The logical fallacy behind "Bonds Are For Safety"

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 10:33 am

Few aphorisms are so blithely repeated as "bonds are for safety", and I've always wondered how such an illogical piece of advice came to become so commonly accepted.

Some of the popularity that this "bonds are for safety" has to do with the weak grasp that many investors, even professional investors, have on the concepts of portfolio diversification and derisking.

Another important factor, I suspect, is a logical fallacy known a the Fallacy of Composition.

Before presenting the actual definitions, let me present an example.

1) Fruit is tasty
2) Cake is tasty
3) Therefore, fruitcake must also be tasty. <h/t to Jim Gaffigan>
Fallacy of Composition: (also known as: composition fallacy, exception fallacy, faulty induction)

Description: Inferring that something is true of the whole from the fact that it is true of some part of the whole.
This fallacy, which I think manifests among investors as a behavioral bias towards short-term bonds, manifests when people erroneously extrapolate the characteristics of the component (e.g. the riskiness of the bond fund) up to the level of the aggregate (e.g. the riskiness of the portfolio) without taking into account how the components are arranged within the aggregate.

Example of the fallacy: bond fund A is less volatile than bond fund B, therefore a portfolio containing bond fund A must be less volatile than a portfolio containing bond fund B.

Students of portfolio construction will spot the error in the example statement, which is that the interaction between the bond fund and the other portfolio components is ignored.

Vanguard published a short note on behavioral finance in which what they call "framing" is really about to the Fallacy of Composition:
Finance theory recommends we treat all of our investments as a single pool, or portfolio, and consider how the risks of each investment offset the risks of others within the portfolio. We’re supposed to think comprehensively about our wealth. Rather than focusing on individual securities or simply our financial assets, traditional financial theory believes that we consider our wealth comprehensively, including our house, company pensions, government benefits and our ability to produce income.

However, human beings tend to focus overwhelmingly on the behaviour of individual investments or securities. As a result, in reviewing portfolios investors tend to fret over the poor performance of a specific asset class or security or mutual fund. These ‘narrow’ frames tend to increase investor sensitivity to loss. By contrast, by evaluating investments and performance at the aggregate level, with a ‘wide’ frame, investors tend to exhibit a greater tendency to accept short-term losses and their effects.
Virtus Investment Partners published a note called Diversification Means Always Having To Say You’re Sorry that also touches on this:
That’s where the second bias comes into play. Our cognitive tendency is to focus on pieces of the whole rather than the whole itself. Not only is it easier to focus on one versus many, it’s then that much more taxing to take on the relationships between the various pieces. When given transparency, we are not wired to see the one portfolio; we instead home in on the individual investments that comprise it.
What often flows as an outcome from the "bonds are for safety" heuristic is a portfolio that has a lower risk-adjusted return and less diversification than would be produced by a more rational heuristic.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

X528
Posts: 52
Joined: Fri Nov 30, 2018 8:51 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by X528 » Fri Jun 07, 2019 10:37 am

What happens to stocks and bonds in deflation?

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 10:42 am

X528 wrote:
Fri Jun 07, 2019 10:37 am
What happens to stocks and bonds in deflation?
The important question is "what happens to the portfolio in deflation", right?

And that will depend on which stocks it holds , which bonds it holds, and how those components interact.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

Leesbro63
Posts: 5877
Joined: Mon Nov 08, 2010 4:36 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Leesbro63 » Fri Jun 07, 2019 10:44 am

What’s your definition of “safety”? What’s your definition of “bond” ?

randomguy
Posts: 7822
Joined: Wed Sep 17, 2014 9:00 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by randomguy » Fri Jun 07, 2019 10:47 am

vineviz wrote:
Fri Jun 07, 2019 10:33 am
What often flows as an outcome from the "bonds are for safety" heuristic is a portfolio that has a lower risk-adjusted return and less diversification than would be produced by a more rational heuristic.
Care to give some example portfolios? And what is the definition of "safety" that we are optimizing for. Are we just concerned with short term volatility or do we care about actually having money. I have a feeling most people talking about safety only care about reduced volatility and don't care about returns as much.

azanon
Posts: 2430
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: The logical fallacy behind "Bonds Are For Safety"

Post by azanon » Fri Jun 07, 2019 10:51 am

You like logical fallacies too? I'm starting to form a crush. :mrgreen:

- Seriously, good post. You already know I agree with all of that, reference a conversation we had a few hrs ago.

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 11:17 am

randomguy wrote:
Fri Jun 07, 2019 10:47 am
vineviz wrote:
Fri Jun 07, 2019 10:33 am
What often flows as an outcome from the "bonds are for safety" heuristic is a portfolio that has a lower risk-adjusted return and less diversification than would be produced by a more rational heuristic.
Care to give some example portfolios? And what is the definition of "safety" that we are optimizing for. Are we just concerned with short term volatility or do we care about actually having money. I have a feeling most people talking about safety only care about reduced volatility and don't care about returns as much.
Sure.

My experience is that most investors equate "risk" with "volatility" consistent with the usage that Markowitz, Sharpe, et al. employed in classical MPT. So for simplicity of discussion, I propose we stick with that for the moment.

So let's look at one Vanguard stock fund and two Vanguard bond funds with reasonably long histories (back to 1987 at least):

• Vanguard 500 Index Investor (VFINX) has had an annualized standard deviation of returns of 14.82%.
• Vanguard Total Bond Market Index Fund (VBMFX) has had an annualized standard deviation of returns of 3.81%.
• Vanguard Long-Term Treasury Fund (VUSTX) has had an annualized standard deviation of returns of 9.69%.


