Utilities - worth another look?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Fri Jun 14, 2019 10:25 pm

vineviz wrote:
Tue Mar 19, 2019 6:24 am
aristotelian wrote:
Mon Mar 18, 2019 5:43 pm

You seem to be reflecting the widely held, but mistaken, belief that increasing the number of funds in a portfolio increases diversification.
That would be absurd, which is why I never suggested any such thing.

The number of funds is irrelevant. All that matters, as far as diversification goes, is their correlation and volatility.
The correlation and volatility of what? The funds or the underlying assets?

User avatar
Sandtrap
Posts: 8634
Joined: Sat Nov 26, 2016 6:32 pm
Location: Hawaii No Ka Oi , N. Arizona

Re: Utilities - worth another look?

Post by Sandtrap » Fri Jun 14, 2019 10:45 pm

HEDGEFUNDIE wrote:
Mon Mar 18, 2019 4:29 pm
Let's back up and review the facts at hand:

1. 100% of our family income is concentrated in tech and healthcare
2. 33% of VTSAX is concentrated in tech and healthcare
3. Utilities exhibit the lowest correlation to both those sectors (around 0.5 correlation)
4. Utilities only comprise 3% of VTSAX

So what's the problem with tilting towards utilities?
Good point about VTSAX's percent in tech and healthcare.

If so, then why not Utilities and REIT's ???

Or, Utilities, REIT's, and Healthcare Index??

j :happy
Wiki Bogleheads Wiki: Everything You Need to Know

User avatar
305pelusa
Posts: 1265
Joined: Fri Nov 16, 2018 10:20 pm

Re: Utilities - worth another look?

Post by 305pelusa » Fri Jun 14, 2019 10:54 pm

vineviz wrote:
Mon Mar 18, 2019 4:38 pm
The second sentence I quoted appears to be rooted in the widely held, but mistaken, belief that a market cap weighted portfolio is the most diversified portfolio possible.
If it's not the most diversified, wouldn't the market efficiently incorporate this information? People would bid up the prices of the assets that would diversify it further (say utilities) until the market resembled the more diversified portfolio and then the cap weighted portfolio would in fact be the most diversified?

I am looking for any reading on factor diversification you might have come across that pits it with market cap weights. Specifically, a text with no historical performance, past correlations, etc. Just a solid, theoretical understanding/proof as to why there can exist a more diversified portfolio than the market cap. Would you have such a reading available?

Thanks!

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Utilities - worth another look?

Post by vineviz » Sat Jun 15, 2019 7:16 am

305pelusa wrote:
Fri Jun 14, 2019 10:54 pm
Just a solid, theoretical understanding/proof as to why there can exist a more diversified portfolio than the market cap. Would you have such a reading available?
If markets are efficient, the market portfolio will optimally reflect the aggregate - or average - of all investor risk tolerances and preferences.

The market portfolio can optimally reflect everyone's risk tolerances and preferences ONLY IF everyone has the same risk tolerances and preferences. Is it reasonable to assume that every investor has the same risk tolerances and preferences? If not, the market portfolio can't be optimal for everyone.

The market portfolio can optimally reflect your risk tolerances and preferences ONLY IF you have the same risk tolerances and preferences as everyone else. Do you think your risk tolerances and preferences are the same as everyone else? If not, the market portfolio can't be optimal for you.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
305pelusa
Posts: 1265
Joined: Fri Nov 16, 2018 10:20 pm

Re: Utilities - worth another look?

Post by 305pelusa » Sat Jun 15, 2019 8:26 am

vineviz wrote:
Sat Jun 15, 2019 7:16 am
305pelusa wrote:
Fri Jun 14, 2019 10:54 pm
Just a solid, theoretical understanding/proof as to why there can exist a more diversified portfolio than the market cap. Would you have such a reading available?
If markets are efficient, the market portfolio will optimally reflect the aggregate - or average - of all investor risk tolerances and preferences.

The market portfolio can optimally reflect everyone's risk tolerances and preferences ONLY IF everyone has the same risk tolerances and preferences. Is it reasonable to assume that every investor has the same risk tolerances and preferences? If not, the market portfolio can't be optimal for everyone.

The market portfolio can optimally reflect your risk tolerances and preferences ONLY IF you have the same risk tolerances and preferences as everyone else. Do you think your risk tolerances and preferences are the same as everyone else? If not, the market portfolio can't be optimal for you.
I follow this so far. But I'm not quite following how that proves that the cap weighted portfolio is not the most diversified.

I take optimal to mean "has the right risk/return trade-off" which means something different than most diversified (exposes itself to as many risks as is sensible, in approximately the right quantities). The market portfolio could still be the most diversified yet not optimal for everyone from a total risk perspective, which would be consistent with your explanation above yet inconsistent with the claim I quoted of yours earlier.

What is the part I'm missing? Thanks!

D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Sat Jun 15, 2019 9:37 am

vineviz wrote:
Mon Mar 18, 2019 4:38 pm
aristotelian wrote:
Mon Mar 18, 2019 4:17 pm
Portfolio 2 is not better diversified. It is concentrated in utilities.
The first sentence I quoted is objectively untrue, and the second sentence is a non sequitur.

No point debating the former, as it just is what it is.

Altering the weights of assets in a portfolio is precisely the mechanism by which you can increase (or decrease) the diversification of the portfolio. The second sentence I quoted appears to be rooted in the widely held, but mistaken, belief that a market cap weighted portfolio is the most diversified portfolio possible.
Portfolio A: 100% TSM
Portfolio B: 50% TSM / 50% VPU (utilities)

Is Portfolio B objectively more diversified? If so, how do you know? Is there a metric you look at to determine this? This is a serious question. I’ve been thinking a lot about diversification and I’m not sure how to measure it in a logical way. Thanks!

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Utilities - worth another look?

Post by vineviz » Sat Jun 15, 2019 10:39 am

D-Dog wrote:
Sat Jun 15, 2019 9:37 am
Portfolio A: 100% TSM
Portfolio B: 50% TSM / 50% VPU (utilities)

Is Portfolio B objectively more diversified? If so, how do you know? Is there a metric you look at to determine this? This is a serious question. I’ve been thinking a lot about diversification and I’m not sure how to measure it in a logical way. Thanks!
Yes, B is objectively more diversified.

One metric you can use is the diversification ratio. This ratio compares the volatility of a portfolio to the weighted average volatility of the underlying assets. This ratio is very simple but accounts for correlations, variances, and relative weights of the assets. In this example the standard deviations of the assets are as follows (using data from 2015 to present):

VTI: 12.46%
VPU: 11.78%
Portfolio B: 9.41%

So the ratio is ((.5 x 12.46%) + (.5 x 11.78%)) / 9.41% = 1.29 ( the higher the ratio, the more diversified the portfolio).

Compare this with Portfolio C which is 50% VTI and 50% VGLT (long term Treasury bonds):

VTI: 12.46%
VGLT: 10.79%
Portfolio B: 6.93%

So the ratio is ((.5 x 12.46%) + (.5 x 10.79%)) / 6.93% = 1.68.

There are other measures, like Shannon entropy, but I think this is easiest to calculate and comprehend.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

sf_tech_saver
Posts: 242
Joined: Sat Sep 08, 2018 9:03 pm

Re: Utilities - worth another look?

Post by sf_tech_saver » Sat Jun 15, 2019 11:19 am

My wife and I also work in tech/biotech.

We also live in SF where PG&E just went Chapter 11 because of their sudden lawsuit/liability claim for the fires last year.

I won't argue with your very sophisticated backtesting, but that approach is a very poor predictor of emergent risks for the sector. 2002-2017 looked like an amazing run for PG&E prior to a cliff in 2019 which took them back to 1997 valuations.

Some tilt seems harmless, but a full 50/50 split with total stock seems like taking on a new fulltime job as a utility sector analyst to me.
VTI is a modern marvel

User avatar
abuss368
Posts: 17030
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Utilities - worth another look?

