[Paul Merriman: This strategy beats a total stock market fund]

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
marcopolo
Posts: 8445
Joined: Sat Dec 03, 2016 9:22 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by marcopolo »

willthrill81 wrote: Tue Oct 22, 2019 5:38 pm
UberGrub wrote: Tue Oct 22, 2019 5:16 pm Any ways, back to this thread. I just saw this article. Do we know if it was in fact a typo?
Yes, it was. Paul clarified it further up the thread.
People seem to read (hear) what they want to believe.

Paul did not say it was a typo. Typographical errors can usually be easily reconciled in the context of an article. In this article he specifically describes the four corners of equity classes, then says a better approach is to own all four of them in equal amounts. Hard to see how that was a typo. If it was a simple typo, i doubt it would have generated as much discussion here as it did, because otherwise it is just another one of his articles advocating mild tilting.

He said that he wrote"growth" as a substitute for "blend" in an effort to simplify things in a short article. I have no idea how that actually simplifies things when it clearly gives the wrong impression.

I do stand corrected in my assertion that he was changing his recommended strategy. He has indeed corrected that misunderstanding on my part by explaining why the article stated something quite different than his usual recommendations.

Here is what he actually said:

"In an attempt to simplify the article (always trying to limit word count at Marketwatch) we used the word growth rather than blend. There is no risk in the past. We always know what we should have done. We should have used blend and taken a few words to make it clear that blend is a combination of growth and value."
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Uncorrelated »

Elysium wrote: Tue Oct 22, 2019 6:27 pm No, that is what I said in the beginning. There has been no value premium since 2004, or very little of it. This could be because everyone found out about it and flooded the market with products that attempts to capture it. We might need to go through a really long and painful cycle that may last 30 years perhaps before value premium makes a come back. Is this a bet worth taking is up to everyone concerned with their money to think hard about.
That is just cherry picking. According to FF, the monthly standard deviation of the value premium is around 10 times as large as the value premium. This indicates that:
The chance that the value premium is negative over the next 8 years is only 1 standard deviation away from zero.
The chance that the value premium is negative over the next 30 years is only 2 standard deviations away from zero.


What about the market premium? On 1963~1991 the standard deviation of the market premium has been 10.6 times as large as the market premium. Since 1926, 10.25 times as large. Simply put, the chance that the market premium is positive over a 10 year period is identical to the chance that the value premium delivers a positive return. In the 14 years following 2000, the market premium was negative delivered negative real return. We might go through a really long and painful cycle before the market premium makes a comeback. Is that a bet worth taking?


There are good reasons to be skeptical of the value premium, but a 15-year period of negative performance is not among them.

Edit: got some stats from ken french data library. MKT was negative for 186 months (15.5 years) following 1929, for 173 months (14.5 years) following 1968. Three months later the market premium went negative again for a total of 199 months (starting at 1968). Counting from the peak of the 2000' bubble, the market premium was negative for 154 months (12.8 years).

The value premium was negative for 164 months (13.6 years) following 1929 and for 153 months (12.7 years, and counting) since 2006. There have been no other periods longer than 100 months where value underperformed (or I can't find them).

The small cap premium went negative for 328 months (27 years) following 1938, which is to be expected since the small cap premium is much smaller than value. (premium 15 and 20 times smaller than the standard deviation)
Last edited by Uncorrelated on Wed Oct 23, 2019 9:25 am, edited 4 times in total.
User avatar
Forester
Posts: 2400
Joined: Sat Jan 19, 2019 1:50 pm
Location: UK

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Forester »

"TSM is diversified because it's the wisdom of the masses. 0.001% smallcap is what the market has deigned to be correct"

- if TSM is so diversified why did it follow the Nasdaq off a cliff and go nowhere from 1999 to 2013? Megacap indexing does have flaws which can be managed.
Amateur Self-Taught Senior Macro Strategist
JustinR
Posts: 1451
Joined: Tue Apr 27, 2010 11:43 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by JustinR »

Is he some sort of psychic?
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by YRT70 »

Elysium wrote: Tue Oct 22, 2019 6:33 pm Profession that requires advanced degrees in Math and Computer Science.
May I ask what education in math you had? And how old are you?
donaldfair71
Posts: 1241
Joined: Wed Mar 06, 2013 3:15 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by donaldfair71 »

Forester wrote: Wed Oct 23, 2019 2:33 am "TSM is diversified because it's the wisdom of the masses. 0.001% smallcap is what the market has deigned to be correct"

- if TSM is so diversified why did it follow the Nasdaq off a cliff and go nowhere from 1999 to 2013? Megacap indexing does have flaws which can be managed.
How many years of "go nowhere" is necessary before we decide something is diversified?
acegolfer
Posts: 3029
Joined: Tue Aug 25, 2009 9:40 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by acegolfer »

I'm surprised there is no single mentioning of "risk" in OP. Isn't it possible to beat TSM by taking more risk? Seems that what this strategy is.
paul merriman
Posts: 50
Joined: Fri Oct 25, 2013 1:00 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by paul merriman »