Which 70/30 portfolio (VFINX/VBMFX or VFINX/VUSTX) had a lower volatility over the 1987 to 2019 timeframe? Answer.

I'm sure you can guess, but I suspect it is not the one that most "bonds are for safety" proponents would expect.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

dbr
Posts: 29442
Joined: Sun Mar 04, 2007 9:50 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by dbr » Fri Jun 07, 2019 11:19 am

Leesbro63 wrote:
Fri Jun 07, 2019 10:44 am
What’s your definition of “safety”? What’s your definition of “bond” ?
Indeed. The logical fallacy about bonds begins with failing to define "safe."

The other points are also valid, of course. The point that "bonds are safe" does not imply that "bonds are for safety" is an excellent one, assuming we think we know what "safe" might mean.

HEDGEFUNDIE
Posts: 2731
Joined: Sun Oct 22, 2017 2:06 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by HEDGEFUNDIE » Fri Jun 07, 2019 11:23 am

You know I’m with you vineviz, but here are the common objections:

1. Correlations between assets can change over time.

2. Behaviorally, even if the portfolio as a whole was safer, the max diversification approach would mean constituent assets could plunge individually, leading the investor to sell out of those individual assets.

3. A missed gain is not felt in the same way as an experienced loss. Another behavioral issue.

dbr
Posts: 29442
Joined: Sun Mar 04, 2007 9:50 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by dbr » Fri Jun 07, 2019 11:25 am

I think one of the most direct examples for how bonds are not safe is the result of safe withdrawal rate studies that one effective way to sabotage the longevity of a portfolio is to invest entirely in bonds while trying to withdraw at a rate that could be sustained with at least a modicum of stocks. In that case safety is defined as having little chance of running out of money too soon. Note that while a high bond allocation defends against the horrific "sequence of returns" risk it also introduces the even more damaging risk of not getting enough return to sustain income over time.

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 11:35 am

HEDGEFUNDIE wrote:
Fri Jun 07, 2019 11:23 am
You know I’m with you vineviz, but here are the common objections:

1. Correlations between assets can change over time.

2. Behaviorally, even if the portfolio as a whole was safer, the max diversification approach would mean constituent assets could plunge individually, leading the investor to sell out of those individual assets.

3. A missed gain is not felt in the same way as an experienced loss. Another behavioral issue.
Yep, I see those a lot as well. Let's hope no one raises them in earnest in this thread ;)
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

azanon
Posts: 2430
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: The logical fallacy behind "Bonds Are For Safety"

Post by azanon » Fri Jun 07, 2019 11:38 am

HEDGEFUNDIE wrote:
Fri Jun 07, 2019 11:23 am
You know I’m with you vineviz, but here are the common objections:

1. Correlations between assets can change over time.

2. Behaviorally, even if the portfolio as a whole was safer, the max diversification approach would mean constituent assets could plunge individually, leading the investor to sell out of those individual assets.

3. A missed gain is not felt in the same way as an experienced loss. Another behavioral issue.
I have a admittedly harsh opinion regarding #2. If #2 is a problem for an individual, I think the first step for them to try to find an adviser. And if they ARE the adviser, remember it's never too late for a career change.

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 11:41 am

dbr wrote:
Fri Jun 07, 2019 11:19 am
Leesbro63 wrote:
Fri Jun 07, 2019 10:44 am
What’s your definition of “safety”? What’s your definition of “bond” ?
Indeed. The logical fallacy about bonds begins with failing to define "safe."
I totally agree with this, and I think it's an important enough point that it deserves it's own thread.

The "bonds are for safety" argument, it seems to me, is often presented with an implicit assumption that safety is exclusively defined as variance.

In other words, cash is completely safe because a dollar is always worth a dollar. An investor who invests their portfolio entirely in a rolling 30-day Treasury bill ladder might be avoiding variance but they are exposing themselves to all sorts of other risks.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

dbr
Posts: 29442
Joined: Sun Mar 04, 2007 9:50 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by dbr » Fri Jun 07, 2019 11:49 am

vineviz wrote:
Fri Jun 07, 2019 11:41 am
dbr wrote:
Fri Jun 07, 2019 11:19 am
Leesbro63 wrote:
Fri Jun 07, 2019 10:44 am
What’s your definition of “safety”? What’s your definition of “bond” ?
Indeed. The logical fallacy about bonds begins with failing to define "safe."
I totally agree with this, and I think it's an important enough point that it deserves it's own thread.

The "bonds are for safety" argument, it seems to me, is often presented with an implicit assumption that safety is exclusively defined as variance.

In other words, cash is completely safe because a dollar is always worth a dollar. An investor who invests their portfolio entirely in a rolling 30-day Treasury bill ladder might be avoiding variance but they are exposing themselves to all sorts of other risks.
It goes back to the investor understanding his objectives and then risk becomes those things that threaten to defeat reaching those objectives. By the time you get down to what to invest your money in the answer could be anything is or is not risky or some degree of offsetting properties in between.

MnD
Posts: 4287
Joined: Mon Jan 14, 2008 12:41 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by MnD » Fri Jun 07, 2019 11:55 am

dbr wrote:
Fri Jun 07, 2019 11:25 am
Note that while a high bond allocation defends against the horrific "sequence of returns" risk it also introduces the even more damaging risk of not getting enough return to sustain income over time.
All that increasing bond allocations did in the horrific 1966 retiree sequence of returns (when 4% failed) was to accelerate their portfolio going bust by a couple of years.
Last edited by MnD on Fri Jun 07, 2019 11:57 am, edited 1 time in total.