Post by abuss368 » Sat Jun 15, 2019 11:29 am

Sandtrap wrote:
Fri Jun 14, 2019 10:45 pm
HEDGEFUNDIE wrote:
Mon Mar 18, 2019 4:29 pm
Let's back up and review the facts at hand:

1. 100% of our family income is concentrated in tech and healthcare
2. 33% of VTSAX is concentrated in tech and healthcare
3. Utilities exhibit the lowest correlation to both those sectors (around 0.5 correlation)
4. Utilities only comprise 3% of VTSAX

So what's the problem with tilting towards utilities?
Good point about VTSAX's percent in tech and healthcare.

If so, then why not Utilities and REIT's ???

Or, Utilities, REIT's, and Healthcare Index??

j :happy
Agreed. I would consider both the U.S. and International REIT index funds.
John C. Bogle - Two Fund Portfolio: Total Stock & Total Bond. "Simplicity is the master key to financial success."

User avatar
TomatoTomahto
Posts: 9742
Joined: Mon Apr 11, 2011 1:48 pm

Re: Utilities - worth another look?

Post by TomatoTomahto » Sat Jun 15, 2019 11:39 am

aristotelian wrote:
Mon Mar 18, 2019 4:23 pm
HEDGEFUNDIE wrote:
Mon Mar 18, 2019 4:21 pm
What possible black swans could exist for local utilities on a nationwide level?
Energy input costs are passed to the ratepayer.
Nationalization?
No idea. It wouldn't be a black swan if I could predict it. Cyberattack?
Hmmmm. So, with solar and batteries and an investment in utilities, I’m golden!!! Wheeeee
Last edited by TomatoTomahto on Sat Jun 15, 2019 11:59 am, edited 1 time in total.
Okay, I get it; I won't be political or controversial. The Earth is flat.

User avatar
Time2Quit
Posts: 244
Joined: Thu Nov 01, 2018 9:47 am
Location: Fridgid State

Re: Utilities - worth another look?

Post by Time2Quit » Sat Jun 15, 2019 11:52 am

Rus In Urbe wrote:
Tue Mar 19, 2019 8:24 am
Disruption about to happen in the utilities sector. Of course, where it will all end up is anyone's guess, but utilities are no longer "bulletproof" for sure.

This from Bill McKibben's excellent essay on energy from the latest issue of The New York Review of Books:
Over the last decade, there has been a staggering fall in the price of solar and wind power, and of the lithium-ion batteries used to store energy. This has led to rapid expansion of these technologies, even though they are still used much less than fossil fuels: in 2017, for instance, sun and wind produced just 6 percent of the world’s electric supply, but they made up 45 percent of the growth in supply, and the cost of sun and wind power continues to fall by about 20 percent with each doubling of capacity. Bond’s analysis suggests that in the next few years, they will represent all the growth. We will then reach peak use of fossil fuels, not because we’re running out of them but because renewables will have become so cheap that anyone needing a new energy supply will likely turn to solar or wind power.
Bond writes that in the 2020s—probably the early 2020s—the demand for fossil fuels will stop growing. The turning point in such transitions “is typically the moment when the impact is felt in financial markets”—when stock prices tumble and never recover. Who is going to invest in an industry that is clearly destined to shrink? Though we’ll still be using lots of oil, its price should fall if it has to compete with the price of sunshine. Hence the huge investments in pipelines and tankers and undersea exploration will be increasingly unrecoverable. Precisely how long it will take is impossible to predict, but the outcome seems clear.

and

“In 2017, the price in India of wind and solar power dropped 50 percent to $35–40 a megawatt hour,” said Tim Buckley, who analyzes Australasia/South Asia for the Institute for Energy Economics and Financial Analysis. “Fifty percent in one year. And a zero inflation indexation for the next twenty-five years. Just amazing.” This price drop occurred not because India subsidizes renewable energy (it doesn’t), but because engineers did such a good job of making solar panels more efficient. The cost of power from a newly built coal plant using Indian coal is, by comparison, about $60 a megawatt hour. If you have to import the coal, the price of power is $70/megawatt hour. And solar’s $40/megawatt hour price is guaranteed not to rise over the thirty-year life of the contract the suppliers sign—their bids are based on building and then running a facility for the life of the contract. No wonder that over the first nine months of 2018, India installed forty times more capacity for renewable than for coal-fired power.

and

You get some sense of the future from the stunning fall of General Electric. “They were the world leader, the thought leader, the finance leader, the IT leader,” said Buckley. “And their share price is down 70 percent in the last two and a half years, in a market that’s up 50 percent. It’s a thermal power–reliant basket case.” That’s in large measure because manufacturing turbines for coal- and gas-fired power plants was a significant part of the company’s business; in 2015, it hugely expanded that capacity by buying its largest European competitor, Alstom. But then the bottom dropped out of the industry as proposed new generating plants couldn’t find financing. GE makes wind turbines, too, but that’s a lower-margin business with many more competitors. The fall in GE’s stock has meant “hundreds of billions of dollars of shareholder value reduction,” according to Buckley. Last June, after more than a century, General Electric was dropped from the Dow Industrial Index, replaced by a drugstore chain.

and

The recent history of European utilities may provide a more realistic preview of what will happen in the rest of the world. In the early years of this century the German government increased the pace of decarbonization, subsidizing solar and wind energy. As more and cheaper renewable supplies became available, the existing utilities were slow to react. They had built new gas plants to account for what they assumed would be rising demand, but solar and wind cut into that demand, and the price of electricity began to fall. So far, European utilities have written down about $150 billion in stranded assets: fossil fuel installations that are no longer needed. “In the Netherlands, by the time the last three coal plants were turned on, their owners had already written them down by 70 percent,” said Buckley. And they’re scheduled to close by 2030.
Read the article if you have questions about what the legacy utilities are doing about investing in renewables. Hint: they are not; the business model is not as profitable as extraction and distribution of fossil fuels.

So, this is the forward-look at energy and utilities. Could be a great place to invest. Or not.

For me. Too risky.
I disagree the utilities are not investing renewables. One of my last contracts were for 2 large utilities. They were massively expanding their solar and wind operations. I am friends with the CEO of a medium size utility, I can’t repeat the stories here, but as someone posted any increased operating costs are past on to the customers. Utilites are guarantteed a certain profit margin by state regulatory agencies. They may not blow the doors off profits but they are steady.

Hint: You can google any major utilities website and see what they are up to and google state regulatory agencies who regulate utilities.
"It is not the man who has too little, but the man who craves more, that is poor." --Seneca

D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Sat Jun 15, 2019 12:01 pm

vineviz wrote:
Sat Jun 15, 2019 10:39 am
D-Dog wrote:
Sat Jun 15, 2019 9:37 am
Portfolio A: 100% TSM
Portfolio B: 50% TSM / 50% VPU (utilities)

Is Portfolio B objectively more diversified? If so, how do you know? Is there a metric you look at to determine this? This is a serious question. I’ve been thinking a lot about diversification and I’m not sure how to measure it in a logical way. Thanks!
Yes, B is objectively more diversified.

One metric you can use is the diversification ratio. This ratio compares the volatility of a portfolio to the weighted average volatility of the underlying assets. This ratio is very simple but accounts for correlations, variances, and relative weights of the assets. In this example the standard deviations of the assets are as follows (using data from 2015 to present):

VTI: 12.46%
VPU: 11.78%
Portfolio B: 9.41%

So the ratio is ((.5 x 12.46%) + (.5 x 11.78%)) / 9.41% = 1.29 ( the higher the ratio, the more diversified the portfolio).

Compare this with Portfolio C which is 50% VTI and 50% VGLT (long term Treasury bonds):

VTI: 12.46%
VGLT: 10.79%
Portfolio B: 6.93%

So the ratio is ((.5 x 12.46%) + (.5 x 10.79%)) / 6.93% = 1.68.

