100% defensive stocks vs 60/40

[Charles Joseph]

There's no free lunch. If there was a sector reliably returning more than Treasury bonds with less risk, don't you think everyone would have moved their money there already?

Diversification can be a free lunch (relative to a view of markets that's incomplete – e.g. only factoring in stocks and bonds).

You are proposing the opposite.

Just pondering, but what are the views on managing risk by only investing in defensive sectors (in this case the three classic recession hedges: Consumer Staples, Healthcare and Utilities) vs stocks and bonds?

Logan Roy,

Until Telecom Bust, Telecom stocks were considered as defensive sector. How well did it works out during Telecom Bust?

KlangFool

Somehow I don't see a Big Bust in toothpaste and toilet paper.

Have you looked at what counts for consumer staples? Coke & Pepsi, Philip Morris and Altria. I don't see an argument for exchanging US treasury bonds for stock in soft drink and cigarette companies, even if they are large and diversified from their flagship products.

I totally agree. I just don't see a "soft drink, cigarette and diaper Bubble." I was half joking around.

Your point is taken. And I digress from the OP's post.

[Charles Joseph]

There's no free lunch. If there was a sector reliably returning more than Treasury bonds with less risk, don't you think everyone would have moved their money there already?

Diversification can be a free lunch (relative to a view of markets that's incomplete – e.g. only factoring in stocks and bonds).

You are proposing the opposite.

Just pondering, but what are the views on managing risk by only investing in defensive sectors (in this case the three classic recession hedges: Consumer Staples, Healthcare and Utilities) vs stocks and bonds?

Logan Roy,

Until Telecom Bust, Telecom stocks were considered as defensive sector. How well did it works out during Telecom Bust?

KlangFool

Somehow I don't see a Big Bust in toothpaste and toilet paper.

Have you looked at what counts for consumer staples? Coke & Pepsi, Philip Morris and Altria. I don't see an argument for exchanging US treasury bonds for stock in soft drink and cigarette companies, even if they are large and diversified from their flagship products.

I totally agree. I just don't see a "soft drink, cigarette and diaper Bubble." I was half joking around.

Your point is taken. And I digress from the OP's post.

But, if enough people moved into Consumer Staples for the illusion of safety, anything can happen.

[nigel_ht]

Quick and dirty run using annual 12 industry.



Survived 4% in 1929 pretty well.



1966 as well

[nigel_ht]

It’s a poor strategy to classify any industry or sector to be immune or impervious to market dislocation or permanent reduction of valuation especially when referring to past performance. As an example, there have been entire industries which either no longer exist or whose importance has been diminished over time. In 2009, all equities fell.

Do you really want to place 50% of your chips on just 3 industries?

It did well in 2008...and unless I did my math wrong this strategy did well in 1929 and 1966 too...

[Grt2bOutdoors]

It’s a poor strategy to classify any industry or sector to be immune or impervious to market dislocation or permanent reduction of valuation especially when referring to past performance. As an example, there have been entire industries which either no longer exist or whose importance has been diminished over time. In 2009, all equities fell.

Do you really want to place 50% of your chips on just 3 industries?

It did well in 2008...and unless I did my math wrong this strategy did well in 1929 and 1966 too...

As I mentioned above, industries and sectors can and do change over time. I would not think the utility sector of 1929 is quite the same as 2022. You are cherry picking specific years but choose 1930, 1931 it may not have done so well. My point is you are not immune from market volatility.

[nigel_ht]

It’s a poor strategy to classify any industry or sector to be immune or impervious to market dislocation or permanent reduction of valuation especially when referring to past performance. As an example, there have been entire industries which either no longer exist or whose importance has been diminished over time. In 2009, all equities fell.

Do you really want to place 50% of your chips on just 3 industries?

It did well in 2008...and unless I did my math wrong this strategy did well in 1929 and 1966 too...

As I mentioned above, industries and sectors can and do change over time. I would not think the utility sector of 1929 is quite the same as 2022. You are cherry picking specific years but choose 1930, 1931 it may not have done so well. My point is you are not immune from market volatility.

Yes, I “cherry picked” the worst years in US history since 1920 to retire in…

The data is there…so if you want to find years where it did worse, feel free but from just a quick pass it seems like the strategy has promise.

I’d probably go with my more diversified portfolio I posted that came in between his two sample portfolios but I don’t get the resistance if the data shows it actually worked when 60/40 didn’t.

[Beensabu]

It’s a poor strategy to classify any industry or sector to be immune or impervious to market dislocation or permanent reduction of valuation especially when referring to past performance. As an example, there have been entire industries which either no longer exist or whose importance has been diminished over time. In 2009, all equities fell.

Do you really want to place 50% of your chips on just 3 industries?

It did well in 2008...and unless I did my math wrong this strategy did well in 1929 and 1966 too...

As I mentioned above, industries and sectors can and do change over time. I would not think the utility sector of 1929 is quite the same as 2022. You are cherry picking specific years but choose 1930, 1931 it may not have done so well. My point is you are not immune from market volatility.

Yes, I “cherry picked” the worst years in US history since 1920 to retire in…

The data is there…so if you want to find years where it did worse, feel free but from just a quick pass it seems like the strategy has promise.

You're missing the point. And the point is that sectors change. Slowly, over time. The composition of the types of companies that predominate a sector changes. Regulations change. Things change. For example, if utilities over the course of a few years suddenly all went public... that would be a change. If the "stability" of healthcare has been due to pharma, and pharma happens to now make up a much smaller percentage of that sector pie than it did even back in 2008, that is a change. If everyone spends three decades bidding up the price of consumer staples because of how well they held up in the late 70s / early 80s, that's a change.

You cannot simply go "let's backtest this sector as far as the data allows'. Because the sector you're backtesting was different back then (it was different along various stretches of the back then, probably) than it is now. So... is it different? How? And is that difference likely to make a difference?

Way way back in the day, it was all about railroads, right? For decades upon decades. And then things changed.

[nigel_ht]

It’s a poor strategy to classify any industry or sector to be immune or impervious to market dislocation or permanent reduction of valuation especially when referring to past performance. As an example, there have been entire industries which either no longer exist or whose importance has been diminished over time. In 2009, all equities fell.

Do you really want to place 50% of your chips on just 3 industries?

It did well in 2008...and unless I did my math wrong this strategy did well in 1929 and 1966 too...

As I mentioned above, industries and sectors can and do change over time. I would not think the utility sector of 1929 is quite the same as 2022. You are cherry picking specific years but choose 1930, 1931 it may not have done so well. My point is you are not immune from market volatility.

Yes, I “cherry picked” the worst years in US history since 1920 to retire in…

The data is there…so if you want to find years where it did worse, feel free but from just a quick pass it seems like the strategy has promise.

You're missing the point. And the point is that sectors change. Slowly, over time. The composition of the types of companies that predominate a sector changes. Regulations change. Things change. For example, if utilities over the course of a few years suddenly all went public... that would be a change. If the "stability" of healthcare has been due to pharma, and pharma happens to now make up a much smaller percentage of that sector pie than it did even back in 2008, that is a change. If everyone spends three decades bidding up the price of consumer staples because of how well they held up in the late 70s / early 80s, that's a change.

You cannot simply go "let's backtest this sector as far as the data allows'. Because the sector you're backtesting was different back then (it was different along various stretches of the back then, probably) than it is now. So... is it different? How? And is that difference likely to make a difference?

