD pulling money from whichever account is higher to get me back to 50/50.
The difference might be that if instead of only drawing from the up asset like in a bucket approach to achieve a 50/50 allocation, you actually also rebalance into a down asset, for example selling bonds to buy stock to achieve a 50/50 allocation. This way you rebalance in BOTH directions. Not just selling high, but also buying low. The article I pinned as well as MnD, points this out. The suggested withdraw strategy was to take out necessary expenses at whatever SWR is determined (4% perhaps), then in a second step at some point yearly, rebalance back into the AA you have chosen (50/50 for example). It is a subtle difference that forces 2 directional rebalancing with better outcomes then a bucket approach. For me, as I am in the accumulation phase, I have X cash for emergencies to sleep well at night with the rest in equities. However, at or near(5 years) retirement, I plan to determine an AA and stick with it using the above rebalancing technique.
“You only find out who is swimming naked when the tide goes out.“ — Warren Buffett

User avatar
willthrill81
Posts: 11984
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What worst case scenario should bonds cover?

Post by willthrill81 » Sun May 05, 2019 10:55 am

dbr wrote:
Sun May 05, 2019 9:58 am
JoMoney wrote:
Sun May 05, 2019 9:55 am
Individuals will have different risk preferences, personally I think 10 years of expenses in bonds is more than I would want (especially at current rates :shock: )... but I don't think it's entirely out of line for some.
I find it an interesting coincidence that if one is also considering the "4% SWR" as a guideline, that the classic 60/40 stock/bond allocation effectively has 10 years of 4% withdrawals in bonds.
Which is why one wonders what is the thinking behind the idea that one has to have x% income in bonds.
I've wondered that myself. If you view your portfolio in terms of 'years of spending', then rebalancing may only be one-way (i.e. selling stocks to buy bonds).
dbr wrote:
Sun May 05, 2019 9:58 am
If people are really worried about this kind of thing the answer that truly serves the financial concern is an inflation indexed SPIA or a TIPS ladder poor man's pension. The monkey wrench that is always in the works is inflation, which is an explicit part of the whole SWR study of portfolios.
I really wonder whether we would be discussing something like the '5% rule' if TIPS had been available in periods like the 1940s and 1970s.
“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

DetroitRick
Posts: 711
Joined: Wed Mar 23, 2016 9:28 am
Location: SE Michigan

Re: What worst case scenario should bonds cover?

Post by DetroitRick » Sun May 05, 2019 10:56 am

This is purely an individual decision without single "right" answer. For me (highly risk tolerant, and in drawn-down mode), I try to keep around 6 to 7 years worth of withdrawal needs (rather than expenses as a whole) in fixed income. So I'm not pegging my choice to total spending at all, just to expected portfolio withdrawals.

I'm balancing the risks to fixed income of unexpected inflation (yes, I do TIPS too), and several other major risk categories, against all the equity risks that exist. With a decent income stream from social security and other sources, coupled with a willingness to expand and contract spending as needed, this works for me. But I wouldn't necessarily recommend it to others, and it took me a lot of thought and analysis over several years to arrive at this with a decent comfort level. The issue really requires a close look at your own risk capacity and tolerance. One of the few advantages of experiencing some pretty severe downturns lies in getting to know yourself and how you react (as opposed to those brokerage house risk questionnaires). And while I'm now very comfortable with my choices, this was one of the more difficult investment issues that I've faced.

dbr
Posts: 29798
Joined: Sun Mar 04, 2007 9:50 am

Re: What worst case scenario should bonds cover?

Post by dbr » Sun May 05, 2019 11:04 am

willthrill81 wrote:
Sun May 05, 2019 10:55 am
dbr wrote:
Sun May 05, 2019 9:58 am
JoMoney wrote:
Sun May 05, 2019 9:55 am
Individuals will have different risk preferences, personally I think 10 years of expenses in bonds is more than I would want (especially at current rates :shock: )... but I don't think it's entirely out of line for some.
I find it an interesting coincidence that if one is also considering the "4% SWR" as a guideline, that the classic 60/40 stock/bond allocation effectively has 10 years of 4% withdrawals in bonds.
Which is why one wonders what is the thinking behind the idea that one has to have x% income in bonds.
I've wondered that myself. If you view your portfolio in terms of 'years of spending', then rebalancing may only be one-way (i.e. selling stocks to buy bonds).
dbr wrote:
Sun May 05, 2019 9:58 am
If people are really worried about this kind of thing the answer that truly serves the financial concern is an inflation indexed SPIA or a TIPS ladder poor man's pension. The monkey wrench that is always in the works is inflation, which is an explicit part of the whole SWR study of portfolios.
I really wonder whether we would be discussing something like the '5% rule' if TIPS had been available in periods like the 1940s and 1970s.
I suppose Pfau or somebody must have run a simulation of that, but I agree I don't remember seeing that result as such. The worst retirement years in the standard analysis are 1965-1970 and I suppose that is firstly as result of poor stock returns for a long time and then a whammy half way through due to inflation. But I am just supposing. It is a fair comment that early SWR studies and probably some models today don't offer what TIPS might have done if they had been around.