There are other measures, like Shannon entropy, but I think this is easiest to calculate and comprehend.
Thanks. That’s helpful. I’ve been studying the diversification ratio as well, but there seem to be some logical issues with it. Let’s create a third portfolio:

Portfolio C: 50% Portfolio B / 50% VPU

Remember, Portfolio B is itself 50% TSM / 50% VPU. So Portfolio B and VPU will have a correlation less than 1.0. Therefore the diversification ratio tells us that by adding VPU to Portfolio B we will increase the diversification. In other words, Portfolio C is more diversified than Portfolio B.

On the other hand, Portfolio C can equivalently be constructed as follows:

Portfolio C: 25% TSM / 75% VPU

For this construction, I calculate a diversification ratio of about 1.21 using the same data you used above. Since this is less than your 1.29 for Portfolio B, it must be that Portfolio C is less diversified than Portfolio B.

So which is it? Is Portfolio C more or less diversified than Portfolio B?

User avatar
unclescrooge
Posts: 4167
Joined: Thu Jun 07, 2012 7:00 pm

Re: Utilities - worth another look?

Post by unclescrooge » Sat Jun 15, 2019 3:34 pm

vineviz wrote:
Tue Mar 19, 2019 8:58 am
EfficientInvestor wrote:
Tue Mar 19, 2019 8:34 am
The backtest below includes the performance of a variety of Fidelity Select Sector Funds dating back to 1987. I have also included a long term treasury fund and gold fund for grins. If you click on the “assets” tab, you can view the performance of each fund over the time period and also view correlation to the market and correlation between each of the funds. I think the biggest thing that sticks out to me when considering the use of utilities is that the consumer staples sector has had a comparable correlation to the market as utilities but with better return and less volatility. I would expect that trend to continue given the nature of consumer staples. Perhaps you should consider it in addition to utilities?

https://www.portfoliovisualizer.com/bac ... n13_1=7.69
Part of the answer is that past performance isn’t a terribly good proxy for expected future returns. Indeed, it’d be illogical to expect one sector to have lower risk and higher return than another.

Another part of the answer is that it’s both the lower correlation and higher volatility that makes the utility sector a better source of diversification than staples.
Wouldn't gold also fit this criteria?

skeptic42
Posts: 81
Joined: Mon Feb 11, 2019 5:27 pm

Re: Utilities - worth another look?

Post by skeptic42 » Sun Jun 16, 2019 3:32 am

D-Dog wrote:
Sat Jun 15, 2019 12:01 pm
vineviz wrote:
Sat Jun 15, 2019 10:39 am
D-Dog wrote:
Sat Jun 15, 2019 9:37 am
Portfolio A: 100% TSM
Portfolio B: 50% TSM / 50% VPU (utilities)

Is Portfolio B objectively more diversified? If so, how do you know? Is there a metric you look at to determine this? This is a serious question. I’ve been thinking a lot about diversification and I’m not sure how to measure it in a logical way. Thanks!
Yes, B is objectively more diversified.

One metric you can use is the diversification ratio. This ratio compares the volatility of a portfolio to the weighted average volatility of the underlying assets. This ratio is very simple but accounts for correlations, variances, and relative weights of the assets. In this example the standard deviations of the assets are as follows (using data from 2015 to present):

VTI: 12.46%
VPU: 11.78%
Portfolio B: 9.41%

So the ratio is ((.5 x 12.46%) + (.5 x 11.78%)) / 9.41% = 1.29 ( the higher the ratio, the more diversified the portfolio).

Compare this with Portfolio C which is 50% VTI and 50% VGLT (long term Treasury bonds):

VTI: 12.46%
VGLT: 10.79%
Portfolio B: 6.93%

So the ratio is ((.5 x 12.46%) + (.5 x 10.79%)) / 6.93% = 1.68.

There are other measures, like Shannon entropy, but I think this is easiest to calculate and comprehend.
Thanks. That’s helpful. I’ve been studying the diversification ratio as well, but there seem to be some logical issues with it. Let’s create a third portfolio:

Portfolio C: 50% Portfolio B / 50% VPU

Remember, Portfolio B is itself 50% TSM / 50% VPU. So Portfolio B and VPU will have a correlation less than 1.0. Therefore the diversification ratio tells us that by adding VPU to Portfolio B we will increase the diversification. In other words, Portfolio C is more diversified than Portfolio B.

On the other hand, Portfolio C can equivalently be constructed as follows:

Portfolio C: 25% TSM / 75% VPU

For this construction, I calculate a diversification ratio of about 1.21 using the same data you used above. Since this is less than your 1.29 for Portfolio B, it must be that Portfolio C is less diversified than Portfolio B.

So which is it? Is Portfolio C more or less diversified than Portfolio B?
That's a good observation. A similar logical issue occurs with risk parity. IMO this logical issue makes the diversification ratio less useful as an objective measurement of diversification, because the DR of TSM alone is 1, adding VPU increases the DR, but VPU is already part of TSM, so you are adding it twice.

Just as a side note, the Sharpe Ratio doesn't have this logical issue. If one optimizes a portfolio for a maximized SR, adding more of an already included asset doesn't increase the SR any further.

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Utilities - worth another look?

Post by vineviz » Sun Jun 16, 2019 7:28 am

skeptic42 wrote:
Sun Jun 16, 2019 3:32 am
That's a good observation. A similar logical issue occurs with risk parity. IMO this logical issue makes the diversification ratio less useful as an objective measurement of diversification, because the DR of TSM alone is 1, adding VPU increases the DR, but VPU is already part of TSM, so you are adding it twice.
I wouldn't say this is a logical problem. Adding more of certain stocks/sectors to a 100% TSM portfolio should increase the diversification ratio unless TSM is already optimized to be maximally diverse, which we know it is not.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

FrankLUSMC
Posts: 174
Joined: Tue Nov 28, 2017 1:43 pm

Re: Utilities - worth another look?

Post by FrankLUSMC » Sun Jun 16, 2019 7:34 am

HEDGEFUNDIE wrote:
Mon Mar 18, 2019 4:21 pm
aristotelian wrote:
Mon Mar 18, 2019 4:17 pm
Portfolio 2 is not better diversified. It is concentrated in utilities. It may have had lower volatility in the past but there is the risk of a black swan in utilities. There may be good reasons to concentrate in a particular sector but you should be aware of the risks in doing so.
What possible black swans could exist for local utilities on a nationwide level?

Energy input costs are passed to the ratepayer.

Nationalization?
Also prime targets of hackers both national and international.

D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Sun Jun 16, 2019 7:50 am

vineviz wrote:
Sun Jun 16, 2019 7:28 am
skeptic42 wrote:
Sun Jun 16, 2019 3:32 am
That's a good observation. A similar logical issue occurs with risk parity. IMO this logical issue makes the diversification ratio less useful as an objective measurement of diversification, because the DR of TSM alone is 1, adding VPU increases the DR, but VPU is already part of TSM, so you are adding it twice.
I wouldn't say this is a logical problem. Adding more of certain stocks/sectors to a 100% TSM portfolio should increase the diversification ratio unless TSM is already optimized to be maximally diverse, which we know it is not.
If it is not a logical problem, then kindly answer my question from above. Is Portfolio C more or less diversified than Portfolio B?

Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: Utilities - worth another look?

Post by HEDGEFUNDIE » Sun Jun 16, 2019 7:55 am

D-Dog wrote:
Sun Jun 16, 2019 7:50 am
vineviz wrote:
Sun Jun 16, 2019 7:28 am
skeptic42 wrote:
Sun Jun 16, 2019 3:32 am
That's a good observation. A similar logical issue occurs with risk parity. IMO this logical issue makes the diversification ratio less useful as an objective measurement of diversification, because the DR of TSM alone is 1, adding VPU increases the DR, but VPU is already part of TSM, so you are adding it twice.
I wouldn't say this is a logical problem. Adding more of certain stocks/sectors to a 100% TSM portfolio should increase the diversification ratio unless TSM is already optimized to be maximally diverse, which we know it is not.
If it is not a logical problem, then kindly answer my question from above. Is Portfolio C more or less diversified than Portfolio B?
You did the math wrong. B and C should have the exact same diversification ratio.