Way way back in the day, it was all about railroads, right? For decades upon decades. And then things changed.

So the argument here is that there are no defensive sectors available in 2022 even though some are known today as defensive sectors with low volatility and also seemed to work as defensive sectors 100 years ago…because sectors change slowly over time…

Mkay.

Amusingly I bet if we talked factors it probably would be suddenly okay…even factors like “quality” have no real explanation why they work, just that they do.

Sector diversification seems to have value. A few sectors see lower volatility than others. Buying these seem to reduce your max drawdown especially when coupled with gold. They also seem to do okay in high inflation decades.

[Grt2bOutdoors]

It did well in 2008...and unless I did my math wrong this strategy did well in 1929 and 1966 too...

As I mentioned above, industries and sectors can and do change over time. I would not think the utility sector of 1929 is quite the same as 2022. You are cherry picking specific years but choose 1930, 1931 it may not have done so well. My point is you are not immune from market volatility.

Yes, I “cherry picked” the worst years in US history since 1920 to retire in…

The data is there…so if you want to find years where it did worse, feel free but from just a quick pass it seems like the strategy has promise.

You're missing the point. And the point is that sectors change. Slowly, over time. The composition of the types of companies that predominate a sector changes. Regulations change. Things change. For example, if utilities over the course of a few years suddenly all went public... that would be a change. If the "stability" of healthcare has been due to pharma, and pharma happens to now make up a much smaller percentage of that sector pie than it did even back in 2008, that is a change. If everyone spends three decades bidding up the price of consumer staples because of how well they held up in the late 70s / early 80s, that's a change.

You cannot simply go "let's backtest this sector as far as the data allows'. Because the sector you're backtesting was different back then (it was different along various stretches of the back then, probably) than it is now. So... is it different? How? And is that difference likely to make a difference?

Way way back in the day, it was all about railroads, right? For decades upon decades. And then things changed.

So the argument here is that there are no defensive sectors available in 2022 even though some are known today as defensive sectors with low volatility and also seemed to work as defensive sectors 100 years ago…because sectors change slowly over time…

Mkay.

Amusingly I bet if we talked factors it probably would be suddenly okay…even factors like “quality” have no real explanation why they work, just that they do.

Sector diversification seems to have value. A few sectors see lower volatility than others. Buying these seem to reduce your max drawdown especially when coupled with gold. They also seem to do okay in high inflation decades.

There was a time not too long ago, maybe about 15-16 years back when you could have picked up utilities for below book, today they are all trading above book and some of them are trading far above normal p/e's, if rates normalize, do you think your stodgy utility will be immune from the higher cost of capital? If certain regulations currently in force today were to be changed with the removal of certain incentives, do you think any industry wold be immune from that? The point is, what is lower volatility today, could certainly be higher volatility tomorrow and vice versa. If you say Coke/Pepsi is it because it seems steady, I offer for you to go back in time, like 30 years worth - I can tell you first hand that one or both of the two have run into operational issues during that time that affected their stock price and the growth in dividends was much less. If everyone started drinking tap water tomorrow, how low do you think the volatility would be? How many people are willing to pay $10 for a bag of potato chips? Or $5 for a bag that is now shrunk by half? Pepsi makes more than 60% of their operating profits from the sale of salty snacks. They are the market leader but raise the price high enough and they too will feel it.

[Beensabu]

So the argument here is that there are no defensive sectors available in 2022 even though some are known today as defensive sectors with low volatility and also seemed to work as defensive sectors 100 years ago…because sectors change slowly over time…

It's not that they're not. It's that they might not be anymore. But you don't know. Unless you do. Do you?

Not a great kind of surprise to have happen. And yet it might to many people.

[nigel_ht]

So the argument here is that there are no defensive sectors available in 2022 even though some are known today as defensive sectors with low volatility and also seemed to work as defensive sectors 100 years ago…because sectors change slowly over time…

Mkay.

Amusingly I bet if we talked factors it probably would be suddenly okay…even factors like “quality” have no real explanation why they work, just that they do.

Sector diversification seems to have value. A few sectors see lower volatility than others. Buying these seem to reduce your max drawdown especially when coupled with gold. They also seem to do okay in high inflation decades.

There was a time not too long ago, maybe about 15-16 years back when you could have picked up utilities for below book, today they are all trading above book and some of them are trading far above normal p/e's, if rates normalize, do you think your stodgy utility will be immune from the higher cost of capital? If certain regulations currently in force today were to be changed with the removal of certain incentives, do you think any industry wold be immune from that? The point is, what is lower volatility today, could certainly be higher volatility tomorrow and vice versa.

That’s true of anything.

If China invaded Taiwan global TSM would crash. If the USD loses reserve currency status US Treasuries would crash.

So what? Those seems a lot more likely than Americans actually stop buying sugar water and chips.

These sectors score well for volatility and quality factors and they are mostly stodgy with the exception of pharma.

And it’s not like most of what we do isn’t based on analysis of historical performance…

[nisiprius]

Stocks are not bonds.
"Defensive" stocks are not bonds.
Dividend stocks from companies with a tradition of dividend-paying are not bonds.
Specific stock sectors are not bonds.
Blue-chip stocks from companies that are household names are not bonds.
Market neutral long-short stock portfolios are not bonds.
Preferred stocks are not bonds.

tocks are participation in business risk.

Bonds are (often) legally enforceable contracts to pay specified numbers of dollars on specified dates.

Investment-grade bonds are bonds from issuers whose books have been examined and are thought by ratings agencies to have enough safety margin to make those contractual payments through business downturns.

A Schwab equity rating of "A" is not the same thing as a Moody's bond rating of "A."

Stocks and bonds are just fundamentally different things, and it's a mistake to look at them as if the only things we know about them are pragmatic wiggly lines of past behavior. We know more than that. They make their money in completely different ways.

100% defensive stocks is not 60/40, and it's not like 60/40. It's 100% stocks. 100% stocks is what it is, and may be suitable for some investors, but it does nothing but harm to think of them as different from what they are. My hunch is that people for whom 100% stocks is suitable are people who can look 100% stocks right in the eye and say "100% stocks."

Like the old joke, how many legs does a horse have if you call a tail a leg? Answer, four, because a tail is not a leg.

[Dude2]

Very early in an individual stock-picker's journey, they hopefully discover that, although Starbucks coffee is the greatest thing ever or HBO is the most wonderful purveyor of entertainment around, etc. this does not directly translate to stock price rising in accordance with their perception of how great it is. There's nothing that says that even if earnings are through the roof that that translates into stock price rising -- the flow of capital just as easily going elsewhere. Extend that to favored sectors. Nothing says that in times of distress that consumer staples or utilities must do well -- as if this were a physical law. Bonds on the other hand will do what their contract says they will do.

[nigel_ht]

Stocks are not bonds.
"Defensive" stocks are not bonds.
Dividend stocks from companies with a tradition of dividend-paying are not bonds.
Specific stock sectors are not bonds.
Blue-chip stocks from companies that are household names are not bonds.
Market neutral long-short stock portfolios are not bonds.
Preferred stocks are not bonds.

Stocks are participation in business risk.

Bonds are (often) legally enforceable contracts to pay specified numbers of dollars on specified dates.

Investment-grade bonds are bonds from issuers whose books have been examined and are thought by ratings agencies to have enough safety margin to make those contractual payments through business downturns.