Whakamole
Posts: 983
Joined: Wed Jan 13, 2016 9:59 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Whakamole » Fri Jun 07, 2019 11:56 am

vineviz wrote:
Fri Jun 07, 2019 11:41 am
dbr wrote:
Fri Jun 07, 2019 11:19 am
Leesbro63 wrote:
Fri Jun 07, 2019 10:44 am
What’s your definition of “safety”? What’s your definition of “bond” ?
Indeed. The logical fallacy about bonds begins with failing to define "safe."
I totally agree with this, and I think it's an important enough point that it deserves it's own thread.

The "bonds are for safety" argument, it seems to me, is often presented with an implicit assumption that safety is exclusively defined as variance.

In other words, cash is completely safe because a dollar is always worth a dollar. An investor who invests their portfolio entirely in a rolling 30-day Treasury bill ladder might be avoiding variance but they are exposing themselves to all sorts of other risks.
Perhaps more likely, investors may look at the SEC yield and take on additional risk (junk bonds, extended duration bonds, etc.) thinking that they are getting a better return without additional risk.

randomguy
Posts: 7822
Joined: Wed Sep 17, 2014 9:00 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by randomguy » Fri Jun 07, 2019 12:02 pm

vineviz wrote:
Fri Jun 07, 2019 11:17 am
randomguy wrote:
Fri Jun 07, 2019 10:47 am
vineviz wrote:
Fri Jun 07, 2019 10:33 am
What often flows as an outcome from the "bonds are for safety" heuristic is a portfolio that has a lower risk-adjusted return and less diversification than would be produced by a more rational heuristic.
Care to give some example portfolios? And what is the definition of "safety" that we are optimizing for. Are we just concerned with short term volatility or do we care about actually having money. I have a feeling most people talking about safety only care about reduced volatility and don't care about returns as much.
Sure.

My experience is that most investors equate "risk" with "volatility" consistent with the usage that Markowitz, Sharpe, et al. employed in classical MPT. So for simplicity of discussion, I propose we stick with that for the moment.

So let's look at one Vanguard stock fund and two Vanguard bond funds with reasonably long histories (back to 1987 at least):

• Vanguard 500 Index Investor (VFINX) has had an annualized standard deviation of returns of 14.82%.
• Vanguard Total Bond Market Index Fund (VBMFX) has had an annualized standard deviation of returns of 3.81%.
• Vanguard Long-Term Treasury Fund (VUSTX) has had an annualized standard deviation of returns of 9.69%.


Which 70/30 portfolio (VFINX/VBMFX or VFINX/VUSTX) had a lower volatility over the 1987 to 2019 timeframe? Answer.

I'm sure you can guess, but I suspect it is not the one that most "bonds are for safety" proponents would expect.
And I add a few more bonds

https://www.portfoliovisualizer.com/bac ... total3=100

and I get a safer (lower drawdown, lower volatility) than your portfolio. That is adding safety.

Your portfolio looks great. But it is also a period where duration risk didn't show up. Plot up 1970-1980 and decide if taking on duration risk is ok. The risk is by adding high risk/high return assets is that unless you have a lot of negative correlation, you run the risk of having both risks show up a the same time and you end up with large drop.

User avatar
Phineas J. Whoopee
Posts: 8390
Joined: Sun Dec 18, 2011 6:18 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Phineas J. Whoopee » Fri Jun 07, 2019 12:11 pm

The definitional fallacy behind "Bonds Are For Safety:"

Safety is an emotive word, not a financial term of art. Let's stop using it on this forum.

PJW

jmk
Posts: 462
Joined: Tue Nov 01, 2011 7:48 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by jmk » Fri Jun 07, 2019 12:12 pm

You are not correct: there is no "fallacy of composition" in saying "Bonds are for Safety".
Of course there are different conceptions of what the term safety means (volatility, versus ability to not run out of money, etc).
(Your fallacy is using a vague word "safety" instead of a precise financial term.)
But in the sense used in the phrase, bonds are safer than stocks because of legal aspects (get paid first in bankruptcies) and in terms of their volatility.
If you're trying to match a liability with a known amount on a certain date, there is nothing safer than some sort of bond (or bond-like investment).
Each bond has these qualities, and therefore a sum of a collection of bonds share these qualities.

Are there times when stocks did better than bonds? Yes, but they are rare.
Last edited by jmk on Fri Jun 07, 2019 12:31 pm, edited 1 time in total.

DesertDiva
Posts: 412
Joined: Thu Mar 01, 2018 12:49 pm
Location: In the desert

Re: The logical fallacy behind "Bonds Are For Safety"

Post by DesertDiva » Fri Jun 07, 2019 12:21 pm

Phineas J. Whoopee wrote:
Fri Jun 07, 2019 12:11 pm
The definitional fallacy behind "Bonds Are For Safety:"

Safety is an emotive word, not a financial term of art. Let's stop using it on this forum.

PJW
I like the idea of bonds as a ballast.

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 12:27 pm

randomguy wrote:
Fri Jun 07, 2019 12:02 pm
And I add a few more bonds
and I get a safer (lower drawdown, lower volatility) than your portfolio. That is adding safety.
I think you're illustrating my point quite well.

The logical approach is to restrict the goals to the portfolio in aggregate, being indifferent to the individual components.