D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Sun Jun 16, 2019 8:08 am

HEDGEFUNDIE wrote:
Sun Jun 16, 2019 7:55 am
D-Dog wrote:
Sun Jun 16, 2019 7:50 am
vineviz wrote:
Sun Jun 16, 2019 7:28 am
skeptic42 wrote:
Sun Jun 16, 2019 3:32 am
That's a good observation. A similar logical issue occurs with risk parity. IMO this logical issue makes the diversification ratio less useful as an objective measurement of diversification, because the DR of TSM alone is 1, adding VPU increases the DR, but VPU is already part of TSM, so you are adding it twice.
I wouldn't say this is a logical problem. Adding more of certain stocks/sectors to a 100% TSM portfolio should increase the diversification ratio unless TSM is already optimized to be maximally diverse, which we know it is not.
If it is not a logical problem, then kindly answer my question from above. Is Portfolio C more or less diversified than Portfolio B?
You did the math wrong. B and C should have the exact same diversification ratio.
I don’t think so. Show me where the math is wrong.

Portfolio B and C are different portfolios.

Portfolio B: 50% TSM / 50% VPU

Portfolio C: 50% Portfolio B / 50% VPU
Or equivalently,
Portfolio C: 25% TSM / 75% VPU

Why are you claiming that two different portfolios should have the same diversification ratio?

Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: Utilities - worth another look?

Post by HEDGEFUNDIE » Sun Jun 16, 2019 8:16 am

D-Dog wrote:
Sun Jun 16, 2019 8:08 am
HEDGEFUNDIE wrote:
Sun Jun 16, 2019 7:55 am
D-Dog wrote:
Sun Jun 16, 2019 7:50 am
vineviz wrote:
Sun Jun 16, 2019 7:28 am
skeptic42 wrote:
Sun Jun 16, 2019 3:32 am
That's a good observation. A similar logical issue occurs with risk parity. IMO this logical issue makes the diversification ratio less useful as an objective measurement of diversification, because the DR of TSM alone is 1, adding VPU increases the DR, but VPU is already part of TSM, so you are adding it twice.
I wouldn't say this is a logical problem. Adding more of certain stocks/sectors to a 100% TSM portfolio should increase the diversification ratio unless TSM is already optimized to be maximally diverse, which we know it is not.
If it is not a logical problem, then kindly answer my question from above. Is Portfolio C more or less diversified than Portfolio B?
You did the math wrong. B and C should have the exact same diversification ratio.
I don’t think so. Show me where the math is wrong.

Portfolio B and C are different portfolios.

Portfolio B: 50% TSM / 50% VPU

Portfolio C: 50% Portfolio B / 50% VPU
Or equivalently,
Portfolio C: 25% TSM / 75% VPU

Why are you claiming that two different portfolios should have the same diversification ratio?
They are the same portfolio because they hold the same underlying stocks.

The weighted average volatilities of B and C are exactly the same. They are just weighted differently between the two pieces of each portfolio, which is where you went wrong.

User avatar
305pelusa
Posts: 1265
Joined: Fri Nov 16, 2018 10:20 pm

Re: Utilities - worth another look?

Post by 305pelusa » Sun Jun 16, 2019 8:30 am

HEDGEFUNDIE wrote:
Sun Jun 16, 2019 8:16 am
D-Dog wrote:
Sun Jun 16, 2019 8:08 am
HEDGEFUNDIE wrote:
Sun Jun 16, 2019 7:55 am
D-Dog wrote:
Sun Jun 16, 2019 7:50 am
vineviz wrote:
Sun Jun 16, 2019 7:28 am


I wouldn't say this is a logical problem. Adding more of certain stocks/sectors to a 100% TSM portfolio should increase the diversification ratio unless TSM is already optimized to be maximally diverse, which we know it is not.
If it is not a logical problem, then kindly answer my question from above. Is Portfolio C more or less diversified than Portfolio B?
You did the math wrong. B and C should have the exact same diversification ratio.
I don’t think so. Show me where the math is wrong.

Portfolio B and C are different portfolios.

Portfolio B: 50% TSM / 50% VPU

Portfolio C: 50% Portfolio B / 50% VPU
Or equivalently,
Portfolio C: 25% TSM / 75% VPU

Why are you claiming that two different portfolios should have the same diversification ratio?
They are the same portfolio because they hold the same underlying stocks.

The weighted average volatilities of B and C are exactly the same. They are just weighted differently between the two pieces of each portfolio, which is where you went wrong.
I think there's a miscommunication here. Portfolio C and B are most definitely not the same portfolio and will have different diversification ratios. Not doubt about that.

@D-Dog: There is an inconsistency of the math here though. The diversification ratio of 50% portfolio B/50% VPU comes out to be different than 25% VTI/75% VPU. Go ahead and try out the math and you'll see.

I believe this is what Hedgefundie is trying to say.

Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: Utilities - worth another look?

Post by HEDGEFUNDIE » Sun Jun 16, 2019 8:37 am

305pelusa wrote:
Sun Jun 16, 2019 8:30 am
HEDGEFUNDIE wrote:
Sun Jun 16, 2019 8:16 am
D-Dog wrote:
Sun Jun 16, 2019 8:08 am
HEDGEFUNDIE wrote:
Sun Jun 16, 2019 7:55 am
D-Dog wrote:
Sun Jun 16, 2019 7:50 am


If it is not a logical problem, then kindly answer my question from above. Is Portfolio C more or less diversified than Portfolio B?
You did the math wrong. B and C should have the exact same diversification ratio.
I don’t think so. Show me where the math is wrong.

Portfolio B and C are different portfolios.

Portfolio B: 50% TSM / 50% VPU

Portfolio C: 50% Portfolio B / 50% VPU
Or equivalently,
Portfolio C: 25% TSM / 75% VPU

Why are you claiming that two different portfolios should have the same diversification ratio?
They are the same portfolio because they hold the same underlying stocks.

The weighted average volatilities of B and C are exactly the same. They are just weighted differently between the two pieces of each portfolio, which is where you went wrong.
I think there's a miscommunication here. Portfolio C and B are most definitely not the same portfolio and will have different diversification ratios. Not doubt about that.

@D-Dog: There is an inconsistency of the math here though. The diversification ratio of 50% portfolio B/50% VPU comes out to be different than 25% VTI/75% VPU. Go ahead and try out the math and you'll see.

I believe this is what Hedgefundie is trying to say.
Correct.

D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Sun Jun 16, 2019 8:44 am

Removed
Last edited by D-Dog on Sun Jun 16, 2019 8:59 am, edited 1 time in total.

D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Sun Jun 16, 2019 8:51 am

HEDGEFUNDIE wrote:
Sun Jun 16, 2019 8:37 am
305pelusa wrote:
Sun Jun 16, 2019 8:30 am
HEDGEFUNDIE wrote:
Sun Jun 16, 2019 8:16 am
D-Dog wrote:
Sun Jun 16, 2019 8:08 am
HEDGEFUNDIE wrote:
Sun Jun 16, 2019 7:55 am


You did the math wrong. B and C should have the exact same diversification ratio.
I don’t think so. Show me where the math is wrong.

Portfolio B and C are different portfolios.

Portfolio B: 50% TSM / 50% VPU

Portfolio C: 50% Portfolio B / 50% VPU
Or equivalently,
Portfolio C: 25% TSM / 75% VPU

Why are you claiming that two different portfolios should have the same diversification ratio?
They are the same portfolio because they hold the same underlying stocks.

The weighted average volatilities of B and C are exactly the same. They are just weighted differently between the two pieces of each portfolio, which is where you went wrong.
I think there's a miscommunication here. Portfolio C and B are most definitely not the same portfolio and will have different diversification ratios. Not doubt about that.