A Schwab equity rating of "A" is not the same thing as a Moody's bond rating of "A."

Stocks and bonds are just fundamentally different things, and it's a mistake to look at them as if the only things we know about them are pragmatic wiggly lines of past behavior. We know more than that. They make their money in completely different ways.

100% defensive stocks is not 60/40, and it's not like 60/40. It's 100% stocks. 100% stocks is what it is, and may be suitable for some investors, but it does nothing but harm to think of them as different from what they are. My hunch is that people for whom 100% stocks is suitable are people who can look 100% stocks right in the eye and say "100% stocks."

Like the old joke, how many legs does a horse have if you call a tail a leg? Answer, four, because a tail is not a leg.

If you wanted to do 80/20 with 40% defensive stocks 40% VT and 20% LTT vs 60/40 VT/BND the risk profiles probably aren’t significantly different and you have higher diversification across sectors and maybe size (I think you get a little more mid cap). It probably also holds more value than growth.

Bonds are ballast but they aren’t necessarily the only ballast. Nor do they always provide downside protection as we’ve seen. Bonds, Stocks, RE, cash, commodities, etc are all just tools to achieve our financial goals.

https://www.portfoliovisualizer.com/bac ... tion8_2=25

If you have lower volatility, lower max drawdown and a higher sustainable withdrawal rate from adding defensive sectors vs classic 60/40 why isn’t that worth consideration?

Especially if it seems to have worked in 1966 and 1929.

[nigel_ht]

Bonds on the other hand will do what their contract says they will do.

Unless they default…Russian debt anyone? Anyone had any Worldcom or Enron bonds? I’m sure some was in BND.

Nothing says that in times of distress that consumer staples or utilities must do well -- as if this were a physical law.

No physical law says flight to safety will end up in US treasuries but I’m holding treasuries partly because that behavior has made treasuries move opposite from the market in the past.

Otherwise why not hold any other AAA rated sovereign debt instead? Heck US is only AA+ with S&P.

If the USD loses reserve currency status that behavior will most likely be greatly reduced.

And for sure, nothing says bonds will do well in times of distress either…anyone who thought so learned differently in 2022…

[Dude2]

No physical law says flight to safety will end up in US treasuries but I’m holding treasuries partly because that behavior has made treasuries move opposite from the market in the past.
And for sure, nothing says bonds will do well in times of distress either…anyone who thought so learned differently in 2022…

I don't count on anything related to an extra bonus push from bonds when stocks tank. Maybe that is irrational exuberance in action where people are suddenly willing to pay more for something today than it was valued yesterday, making the items I already own seem more valuable. However, these are short-lived things, and the tide goes in and out. I buy bonds for the fundamentals of bonds, and I buy stocks hoping that they will meet their obligations to pay me more than what bonds are paying. That's the risk. I think this thread addresses if there is such a thing as less risky stocks that you can know about with foresight, and so if markets tank, you will do better than the total market. On the other hand, when stocks shine, you'd do worse than the total market. All of that may be what we're after here, but the point is to not believe for an instant that there is certainty here like there is certainty with bonds.

[vineviz]

If you wanted to do 80/20 with 40% defensive stocks 40% VT and 20% LTT vs 60/40 VT/BND the risk profiles probably aren’t significantly different and you have higher diversification across sectors and maybe size (I think you get a little more mid cap). It probably also holds more value than growth.

Yeah, I think the primary red flag in the OP is the "100%" part of " 100% defensive stocks vs 60/40".

I can think of any number of reasons that increasing the weight of "defensive" sectors relative to TSM might improve outcomes but, as I pointed out before, there are ways to achieve similar outcomes without schemes as crude and arbitrary as sector or industry classification.

IMHO it would take a combination of hubris and naïveté to extend that all the way toward owning ONLY stocks from three sectors based solely on backtests, without reference to any fundamental or theoretical explanation for assessing whether that result might continue to hold in the future.

[nisiprius]

...If you have lower volatility, lower max drawdown and a higher sustainable withdrawal rate from adding defensive sectors vs classic 60/40 why isn’t that worth consideration?...

Because you just used the present tense (although you were careful to qualify it with "if.") Perhaps they have had these, but they are not intrinsic characteristics, they are just observed past behavior.

For example, they do not "have" lower volatility than the whole stock market. All you can say is that over some specified time period they have had lower volatility.

The difference between equity and debt is a fundamental characteristic. Stocks and bonds are different things. They make their money in different ways. Bonds are not riskless, but their risk profile is fundamentally different, and that doesn't rest just on observation of past performance. Stocks and bonds even have different Fama-French factors.

[seajay]

All these back tests are all looking backwards.

Unfortunately they may or may not pan out in the future.

If it does not pan out then You will constantly second guess your tilt.

This is the reason why you just go with Total Stock Market preferably global.

Then for your fixed income you decide how much risk you want to take.

I choose corporate bonds but some might stick with treasuries.

Do you buy the corporate bonds from the same stocks that you buy?
Ah! Yes I see you buy a total stock market fund, so if the average firm borrows amounts comparable to around half it its book-value, and share prices are typically bid up to twice-book-value, each $1 of stock purchase buys into around 25c of short (corporate) bond exposure. So 80/20 TSM/corporate is broadly like you buying the corporate bonds that firms/stocks issue.

That used to work when lending typically payed (lenders won), but since fiat currency (1933) its broadly neutral (sometimes borrows win, sometimes lenders win, overall balanced). As short bonds broadly = long bonds, just buying the 80 stock, sticking 20 under the mattress likely broadly averages out much the same overall. PV (as PV doesn't cater for mattress cash, I used a equal split of long and short Dow as a proxy for 'hard cash').

As stocks do well when real yields are positive, short bonds do well when real yields are negative (as can gold), then instead of mattress money - gold instead PV adds a bit more 'short bond' type - does well when real yields are negative exposure.

If 80 stock halve in real terms, 20 gold doubles in real terms, 80/20 -> 40/40 and dumping all of gold into stocks at that time has you double up on the number of shares being held. Martingale/d'Alembert betting sequence, after a lose, double up your stake.

Bet $1, lose (stocks down 50%)
Bet $2, win (stocks +100% rebound).
$3 out, $4 in (whilst stocks compounded 0%)

But where that's limited to just one 50% down, after which you're all-in with 100% stock (and twice as many shares as someone who lumped all-in from day zero), awaiting a recover, but maybe enduring even deeper declines beforehand.

[Logan Roy]

Well LTCM thought their portfolio was low risk, on account of its equity-market-like Sharpe ratio.

LTCM was playing a completely different game than we're talking about here, and if you were talking about leveraging your defensive stock portfolio by 25x or 30x I'm pretty sure my primary objection wouldn't be the choice of utility stocks etc. as the base investment.

Sure. But the thing with financial theory is adding leverage doesn't fundamentally change the maths. LTCM is used (e.g. in Shiller's Yale course) to demonstrate the nonlinearity of tail risk. You can have a 50:50 chance of losing half, or a 1 in 100 chance of losing everything – and standard deviation won't really describe that.

For the past 40 years, there's been a linear reduction in risk as we add bonds to a portfolio. But as soon as that relationship flips back to a positive correlation, you've got the LifeStrategy funds falling in unison. So the extent to which we're ever really reducing risk, or just pushing it further down the tail, is at very least multivariate. The Hedgefundie thread was another example of failing to represent risk as a distribution. So there's 'conventional', then there's 'correct'.