I think the following comment may help move this debate toward a close. In the article we list the 4 great asset classes. In fact, large and small cap growth are both very important asset classes. The question becomes how much you should have in your portfolio. My recommendation of using large and small cap blend, instead of large and small cap growth, is simply an attempt to build a portfolio that leans to value rather than growth. When you do this with funds or ETFs, the percentages of each asset class will vary as the funds and ETFs are built with more or less of each of the asset classes. My own buy and hold portfolio is built with DFA funds, rather than Vanguard. The outcome of that is a smaller average size company and more deeply discounted value. That doesn't make DFA better than Vanguard but it is likely to generate higher returns over the long term. How long is the long term? In ny case the long term is however long I live. When people ask me how my portfolio has done, I suggest they ask my kids after I've died as that is the point at which my investment decisions should be judged. If Vanguard would have been the better decision it suggests that large produced better returns than small and growth produced better returns than value. I won't be shocked if that's the outcome as no one knows if the future will be different than the past. At this point I am trusting the past but many people, much smarter than I, believe the future will be different. I don't envy the young but I do envy the fact that 50 years from now they will know the outcome. We have to keep in mind that no matter how loud we yell our beliefs and no matter how many facts we have to support our beliefs, it doesn't mean our beliefs are right. Good luck to you all. I wish you could all be right.
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Uncorrelated »

acegolfer wrote: Wed Oct 23, 2019 7:43 am I'm surprised there is no single mentioning of "risk" in OP. Isn't it possible to beat TSM by taking more risk? Seems that what this strategy is.
The approach in the OP indeed takes more risk, but that is a choice. It is also possible to get the same returns as TSM with less risk, or higher returns than TSM with the same risk.
rascott
Posts: 2957
Joined: Wed Apr 15, 2015 10:53 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by rascott »

Uncorrelated wrote: Wed Oct 23, 2019 2:00 am
Elysium wrote: Tue Oct 22, 2019 6:27 pm No, that is what I said in the beginning. There has been no value premium since 2004, or very little of it. This could be because everyone found out about it and flooded the market with products that attempts to capture it. We might need to go through a really long and painful cycle that may last 30 years perhaps before value premium makes a come back. Is this a bet worth taking is up to everyone concerned with their money to think hard about.
That is just cherry picking. According to FF, the monthly standard deviation of the value premium is around 10 times as large as the value premium. This indicates that:
The chance that the value premium is negative over the next 8 years is only 1 standard deviation away from zero.
The chance that the value premium is negative over the next 30 years is only 2 standard deviations away from zero.


What about the market premium? On 1963~1991 the standard deviation of the market premium has been 10.6 times as large as the market premium. Since 1926, 10.25 times as large. Simply put, the chance that the market premium is positive over a 10 year period is identical to the chance that the value premium delivers a positive return. In the 14 years following 2000, the market premium was negative. We might go through a really long and painful cycle before the market premium makes a comeback. Is that a bet worth taking?


There are good reasons to be skeptical of the value premium, but a 15-year period of negative performance is not among them.

Exit: got some stats from ken french data library. MKT was negative for 186 months (15.5 years) following 1929, for 173 months (14.5 years) following 1968. Three months later the market premium went negative again for a total of 199 months (starting at 1968). Counting from the peak of the 2000' bubble, the market premium was negative for 154 months (12.8 years).

The value premium was negative for 164 months (13.6 years) following 1929 and for 153 months (12.7 years, and counting) since 2006. There have been no other periods longer than 100 months where value underperformed (or I can't find them).

The small cap premium went negative for 328 months (27 years) following 1938, which is to be expected since the small cap premium is much smaller than value. (premium 15 and 20 times smaller than the standard deviation)

Thank you for some real math and data
dknightd
Posts: 3727
Joined: Wed Mar 07, 2018 10:57 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by dknightd »

paul merriman wrote: Wed Oct 23, 2019 8:07 am I don't envy the young but I do envy the fact that 50 years from now they will know the outcome. We have to keep in mind that no matter how loud we yell our beliefs and no matter how many facts we have to support our beliefs, it doesn't mean our beliefs are right. Good luck to you all. I wish you could all be right.
I have to respond. Even though I did not properly read all comments and links.

I envy the young, because they are young!

I've never figured out this factor investing thing. When does a LCG turn into a LCV. When does small cap turn into large cap?
When does an emerging market turn into a developed market? I suspect all those decisions are made by others, using different numbers and criteria. My numbers and criteria might be different than what the factor/indexers think. I expect that only very small companies are not effected by global markets. I do agree that market capitalization is probably not the best way to define an index. I'm not sure there is a better way.

I do believe that telling our kids to save money, spend less than they earn, is a good thing. I don't think we need to yell about it. They will learn by example! Hopefully ;)
Retired 2019. So far, so good. I want to wake up every morning. But I want to die in my sleep. Just another conundrum. I think the solution might be afternoon naps ;)
acegolfer
Posts: 3029
Joined: Tue Aug 25, 2009 9:40 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by acegolfer »

Uncorrelated wrote: Wed Oct 23, 2019 8:11 am It is also possible to get the same returns as TSM with less risk, or higher returns than TSM with the same risk.
1. consistently?
2. are you measuring all the risks by stdev?
MotoTrojan
Posts: 11259
Joined: Wed Feb 01, 2017 7:39 pm

Re: Paul Merriman article

Post by MotoTrojan »

willthrill81 wrote: Mon Oct 21, 2019 3:12 pm
greg24 wrote: Mon Oct 21, 2019 3:11 pm He may not have mentioned SCG on his podcast, but the linked article we are discussing includes these quotes:

"The four most important U.S. equity asset classes are large-cap growth, large-cap value, small-cap growth, and small-cap value."