@D-Dog: There is an inconsistency of the math here though. The diversification ratio of 50% portfolio B/50% VPU comes out to be different than 25% VTI/75% VPU. Go ahead and try out the math and you'll see.

I believe this is what Hedgefundie is trying to say.
Correct.
I agree. I get different answers when I calculate the diversification ratio of these two things:

50% Portfolio B / 50% VPU
25% TSM / 75% VPU

But these two things are the same thing. That's the logical inconsistency I've been trying to point out.

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Utilities - worth another look?

Post by vineviz » Sun Jun 16, 2019 9:43 am

D-Dog wrote:
Sun Jun 16, 2019 7:50 am
vineviz wrote:
Sun Jun 16, 2019 7:28 am
skeptic42 wrote:
Sun Jun 16, 2019 3:32 am
That's a good observation. A similar logical issue occurs with risk parity. IMO this logical issue makes the diversification ratio less useful as an objective measurement of diversification, because the DR of TSM alone is 1, adding VPU increases the DR, but VPU is already part of TSM, so you are adding it twice.
I wouldn't say this is a logical problem. Adding more of certain stocks/sectors to a 100% TSM portfolio should increase the diversification ratio unless TSM is already optimized to be maximally diverse, which we know it is not.
If it is not a logical problem, then kindly answer my question from above. Is Portfolio C more or less diversified than Portfolio B?
First, I hope that sounded much less entitled and demanding in your head that it comes across in writing.

Second, the diversification ratio is a differential (or relative) measure not an absolute one. As a result, you cannot compare DR across portfolios that include different assets.

For instance, a portfolio that is 100% VBINX has a DR=1 but a portfolio that is 60% VTI and 40% BND does not, even though they are functionally equivalent portfolios. Comparing those two portfolios is not a use case for the diversification, but comparing a portfolio that is 60/40 with one that is 70/30 is a use case.

In other words, because portfolio B and portfolio C are composed of assets with different levels of aggregation, their DRs aren't directly comparable. The Choueifaty implementation of DR includes an assumption that the combined assets are single assets, but in portfolio C we've violated that assumption because the 2nd asset (VTI) is 50% of the first asset (portfolio B). You can trace this implicitly in the original Choueifaty paper in their references to the concentration ratio.

Now technically this assumption is also violated in the first use case (VPU is about 3% of VTI), but the impact is relatively trivial (50%^2 is a lot more than 3%^2).
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Sun Jun 16, 2019 10:44 am

vineviz wrote:
Sun Jun 16, 2019 9:43 am
D-Dog wrote:
Sun Jun 16, 2019 7:50 am
vineviz wrote:
Sun Jun 16, 2019 7:28 am
skeptic42 wrote:
Sun Jun 16, 2019 3:32 am
That's a good observation. A similar logical issue occurs with risk parity. IMO this logical issue makes the diversification ratio less useful as an objective measurement of diversification, because the DR of TSM alone is 1, adding VPU increases the DR, but VPU is already part of TSM, so you are adding it twice.
I wouldn't say this is a logical problem. Adding more of certain stocks/sectors to a 100% TSM portfolio should increase the diversification ratio unless TSM is already optimized to be maximally diverse, which we know it is not.
If it is not a logical problem, then kindly answer my question from above. Is Portfolio C more or less diversified than Portfolio B?
First, I hope that sounded much less entitled and demanding in your head that it comes across in writing.

Second, the diversification ratio is a differential (or relative) measure not an absolute one. As a result, you cannot compare DR across portfolios that include different assets.

For instance, a portfolio that is 100% VBINX has a DR=1 but a portfolio that is 60% VTI and 40% BND does not, even though they are functionally equivalent portfolios. Comparing those two portfolios is not a use case for the diversification, but comparing a portfolio that is 60/40 with one that is 70/30 is a use case.

In other words, because portfolio B and portfolio C are composed of assets with different levels of aggregation, their DRs aren't directly comparable. The Choueifaty implementation of DR includes an assumption that the combined assets are single assets, but in portfolio C we've violated that assumption because the 2nd asset (VTI) is 50% of the first asset (portfolio B). You can trace this implicitly in the original Choueifaty paper in their references to the concentration ratio.

Now technically this assumption is also violated in the first use case (VPU is about 3% of VTI), but the impact is relatively trivial (50%^2 is a lot more than 3%^2).
I sounded like a schmuck in my last comment. I didn’t mean it that way. I’m just trying to learn, and working through examples is how I do that best. Sorry about that vineviz.

I think this means that according to the diversification ratio Portfolio C is less diversified than Portfolio B. I need to do the calculation using 25% TSM and 75% VPU rather than the other way. I need to do it this way because Portfolio B and VPU have a high degree of overlapping assets.

Thanks for helping me work through this example. I now better understand the situations in which the diversification ratio can be used and in which situations it cannot.

skeptic42
Posts: 81
Joined: Mon Feb 11, 2019 5:27 pm

Re: Utilities - worth another look?

Post by skeptic42 » Sun Jun 16, 2019 3:26 pm

vineviz wrote:
Sun Jun 16, 2019 7:28 am
skeptic42 wrote:
Sun Jun 16, 2019 3:32 am
That's a good observation. A similar logical issue occurs with risk parity. IMO this logical issue makes the diversification ratio less useful as an objective measurement of diversification, because the DR of TSM alone is 1, adding VPU increases the DR, but VPU is already part of TSM, so you are adding it twice.
I wouldn't say this is a logical problem. Adding more of certain stocks/sectors to a 100% TSM portfolio should increase the diversification ratio unless TSM is already optimized to be maximally diverse, which we know it is not.
Yes, but even if TSM would be maximally diverse, adding more of certain stocks/sectors to this TSM would still lead to a diversification ratio greater 1, suggesting better diversification.
Unfortunately, portfolio visualizer dropped the diversification ratio from its optimization tool.
Do you know what weights would be assigned to each sector, if you maximize the diversification ratio of a portfolio built of all sectors which are part of TSM?

skeptic42
Posts: 81
Joined: Mon Feb 11, 2019 5:27 pm

Re: Utilities - worth another look?

Post by skeptic42 » Mon Jun 17, 2019 9:46 am

Mathematically speaking, the logical issue is due to the diversification ratio is not associative. That means the DR of assets A,B,C is not the same as the DR of (A,B),C.

For example, three assets with the exact same volatilities and expected returns and zero correlations, it is logical from symmetry that a portfolio of all three assets equally weighted (33%) would be optimal.
1. If you combine two of these assets equally weighted and add the third asset to this combined asset, the diversification ratio gets maximized at a weight of 41% for the third asset. (logically it should be 33%)
2. If you combine two of these assets equally weighted and add one of these two assets again (the correlation is 71%), the diversification ratio gets maximized at a weight of 41% again. (logically it should be 0%)

This behavior is why I think the diversification ratio is less useful as an objective measurement of diversification than suggested, especially when adding something to TSM which is already included in TSM (stocks, sectors).
By the way, the Sharpe Ratio is maximized in both cases at the logical weights, 33% and 0% respectively.

User avatar
elcadarj
Posts: 179
Joined: Tue Aug 30, 2016 8:21 am

Re: Utilities - worth another look?

Post by elcadarj » Mon Jun 17, 2019 6:24 pm

vineviz wrote:
Sat Jun 15, 2019 7:16 am
305pelusa wrote:
Fri Jun 14, 2019 10:54 pm
Just a solid, theoretical understanding/proof as to why there can exist a more diversified portfolio than the market cap. Would you have such a reading available?
If markets are efficient, the market portfolio will optimally reflect the aggregate - or average - of all investor risk tolerances and preferences.

The market portfolio can optimally reflect everyone's risk tolerances and preferences ONLY IF everyone has the same risk tolerances and preferences. Is it reasonable to assume that every investor has the same risk tolerances and preferences? If not, the market portfolio can't be optimal for everyone.