Turbo29
Posts: 541
Joined: Tue May 01, 2018 7:12 am

Re: What worst case scenario should bonds cover?

Post by Turbo29 » Sun May 05, 2019 11:36 am

willthrill81 wrote:
Sat May 04, 2019 10:13 pm
TomCat96 wrote:
Sat May 04, 2019 10:06 pm
while assuming a functioning society, financial system, and governance.

Bonds should not cover situations of prolonged warfare, martial law, breakdown of society, hyperinflation.
In the latter situations, the best investments may be in things like water, food, medical supplies, gold, brass, and lead. Some might look at that with a smirk and think "yeah, right," but I personally know of at least one deca-millionaire who's got all of that and more, and he's not at all a 'survivalist'. He can easily afford to use a tiny percentage of his net worth to at least partially hedge against those kinds of risk, and history has shown that those risks are very real and not as tiny as many of us think.
Bernstein has a take on this, "Back-of-the-envelope, that’s about an 80% survival rate over the next 40 years. Thus, any estimate of long-term financial success greater than about 80% is meaningless. "

http://www.efficientfrontier.com/ef/901/hell3.htm

User avatar
willthrill81
Posts: 11984
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What worst case scenario should bonds cover?

Post by willthrill81 » Sun May 05, 2019 11:49 am

Turbo29 wrote:
Sun May 05, 2019 11:36 am
willthrill81 wrote:
Sat May 04, 2019 10:13 pm
TomCat96 wrote:
Sat May 04, 2019 10:06 pm
while assuming a functioning society, financial system, and governance.

Bonds should not cover situations of prolonged warfare, martial law, breakdown of society, hyperinflation.
In the latter situations, the best investments may be in things like water, food, medical supplies, gold, brass, and lead. Some might look at that with a smirk and think "yeah, right," but I personally know of at least one deca-millionaire who's got all of that and more, and he's not at all a 'survivalist'. He can easily afford to use a tiny percentage of his net worth to at least partially hedge against those kinds of risk, and history has shown that those risks are very real and not as tiny as many of us think.
Bernstein has a take on this, "Back-of-the-envelope, that’s about an 80% survival rate over the next 40 years. Thus, any estimate of long-term financial success greater than about 80% is meaningless. "

http://www.efficientfrontier.com/ef/901/hell3.htm
This is one area where Bernstein and I definitely agree. Many folks' obsession with moving from a '94% success rate' to a '98% success rate' strikes me as squinting at a gnat while swallowing a camel.
“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

chambers136
Posts: 232
Joined: Tue Feb 28, 2017 9:49 am

Re: What worst case scenario should bonds cover?

Post by chambers136 » Sun May 05, 2019 12:03 pm

We keep 8-10 years of expenses in bonds and cash. This was chosen based on the time it takes for markets to recover, plus a cushion. I don’t think there’s much that could be done to deal with the market nationalization that people mention. If they were, there’s probably not much that would be safe, including bonds and real estate.

Valuethinker
Posts: 38436
Joined: Fri May 11, 2007 11:07 am

Re: What worst case scenario should bonds cover?

Post by Valuethinker » Sun May 05, 2019 12:38 pm

willthrill81 wrote:
Sat May 04, 2019 10:13 pm
TomCat96 wrote:
Sat May 04, 2019 10:06 pm
while assuming a functioning society, financial system, and governance.

Bonds should not cover situations of prolonged warfare, martial law, breakdown of society, hyperinflation.
In the latter situations, the best investments may be in things like water, food, medical supplies, gold, brass, and lead. Some might look at that with a smirk and think "yeah, right," but I personally know of at least one deca-millionaire who's got all of that and more, and he's not at all a 'survivalist'. He can easily afford to use a tiny percentage of his net worth to at least partially hedge against those kinds of risk, and history has shown that those risks are very real and not as tiny as many of us think.
Billionaires going all survivalist is a thing. Theres been a lot of stuff about it since 2016.

The most notable is Peter Theil a founder of ?PayPal?. Who has built a self sufficient bunker in New Zealand. And invested sufficient money and spent 20 days in the country and received NZ citizenship.