"But here’s an even better idea for the equity part of your portfolio: Ditch the TMI or S&P 500 fund, and diversify equally among large-cap growth, large-cap value, small-cap growth, and small-cap value."
What's the advantage to owning both LCG and LCV rather than just LC? Ditto for SCG and SCV.

IIRC, owning LC and SCV give you the greatest diversification.
+1. Makes no sense to me, unless you’re splitting between taxable and IRA for tax-efficiency.

I’m happy with my Total US plus strong small-value tilt.
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Uncorrelated »

acegolfer wrote: Wed Oct 23, 2019 9:33 am
Uncorrelated wrote: Wed Oct 23, 2019 8:11 am It is also possible to get the same returns as TSM with less risk, or higher returns than TSM with the same risk.
1. consistently?
I don't know what the meaning of this word is in this context. When playing roulette, does abstaining consistently results in higher returns than betting on black, or just 19/37 (51.3%) of the time?


2. are you measuring all the risks by stdev?
Yes. I'm sure there is a better way, but this is the best way that I understand.
acegolfer
Posts: 3029
Joined: Tue Aug 25, 2009 9:40 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by acegolfer »

Uncorrelated wrote: Wed Oct 23, 2019 9:44 am
acegolfer wrote: Wed Oct 23, 2019 9:33 am 2. are you measuring all the risks by stdev?
Yes. I'm sure there is a better way, but this is the best way that I understand.
Under multi-factor asset pricing model, it is certainly possible to have higher expected return than the market without increasing stdev. IOW, higher Sharpe ratio is possible. But it's going to increase other risks that you are not measuring.
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Uncorrelated »

acegolfer wrote: Wed Oct 23, 2019 10:13 am
Uncorrelated wrote: Wed Oct 23, 2019 9:44 am
acegolfer wrote: Wed Oct 23, 2019 9:33 am 2. are you measuring all the risks by stdev?
Yes. I'm sure there is a better way, but this is the best way that I understand.
Under multi-factor asset pricing model, it is certainly possible to have higher expected return than the market without increasing stdev. IOW, higher Sharpe ratio is possible. But it's going to increase other risks that you are not measuring.
Such as? Data mining bias? Reduce the expected factor premium with 50% and you're pretty safe. Tracking error from long-only funds? Easy to account for that by increasing the standard deviation of the expected premium a bit.

I suppose you could add management risk to the list of risks, but there are a few small cap value funds available where the risk of mismanagement appears to be no larger than that of the S&P 500.
texasdiver
Posts: 3937
Joined: Thu Jun 25, 2009 12:50 am
Location: Vancouver WA

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by texasdiver »

I'm just a layman but I'm unconvinced.

1. If I own Total Stock Market and Total Int'l Stock Market then I already own every single equity in Merriman's more complicated portfolio, just in somewhat different weights.

2. The market has already valued every single equity at their current values. Any other tilts such as those proposed by Merriman are second-guessing for no apparent rational reason. There may be a reason to tilt towards small cap value which is the smallest of the four boxes, but I don't understand what it is.

3. Dividing a portfolio equally between four unequal sectors makes no rational sense. The Morningstar 4-corner box thing is simply a visual construct. The actual market doesn't look like that. It would be like dividing up a global portfolio into 7 equal funds, one for each continent simply because there are 7 continents.
donaldfair71
Posts: 1241
Joined: Wed Mar 06, 2013 3:15 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by donaldfair71 »

texasdiver wrote: Wed Oct 23, 2019 11:25 am I'm just a layman but I'm unconvinced.

1. If I own Total Stock Market and Total Int'l Stock Market then I already own every single equity in Merriman's more complicated portfolio, just in somewhat different weights.

2. The market has already valued every single equity at their current values. Any other tilts such as those proposed by Merriman are second-guessing for no apparent rational reason. There may be a reason to tilt towards small cap value which is the smallest of the four boxes, but I don't understand what it is.

3. Dividing a portfolio equally between four unequal sectors makes no rational sense. The Morningstar 4-corner box thing is simply a visual construct. The actual market doesn't look like that. It would be like dividing up a global portfolio into 7 equal funds, one for each continent simply because there are 7 continents.
You're not wrong (in so far as having a "good enough" plan). You may outperform or underperform Merriman (or any tilt).

What is more important than all of this is sticking to whatever plan you have and not changing because of market conditions/what you read here. Or selling at the next bottom.
rascott
Posts: 2957
Joined: Wed Apr 15, 2015 10:53 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by rascott »

texasdiver wrote: Wed Oct 23, 2019 11:25 am I'm just a layman but I'm unconvinced.

1. If I own Total Stock Market and Total Int'l Stock Market then I already own every single equity in Merriman's more complicated portfolio, just in somewhat different weights.

2. The market has already valued every single equity at their current values. Any other tilts such as those proposed by Merriman are second-guessing for no apparent rational reason. There may be a reason to tilt towards small cap value which is the smallest of the four boxes, but I don't understand what it is.

3. Dividing a portfolio equally between four unequal sectors makes no rational sense. The Morningstar 4-corner box thing is simply a visual construct. The actual market doesn't look like that. It would be like dividing up a global portfolio into 7 equal funds, one for each continent simply because there are 7 continents.