The market portfolio can optimally reflect your risk tolerances and preferences ONLY IF you have the same risk tolerances and preferences as everyone else. Do you think your risk tolerances and preferences are the same as everyone else? If not, the market portfolio can't be optimal for you.
+1 indeed. The OP is building his portfolio based on his specific risk profile.

User avatar
G12
Posts: 1526
Joined: Mon Apr 16, 2007 2:35 pm

Re: Utilities - worth another look?

Post by G12 » Mon Jun 17, 2019 7:48 pm

OP, I have almost 5x the utility sector exposure to the US market of 3%. Very long term holder of VPU, have huge (for a utility) embedded LT gain in WEC, plus D and DUK. I owned SO for a long time but their ineptitude with the new nuke build out got old quickly with me so sold while I still had significant gain. Fair warning I also hold a higher multiple of energy than the market so I may be somewhat irrational. I also owned the Vanguard REIT fund in a tax advantaged account when it was cheap long ago and the distribution rates were ~ 7%, sold long ago. There are not as many utes in existence that have what you implied as guaranteed rate increases, and if companies keep going the way of SO and there may be fewer :twisted: . Also hold a lot of VMNVX acquired soon after it became available. I don't own any VTI, maybe I'm an atheist? Sometimes it's ok to march to the beat of a different drummer as long as you know when to rein it in. Also doesn't mean one should jump all-in into a new investment of what I assume represents significant $.

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Utilities - worth another look?

Post by vineviz » Tue Jun 18, 2019 6:58 am

skeptic42 wrote:
Sun Jun 16, 2019 3:26 pm
Do you know what weights would be assigned to each sector, if you maximize the diversification ratio of a portfolio built of all sectors which are part of TSM?
I was out of town for funeral this weekend, but I just ran the numbers and got this:

38.88% Vanguard Utilities ETF ( VPU )
18.04% Vanguard Energy ETF ( VDE )
15.07% Vanguard Health Care ETF ( VHT )
12.70% Vanguard Financials ETF ( VFH )
5.70% Vanguard Information Technology ETF ( VGT )
5.52% Vanguard Real Estate ETF ( VNQ )
3.41% Vanguard Communication Services ETF ( VOX )
0.68% Vanguard Consumer Staples ETF ( VDC )
0.00% Vanguard Industrials ETF ( VIS )
0.00% Vanguard Consumer Discretionary ETF ( VCR )
0.00% Vanguard Materials ETF ( VAW )
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

Hydromod
Posts: 283
Joined: Tue Mar 26, 2019 10:21 pm

Re: Utilities - worth another look?

Post by Hydromod » Tue Jun 18, 2019 9:51 am

Adding in TMF to the maximum efficiency mix, say weighted at 37 TMF/63 max efficient, does well to smooth out some bumps. I get for 2005 to present

CAGR: 10.86% versus 8.61%
worst year: -5.99% versus -32.75%
Sharpe ratio: 0.71 versus 0.62
Sortino ratio: 1.14 versus 0.86

Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: Utilities - worth another look?

Post by HEDGEFUNDIE » Tue Jun 18, 2019 10:17 am

Hydromod wrote:
Tue Jun 18, 2019 9:51 am
Adding in TMF to the maximum efficiency mix, say weighted at 37 TMF/63 max efficient, does well to smooth out some bumps. I get for 2005 to present

CAGR: 10.86% versus 8.61%
worst year: -5.99% versus -32.75%
Sharpe ratio: 0.71 versus 0.62
Sortino ratio: 1.14 versus 0.86
Trying to keep the leveraged products out of my main portfolio, thanks.

Hydromod
Posts: 283
Joined: Tue Mar 26, 2019 10:21 pm

Re: Utilities - worth another look?

Post by Hydromod » Tue Jun 18, 2019 10:57 am

HEDGEFUNDIE wrote:
Tue Jun 18, 2019 10:17 am
Hydromod wrote:
Tue Jun 18, 2019 9:51 am
Adding in TMF to the maximum efficiency mix, say weighted at 37 TMF/63 max efficient, does well to smooth out some bumps. I get for 2005 to present

CAGR: 10.86% versus 8.61%
worst year: -5.99% versus -32.75%
Sharpe ratio: 0.71 versus 0.62
Sortino ratio: 1.14 versus 0.86
Trying to keep the leveraged products out of my main portfolio, thanks.
Was just intended as an interesting observation as I explore aspects of your strategies, I'm not necessarily suggesting doing this. It seemed nice that a relatively small fraction of the portfolio with leveraged treasuries could be used to give good ballast to the remainder of the portfolio with unleveraged stocks.

User avatar
Electron
Posts: 1932
Joined: Sat Mar 10, 2007 8:46 pm

Re: Utilities - worth another look?

Post by Electron » Tue Jun 18, 2019 2:20 pm

I hold a selection of utility stocks in a taxable account for the qualified dividends. The performance overall has far exceeded my expectations with both regular dividend increases and capital appreciation.

I'm not sure this is the best time to get into the sector unless the time horizon is quite long. The utility sector is actually quite risky which can be seen in longer term charts. I have added to positions in the past on weakness but at much lower prices than today.
Electron

aristotelian
Posts: 6668
Joined: Wed Jan 11, 2017 8:05 pm

Re: Utilities - worth another look?

Post by aristotelian » Tue Jun 18, 2019 2:24 pm

vineviz wrote:
Sat Jun 15, 2019 10:39 am
D-Dog wrote:
Sat Jun 15, 2019 9:37 am
Portfolio A: 100% TSM
Portfolio B: 50% TSM / 50% VPU (utilities)

Is Portfolio B objectively more diversified? If so, how do you know? Is there a metric you look at to determine this? This is a serious question. I’ve been thinking a lot about diversification and I’m not sure how to measure it in a logical way. Thanks!
Yes, B is objectively more diversified.

One metric you can use is the diversification ratio. This ratio compares the volatility of a portfolio to the weighted average volatility of the underlying assets. This ratio is very simple but accounts for correlations, variances, and relative weights of the assets. In this example the standard deviations of the assets are as follows (using data from 2015 to present):

VTI: 12.46%
VPU: 11.78%
Portfolio B: 9.41%

So the ratio is ((.5 x 12.46%) + (.5 x 11.78%)) / 9.41% = 1.29 ( the higher the ratio, the more diversified the portfolio).

Compare this with Portfolio C which is 50% VTI and 50% VGLT (long term Treasury bonds):

VTI: 12.46%
VGLT: 10.79%
Portfolio B: 6.93%

So the ratio is ((.5 x 12.46%) + (.5 x 10.79%)) / 6.93% = 1.68.

There are other measures, like Shannon entropy, but I think this is easiest to calculate and comprehend.
No, B is objectively concentrated in Utilities. I don't see how that meets any common sense definition of "diversified". It may have better risk-adjusted expected return or lower expected volatility than 100% total stock.

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Utilities - worth another look?

Post by vineviz » Tue Jun 18, 2019 3:01 pm

aristotelian wrote:
Tue Jun 18, 2019 2:24 pm
No, B is objectively concentrated in Utilities. I don't see how that meets any common sense definition of "diversified".
I'm sorry that you don't see it.

There are several threads on diversification on this board that might explain it in a way that makes sense to you, so you might start by searching for those. If it still doesn't make sense to you, start a thread asking for an explanation and I'll keep an eye open for it.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

skeptic42
Posts: 81
Joined: Mon Feb 11, 2019 5:27 pm

Re: Utilities - worth another look?