[Logan Roy]

Stocks are not bonds.
"Defensive" stocks are not bonds.
Dividend stocks from companies with a tradition of dividend-paying are not bonds.
Specific stock sectors are not bonds.
Blue-chip stocks from companies that are household names are not bonds.
Market neutral long-short stock portfolios are not bonds.
Preferred stocks are not bonds.

tocks are participation in business risk.

Bonds are (often) legally enforceable contracts to pay specified numbers of dollars on specified dates.

Investment-grade bonds are bonds from issuers whose books have been examined and are thought by ratings agencies to have enough safety margin to make those contractual payments through business downturns.

A Schwab equity rating of "A" is not the same thing as a Moody's bond rating of "A."

Stocks and bonds are just fundamentally different things, and it's a mistake to look at them as if the only things we know about them are pragmatic wiggly lines of past behavior. We know more than that. They make their money in completely different ways.

100% defensive stocks is not 60/40, and it's not like 60/40. It's 100% stocks. 100% stocks is what it is, and may be suitable for some investors, but it does nothing but harm to think of them as different from what they are. My hunch is that people for whom 100% stocks is suitable are people who can look 100% stocks right in the eye and say "100% stocks."

Like the old joke, how many legs does a horse have if you call a tail a leg? Answer, four, because a tail is not a leg.

I'd suggest it's perhaps not entirely clearcut.

A perpetual bond can be regarded as a stock. In industries where governments subsidise and provide safety nets to businesses, there's a guarantee on future cashflows that at least has something bond-like about it. There's to what extent a government can be viewed as a business, and at what point a business might become more like a state. A business can fail, but if a whole market were to fail, might one not already be at the point where a government's failed? Wouldn't we default on debt before we let the Utilities sector go to zero, and plunge us all into the middle ages?

[vineviz]

Well LTCM thought their portfolio was low risk, on account of its equity-market-like Sharpe ratio.

LTCM was playing a completely different game than we're talking about here, and if you were talking about leveraging your defensive stock portfolio by 25x or 30x I'm pretty sure my primary objection wouldn't be the choice of utility stocks etc. as the base investment.

Sure. But the thing with financial theory is adding leverage doesn't fundamentally change the maths. LTCM is used (e.g. in Shiller's Yale course) to demonstrate the nonlinearity of tail risk. You can have a 50:50 chance of losing half, or a 1 in 100 chance of losing everything – and standard deviation won't really describe that.

The point is that without leverage, LTCM wouldn’t have failed. Much less failed as spectacularly as it did.

If you’ve got a highly leveraged and high-frequency trading strategy, a 4 standard deviation day can impart a lot of damage.

An unleveraged investor with a 20 year investment horizon can shrug off a day like that

[Capster1]

I don't see how you circumvent market risk by buying different equities, regardless of which sector they are.

I buy bonds to reduce my exposure to the stock market. One can define 'risk' in a number of different ways. However, for me, having X percentage of my net worth in equities is a huge risk to me. Go past that point and I want the money to be somewhere else. Changing the sectors doesn't change the fact it is still equities. I'm not trying to get the highest performance, I'm trying to get the highest performance within my risk profile.

As someone has already said, dividends and stocks aren't bonds. They aren't guaranteed payments. Part of my own risk profile is to have a certain % in GUARANTEED items. I don't have to worry about getting paid -20% less on my bond if I hold them to maturity. It doesn't matter how many how safe a sector might appear, it can still go down a large percentage.

Playing things by sector also comes with issues of more expense fees, more rebalancing, etc. Not to mention you are back to timing the market. You buy in to the sectors but do you know when to sell them? If you play the game of buy and hold, you are going to underperform long-term.
Lastly, it's all based on historical performance, backwards looking. Look at AT&T and Verizon right now, I'm glad I'm holding VTSAX instead of either one of those.

[nigel_ht]

I don't see how you circumvent market risk by buying different equities, regardless of which sector they are.

You’re reducing sector risk…which being heavily invested in tech you have a lot of right now.

Of course that was a big driver of gains the last decade it’s been a winning bet.

I buy bonds to reduce my exposure to the stock market. One can define 'risk' in a number of different ways. However, for me, having X percentage of my net worth in equities is a huge risk to me. Go past that point and I want the money to be somewhere else. Changing the sectors doesn't change the fact it is still equities. I'm not trying to get the highest performance, I'm trying to get the highest performance within my risk profile.

As someone has already said, dividends and stocks aren't bonds. They aren't guaranteed payments. Part of my own risk profile is to have a certain % in GUARANTEED items.

BND YTD return: -11.04%
VGLT YTD return: -23.30%

I don't have to worry about getting paid -20% less on my bond if I hold them to maturity.

If you are going holding stocks for the same duration as LTT it’s highly unlikely to be 20% down…

Playing things by sector also comes with issues of more expense fees, more rebalancing, etc. Not to mention you are back to timing the market. You buy in to the sectors but do you know when to sell them? If you play the game of buy and hold, you are going to underperform long-term.

You aren’t likely to underperform bonds over the long term.

Lastly, it's all based on historical performance, backwards looking. Look at AT&T and Verizon right now, I'm glad I'm holding VTSAX instead of either one of those.

VTSAX holds 118,099,130 of VZ…which is down -21.24% YTD.

It also holds even more nvidia which is down -54.69% YTD…

[Logan Roy]

Well LTCM thought their portfolio was low risk, on account of its equity-market-like Sharpe ratio.

LTCM was playing a completely different game than we're talking about here, and if you were talking about leveraging your defensive stock portfolio by 25x or 30x I'm pretty sure my primary objection wouldn't be the choice of utility stocks etc. as the base investment.

Sure. But the thing with financial theory is adding leverage doesn't fundamentally change the maths. LTCM is used (e.g. in Shiller's Yale course) to demonstrate the nonlinearity of tail risk. You can have a 50:50 chance of losing half, or a 1 in 100 chance of losing everything – and standard deviation won't really describe that.

The point is that without leverage, LTCM wouldn’t have failed. Much less failed as spectacularly as it did.

If you’ve got a highly leveraged and high-frequency trading strategy, a 4 standard deviation day can impart a lot of damage.

An unleveraged investor with a 20 year investment horizon can shrug off a day like that

Absolutely. I bring up LTCM as an example of how hedge funds had to change the way they quantified risk – with regards to there being a correct way to do so. Leverage is what made LTCM such a public example, but we can all intuitively know standard deviation is a fairly crude analog for risk.

[rockstar]

Well LTCM thought their portfolio was low risk, on account of its equity-market-like Sharpe ratio.

LTCM was playing a completely different game than we're talking about here, and if you were talking about leveraging your defensive stock portfolio by 25x or 30x I'm pretty sure my primary objection wouldn't be the choice of utility stocks etc. as the base investment.

Sure. But the thing with financial theory is adding leverage doesn't fundamentally change the maths. LTCM is used (e.g. in Shiller's Yale course) to demonstrate the nonlinearity of tail risk. You can have a 50:50 chance of losing half, or a 1 in 100 chance of losing everything – and standard deviation won't really describe that.

The point is that without leverage, LTCM wouldn’t have failed. Much less failed as spectacularly as it did.

If you’ve got a highly leveraged and high-frequency trading strategy, a 4 standard deviation day can impart a lot of damage.