Does it make rational sense to put 20c of every dollar invested into only 10 stocks?

For some it does, for others it doesn't.

I'd actually prefer to own an equal weighting of every publicly traded company available, it's just that it's not functionally feasible. If markets are efficient, and every equity is fairly priced.... why do I want 20% of my money in the largest 10 companies, and virtually zero in the bottom 2500 or so?
Topic Author
garlandwhizzer
Posts: 3565
Joined: Fri Aug 06, 2010 3:42 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by garlandwhizzer »

Paul Merriman wrote:

I won't be shocked if that's the outcome as no one knows if the future will be different than the past. At this point I am trusting the past but many people, much smarter than I, believe the future will be different. I don't envy the young but I do envy the fact that 50 years from now they will know the outcome. We have to keep in mind that no matter how loud we yell our beliefs and no matter how many facts we have to support our beliefs, it doesn't mean our beliefs are right. Good luck to you all. I wish you could all be right.
1+

Mr. Merriman is IMO brilliant, extremely knowledgeable, scrupulously honest, and has been investing long enough to approach the market's future admitting to some degree of uncertainty. That, to me at least, that demonstrates a rare quality these days, investing wisdom. After you've been listening to expert predictions of the market's future for decades it dawns on you that predictions may not always be reliable oracles of the future no matter who makes them. Humility is IMO an appropriate attitude when considering the market's future. Like Mr. Merriman, I believe that in the end value SCV will rise again but I do no have sufficient faith in that opinion to stake all my equity portfolio on it. It seems entirely rational to me add some SCV to TSM which will certainly increase diversification and may (or may not) increase risk adjusted long term return. In addition to the 4 fund portfolio that the Forum is arguing over presently, Mr. Merriman suggested in his article using that combination of TSM which leans to LCG and counterbalancing that with at least 10% SCV which seems to me a very attractive option. I do that with a 25% SCV/75% TSM in my US equity portfolio. Holding both in those weights suits me, but others may see it entirely differently, 100% one way or the other, and turn out in the end to be right. I prefer to hedge my bets a bit just like Mr. Merriman.

Garland Whizzer
robertmcd
Posts: 554
Joined: Tue Aug 09, 2016 9:06 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by robertmcd »

rascott wrote: Wed Oct 23, 2019 12:16 pm
texasdiver wrote: Wed Oct 23, 2019 11:25 am I'm just a layman but I'm unconvinced.

1. If I own Total Stock Market and Total Int'l Stock Market then I already own every single equity in Merriman's more complicated portfolio, just in somewhat different weights.

2. The market has already valued every single equity at their current values. Any other tilts such as those proposed by Merriman are second-guessing for no apparent rational reason. There may be a reason to tilt towards small cap value which is the smallest of the four boxes, but I don't understand what it is.

3. Dividing a portfolio equally between four unequal sectors makes no rational sense. The Morningstar 4-corner box thing is simply a visual construct. The actual market doesn't look like that. It would be like dividing up a global portfolio into 7 equal funds, one for each continent simply because there are 7 continents.

Does it make rational sense to put 20c of every dollar invested into only 10 stocks?

For some it does, for others it doesn't.

I'd actually prefer to own an equal weighting of every publicly traded company available, it's just that it's not functionally feasible. If markets are efficient, and every equity is fairly priced.... why do I want 20% of my money in the largest 10 companies, and virtually zero in the bottom 2500 or so?
I look at it in terms of me owning an equal percentage of every company in the US (not dollar amount of every company). Also, I really don't want to deviate from the total market/S&P 500 because I know that is what monetary policy is based off. My basis for risk parity in US stocks/US treasury bonds comes down to what happened in Fall of 2018 and ever since. It stocks don't go up, cash gets devalued to push treasury bond prices up to make money flow into stocks again.
gtt561
Posts: 216
Joined: Tue Jan 14, 2014 10:14 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by gtt561 »

rascott wrote: Wed Oct 23, 2019 12:16 pm
texasdiver wrote: Wed Oct 23, 2019 11:25 am I'm just a layman but I'm unconvinced.

1. If I own Total Stock Market and Total Int'l Stock Market then I already own every single equity in Merriman's more complicated portfolio, just in somewhat different weights.

2. The market has already valued every single equity at their current values. Any other tilts such as those proposed by Merriman are second-guessing for no apparent rational reason. There may be a reason to tilt towards small cap value which is the smallest of the four boxes, but I don't understand what it is.

3. Dividing a portfolio equally between four unequal sectors makes no rational sense. The Morningstar 4-corner box thing is simply a visual construct. The actual market doesn't look like that. It would be like dividing up a global portfolio into 7 equal funds, one for each continent simply because there are 7 continents.

Does it make rational sense to put 20c of every dollar invested into only 10 stocks?

For some it does, for others it doesn't.