Post by skeptic42 » Tue Jun 18, 2019 6:42 pm

vineviz wrote:
Tue Jun 18, 2019 6:58 am
skeptic42 wrote:
Sun Jun 16, 2019 3:26 pm
Do you know what weights would be assigned to each sector, if you maximize the diversification ratio of a portfolio built of all sectors which are part of TSM?
I was out of town for funeral this weekend, but I just ran the numbers and got this:

38.88% Vanguard Utilities ETF ( VPU )
18.04% Vanguard Energy ETF ( VDE )
15.07% Vanguard Health Care ETF ( VHT )
12.70% Vanguard Financials ETF ( VFH )
5.70% Vanguard Information Technology ETF ( VGT )
5.52% Vanguard Real Estate ETF ( VNQ )
3.41% Vanguard Communication Services ETF ( VOX )
0.68% Vanguard Consumer Staples ETF ( VDC )
0.00% Vanguard Industrials ETF ( VIS )
0.00% Vanguard Consumer Discretionary ETF ( VCR )
0.00% Vanguard Materials ETF ( VAW )
Thank you very much for running the numbers, vineviz!

With these maximally diverse sector weights, I get a diversification ratio of about 1.31.
With the current sector weights in VTI, I get a diversification ratio of about 1.17.

I used the correlations and standard deviations from PV for the calculation:
https://www.portfoliovisualizer.com/ass ... &months=36

Backtest with maximally diverse sector weights (portfolio 1) and with current sector weights (portfolio 2):
https://www.portfoliovisualizer.com/bac ... 0&total3=0

D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Wed Jun 19, 2019 6:21 am

vineviz wrote:
Tue Jun 18, 2019 3:01 pm
aristotelian wrote:
Tue Jun 18, 2019 2:24 pm
No, B is objectively concentrated in Utilities. I don't see how that meets any common sense definition of "diversified".
I'm sorry that you don't see it.

There are several threads on diversification on this board that might explain it in a way that makes sense to you, so you might start by searching for those. If it still doesn't make sense to you, start a thread asking for an explanation and I'll keep an eye open for it.
What if instead of VPU (utilities etf) we used one utility stock. NextEra (NEE) is the largest holding in VPU. Consider this portfolio:

50% VTI / 50% NEE

According to portfolio visualizer, NEE and VTI have a correlation of 0.4. Therefore, the above portfolio should be objectively more diversified than VTI according to the diversification ratio. But intuitively, having 50% of your portfolio in one stock sure doesn’t feel more diversified. What are your thoughts on this one?

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Utilities - worth another look?

Post by vineviz » Wed Jun 19, 2019 9:04 am

D-Dog wrote:
Wed Jun 19, 2019 6:21 am
But intuitively, having 50% of your portfolio in one stock sure doesn’t feel more diversified. What are your thoughts on this one?
I guess I have a couple of thoughts.

One is that I have the same intuitive reaction that you seem to (i.e. that having 50% of the portfolio in one stock doesn't feel more diversified), but I also feel strongly that portfolio construction is an endeavor where intuition can often lead to poor choices. So I guess I'd say that I'd be slow assume that the portfolio with NEE is necessarily less diversified than the one with only total stock market.

A second thought is that (and I think I've said this before) that the diversification ratio is neither an ideal nor a complete measure of diversification. It has a specific use case, for which it works pretty well, and its ease of calculation means it can used much more quickly and readily than some other measures. Sometimes these other measures can provide some useful insight, and in this case something like the concentration ratio (i.e. sum of squared weights aka Herfindahl-Hirschman Index) would highlight how much more concentrated the VTI/NEE portfolio is than the VTI/VPU portfolio.

A third thought is related to the second, and is that diversification ratio should be used as part of process that reflects its formulation and limitations. It's not an absolute metric, for instance, but differential or relative one. And the differential comparison is predicated on the assets being compared not being subsets of each other. This predication isn't necessarily debilitating, but it does mean that utilizing a bit of judgement will be rewarded.

For instance, pulling out the top stock from VPU and calculating the DR with just VTI and NEE does give you a diversification ratio above 1.0. But as a differential measure, the question the diversification ratio is actually well suited to answer is whether that combination of VTI and utility stocks is the best you can do.

What if instead of 50% each in VTI and NEE you construct a portfolio that is 50% VTI plus 25% each of NEE and DUK? The DR goes up significantly. Continue splitting the utility portion equally between the other top stocks in VPU and you continue to improve diversification.

Image

So now we're back to a place that makes intuitive sense: having 5 or 8 different utility stocks is more diversified than just having one, but the marginal diversification benefits start to dwindle pretty quickly.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

aristotelian
Posts: 6668
Joined: Wed Jan 11, 2017 8:05 pm

Re: Utilities - worth another look?

Post by aristotelian » Wed Jun 19, 2019 9:11 am

I was playing with PV this morning to backtest some of the ideas in this thread, and made the mistake of using BND instead of VBMFX as the benchmark comparing to XLU. Doing this gives a shorter dataset from 2008 to the present. The argument for XLU hinges on providing stock-like returns less correlated to the market, but during this period XLU significantly underperformed, 8.08% to 6.69%.

During this period, a portfolio of 75 VTSAX/25 BND achieves about the same CAGR of XLU but with lower volatility due to the stronger non-correlation of bonds. If you were looking to reduce volatility, you would be better off with a smaller % of bonds than concentrating stocks in Utilities.

XLU becomes competitive when compared to the larger dataset going back to 2000, primarily due to a big run of outperformance (both higher highs and lower losses) in 2005-8. Regardless of whether you put greater trust in the larger dataset, this does highlight at the very least that concentrating in a single sector risks significantly long periods of underperformance and/or uncompensated risk.

I also have to wonder whether XLU will "progress" to the mean by outperforming the market during some future period, or if 2008-present is the "new normal". Usually I would trust the larger dataset, but I would note some reason to think the latter. That is, if Utilities as a sector have lower volatility, one would expect lower returns.

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: Utilities - worth another look?

Post by vineviz » Wed Jun 19, 2019 9:22 am

aristotelian wrote:
Wed Jun 19, 2019 9:11 am
I was playing with PV this morning to backtest some of the ideas in this thread, and made the mistake of using BND instead of VBMFX as the benchmark comparing to XLU. Doing this gives a shorter dataset from 2008 to the present. The argument for XLU hinges on providing stock-like returns less correlated to the market, but during this period XLU significantly underperformed, 8.08% to 6.69%.
If you want to extend your tests back further I PV you can use Franklin Utilities Fund Class A1 (FKUTX). It actually goes back earlier than VFINX does, and while it is definite not an index fund it has historically tracked VPU reasonably well.

https://www.portfoliovisualizer.com/eff ... X&total1=0
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

Oddball
Posts: 66
Joined: Thu May 03, 2018 9:35 am

Re: Utilities - worth another look?

Post by Oddball » Wed Jun 19, 2019 9:47 am

So, it is a good idea to shift some funds to Utilities? All of this diversification discussion is interesting but not that helpful to me in deciding to make some changes or not.

If you have common 3-fund portfolio of 50% US stocks (VTSAX), 25% VTI, and 25% BND, would shifting to 40/25/25/10 VPU (for example) smooth out some of the rough times. Or is it likely not going to matter too much in the greater scheme of things. I am looking at long term (20 year) timeline, and of course I know that nothing is guaranteed moving forward.

Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: Utilities - worth another look?

Post by HEDGEFUNDIE » Wed Jun 19, 2019 9:56 am

Oddball wrote:
Wed Jun 19, 2019 9:47 am
So, it is a good idea to shift some funds to Utilities? All of this diversification discussion is interesting but not that helpful to me in deciding to make some changes or not.

If you have common 3-fund portfolio of 50% US stocks (VTSAX), 25% VTI, and 25% BND, would shifting to 40/25/25/10 VPU (for example) smooth out some of the rough times. Or is it likely not going to matter too much in the greater scheme of things. I am looking at long term (20 year) timeline, and of course I know that nothing is guaranteed moving forward.
10% allocation will make no difference.

You will need at least 20% to really feel the impact.

aristotelian
Posts: 6668
Joined: Wed Jan 11, 2017 8:05 pm

Re: Utilities - worth another look?