Yes, it is possible to lower the volatility of a portfolio merely by allocating more to bonds (or to cash, for that matter). That's called derisking, and it's totally legitimate.

Your 50/50 VFINX/VBMFX portfolio had lower variance than my examples, and that's fine. But it achieves that lower variance in naive manner (just throw in more of some random bond fund that looks safe) rather than in a manner that is most likely to meet your investment goals.

You prefer a portfolio that with an expected volatility closer to 8% than to 10.5%? Fine, you can do that with even LESS than 50% stocks if you're smart about it.

https://www.portfoliovisualizer.com/bac ... total3=100

Remember, if markets are reasonable efficient then the following must be true: for any given level of expected portfolio variance the most highly diversified portfolio will have the highest expected return.

The "bonds are for safety" approach has the effect of engineering the LEAST efficient and diversified portfolio possible: it pushes ALL the risk onto a single source like the equity risk premium instead of diversifying it across the equity risk premium and the fixed income term/duration premium.

Your 50/50 portfolio puts 92% of the risk exposure on the equities. Why would you do that if you could instead built a portfolio with higher expected return, the same portfolio variance, and much higher diversification (only 68% of risk on equities)?
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 12:31 pm

jmk wrote:
Fri Jun 07, 2019 12:12 pm
You are not correct: there is no "fallacy of composition" in saying "Bonds are for Safety".
That's not my saying: I've lost track of how many times I've seen that exact phrase here, but Google tells me it's over 500.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

bgf
Posts: 958
Joined: Fri Nov 10, 2017 9:35 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by bgf » Fri Jun 07, 2019 12:31 pm

any argument founded on the basis of predicting future correlation gets thrown into the same garbage bin as the argument founded on the basis of predicting future return.

i see no reason why one should be able to predict future correlation with any more accuracy or precision than future return.

this is like kind of a problem for any argument that portfolio's should be constructed in a manner that requires, and is in fact entirely built upon, assumptions of great precision as to these two variables.

it a lot easier to just understand what a bond is fundamentally and what a share in a company is fundamentally, and then make the best decision you can based on your time horizon, wealth, and future liabilities.

i've noticed vineviz uses the phrase 'if you're smart' in managing a portfolio. this is a code for 'if you predict unpredictable things ahead of time'

edit* id also like to make clear that this is fundamentally a margin of error problem. in other words, i think that optimization by means of this strategy requires precision far greater than one should reasonably anticipate being able to hit, especially over like time periods.

this just makes it a nonstarter from the get go.
“TE OCCIDERE POSSUNT SED TE EDERE NON POSSUNT NEFAS EST"

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 12:43 pm

bgf wrote:
Fri Jun 07, 2019 12:31 pm
any argument founded on the basis of predicting future correlation gets thrown into the same garbage bin as the argument founded on the basis of predicting future return.
There's no argument here base on any sort of prediction: "Bonds Are For Safety" depends on a logical fallacy, not on any sort of prediction of correlations.
bgf wrote:
Fri Jun 07, 2019 12:31 pm
i see no reason why one should be able to predict future correlation with any more accuracy or precision than future return.
However, I will address this point: it is unquestionably true (and empirically observable) that predicting future correlations can be done with MUCH MORE accuracy and precision than predicting future returns.

No one is claiming that correlations are invariant. They are not. They are, nonetheless, more predictable than returns.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

bgf
Posts: 958
Joined: Fri Nov 10, 2017 9:35 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by bgf » Fri Jun 07, 2019 12:53 pm

vineviz wrote:
Fri Jun 07, 2019 12:43 pm
However, I will address this point: it is unquestionably true (and empirically observable) that predicting future correlations can be done with MUCH MORE accuracy and precision than predicting future returns.

No one is claiming that correlations are invariant. They are not. They are, nonetheless, more predictable than returns.
that's interesting. over what time periods? which asset classes? what is the ultimate source of that robustness in correlation? also, i guess, fundamentally, it makes sense that correlation would be more predictable than returns without providing a further clarification.

correlation is just hitting the up or down, without mention of to what degree. return is not that simple. we aren't, for example, just talking about whether return is negative or positive, but across a broad range of possible returns.

so, if we include correlation and degree, how does that affect the comparison?
“TE OCCIDERE POSSUNT SED TE EDERE NON POSSUNT NEFAS EST"

User avatar
Mountain Doc
Posts: 89
Joined: Tue Aug 08, 2017 3:15 pm
Location: Life Elevated

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Mountain Doc » Fri Jun 07, 2019 1:20 pm

I have little doubt that a portfolio of stocks and long-term bonds has been the most efficient. Honestly, it has been a thing of beauty to see MPT at work.

However, I have little confidence that our historical record includes all possible crises. It is entirely possible that a future crisis might be one that is uniquely detrimental to both stocks and long-term bonds in favor of short-term bonds. Since I try to choose a portfolio that minimizes the risk of severe regret, and therefore one I can stick with through all circumstances, I default to broad diversification (in the sense of more asset classes - even within bonds).

So, yes, I guess I am raising objection #1 in earnest :D

randomguy
Posts: 7822
Joined: Wed Sep 17, 2014 9:00 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by randomguy » Fri Jun 07, 2019 1:43 pm

vineviz wrote:
Fri Jun 07, 2019 12:27 pm
randomguy wrote:
Fri Jun 07, 2019 12:02 pm
And I add a few more bonds
and I get a safer (lower drawdown, lower volatility) than your portfolio. That is adding safety.
I think you're illustrating my point quite well.