I'd actually prefer to own an equal weighting of every publicly traded company available, it's just that it's not functionally feasible. If markets are efficient, and every equity is fairly priced.... why do I want 20% of my money in the largest 10 companies, and virtually zero in the bottom 2500 or so?
That is the question I have been trying to get an answer on since starting this thread. A lot of discussion on to tilt or not to tilt, however, as someone who is trying to learn as much as I can about investing, I am curious in just that. How does a cap-weighted index like TSM where the majority of your money is invested in the largest 10 companies and very little in the remaining companies, benefit or not benefit the end investor? Seems like it would be more beneficial to have your money equally spread throughout all the different stock asset classes (i.e. SCV, SCG, LCG, LCV or even the blends) as the article suggests?
"Anything less than a 10-year horizon is just speculation." John Bogle
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by willthrill81 »

gtt561 wrote: Wed Oct 23, 2019 1:58 pm
rascott wrote: Wed Oct 23, 2019 12:16 pm
texasdiver wrote: Wed Oct 23, 2019 11:25 am I'm just a layman but I'm unconvinced.

1. If I own Total Stock Market and Total Int'l Stock Market then I already own every single equity in Merriman's more complicated portfolio, just in somewhat different weights.

2. The market has already valued every single equity at their current values. Any other tilts such as those proposed by Merriman are second-guessing for no apparent rational reason. There may be a reason to tilt towards small cap value which is the smallest of the four boxes, but I don't understand what it is.

3. Dividing a portfolio equally between four unequal sectors makes no rational sense. The Morningstar 4-corner box thing is simply a visual construct. The actual market doesn't look like that. It would be like dividing up a global portfolio into 7 equal funds, one for each continent simply because there are 7 continents.

Does it make rational sense to put 20c of every dollar invested into only 10 stocks?

For some it does, for others it doesn't.

I'd actually prefer to own an equal weighting of every publicly traded company available, it's just that it's not functionally feasible. If markets are efficient, and every equity is fairly priced.... why do I want 20% of my money in the largest 10 companies, and virtually zero in the bottom 2500 or so?
That is the question I have been trying to get an answer on since starting this thread. A lot of discussion on to tilt or not to tilt, however, as someone who is trying to learn as much as I can about investing, I am curious in just that. How does a cap-weighted index like TSM where the majority of your money is invested in the largest 10 companies and very little in the remaining companies, benefit or not benefit the end investor? Seems like it would be more beneficial to have your money equally spread throughout all the different stock asset classes (i.e. SCV, SCG, LCG, LCV or even the blends) as the article suggests?
Historically, a portfolio with one-third each in LC, MC, and SC has had only very slightly higher volatility than TSM but has had significantly higher returns, about 1% annualized. But again, it makes no sense to own both growth and value since doing so is, for all practical purposes, equivalent to owning a blend.
The Sensible Steward
Elysium
Posts: 4120
Joined: Mon Apr 02, 2007 6:22 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Elysium »

Uncorrelated wrote: Wed Oct 23, 2019 2:00 am
Elysium wrote: Tue Oct 22, 2019 6:27 pm No, that is what I said in the beginning. There has been no value premium since 2004, or very little of it. This could be because everyone found out about it and flooded the market with products that attempts to capture it. We might need to go through a really long and painful cycle that may last 30 years perhaps before value premium makes a come back. Is this a bet worth taking is up to everyone concerned with their money to think hard about.
That is just cherry picking. According to FF, the monthly standard deviation of the value premium is around 10 times as large as the value premium. This indicates that:
The chance that the value premium is negative over the next 8 years is only 1 standard deviation away from zero.
The chance that the value premium is negative over the next 30 years is only 2 standard deviations away from zero.


What about the market premium? On 1963~1991 the standard deviation of the market premium has been 10.6 times as large as the market premium. Since 1926, 10.25 times as large. Simply put, the chance that the market premium is positive over a 10 year period is identical to the chance that the value premium delivers a positive return. In the 14 years following 2000, the market premium was negative delivered negative real return. We might go through a really long and painful cycle before the market premium makes a comeback. Is that a bet worth taking?


There are good reasons to be skeptical of the value premium, but a 15-year period of negative performance is not among them.

Edit: got some stats from ken french data library. MKT was negative for 186 months (15.5 years) following 1929, for 173 months (14.5 years) following 1968. Three months later the market premium went negative again for a total of 199 months (starting at 1968). Counting from the peak of the 2000' bubble, the market premium was negative for 154 months (12.8 years).

The value premium was negative for 164 months (13.6 years) following 1929 and for 153 months (12.7 years, and counting) since 2006. There have been no other periods longer than 100 months where value underperformed (or I can't find them).

The small cap premium went negative for 328 months (27 years) following 1938, which is to be expected since the small cap premium is much smaller than value. (premium 15 and 20 times smaller than the standard deviation)
Its quite simple. You just described what was already established and accepted by respected academics on peer reviewed work. Who am I to find fault with the work of fama-French? What you are missing is the obvious. FF work is well regarded for its explanatory power of factors affecting returns post-fact for the 100 years or so when no real mutual funds existed that attempted to capture this premia. What I am saying, and many others who are smarter than i am are saying this too, is that perhaps the premium do not exist anymore, because once an anomaly has been discovered it will be quickly exploited by the market.

What is not well accepted is the power of those models in predicting future returns. Future returns may or may not have been affected by the discovery of the SV anomaly which doesn't have a lot of rationale behind it. FF suggested perhaps there is a cost of capital rationale behind it, and it wasn't their job to defend that, so they left it at that.

No amount of past data or number crunching will satisfy the ability to predict future premiums. There is a false equivalency attempted between ERP and value premia, but that is not real, only a feeble attempt in defending the so called predictive power of past data.