Post by aristotelian » Wed Jun 19, 2019 10:05 am

Oddball wrote:
Wed Jun 19, 2019 9:47 am
So, it is a good idea to shift some funds to Utilities? All of this diversification discussion is interesting but not that helpful to me in deciding to make some changes or not.

If you have common 3-fund portfolio of 50% US stocks (VTSAX), 25% VTI, and 25% BND, would shifting to 40/25/25/10 VPU (for example) smooth out some of the rough times. Or is it likely not going to matter too much in the greater scheme of things. I am looking at long term (20 year) timeline, and of course I know that nothing is guaranteed moving forward.
I don't think there is a consensus. There are some theories as to why Utilities might be good and some backtesting to support it. However, you are concentrating in a single sector and putting yourself at risk of underperforming.

skeptic42
Posts: 81
Joined: Mon Feb 11, 2019 5:27 pm

Re: Utilities - worth another look?

Post by skeptic42 » Wed Jun 19, 2019 11:57 am

D-Dog wrote:
Wed Jun 19, 2019 6:21 am
What if instead of VPU (utilities etf) we used one utility stock. NextEra (NEE) is the largest holding in VPU. Consider this portfolio:

50% VTI / 50% NEE

According to portfolio visualizer, NEE and VTI have a correlation of 0.4. Therefore, the above portfolio should be objectively more diversified than VTI according to the diversification ratio. But intuitively, having 50% of your portfolio in one stock sure doesn’t feel more diversified. What are your thoughts on this one?
The diversification ratio is always larger than 1 when adding something not perfectly correlated to a portfolio, even if the portfolio is already maximally diversified resulting in diworsification/concentration. The DR doesn't have the associative property, so you will get different results depending on how you slice and dice the portfolio (e.g. single stocks, sectors, styles). Thus, the DR doesn't give an absolute answer and can be misleading like in your example. A relative comparison makes more sense, e.g. splitting VTI into all single stocks and then assessing whether over/under-weighting NEE increases the DR.

D-Dog
Posts: 74
Joined: Sun Dec 02, 2007 12:44 am

Re: Utilities - worth another look?

Post by D-Dog » Wed Jun 19, 2019 10:41 pm

skeptic42 wrote:
Wed Jun 19, 2019 11:57 am
D-Dog wrote:
Wed Jun 19, 2019 6:21 am
What if instead of VPU (utilities etf) we used one utility stock. NextEra (NEE) is the largest holding in VPU. Consider this portfolio:

50% VTI / 50% NEE

According to portfolio visualizer, NEE and VTI have a correlation of 0.4. Therefore, the above portfolio should be objectively more diversified than VTI according to the diversification ratio. But intuitively, having 50% of your portfolio in one stock sure doesn’t feel more diversified. What are your thoughts on this one?
The diversification ratio is always larger than 1 when adding something not perfectly correlated to a portfolio, even if the portfolio is already maximally diversified resulting in diworsification/concentration. The DR doesn't have the associative property, so you will get different results depending on how you slice and dice the portfolio (e.g. single stocks, sectors, styles). Thus, the DR doesn't give an absolute answer and can be misleading like in your example. A relative comparison makes more sense, e.g. splitting VTI into all single stocks and then assessing whether over/under-weighting NEE increases the DR.
After working through the various examples and considering the mathematical issues you've pointed out, my opinion is that the diversification ratio does not have a lot of practical value. I certainly don't consider it an objective measure of diversification. For me, diversification requires a certain amount of common sense.

User avatar
nedsaid
Posts: 12971
Joined: Fri Nov 23, 2012 12:33 pm

Re: Utilities - worth another look?

Post by nedsaid » Wed Jun 19, 2019 11:07 pm

It has been known for a long time that different sectors of the market act differently from each other. So we chop it up and recombine with different weightings and tilts to add diversification benefits and hopefully boost performance.

We have divided the stock market by industry groups or sectors. Technology acts differently than Utilities, Real Estate acts differently than Consumer Staples. Folks gravitate to the hot sectors. Energy would have been the place to be in the 1970's. The 1990's were a Technology decade as are the 2010's. It seems lots of folks loved the Healthcare sector, the Vanguard fund was a big favorite here.

Sectors behave differently from each other because each has an outside force that drives it. Real Estate, Utilities, and Financials are interest rate sensitive. Basic materials and consumer cyclical rise and fall with the economy itself. Technology is driven by innovation and the public's thirst for the latest and greatest gadget, what I call the cool factor. Utilities and Consumer Staples are seen as defensive stocks, these sectors compose much of Low Volatility, these are lower beta stocks. People still need such things as electricity, water, toothpaste, toilet paper, mouthwash during recessions. These type of stocks are not as economically sensitive.

We also have divided the market by Large-Cap, Mid-Cap, and Small-Cap. Large stocks and smaller stocks take turns leading the market. We have further parsed the market by Value, Blend, and Growth. Value stocks and Growth stocks also take turns outperforming each other. Large stocks tend to be more multi-national companies, deriving much of their revenue abroad. Smaller stocks tend to be more domestic oriented, most of their revenue being generated in the United States.

Sector funds used to be the rage and now investors are chasing the factors. A cynic could say that factor funds are sector funds in drag. Slicing up the market differently but much the same thing. Sector funds were the rage in the 1990's.

We haven't even started discussing US Stocks vs. International Stocks. One can also parse the International Stocks by sectors and factors as well. A big driver for International Stocks is the strength of the US Dollar relative to other currencies. A weakening dollar is a tailwind for International Stocks and a strengthening dollar is a headwind. One big reason that US Stocks have outperformed International Stocks since the 2008-2009 financial crisis is the strengthening of the US Dollar relative to the other strong currencies like the Euro, the Pound, and the Yen.

Some of this is performance chasing, we notice that something has performed well recently and so we want it. So we chase factors and sectors. Some of the preference with certain sectors are the diversification benefits. For example, the pitch for Real Estate stocks (REITs) was that they outperformed the S&P 500 a bit but with low correlation to the index. So the S&P 500 often zigged while REITs zagged, so in theory a portfolio with both should have about the same returns with less volatility.

As far as the diversification benefits of Utilities, a good case has been made. The Achilles heel of Utilities are rising interest rates. Two reasons, utilities carry a lot of debt as they have high investment and they also compete with bonds in the chase for income. So as long as interest rates are well behaved, you are good. An inflation spike with resulting interest rate hikes will cause lots of heartburn for this sector. I would say these are a rather imperfect diversifier.
Last edited by nedsaid on Wed Jun 19, 2019 11:45 pm, edited 1 time in total.
A fool and his money are good for business.

User avatar
nedsaid
Posts: 12971
Joined: Fri Nov 23, 2012 12:33 pm

Re: Utilities - worth another look?

Post by nedsaid » Wed Jun 19, 2019 11:29 pm

Maybe I am biased against utilities because of my experience with individual stocks. I had some success with Telecom. But electric utilities just never seemed to work for me, I bought the two local utilities at different times and they were just dogs. Ditto for the auto sector, never seemed to be able to make much money on these. My favorite industry groups were forest products, banks, insurance companies, oil, and pharmaceuticals. Also have had some luck with technology.

So perhaps my bad experiences with electric utilities has colored my viewpoint. I probably would not buy a Utility sector fund or ETF.
A fool and his money are good for business.

aristotelian
Posts: 6668
Joined: Wed Jan 11, 2017 8:05 pm

Re: Utilities - worth another look?

Post by aristotelian » Wed Jul 24, 2019 9:11 am

I don't know if this has been mentioned, but I just had a random thought. If you were considering a tilt toward Utilities and/or REIT, one advantage of Utilities over REIT would be tax efficiency. You could hold long term in taxable account and dividends would be qualified. This would allow you additional space in Roth for REIT, or to keep Roth diversified if you did not want to risk overconcentration in Utilities/REIT with your valuable Roth space.

Post Reply