The logical approach is to restrict the goals to the portfolio in aggregate, being indifferent to the individual components.

Yes, it is possible to lower the volatility of a portfolio merely by allocating more to bonds (or to cash, for that matter). That's called derisking, and it's totally legitimate.

Your 50/50 VFINX/VBMFX portfolio had lower variance than my examples, and that's fine. But it achieves that lower variance in naive manner (just throw in more of some random bond fund that looks safe) rather than in a manner that is most likely to meet your investment goals.

You prefer a portfolio that with an expected volatility closer to 8% than to 10.5%? Fine, you can do that with even LESS than 50% stocks if you're smart about it.

https://www.portfoliovisualizer.com/bac ... total3=100

Remember, if markets are reasonable efficient then the following must be true: for any given level of expected portfolio variance the most highly diversified portfolio will have the highest expected return.

The "bonds are for safety" approach has the effect of engineering the LEAST efficient and diversified portfolio possible: it pushes ALL the risk onto a single source like the equity risk premium instead of diversifying it across the equity risk premium and the fixed income term/duration premium.

Your 50/50 portfolio puts 92% of the risk exposure on the equities. Why would you do that if you could instead built a portfolio with higher expected return, the same portfolio variance, and much higher diversification (only 68% of risk on equities)?
You are arguing for risk-adjust performance not saftey. But ignore that. The questions is what evidence do you have that taking risk on the bond side works out? Again how did your portfolio do in the rising interest rate environments of the 60s/70s. Were you getting higher returns and lower variance or are you just posting a chart of 1 time period were your portfolio happened to do great?

Fallible
Posts: 6885
Joined: Fri Nov 27, 2009 4:44 pm
Contact:

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Fallible » Fri Jun 07, 2019 1:45 pm

Phineas J. Whoopee wrote:
Fri Jun 07, 2019 12:11 pm
The definitional fallacy behind "Bonds Are For Safety:"

Safety is an emotive word, not a financial term of art. Let's stop using it on this forum.

PJW
Good point, good idea.

Also, when recommending some bonds in a portfolio, top pros who obviously know the downsides of bonds, clearly mean relatively safe compared with stocks.
"John Bogle has changed a basic industry in the optimal direction. Of very few can this be said." ~Paul A. Samuelson

stlutz
Posts: 5132
Joined: Fri Jan 02, 2009 1:08 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by stlutz » Fri Jun 07, 2019 2:01 pm

vineviz wrote:
Fri Jun 07, 2019 10:33 am
Few aphorisms are so blithely repeated as "bonds are for safety", and I've always wondered how such an illogical piece of advice came to become so commonly accepted.

Some of the popularity that this "bonds are for safety" has to do with the weak grasp that many investors, even professional investors, have on the concepts of portfolio diversification and derisking.
A while back I laid out what I view to be the 4 main approaches to fixed income that are in use on this board:

a) The "bonds are for safety" approach. For these investors, they are seeking returns from domestic and international stocks and using bonds to lever down their exposure. With this approach, short term bonds are most appropriate.

b) The "non-correlated investments" approach. For these investors, they are looking for multiple sources of risk/volatility in their portfolio. Stocks and long-term bonds are both volatile. Sometimes they correlate positively; sometimes they don't. If you can't handle the drawdown that could result from the two having positively correlation, then the portfolio should be levered down using short-term bonds.

c) Then there are people who are looking for bonds to generate income. In this case, the duration of the portfolio should be one-half of the period you need the income to be spread over. For the income investor, fluctuation in principal value is irrelevant. What matters is stability of income. This is achieved by matching duration with when you need the money.

d) Finally there are the people who just want to hold a balanced portfolio of assets and don't want to think about it any further. In that case, a total bond market fund probably makes sense. It will provide some portfolio stability, possibly some negative correlation, and hopefully some modest positive return.
Are you really suggesting that using bonds to lever down your stock portfolio is really a completely illogical approach to portfolio construction?

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 2:13 pm

randomguy wrote:
Fri Jun 07, 2019 1:43 pm
You are arguing for risk-adjust performance not saftey. But ignore that. The questions is what evidence do you have that taking risk on the bond side works out? Again how did your portfolio do in the rising interest rate environments of the 60s/70s. Were you getting higher returns and lower variance or are you just posting a chart of 1 time period were your portfolio happened to do great?
You keep dragging returns into the discussion, but I can't understand why. My original post said nothing about higher returns (risk-adjusted or otherwise), and that was clearly NOT the point I was making.

My point is that the common aphorism "bonds are for safety" is predicated on a logically fallacious premise.

That premise is this: part of the portfolio has property P therefore the portfolio must have property P.

I think it would be helpful if all investors understood why that premise is false. Don't you?
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
stockpickerted
Posts: 52
Joined: Thu Dec 11, 2008 10:14 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by stockpickerted » Fri Jun 07, 2019 2:20 pm

Simply put, Bonds are safe, they pay interest, they mature and you get your investment back! Enough said!

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 2:22 pm

Mountain Doc wrote:
Fri Jun 07, 2019 1:20 pm
So, yes, I guess I am raising objection #1 in earnest :D
So here's the thing: the optimal response to a realization that you don't have perfect foresight is not to discard all the good information you do have.

Right?

You can't be 100% sure that wearing your seatbelt will protect you in a car crash, but it would be clearly irrational to use that as a justification for never wearing your seatbelt at all.