We can go over and over about this and the outcome remains same. There is no amount of math and data in the world to prove what we dont know, except as Mr.Merriman said above, he believes the past will continue but many others disagree. We have to just leave it at that.
User avatar
Taylor Larimore
Posts: 32842
Joined: Tue Feb 27, 2007 7:09 pm
Location: Miami FL

Adding Small Cap Value Stocks?

Post by Taylor Larimore »

It seems entirely rational to me add some SCV to TSM which will certainly increase diversification.
Bogleheads:

In my opinion, a total market index fund, which holds nearly all U.S. stocks, is about as diversified as we can get in U.S. stocks. Overweighting one corner of the market does not add diversification. If anything, it decreases diversification. Overweighting the market portfolio with SCV (and increasing costs and complexity), advocated by the financial industry, has significantly reduced total market returns during the past 10 years.

Three Proofs That TSM Is Efficient

Best wishes
Taylor
Jack Bogle's Words of Wisdom: "SCV underperformed from 1926 to 1942 (14 years), did nothing from 1944 to 1963 (another 19 years) and again from 1968 to 1976 (8 years) and yet again from 1979 to 1999 (another 20 years). That's a mere 61 years of the 79-year period where SCV was not a winning strategy."
"Simplicity is the master key to financial success." -- Jack Bogle
donaldfair71
Posts: 1241
Joined: Wed Mar 06, 2013 3:15 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by donaldfair71 »

gtt561 wrote: Wed Oct 23, 2019 1:58 pm
rascott wrote: Wed Oct 23, 2019 12:16 pm
texasdiver wrote: Wed Oct 23, 2019 11:25 am I'm just a layman but I'm unconvinced.

1. If I own Total Stock Market and Total Int'l Stock Market then I already own every single equity in Merriman's more complicated portfolio, just in somewhat different weights.

2. The market has already valued every single equity at their current values. Any other tilts such as those proposed by Merriman are second-guessing for no apparent rational reason. There may be a reason to tilt towards small cap value which is the smallest of the four boxes, but I don't understand what it is.

3. Dividing a portfolio equally between four unequal sectors makes no rational sense. The Morningstar 4-corner box thing is simply a visual construct. The actual market doesn't look like that. It would be like dividing up a global portfolio into 7 equal funds, one for each continent simply because there are 7 continents.

Does it make rational sense to put 20c of every dollar invested into only 10 stocks?

For some it does, for others it doesn't.

I'd actually prefer to own an equal weighting of every publicly traded company available, it's just that it's not functionally feasible. If markets are efficient, and every equity is fairly priced.... why do I want 20% of my money in the largest 10 companies, and virtually zero in the bottom 2500 or so?
That is the question I have been trying to get an answer on since starting this thread. A lot of discussion on to tilt or not to tilt, however, as someone who is trying to learn as much as I can about investing, I am curious in just that. How does a cap-weighted index like TSM where the majority of your money is invested in the largest 10 companies and very little in the remaining companies, benefit or not benefit the end investor? Seems like it would be more beneficial to have your money equally spread throughout all the different stock asset classes (i.e. SCV, SCG, LCG, LCV or even the blends) as the article suggests?
Nothing wrong with that. Just stick with it, and that means sticking with it when the simple broad indexes (S&P500 for example) are performing better for long periods of time (and they will). Tilted portfolios will also perform better for long periods. It ebbs. It flows. I just knew, and I tried both strategies, but I just know that when using TSM/TISM performed worse, I could justify using it (it was cheaper and was the entire world). I couldn't justify using the Merriman approach when it didn't perform as well as the alternative. But others can.
User avatar
Taylor Larimore
Posts: 32842
Joined: Tue Feb 27, 2007 7:09 pm
Location: Miami FL

Investing in the largest 10 companies?

Post by Taylor Larimore »

Why do I want 20% of my money in the largest 10 companies?
rascott:

It makes me happy to know that at least 20% of my money is invested in the largest and most successful companies in the United States. :happy

Best wishes
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
Jags4186
Posts: 8198
Joined: Wed Jun 18, 2014 7:12 pm

Re: Adding Small Cap Value Stocks?

Post by Jags4186 »

Taylor Larimore wrote: Wed Oct 23, 2019 2:29 pm
It seems entirely rational to me add some SCV to TSM which will certainly increase diversification.
Bogleheads:

In my opinion, a total market index fund, which holds nearly all U.S. stocks, is about as diversified as we can get in U.S. stocks. Overweighting one corner of the market does not add diversification. If anything, it decreases diversification. Overweighting the market portfolio with SCV (and increasing costs and complexity), advocated by the financial industry, has significantly reduced total market returns during the past 10 years.