To which I am reminded of the man who foresaw WW2 coming and moved to the isolation and safety if the South Pacific and an island called....

Guadacanal

The very rich are kidding themselves if they think they can escape societal collapse.

Thesaints
Posts: 2828
Joined: Tue Jun 20, 2017 12:25 am

Re: What worst case scenario should bonds cover?

Post by Thesaints » Sun May 05, 2019 1:44 pm

The only bonds that can protect against the second coming of Lenin are those issued by another country’s government.

Topic Author
Progman
Posts: 24
Joined: Mon Apr 29, 2019 8:08 am

Re: What worst case scenario should bonds cover?

Post by Progman » Sun May 05, 2019 1:48 pm

Thanks for the discussion! Some very good points raised here. There were a few that use the concept of determining how much Bond's are appropriate for them based on how many years of draw-down that would provide. I did not mention that I am 2 years into retirement (retired at 58), so currently in draw-down (3% range). I am also fortunate to have a pension that covers over 50% of my budget + taxes. Now that I'm a couple years in I am relooking at things and am more in line with DetroitRick's thinking. I really like the TIPS discussion points which brings to mind some other specific questions about selections for my portfolio. I'll head over to the Personal Investments sub-Forum for more discussion - I'll try search first though.....

dbr
Posts: 29798
Joined: Sun Mar 04, 2007 9:50 am

Re: What worst case scenario should bonds cover?

Post by dbr » Sun May 05, 2019 2:10 pm

Thesaints wrote:
Sun May 05, 2019 1:44 pm
The only bonds that can protect against the second coming of Lenin are those issued by another country’s government.
Hopefully not the Weimar Republic. But then there is this: https://www.wsj.com/articles/SB860708223743220000 sorry for the paywall.

Almond
Posts: 38
Joined: Wed May 01, 2019 3:11 pm

Re: What worst case scenario should bonds cover?

Post by Almond » Mon May 06, 2019 2:23 pm

willthrill81 wrote:
Sat May 04, 2019 9:45 pm
Progman wrote:
Sat May 04, 2019 5:57 pm
I think the investment approach should be a based on risk. The whole thing about AA and how much % stocks/bonds in general is based on risk - usually via a monte carlo simulation or age. I agree that approach is OK but one can have the same risk and better long term portfolio performance - if you believe there is a 100% chance that equities will eventually recover. History shows this to be true of course - and I do not worry particularly that society will fall apart catastrophically worse than we have seen over the 100 years. There is even a decent argument that we will not suffer another 1930's great depression (based on our recent history).

Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression, I have come to realize that drawing from purely bonds during depression is the efficient approach. The question is how much should one have in bonds to cover a worst case scenario. After some historical review, I am thinking that having 10 years of expenses (beyond annuities) in bonds might be a solid approach. Interested in others perspective.



Russia's and China's stock markets were both nationalized and went to zero with no recovery at all. No matter how remote you may believe it to be for a given country, it is a possibility.

And even if stocks do eventually recover, as a retiree, you can only wait so long before your assets are depleted. For instance, the Japanese stock market has not yet recovered to the highs it sent back in 1989. It may set new highs in the future, but if you only owned Japanese stock, could you have endured 30 years of cumulative losses?

Bonds are no sure thing either. In the period from 1978-1981, long-term U.S. treasuries had an inflation-adjusted maximum drawdown of 46.5%. Intermediate-term Treasuries were down by a third in real dollars. The 1940s were also very hard on U.S. bonds. Very few of the 'X years in bonds' plans that retirees craft account for the possibility that their bonds may lose real and significant value over time.

As such, all roads carry risk, and there's no way to eliminate this. The best that you can do is to balance out the various risks in play in a way that is logical and you can stick with.

Your strategy isn't unreasonable, but if I was going down this path, I'd strongly consider putting half of my bonds into a TIPS ladder using individual TIPS bought directly from the Treasury. That would help to reduce the risk of unexpected inflation, and owning individual TIPS means that if real yields spike or plummet, you can adjust your TIPS allocation accordingly.
I just looked at the time period you said for long term treasury, seems max down was -5%. the following year or the year thereafter returned 40+%. What am I missing?

User avatar
willthrill81
Posts: 11984
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What worst case scenario should bonds cover?