So we know that past returns don't predict future returns very well at all. We also know that past correlations do predict future correlations quite well, and that past variance predicts future variance quite well. Not 100% perfectly, granted, but darn close. What do you do with that information? The only logical thing to do, in my mind, is to incorporate it as much as you reasonably can into your investment process.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
305pelusa
Posts: 286
Joined: Fri Nov 16, 2018 10:20 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by 305pelusa » Fri Jun 07, 2019 2:23 pm

vineviz wrote:
Fri Jun 07, 2019 2:13 pm

That premise is this: part of the portfolio has property P therefore the portfolio must have property P.
If the statement in question was "a portfolio with bonds must be safe because bonds are safe", then you're correct. That's fallacy by composition.

"Bonds are for safety" argues something else entirely. That the main (and perhaps sole) purpose of a bond is to derisk a portfolio. I don't agree with it, but I'm not sure the fallacy is by composition

If people assumed their portfolios were safe because it had bonds (even just one singular bond), I agree with you. But what people are assuming is that the more bonds they add, the less risk they have.

Which is not an illogical assumption.

Either way, I like your points. I agree with them. Still unconvinced it's that kind of fallacy you claim, at leas not in the traditional sense I've heard that fallacy being used.

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 2:24 pm

stockpickerted wrote:
Fri Jun 07, 2019 2:20 pm
Simply put, Bonds are safe, they pay interest, they mature and you get your investment back! Enough said!
Unless they default. Or you need to sell before they mature. Or you need a higher rate of return than they offer. Or . . .
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

abc132
Posts: 253
Joined: Thu Oct 18, 2018 1:11 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by abc132 » Fri Jun 07, 2019 2:29 pm

I agree that there are those that seem to equate volatility to risk, and that this is can be inappropriate, particularly in the accumulation phase.

If we can find several assets that are not completely correlated with each other or that vary in correlation, we will be taking on less asset specific risk than using only one or two assets.

A portfolio with short, medium, and long term bonds can handle more worst case scenarios.

A portfolio with stocks, bonds AND other assets can handle more worst case scenarios.

For most real portfolio's, some risks can probably be lowered without harming performance and/or expected performance.


From these ideas, we have to consider what works for the individual investor:

1. Boglehead investors value simplicity

For that investor that can't stop second guessing those 5 things in their portfolio, having one or two things might be an optimal choice based on their preferences and/or emotional make up. For the investor that wants to make zero decisions, a 50/50 stock/bond portfolio is a way to rationalize not making a decision, even though they have made a decision.

2. Diversification

We don't know in advance what composition of diversification will be beneficial. How to diversify becomes tricky. We know we should do it, but we don't know in advance how to do it well.

3. Risk vs Return

We all would like to avoid risk, but many of us also invest for return. Taking two ideas that are often counter to each other, and finding an optimum is very difficult for many people. Taking one idea way too far is much more common than being able to incorporate several ideas to make a good decision. There are tools to do this, but the average person is just not wired to do this on their own. One idea is going to be emotionally satisfying, and that emotion often drives decisions.

User avatar
Mountain Doc
Posts: 89
Joined: Tue Aug 08, 2017 3:15 pm
Location: Life Elevated

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Mountain Doc » Fri Jun 07, 2019 2:35 pm

vineviz wrote:
Fri Jun 07, 2019 2:22 pm
So here's the thing: the optimal response to a realization that you don't have perfect foresight is not to discard all the good information you do have.

Right?

You can't be 100% sure that wearing your seatbelt will protect you in a car crash, but it would be clearly irrational to use that as a justification for never wearing your seatbelt at all.

So we know that past returns don't predict future returns very well at all. We also know that past correlations do predict future correlations quite well, and that past variance predicts future variance quite well. Not 100% perfectly, granted, but darn close. What do you do with that information? The only logical thing to do, in my mind, is to incorporate it as much as you reasonably can into your investment process.
Maybe a better analogy is this: suppose the historical data show that seatbelts were the lifesaving measure in 95% of car crashes, but airbags were the lifesaving measure in the other 5%. If I can only choose one, then by all means I will choose the seatbelt. But I don't have to choose one or the other, I can have both.

I think your choice of using only long-term treasuries will likely lead to the most efficient portfolio. But I am not confident enough in that outcome to go all in. I want some protection against the unlikely scenario that long-term bonds and stocks both get crushed. It is a choice about managing deep risk, not about optimizing my portfolio for the most likely outcome.

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 2:36 pm

305pelusa wrote:
Fri Jun 07, 2019 2:23 pm
"Bonds are for safety" argues something else entirely. That the main (and perhaps sole) purpose of a bond is to derisk a portfolio. I don't agree with it, but I'm not sure the fallacy is by composition.
It's always tricky to understand what people mean when they say something like ""bonds are for safety", so maybe you're right.

One context I often see ""bonds are for safety" employed is as justification for using short-term bonds (or even cash vehicles, like CDs or stable value funds) to derisk a portfolio, even when intermediate or long term bonds would not only derisk the portfolio even more while also improving its diversification and expected return.

The mathematics of diversification are pretty cut-and-dry: among assets with similar correlations to a portfolio, the asset with the highest variance which is capable of derisking the portfolio to a target volatility should be preferred.