Three Proofs That TSM Is Efficient

Best wishes
Taylor
Jack Bogle's Words of Wisdom: "SCV underperformed from 1926 to 1942 (14 years), did nothing from 1944 to 1963 (another 19 years) and again from 1968 to 1976 (8 years) and yet again from 1979 to 1999 (another 20 years). That's a mere 61 years of the 79-year period where SCV was not a winning strategy."
It appears the information Bogle had was wrong:

Image
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Uncorrelated »

Elysium wrote: Wed Oct 23, 2019 2:09 pm
Uncorrelated wrote: Wed Oct 23, 2019 2:00 am
Elysium wrote: Tue Oct 22, 2019 6:27 pm No, that is what I said in the beginning. There has been no value premium since 2004, or very little of it. This could be because everyone found out about it and flooded the market with products that attempts to capture it. We might need to go through a really long and painful cycle that may last 30 years perhaps before value premium makes a come back. Is this a bet worth taking is up to everyone concerned with their money to think hard about.
That is just cherry picking. According to FF, the monthly standard deviation of the value premium is around 10 times as large as the value premium. This indicates that:
The chance that the value premium is negative over the next 8 years is only 1 standard deviation away from zero.
The chance that the value premium is negative over the next 30 years is only 2 standard deviations away from zero.


What about the market premium? On 1963~1991 the standard deviation of the market premium has been 10.6 times as large as the market premium. Since 1926, 10.25 times as large. Simply put, the chance that the market premium is positive over a 10 year period is identical to the chance that the value premium delivers a positive return. In the 14 years following 2000, the market premium was negative delivered negative real return. We might go through a really long and painful cycle before the market premium makes a comeback. Is that a bet worth taking?


There are good reasons to be skeptical of the value premium, but a 15-year period of negative performance is not among them.

Edit: got some stats from ken french data library. MKT was negative for 186 months (15.5 years) following 1929, for 173 months (14.5 years) following 1968. Three months later the market premium went negative again for a total of 199 months (starting at 1968). Counting from the peak of the 2000' bubble, the market premium was negative for 154 months (12.8 years).

The value premium was negative for 164 months (13.6 years) following 1929 and for 153 months (12.7 years, and counting) since 2006. There have been no other periods longer than 100 months where value underperformed (or I can't find them).

The small cap premium went negative for 328 months (27 years) following 1938, which is to be expected since the small cap premium is much smaller than value. (premium 15 and 20 times smaller than the standard deviation)
Its quite simple. You just described what was already established and accepted by respected academics on peer reviewed work. Who am I to find fault with the work of fama-French? What you are missing is the obvious. FF work is well regarded for its explanatory power of factors affecting returns post-fact for the 100 years or so when no real mutual funds existed that attempted to capture this premia. What I am saying, and many others who are smarter than i am are saying this too, is that perhaps the premium do not exist anymore, because once an anomaly has been discovered it will be quickly exploited by the market.
If that is the argument you want to make, then fine. It is a good argument. But in this context, your other posts on this topic appear surprisingly weak and I'm not surprised that so man people disagreed with your viewpoint.
No amount of past data or number crunching will satisfy the ability to predict future premiums. There is a false equivalency attempted between ERP and value premia, but that is not real, only a feeble attempt in defending the so called predictive power of past data.
I don't see the difference between ERP and factors. Do you happen to have some literature on hand that I can read?
Northern Flicker
Posts: 15363
Joined: Fri Apr 10, 2015 12:29 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Northern Flicker »

Paul Merriman wrote: I generally write/talk about combining large and small cap blend (which, depending how you differentiate growth from value, is mostly growth) with large and small cap value . In an attempt to simplify the article (always trying to limit word count at Marketwatch) we used the word growth rather than blend.
Claiming that the S&P500 or the total market index have a growth tilt is just not correct. By definition, the total market index is an untilted portfolio. The beta coefficient of market return in any factor model is one, and the coefficients of all other factors are zero. In particular, there is zero contribution of the value factor (positively or negatively) to the total market index.

This is also true of the S&P500. A factor regression of VFINX (Vanguard S&P500 fund) from inception in Sept 1976 to present shows a value factor loading of .02, which is a statistically insignificant deviation from zero, likely representing tracking variance of the fund, but in any case certainly is not a growth tilt:

https://www.portfoliovisualizer.com/fac ... e&total1=0

It is even less accurate to describe a small-cap blended index fund as a growth fund. Small-cap index funds have a bit of a value tilt generally. Looking at IJR the small-cap 600 index fund from inception in June 2000, we see a value loading of 0.34, which is not even neutral but reflective of some value loading. This certainly is not a proxy for small growth:

https://www.portfoliovisualizer.com/fac ... e&total1=0

At this point you may ask that if the total market and large caps are neutral with respect to value and growth and small caps have a value tilt, where is the growth tilt that would be needed to make the total market neutral once the value tilt in a small cap blend index is factored in? The answer is that the stocks filtered out of index funds by investability screens have a very strong growth tilt. In a total market index fund, this effect is quite diluted by virtue of the small market cap of the stocks filtered out. But there is enough market cap among the stocks filtered out relative to a small-cap index fund market cap to tilt the small-cap index fund to value.
Last edited by Northern Flicker on Thu Oct 24, 2019 1:36 pm, edited 4 times in total.
Elysium
Posts: 4120
Joined: Mon Apr 02, 2007 6:22 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Elysium »

Uncorrelated wrote: Wed Oct 23, 2019 3:32 pm
Elysium wrote: Wed Oct 23, 2019 2:09 pm
Uncorrelated wrote: Wed Oct 23, 2019 2:00 am
Elysium wrote: Tue Oct 22, 2019 6:27 pm No, that is what I said in the beginning. There has been no value premium since 2004, or very little of it. This could be because everyone found out about it and flooded the market with products that attempts to capture it. We might need to go through a really long and painful cycle that may last 30 years perhaps before value premium makes a come back. Is this a bet worth taking is up to everyone concerned with their money to think hard about.
That is just cherry picking. According to FF, the monthly standard deviation of the value premium is around 10 times as large as the value premium. This indicates that:
The chance that the value premium is negative over the next 8 years is only 1 standard deviation away from zero.
The chance that the value premium is negative over the next 30 years is only 2 standard deviations away from zero.