Post by willthrill81 » Mon May 06, 2019 2:27 pm

Almond wrote:
Mon May 06, 2019 2:23 pm
willthrill81 wrote:
Sat May 04, 2019 9:45 pm
Progman wrote:
Sat May 04, 2019 5:57 pm
I think the investment approach should be a based on risk. The whole thing about AA and how much % stocks/bonds in general is based on risk - usually via a monte carlo simulation or age. I agree that approach is OK but one can have the same risk and better long term portfolio performance - if you believe there is a 100% chance that equities will eventually recover. History shows this to be true of course - and I do not worry particularly that society will fall apart catastrophically worse than we have seen over the 100 years. There is even a decent argument that we will not suffer another 1930's great depression (based on our recent history).

Recognizing that the purpose of bonds is to stabilize income flow during times of equity depression, I have come to realize that drawing from purely bonds during depression is the efficient approach. The question is how much should one have in bonds to cover a worst case scenario. After some historical review, I am thinking that having 10 years of expenses (beyond annuities) in bonds might be a solid approach. Interested in others perspective.



Russia's and China's stock markets were both nationalized and went to zero with no recovery at all. No matter how remote you may believe it to be for a given country, it is a possibility.

And even if stocks do eventually recover, as a retiree, you can only wait so long before your assets are depleted. For instance, the Japanese stock market has not yet recovered to the highs it sent back in 1989. It may set new highs in the future, but if you only owned Japanese stock, could you have endured 30 years of cumulative losses?

Bonds are no sure thing either. In the period from 1978-1981, long-term U.S. treasuries had an inflation-adjusted maximum drawdown of 46.5%. Intermediate-term Treasuries were down by a third in real dollars. The 1940s were also very hard on U.S. bonds. Very few of the 'X years in bonds' plans that retirees craft account for the possibility that their bonds may lose real and significant value over time.

As such, all roads carry risk, and there's no way to eliminate this. The best that you can do is to balance out the various risks in play in a way that is logical and you can stick with.

Your strategy isn't unreasonable, but if I was going down this path, I'd strongly consider putting half of my bonds into a TIPS ladder using individual TIPS bought directly from the Treasury. That would help to reduce the risk of unexpected inflation, and owning individual TIPS means that if real yields spike or plummet, you can adjust your TIPS allocation accordingly.
I just looked at the time period you said for long term treasury, seems max down was -5%. the following year or the year thereafter returned 40+%. What am I missing?
Inflation, which was rampant at the time.
“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

pascalwager
Posts: 1489
Joined: Mon Oct 31, 2011 8:36 pm

Re: What worst case scenario should bonds cover?

Post by pascalwager » Mon May 06, 2019 4:06 pm

Progman wrote:
Sun May 05, 2019 8:56 am
Ferdinand2014 wrote:
Sun May 05, 2019 6:41 am
1.) The idea that drawing from bonds during a depression is an efficient approach sounds a bit like a bucket approach, meaning only one way rebalancing. There is some data suggesting this approach may not actually be ideal despite it’s intuitive sense. In short, the article below found a simple 60/40 balanced fund over the past 115 years using annual rebalancing regardless of market conditions and using a withdrawal of 4% inflation adjusted had a failure rate of 0 vs a 2.3% failure rate using a bucket approach.

Abstract
“A bucket approach, which broadly consists of parking a few years of annual withdrawals safely in cash and investing the rest of the portfolio more aggressively, is a popular strategy often recommended by financial planners and typically embraced by retirees. Although this strategy is not devoid of merit, the comprehensive evidence discussed here, from 21 countries over a 115-year period, questions its effectiveness. In fact, simple static strategies, which by definition involve periodic rebalancing, clearly outperform bucket strategies, and they do so based not just on one but on four different ways of assessing performance.”
October, 2018

2.) Most every stock downturn at least in the U.S. according to Jeremy Siegel (Stocks for the long run) recovers to the previous high within 5 years. The article above however, seems to refute the idea of a number of years of safe savings as a hedge against failure rates in retirement.
On point 1 - you are correct at the 10,000 foot level that I am describing a bucket approach. And the article does elicit a good point that the 60/40 w/annual rebalancing is simple and effective. Details are important though on the bucket approach. The author of the paper had to make specific bucket criteria which I think is doomed to fail - the maximum bond bucket was 5 years expenses - which is getting at the specific question of this thread - how big is the bond bucket?. So I think picking this short of time periods is doomed to fail - this thread seems to be in the 15-20 yr range is more appropriate. I wonder if there is an article somewhere on bigger bond buckets?

On point 2 - As I noted before we had a 15 year period 70's-80's for the S&P 500 to recover.
Estrada discusses the reason for portfolio failure: The cash (not bond) bucket removes portfolio assets from the annual rebalancing procedure and results in reduced equity assets. Enlarging the cash bucket from five to 20 years would only increase the failure rate.

Post Reply