An unleveraged investor with a 20 year investment horizon can shrug off a day like that

Absolutely. I bring up LTCM as an example of how hedge funds had to change the way they quantified risk – with regards to there being a correct way to do so. Leverage is what made LTCM such a public example, but we can all intuitively know standard deviation is a fairly crude analog for risk.

When Genius Failed is a great book covering LTCM. If you haven't read it yet, I recommend it.

With rates going up, the cost of leverage is also likely going up.

I did trade TQQQ in the past. I'm not now.

[Logan Roy]

I don't see how you circumvent market risk by buying different equities, regardless of which sector they are.

I buy bonds to reduce my exposure to the stock market. One can define 'risk' in a number of different ways. However, for me, having X percentage of my net worth in equities is a huge risk to me. Go past that point and I want the money to be somewhere else. Changing the sectors doesn't change the fact it is still equities. I'm not trying to get the highest performance, I'm trying to get the highest performance within my risk profile.

As someone has already said, dividends and stocks aren't bonds. They aren't guaranteed payments. Part of my own risk profile is to have a certain % in GUARANTEED items. I don't have to worry about getting paid -20% less on my bond if I hold them to maturity. It doesn't matter how many how safe a sector might appear, it can still go down a large percentage.

Playing things by sector also comes with issues of more expense fees, more rebalancing, etc. Not to mention you are back to timing the market. You buy in to the sectors but do you know when to sell them? If you play the game of buy and hold, you are going to underperform long-term.
Lastly, it's all based on historical performance, backwards looking. Look at AT&T and Verizon right now, I'm glad I'm holding VTSAX instead of either one of those.

I suppose I'd argue that growing wealth inevitably means taking market risk. It's whether the optimal way to moderate that risk is to take the market as a whole and add water; or take the market as a series of parts, and change ones exposures to them. An analogy might be if you wanted to make a soft drink healthier, would you mix the drink with water, or would you just take some of the sugar or high fructose corn syrup out?

If a combination of staples, healthcare and utilities actually holds up, over the long-term and through bear markets, as well as a portfolio of stocks and bonds, then we'd probably have to conclude that they are doing as good a job of moderating risk as a combination of stocks and bonds.

[Logan Roy]

LTCM was playing a completely different game than we're talking about here, and if you were talking about leveraging your defensive stock portfolio by 25x or 30x I'm pretty sure my primary objection wouldn't be the choice of utility stocks etc. as the base investment.

Sure. But the thing with financial theory is adding leverage doesn't fundamentally change the maths. LTCM is used (e.g. in Shiller's Yale course) to demonstrate the nonlinearity of tail risk. You can have a 50:50 chance of losing half, or a 1 in 100 chance of losing everything – and standard deviation won't really describe that.

The point is that without leverage, LTCM wouldn’t have failed. Much less failed as spectacularly as it did.

If you’ve got a highly leveraged and high-frequency trading strategy, a 4 standard deviation day can impart a lot of damage.

An unleveraged investor with a 20 year investment horizon can shrug off a day like that

Absolutely. I bring up LTCM as an example of how hedge funds had to change the way they quantified risk – with regards to there being a correct way to do so. Leverage is what made LTCM such a public example, but we can all intuitively know standard deviation is a fairly crude analog for risk.

When Genius Failed is a great book covering LTCM. If you haven't read it yet, I recommend it.

With rates going up, the cost of leverage is also likely going up.

I did trade TQQQ in the past. I'm not now.

And something somewhat counterintuitive I heard from a very experienced money manager the other day: that it's usually proven best to leverage returns when the cost of borrowing is very high (e.g. 1982).

Mentioning When Genius Failed actually reminded me of the quote I was looking for on defining risk:

Risk is based on the assumption of the existence of a well-defined and constant objective probability distribution which is assumed to be known subjectively. Under uncertainty, however, this is not the case. As Keynes (1937, p.214 our italics) noted, in matters of true uncertainty, “.... there is no scientific basis on which to form any calculable probability whatever. We simply do not know”.
– LESSONS NOT LEARNT: FROM THE COLLAPSE OF LONG-TERM CAPITAL MANAGEMENT TO THE SUBPRIME CRISIS

[nedsaid]

Just pondering, but what are the views on managing risk by only investing in defensive sectors (in this case the three classic recession hedges: Consumer Staples, Healthcare and Utilities) vs stocks and bonds?

I've never been particularly keen on Treasuries as diversifiers for stocks, as although they often provide useful volatility dampening, the risk-free rate tends to dictate fair value for stocks, meaning their respective valuations generally move in tandem (Bond Equity Earnings-yield Ratio). So rebalancing between stocks and bonds rarely leads to these particularly efficient portfolios, as earnings yields on each don't tend to diverge greatly.

If anyone knows where I can get longer-term US stocks sector data, I'd be the first to say backtests that don't go back at least 50 years can be interesting, but not particularly useful.

I would say that this is actually not a bad idea, just something that I wouldn't do myself.

My friend from years back who went into the Brokerage business and got me started with a Brokerage account and individual stocks is doing something similar to what you are describing. Him and his wife's portfolio are 100% individual stocks and I believe they are all dividend payors. As far as I know, he isn't pursuing a Low Volatility strategy but with a Value orientation and a preference for dividends he is probably accomplishing much the same thing.

The other part of this that gives me pause is concentrating in just three sectors of the economy. If you wanted a lower volatility strategy, you could just overweight the three sectors compared to the index but otherwise have a market portfolio.

[jh]

Just pondering, but what are the views on managing risk by only investing in defensive sectors (in this case the three classic recession hedges: Consumer Staples, Healthcare and Utilities) vs stocks and bonds?

I'm doing something similar to that. I retired early and I'm just going to live off of the dividend income from my taxable account until 60+. I'm using the Schwab US Dividend ETF ticker SCHD.

I'm not strictly limited to specific sectors, but dividend growth etfs usually do overweight the defensive stocks you are talking about.

I am more comfortable with this than I am with a 60/40 portfolio and withdrawals coming from both dividends and capital gains.

The SCHD etf selects stocks based on a composite score coming from the return on equity, debt to cash flow, five years dividend growth rate, and current dividend yield. The stocks that make it passed the initial screens are then bought by market cap.

Due to the selective algorithm it is very likely that the dividend income will grow every year. So far the etf's lowest yearly dividend growth rate was still a positive 6.96% for the year. The mean yearly dividend growth rate over the life of the etf is currently 12.10%.

[Alpha4]

What risk are you trying to reduce?

Defensive sectors got hammered a lot more than Treasury Bonds from 1929 - 1932 [treasuries went UP in real terms] and (probably) a portfolio of 100% defensive stock probably got hammered a lot more than a mix of 60% stocks and 40% Treasury Bonds. Maybe the 100% defensive stocks worked better from 1966 - 1982 (one could check, I have not). But 1929 and 1966 were two different TYPES of risks.

Defensive stocks are not equivalent to bonds

That's why I'd really like to find longer-term estimated data on sector returns, and see how it plays out over a full cycle. This is just me playing around with an idea. And on the other side of that idea, questions about the usefulness of Treasuries as diversifiers, and whether they're likely to become compelling again (as more than Options on deflation/easing, which seem like the least useful things to hedge stocks from).

I think holding only TIPS to diversify stocks makes a lot more sense, as you get the inflation hedging at the short end, and the deflation hedging (with their sensitivity to rates) at the other.