"Bonds are for safety" contains, to me at least, an implicit claim that the characteristics of the bonds in isolation are a legitimate investment goal. In reality, a purely rational investor should be indifferent about the volatility of the portfolio components, caring only about the volatility of the combined portfolio.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 2:47 pm

Mountain Doc wrote:
Fri Jun 07, 2019 2:35 pm
I think your choice of using only long-term treasuries will likely lead to the most efficient portfolio. But I am not confident enough in that outcome to go all in. I want some protection against the unlikely scenario that long-term bonds and stocks both get crushed. It is a choice about managing deep risk, not about optimizing my portfolio for the most likely outcome.
Don't focus on long-term bonds. That's a distraction. Focus instead on the important takeaway (which I think you probably get correctly, by the way): the portfolio matters, not the pieces.

If you are genuinely worried about a scenario I which both long-term bonds and stocks get crushed, you want a portfolio that provides some reasonable protection for you in that scenario. Right? That's not relying on a "bonds are for safety" approach, it seems to me. It's trying to engineer a portfolio that in aggregate isolates you from that risk.

Another analogy: say there are two portfolios, each invested equally in two assets.

Portfolio 1 contains asset A which goes up 60% and asset B which goes down 10%.

Portfolio 2 contains asset A which goes up 20% and asset B which goes up 4%.

The "bonds are for safety" argument is akin to saying "asset B is for safety", so Portfolio 2 is the preferred portfolio. Ignoring the fact that Portfolio 1 had twice the return of Portfolio 2.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

azanon
Posts: 2430
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: The logical fallacy behind "Bonds Are For Safety"

Post by azanon » Fri Jun 07, 2019 2:52 pm

I'm not sure if you follow Tyler's work at portfoliocharts.com viveviz, but based on this recent article of his, ("High Profits at Low Rates : The Benefits of Bond Convexity": https://portfoliocharts.com/2019/05/27/ ... convexity/ ) he certainly gets the concept that Bonds are, or can be, so much more than just for safety.

Seasonal
Posts: 320
Joined: Sun May 21, 2017 1:49 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Seasonal » Fri Jun 07, 2019 2:58 pm

vineviz wrote:
Fri Jun 07, 2019 2:24 pm
stockpickerted wrote:
Fri Jun 07, 2019 2:20 pm
Simply put, Bonds are safe, they pay interest, they mature and you get your investment back! Enough said!
Unless they default. Or you need to sell before they mature. Or you need a higher rate of return than they offer. Or . . .
Default? Debt issued by a G7 country, e.g., US treasuries, are highly unlikely to default. Getting paid a known amount at a known date certainly sounds like a reasonable definition of safety. Dictionary: "the condition of being protected from or unlikely to cause danger, risk, or injury." Note the use of "unlikely".

Sell before they mature? If you want an investment that doesn't change in value, go for a money market fund. If memory serves, these invest in bonds.

You need a higher rate of return? Since when is that a common definition of safety?

User avatar
305pelusa
Posts: 286
Joined: Fri Nov 16, 2018 10:20 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by 305pelusa » Fri Jun 07, 2019 3:01 pm

vineviz wrote:
Fri Jun 07, 2019 2:36 pm

One context I often see ""bonds are for safety" employed is as justification for using short-term bonds (or even cash vehicles, like CDs or stable value funds) to derisk a portfolio, even when intermediate or long term bonds would not only derisk the portfolio even more while also improving its diversification and expected return.
But that's still not a fallacy by composition. You could say "sugar is for sweetening" even though there are other products that would sweeten more.

Would you claim that "sugar is for sweetening" is illogical simply because others sweeten more? Even if so, I don't quite follow the Fallacy of composition here.

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 3:11 pm

305pelusa wrote:
Fri Jun 07, 2019 3:01 pm
vineviz wrote:
Fri Jun 07, 2019 2:36 pm

One context I often see ""bonds are for safety" employed is as justification for using short-term bonds (or even cash vehicles, like CDs or stable value funds) to derisk a portfolio, even when intermediate or long term bonds would not only derisk the portfolio even more while also improving its diversification and expected return.
But that's still not a fallacy by composition. You could say "sugar is for sweetening" even though there are other products that would sweeten more.

Would you claim that "sugar is for sweetening" is illogical simply because others sweeten more? Even if so, I don't quite follow the Fallacy of composition here.
The fallacy is believing that the safer the bond fund is, the safer the portfolio is. Or that the safest bond fund always makes the safest portfolio.

I agree that a statement like "bonds can help lower the risk of a portfolio" would not be fallacious, and that's not the usage I had in mind.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
Mountain Doc
Posts: 89
Joined: Tue Aug 08, 2017 3:15 pm
Location: Life Elevated

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Mountain Doc » Fri Jun 07, 2019 3:20 pm

vineviz wrote:
Fri Jun 07, 2019 2:47 pm
Don't focus on long-term bonds. That's a distraction. Focus instead on the important takeaway (which I think you probably get correctly, by the way): the portfolio matters, not the pieces.

If you are genuinely worried about a scenario I which both long-term bonds and stocks get crushed, you want a portfolio that provides some reasonable protection for you in that scenario. Right? That's not relying on a "bonds are for safety" approach, it seems to me. It's trying to engineer a portfolio that in aggregate isolates you from that risk.

Another analogy: say there are two portfolios, each invested equally in two assets.

Portfolio 1 contains asset A which goes up 60% and asset B which goes down 10%.

Portfolio 2 contains asset A which goes up 20% and asset B which goes up 4%.