Sector returns for 1929-32 may be tough to find. This page has some (claimed ... I have not checked myself) examples of peak-to-trough stock prices for various companies (e.g. Gillette, P&G, AT&T, General Mills) during the 1929-32 excitement.

https://www.joshuakennon.com/a-look-at- ... ket-crash/

It is pretty clear that companies that need earnings to survive behave very differently from government bonds (issued by a government that can print money to make payments ...) during this sort of "recession".

If you found stock data that said that P&G+Gillette+General Mills+whatever stock was "just as good" as 60% Stocks + 40% Us Treasuries from 1929 - 1932 would you believe it?

Ken French's online data library has industry portfolios in addition to factor portfolios.

https://mba.tuck.dartmouth.edu/pages/fa ... brary.html

They are near the bottom of the page.

Danke.

The "Hshld" entry in the 49 industry portfolios looks ... grim ... from 1929 - 1932.
I'm guessing that Hshld means something such as "Household goods" (?). Monthly
returns starting in Oct-1929:
-32.46
-14.73
+1.13
+10.32
-8.89
-0.19
-1.53
-7.45
-24.57
+5.84
+0.63
-10.53
-4.62
-1.84
-6.15

Maybe this isn't the industry sector that is defensive. Eyeballing the data seems to show that
"Smoke" held up pretty well into 1930 but then started to do poorly. Utilities might be the best
bet.

A long time ago when I was trying to backtest "consumer staples" and had to use the data from this site I used (IIRC...it's been a while) a mixture of the FF10 or FF12 "consumer non-durables" sector; the FF17 "consumer" sector" plus a bit each of the FF49 "food", smoke", and "beer" sectors (note that FF10 is the 10 industry portfolio one, FF12 is the 12 industry portfolio one, FF17 is the 17 industry portfolio one, and food/smoke/beer sectors are from the 49 industry portfolio one and are for the foods, tobacco, and alcohol sub-sectors specifically). I tried to use the blend that had the highest monthly and annual correlation to the "consumer staples" sector for the few decades I did have actual consumer staples data for (which consisted of the S&P 500 Consumer Staples TR sector back to late 1989, then the Morningstar "consumer staples" category from 1986 to late 1989, and finally--even though it was only annual data and not monthly--a simulated S&P 500 consumer staples dataset produced by Factset for Goldman Sachs back in 2014 and that was created using the same methodology that S&P Dow Jones used for its consumer staples index from 1989 to the present from 1974 to 1985).

I had to do the above because the Ken French data website doesn't really include a pure "consumer staples" category; "consumer non-durables" and "consumer" have various things like clothes, leather goods, apparel, shoes, textiles, toys, etc.

But yeah, as other posters on this thread have stated, the "Household" category from the 49 industry portfolios isn't really a very close match with consumer staples at all.

[nigel_ht]

...If you have lower volatility, lower max drawdown and a higher sustainable withdrawal rate from adding defensive sectors vs classic 60/40 why isn’t that worth consideration?...

Because you just used the present tense (although you were careful to qualify it with "if.") Perhaps they have had these, but they are not intrinsic characteristics, they are just observed past behavior.

For example, they do not "have" lower volatility than the whole stock market. All you can say is that over some specified time period they have had lower volatility.

If you purchase any basket of stocks smaller than the whole stock market you are basing it on current and prior characteristics.

If I buy an ETF for SCV then these stocks, up until today, meet the characteristics of small cap value.

If I buy an ETF for defensive sectors for low volatility then it’s because these sectors, up until today, meets the characteristics of low volatility factor.

How is this particularly different than buying a basket of stocks based on the volatility factor?

https://www.msci.com/documents/1296102/ ... 2eab715ed6

The difference is that a defensive sector allocation is an “unconstrained” minimum volatility approach:

“However, a naive unconstrained minimum volatility strategy has its own set of challenges, such as biases toward certain sectors and countries, unwanted factor exposures and potentially high turnover at rebalancing.“

Of these, only high turnover at rebalancing is of interest because when we invest in the “whole stock market” we’ve already accepted a large cap bias that heavily favors the technology sector in US, EU, JPN, KOR, TWN and CHN.

I don’t know how to evaluate the turnover concern listed by MSCI but I don’t rebalance as often as an ETF might…so maybe not as important to me as an individual investor.

Finally, as an investor with a finite time horizon I’m not as concerned with whether or not low volatility is an “intrinsic” trait of certain sectors as much as I am concerned with building a total portfolio that is diverse in ways that make sense to me and has historical evidence of doing what I want it to do.

USMV, SPLV, EFAV, etc holds many of the same stocks but are younger without the same ability to see how they performed in the past. And typically their ER is higher than a sector ETF.

If I want to lower my risk, as defined (inaccurately) as volatility, I have several options. I can add more bonds OR I can move some of my stocks from “neutral” to lower volatility OR I can put less overall into securities and into something else (real estate, physical gold, art, etc) considered “safe”.

[Logan Roy]

Just pondering, but what are the views on managing risk by only investing in defensive sectors (in this case the three classic recession hedges: Consumer Staples, Healthcare and Utilities) vs stocks and bonds?

I've never been particularly keen on Treasuries as diversifiers for stocks, as although they often provide useful volatility dampening, the risk-free rate tends to dictate fair value for stocks, meaning their respective valuations generally move in tandem (Bond Equity Earnings-yield Ratio). So rebalancing between stocks and bonds rarely leads to these particularly efficient portfolios, as earnings yields on each don't tend to diverge greatly.

If anyone knows where I can get longer-term US stocks sector data, I'd be the first to say backtests that don't go back at least 50 years can be interesting, but not particularly useful.

I would say that this is actually not a bad idea, just something that I wouldn't do myself.

My friend from years back who went into the Brokerage business and got me started with a Brokerage account and individual stocks is doing something similar to what you are describing. Him and his wife's portfolio are 100% individual stocks and I believe they are all dividend payors. As far as I know, he isn't pursuing a Low Volatility strategy but with a Value orientation and a preference for dividends he is probably accomplishing much the same thing.

The other part of this that gives me pause is concentrating in just three sectors of the economy. If you wanted a lower volatility strategy, you could just overweight the three sectors compared to the index but otherwise have a market portfolio.

It's a semi-real-world experiment for me – I had some spare cash a year back, and put it into a new platform as 'fun money'. And used the opportunity to see what kind of portfolio I'd design from scratch today. And this has had me playing around with Portfolio Visualizer more than I have in a while.

So what I came up with, in August 2021, was 34% split evenly between TIPS and gold, and 66% split between US Large-cap Value and S&P 500. And it's held up well, and I like the platform, so I decided to move some more cash in recently. The aim is for something somewhat All Weather and hopefully capable of riding out stagflation. A fun little experiment. It's also evolved out of using sectors a bit more in portfolio design. I do know of some individual stock investors who do similar things – primarily investing for dividends, or focusing on utilities or staples, where there may be some dependability to cashflows. Obviously such short-term backtests are not likely to be prophetic, but just to test the concept.

[Logan Roy]

EDIT: because I couldn't delete the post. Here's a work in progress. This should be Consumer Non-durables, Healthcare, Utilities, and a portfolio of the three with a monthly rebalance, vs S&P Composite nominal returns with dividends. If Portfolio Visualizer would get its data going back this far, it would save a lot of work.

I think we could truncate the data, to account for the goings on with Utilities in the 30s, 40s. What I'm planning is to get bond market data in there, then run a machine learning algorithm to work out the optimal portfolio (only backward-looking, and educational, of course). Utilities do seem to have a persistently low risk premium – so the market may be treating them a bit more like bonds.

[minesweep]

Just pondering, but what are the views on managing risk by only investing in defensive sectors (in this case the three classic recession hedges: Consumer Staples, Healthcare and Utilities) vs stocks and bonds?

Logan Roy,

Until Telecom Bust, Telecom stocks were considered as defensive sector. How well did it works out during Telecom Bust?

KlangFool

Somehow I don't see a Big Bust in toothpaste and toilet paper.

I didn’t think there would be a Big Bust in toilet paper until Covid-19 came into the picture. It was wiped out (pun intended) on the store shelves.

[Charles Joseph]

Just pondering, but what are the views on managing risk by only investing in defensive sectors (in this case the three classic recession hedges: Consumer Staples, Healthcare and Utilities) vs stocks and bonds?

Logan Roy,

Until Telecom Bust, Telecom stocks were considered as defensive sector. How well did it works out during Telecom Bust?

KlangFool

Somehow I don't see a Big Bust in toothpaste and toilet paper.

I didn’t think there would be a Big Bust in toilet paper until Covid-19 came into the picture. It was wiped out (pun intended) on the store shelves.

You are correct! I have a picture somewhere that my wife took of the empty TP shelf in our local supermarket.

[Logan Roy]

Quick work in progress.

So far it's looking pretty good for broad stocks and bonds. This is a brute force (trial 100,000 random portfolios) to simply to maximise Sharpe ratio. I'm starting the run in the 1940s, to chop off the business with Utilities cos going public, etc. The optimal portfolio so far (monthly rebalance) is:

Consumer NonDurables 0%
Utilities 6%
Healthcare 10%
S&P Comp 24%
Bonds 60%

Reporting an annualised Sharpe ratio of 1.35 (based on monthly returns).

Next will be to optimise Sharpe, but place more weight on achieving a higher return.



EDIT:

Maximise Sharpe for a minimum average monthly return of 1% nominal (switched to an annual rebalance):

Consumer NonDurables 16%
Utilities 19%
Healthcare 22%
S&P Comp 42%
Bonds 1%

Annualised Sharpe 1.1

Note: I'm finding a problem with Sharpe ratio is in using a mean average monthly return, it's not doing a good job of optimising for sequence risk (which also isn't really represented by Standard Deviation). Quite an issue for portfolio design.

[gougou]

I feel that Consumer Staples is a pretty arbitrary definition. And the definition may have changed over time. So the backtesting is probably not as reliable as you think.

[Forester]

Link to the S&P Low Volatility paper, instead of going around in circles discussing industry sectors: https://www.spglobal.com/spdji/en/docum ... istory.pdf. "Defensive stocks" would tend to be those exhibiting less day to day price volatility, therefore in a Dotcom/GFC they sell off less.

[Logan Roy]

Link to the S&P Low Volatility paper, instead of going around in circles discussing industry sectors: https://www.spglobal.com/spdji/en/docum ... istory.pdf. "Defensive stocks" would tend to be those exhibiting less day to day price volatility, therefore in a Dotcom/GFC they sell off less.

I'm an outlier on this, but I found the more I looked into 'factors', the less practical use I could find for them (the exception would be if I had to write a thesis – beyond that, I'm not convinced factors have had a net positive effect on portfolios or market efficiency).

Here, Utilities can be quite volatile. They can also be quite sensitive to rates (with their bond proxy qualities). But the underlying earnings have this stability to them that the average public company probably doesn't. They're more volatile than the S&P, yet their earnings may be more reliable. Long-duration TIPS can be volatile, despite the real return being as certain as anything you get in investing. What I like about sectors is – like other asset classes – they really are linked, in logical ways, to macroeconomics. So if the aim were to produce a more robust All Weather portfolio, or simply a risk-balanced portfolio – in an age in which Treasuries are likely to be stuck on negative real yields at least at the short end – I think sectors may be where to look.

[Logan Roy]

Update:
This portfolio achieves a slightly higher Sharpe ratio than 60:40, and a slightly higher total return than 100% S&P. Results seem consistent enough to suggest a possible effect beyond simply backfitting. Really what I'm looking for at this stage is whether the algorithm naturally overweights defensive sectors in place of bonds.

Consumer Durables – 2%
Manufacturers - 2%
Energy - 19%
Technology - 14%
Shops - 8%
Healthcare - 14%
Utilities - 13%
S&P Comp - 19%
T-Bonds - 9%

Rebalancing every 19 months

[MarkRoulo]

Update:
This portfolio achieves a slightly higher Sharpe ratio than 60:40, and a slightly higher total return than 100% S&P. Results seem consistent enough to suggest a possible effect beyond simply backfitting. Really what I'm looking for at this stage is whether the algorithm naturally overweights defensive sectors in place of bonds.

Consumer Durables – 2%
Manufacturers - 2%
Energy - 19%
Technology - 14%
Shops - 8%
Healthcare - 14%
Utilities - 13%
S&P Comp - 19%
T-Bonds - 9%

Rebalancing every 19 months

This is a textbook example of overfitting. You have at least ten knobs to tweak (the percentage in Consumer Durables ... T-Bonds plus your rebalancing frequency). Being able to find a mix of these THAT WORKED CONSISTENTLY IN THE PAST is not surprising. What would be interesting is if you got similar values using data from the first half of your data set and this replicated in the second.

[Logan Roy]

This is a textbook example of overfitting. You have at least ten knobs to tweak (the percentage in Consumer Durables ... T-Bonds plus your rebalancing frequency). Being able to find a mix of these THAT WORKED CONSISTENTLY IN THE PAST is not surprising. What would be interesting is if you got similar values using data from the first half of your data set and this replicated in the second.

Trust me, you're preaching to the choir. We've got factors, like small minus big, that haven't worked since the 1980s, and even prior to that, seem to have been heavily skewed by issues with old data (survivorship and liquidity). As soon as we go beyond brute force portfolio to machine learning, I'll be using validation data. I'm ironing out issues now so I don't have to deal with them on top of an array of actor-critic algorithms.

What's interesting about testing over a pretty full economic cycle is you've got such a variety of macroeconomic conditions, the tendency to backfit seems much lower. So as a test, here I've simply maximised returns. And in a good backfit – and certainly over a few decades – you'd simply expect it to go all-in on the best performing sector. But it doesn't, because there's so much mean reversion over these kind of periods, it actually gets better results being more balanced and rebalancing more often (I'm letting it rebalance as infrequently as once in 1,000 months). You can also tell, with these clear charts, whether performance over the first 50 years is consistent with the second – which is also happening more often than I'd expect optimising within shorter periods. So I think there's more generalisability already, simply by virtue of using more data.

['NoDur', 'Durbl', 'Manuf', 'Enrgy', 'HiTec', 'Telcm', 'Shops', 'Hlth ', 'Utils', 'Other', 'S&P', 'T. Bill']
[ 3.24 15.94 14.07 16.5 11.95 1.57 12.09 16.99 0.4 0.62 4.69 1.96] %
22 month rebalance

[nigel_ht]

it actually gets better results being more balanced and rebalancing more often (I'm letting it rebalance as infrequently as once in 1,000 months).

Jim Otar has suggested rebalancing at the end of presidential election years.