The "bonds are for safety" argument is akin to saying "asset B is for safety", so Portfolio 2 is the preferred portfolio. Ignoring the fact that Portfolio 1 had twice the return of Portfolio 2.
I'm with you on the portfolio being what matters, not the components. I'm not really defending the "bonds are for safety" mindset either, so I probably chose the wrong thread in which to opine :happy. I just don't think an appreciation for MPT makes long-term treasuries the only rational way to construct a portfolio. Put me in the "diversification is for safety" camp, and I'm not defining safety as volatility.

User avatar
305pelusa
Posts: 286
Joined: Fri Nov 16, 2018 10:20 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by 305pelusa » Fri Jun 07, 2019 3:21 pm

vineviz wrote:
Fri Jun 07, 2019 3:11 pm
305pelusa wrote:
Fri Jun 07, 2019 3:01 pm
vineviz wrote:
Fri Jun 07, 2019 2:36 pm

One context I often see ""bonds are for safety" employed is as justification for using short-term bonds (or even cash vehicles, like CDs or stable value funds) to derisk a portfolio, even when intermediate or long term bonds would not only derisk the portfolio even more while also improving its diversification and expected return.
But that's still not a fallacy by composition. You could say "sugar is for sweetening" even though there are other products that would sweeten more.

Would you claim that "sugar is for sweetening" is illogical simply because others sweeten more? Even if so, I don't quite follow the Fallacy of composition here.
The fallacy is believing that the safer the bond fund is, the safer the portfolio is. Or that the safest bond fund always makes the safest portfolio.

I agree that a statement like "bonds can help lower the risk of a portfolio" would not be fallacious, and that's not the usage I had in mind.
Fair enough. I thought "bonds are for safety" was more in the context of not chasing yield with bonds (I.e. staying away from high yield bonds or emerging markets bonds) since equities are much more tax efficient at giving yield.

But yes, in the context of "the safer the bond, the safer the portfolio of the bond itself" can be incorrect in many instances. I frankly hadn't heard this too much.

Cheers

randomguy
Posts: 7822
Joined: Wed Sep 17, 2014 9:00 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by randomguy » Fri Jun 07, 2019 3:27 pm

vineviz wrote:
Fri Jun 07, 2019 2:47 pm
Mountain Doc wrote:
Fri Jun 07, 2019 2:35 pm
I think your choice of using only long-term treasuries will likely lead to the most efficient portfolio. But I am not confident enough in that outcome to go all in. I want some protection against the unlikely scenario that long-term bonds and stocks both get crushed. It is a choice about managing deep risk, not about optimizing my portfolio for the most likely outcome.
Don't focus on long-term bonds. That's a distraction. Focus instead on the important takeaway (which I think you probably get correctly, by the way): the portfolio matters, not the pieces.

If you are genuinely worried about a scenario I which both long-term bonds and stocks get crushed, you want a portfolio that provides some reasonable protection for you in that scenario. Right? That's not relying on a "bonds are for safety" approach, it seems to me. It's trying to engineer a portfolio that in aggregate isolates you from that risk.

Another analogy: say there are two portfolios, each invested equally in two assets.

Portfolio 1 contains asset A which goes up 60% and asset B which goes down 10%.

Portfolio 2 contains asset A which goes up 20% and asset B which goes up 4%.

The "bonds are for safety" argument is akin to saying "asset B is for safety", so Portfolio 2 is the preferred portfolio. Ignoring the fact that Portfolio 1 had twice the return of Portfolio 2.
So what is the portfolio that is engineered to avoid stock market risk and long bond duration risk? Is it a say a portfolio of stocks and bonds with intermediate duration risk😅

I am pretty confidant that adding duration risk improves average returns. I also think it makes the worse cases worse.

DaftInvestor
Posts: 4555
Joined: Wed Feb 19, 2014 10:11 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by DaftInvestor » Fri Jun 07, 2019 3:28 pm

Personally, I like fruitcake.

GRP
Posts: 97
Joined: Wed Nov 22, 2017 5:35 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by GRP » Fri Jun 07, 2019 3:31 pm

DaftInvestor wrote:
Fri Jun 07, 2019 3:28 pm
Personally, I like fruitcake.
LMAO

User avatar
Topic Author
vineviz
Posts: 4388
Joined: Tue May 15, 2018 1:55 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by vineviz » Fri Jun 07, 2019 3:44 pm

randomguy wrote:
Fri Jun 07, 2019 3:27 pm
I am pretty confidant that adding duration risk improves average returns. I also think it makes the worse cases worse.
Again, totally not the topic of this thread but I'll humor you: can you give us an example of a time when both stocks crashed and long-term bonds crashed?

Say, a year in which both of them had -10% returns or worse? What kind of "worst case" are we talking about?
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

Broken Man 1999
Posts: 2703
Joined: Wed Apr 08, 2015 11:31 am

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Broken Man 1999 » Fri Jun 07, 2019 3:47 pm

Fruitcake soaked with rum is very good!

Me, I treat my bonds with the soft bigotry of low expectations concerning many of their measurements of suitability.

Broken Man 1999
“If I cannot drink Bourbon and smoke cigars in Heaven than I shall not go. " -Mark Twain

Teague
Posts: 1492
Joined: Wed Nov 04, 2015 6:15 pm

Re: The logical fallacy behind "Bonds Are For Safety"

Post by Teague » Fri Jun 07, 2019 4:02 pm

Phineas J. Whoopee wrote:
Fri Jun 07, 2019 12:11 pm
The definitional fallacy behind "Bonds Are For Safety:"

Safety is an emotive word, not a financial term of art. Let's stop using it on this forum.

PJW
That sounds risky. :happy
Semper Augustus

Locked