“There are several known market cycles; the 54-year Kondratieff cycle, 10-year decennial cycle, and the 4-year U.S. Presidential election cycle, to name a few. We will focus on the U.S. Presidential Election cycle as the basis of our rebalancing study. It is the shortest market cycle that is meaningful to retirement planning.”

http://retirementoptimizer.com/articles ... ancing.pdf

You might give that a whirl…

And I’d personally use total world as your core holding…just to have some international diversification given the rest of the portfolio is US only even if the numbers come off worse over the whole period.

Given things are cyclic I assume (pray?) that my vxus holdings will eventually pay out…

[Logan Roy]

it actually gets better results being more balanced and rebalancing more often (I'm letting it rebalance as infrequently as once in 1,000 months).

Jim Otar has suggested rebalancing at the end of presidential election years.

“There are several known market cycles; the 54-year Kondratieff cycle, 10-year decennial cycle, and the 4-year U.S. Presidential election cycle, to name a few. We will focus on the U.S. Presidential Election cycle as the basis of our rebalancing study. It is the shortest market cycle that is meaningful to retirement planning.”

http://retirementoptimizer.com/articles ... ancing.pdf

You might give that a whirl…

And I’d personally use total world as your core holding…just to have some international diversification given the rest of the portfolio is US only even if the numbers come off worse over the whole period.

Given things are cyclic I assume (pray?) that my vxus holdings will eventually pay out…

I think there might be a lot of sense in that. I've probably let it test about 1 million random portfolios and rebalance frequencies so far (within the very limited criteria and breakdown of asset classes it's got), and it's undoubtedly hovering around a 19-22 month rebalance. I wonder if it's picking up the mid-terms?

I fear it may be harder to find reliable international stock market data going back this far (I'd certainly give it a go if there are any sources?). To an extent, I probably side with Bogle and Buffett on this (afaik) on balance: that US listings act as a useful filter of businesses that may not be practicing proper accounting standards and reporting reliably. I can find slightly different figures on this, but as the S&P 500 is about 40% exposed to non-US revenues, it's still a pretty good play on the global economy. Until you get to smaller companies (or markets) where a market starts to become more a play on that domestic economy, I tend to think the primary difference between US and international markets tends to be arbitrary differences in sector composition. But this is probably not a common viewpoint.

[TheDoctor91]

IMO it makes more sense to just use a minimum or low volatility fund rather than pick sectors for many of the reasons already given.

I’d also probably add something like 10% TYA for some leveraged bond exposure.

[Logan Roy]

IMO it makes more sense to just use a minimum or low volatility fund rather than pick sectors for many of the reasons already given.

I’d also probably add something like 10% TYA for some leveraged bond exposure.

I think volatility is a better analog for risk across a portfolio than it is for individual assets. e.g. Utilities have been one of the most volatile sectors in these backtests, despite having an earnings stability somewhere between stocks and bonds. And I think that's partly because their earnings are so predictable, they can be treated like long/infinite-duration gov. bonds (also volatile, on account of their sensitivity to rates, but certainly not risky).

The reason I prefer sectors to factors is because sectors have predictable relationships with things that go on in the economy. Rates, inflation, recession, etc. And these shouldn't fundamentally change over 100-200 years. Factors are these abstract mathematical objects that have a lot to do with how investors assess stock fundamentals – which we know has changed over time, and which even the existence of factor products affects.

I'm new to having sector data that goes back this far, so this was all I managed yesterday evening (with a lot of tidying up code). I devised a geometric Sharpe ratio (using the geometric mean to calculate Sharpe, rather than arithmetic) – the idea being that this factors in sequence risk a bit more (but could be slightly more prone to backfitting). This portfolio cracked stock market returns for 60/40 equivalent risk, with quite reasonable consistency (so the outperformance isn't particularly concentrated in the chart). The question would be whether this approach is stumbling on a useful investment principle – and I think it's favouring a barbell of high earnings stability (e.g. Healthcare), with cyclicality (energy) as an All Weather approach.

What I'm intending next is to take a leaf out of PortfolioCharts' book, and work on a rolling assessment of risk and return (e.g. rolling 15 year performance). Maybe add gold to the mix. The dream would be if I could work out a way to simulate (or obtain the simulation of) TIPS data back to the 1920s – even better long and short duration independently. Also this sector breakdown doesn't seem to include financials.



Consumer NonDur: 14%
Energy: 15%
Telecoms: 8%
Shops: 6%
Healthcare: 12%
Other: 11%
S&P: 14%
T-Bonds: 20%

Rebalance every 47 months

[rule of law guy]

"So rebalancing between stocks and bonds rarely leads to these particularly efficient portfolios, as earnings yields on each don't tend to diverge greatly."

generally yes, and certainly this year...but not so re stocks and cash this year. I have a portfolio of cash and equities, and have been waiting for rates to rise for what seems forever, as I have diversified with cash rather than bonds. I found that my portfolio return was acceptable during the past bull market notwithstanding the sizable allocation to cash, and now I am getting a substantial cushioning effect by my cash allocation, whose return rate appears to be on the rise to above 3%....at which point I may apportion some cash to 2-3 year treasuries.

[Logan Roy]

"So rebalancing between stocks and bonds rarely leads to these particularly efficient portfolios, as earnings yields on each don't tend to diverge greatly."

generally yes, and certainly this year...but not so re stocks and cash this year. I have a portfolio of cash and equities, and have been waiting for rates to rise for what seems forever, as I have diversified with cash rather than bonds. I found that my portfolio return was acceptable during the past bull market notwithstanding the sizable allocation to cash, and now I am getting a substantial cushioning effect by my cash allocation, whose return rate appears to be on the rise to above 3%....at which point I may apportion some cash to 2-3 year treasuries.

Absolutely. I've also had a fair bit in cash through this bull market. And it seems when you've got tightening and inflation, it's hard to find much that does better than the dollar.

Of course, there are reasons to be pessimistic on cash and short-dated treasuries. The reason you can get 3% on 2 and 3 yrs is because we expect inflation to average above 3% over the next 2-3 years (and it could be well above). With the levels of global debt, we'll likely need negative short-term real yields for a long time, so that debt can be refinanced (and slowly inflated away). And it makes a lot of sense to keep devaluing the cash that debt is denominated in.

The only silver lining for cash and treasuries would be unexpected deflation – but that would be so disastrous for governments with huge debt piles, that we'll throw everything at it (as Japan have been), and that will probably be good for stocks and corporate bonds. So I think this is an environment for All Weather investing, and it's an environment that could drag on a long time. We might breathe a sign of relief when inflation cools, but debt really limits where yields can go. So it's: how much of the slack of Treasuries can allocations to things like Utilities, or the rebalancing effect between sectors, pick up? And I think, historically, it's not been a problem to reduce risk without relying on bonds and cash so much. And going forward, I think there's a reasonable argument it could be a necessity.

[Lawrence of Suburbia]

Can this defensive stocks idea be translated into a simple, Bogleheads-style move ...?

A big chunk of my portfolio is in index fund-related stuff (for example, my 401k is in Vanguard Target 2025 fund); might SCHD (Schwab U.S. Dividend Index ETF, in my brokerage account) be seen as defensive? If not, I'm considering trading that for the Vanguard Total Stock Market fund. I've already got bond-y things like Wellesley and Dodge & Cox Balanced in IRA and Roth accounts (balanced funds have always been my comfort zone).

Edit: I see that Vanguard has a Consumer Staples ETF, VDC. I guess that's my answer.