What about the market premium? On 1963~1991 the standard deviation of the market premium has been 10.6 times as large as the market premium. Since 1926, 10.25 times as large. Simply put, the chance that the market premium is positive over a 10 year period is identical to the chance that the value premium delivers a positive return. In the 14 years following 2000, the market premium was negative delivered negative real return. We might go through a really long and painful cycle before the market premium makes a comeback. Is that a bet worth taking?


There are good reasons to be skeptical of the value premium, but a 15-year period of negative performance is not among them.

Edit: got some stats from ken french data library. MKT was negative for 186 months (15.5 years) following 1929, for 173 months (14.5 years) following 1968. Three months later the market premium went negative again for a total of 199 months (starting at 1968). Counting from the peak of the 2000' bubble, the market premium was negative for 154 months (12.8 years).

The value premium was negative for 164 months (13.6 years) following 1929 and for 153 months (12.7 years, and counting) since 2006. There have been no other periods longer than 100 months where value underperformed (or I can't find them).

The small cap premium went negative for 328 months (27 years) following 1938, which is to be expected since the small cap premium is much smaller than value. (premium 15 and 20 times smaller than the standard deviation)
Its quite simple. You just described what was already established and accepted by respected academics on peer reviewed work. Who am I to find fault with the work of fama-French? What you are missing is the obvious. FF work is well regarded for its explanatory power of factors affecting returns post-fact for the 100 years or so when no real mutual funds existed that attempted to capture this premia. What I am saying, and many others who are smarter than i am are saying this too, is that perhaps the premium do not exist anymore, because once an anomaly has been discovered it will be quickly exploited by the market.
If that is the argument you want to make, then fine. It is a good argument. But in this context, your other posts on this topic appear surprisingly weak and I'm not surprised that so man people disagreed with your viewpoint.
I have not tried to refute the existence of value premium in past data, or try to reject FF work, if earlier responses did not mention this, that is only because I was responding to specific claims of Value outperformance by other posters. When I say I am skeptical, I am in the camp of people like Bill Bernstein who has expressed doubts whether value premium has been diluted away, although he hasn't been going around saying that loud I did read it somewhere. My apologies to Dr.BB if this is not correct. Even Mr.Merriman points to the possibility of others disagreeing although he is sticking to the belief it will persist.
My second and third problems are with the deluge of factors lately. This makes the case even weaker imo and hard to implement together. Eugene Fama himself has rejected momentum for instance by itself. Then there are issues with having quality, low vol, value, all together. Some cancel others out. Low vol takes sector bets, so on.
All of this points to just one factor value that seems somewhat reliable, but you ask the question why at this point, because at the outset all of them appears to have merits. But when taken together falls apart imo. At least I have not seen a proper framework for inclusion of all in a portfolio.
That brings me back to the power of simplicity of Jack Bogle and his advice that the enemy of a good plan is pursuit of the perfect plan.
Uncorrelated wrote: Wed Oct 23, 2019 3:32 pm I don't see the difference between ERP and factors. Do you happen to have some literature on hand that I can read?
Actually quite the opposite is true I think. I may not be able to find any literature showing how to calculate value premium going forward, but of course easy to find material on how to calculate ERP. You can calculate the ERP with CAPM by estimating future market returns minus risk free rate. How do you estimate future value premium? CAPM supports beta only for market risk, there is no HmL risk estimate in my understanding as it is only helpful to explain the returns by looking back. If there is a way Io do this systemically without individual security evaluation then would like to know. Therefore I think equity premium can be defined for future estimates but not value or size premiums. That said, there is actually no difference between finding equity premium of any set of securities, they will be whatever that comes out of based on current valuations and future forecasts minus the risk free rate. The larger point is, there is no discernible way to calculate a forward looking value premium, it can only be used as an explanatory factor.
Last edited by Elysium on Wed Oct 23, 2019 8:25 pm, edited 2 times in total.
User avatar
One Ping
Posts: 1087
Joined: Thu Sep 24, 2015 4:53 pm

Re: Adding Small Cap Value Stocks?

Post by One Ping »

Taylor Larimore wrote: Wed Oct 23, 2019 2:29 pm Taylor
Jack Bogle's Words of Wisdom: "SCV underperformed from 1926 to 1942 (14 years), did nothing from 1944 to 1963 (another 19 years) and again from 1968 to 1976 (8 years) and yet again from 1979 to 1999 (another 20 years). That's a mere 61 years of the 79-year period where SCV was not a winning strategy."
I've seen the above presented/mentioned here many times and I have to agree with Jags4186 ...
Jags4186 wrote: Wed Oct 23, 2019 2:59 pm It appears the information Bogle had was wrong:
[Note: Not only does the quote not count the length of the intervals (number of years) correctly, the returns used appear to be just wrong. (I suspect that the reason the quote says 14 years for 1926 through 1942, when that interval is actually 17 years long, is that the comparison actually starts with 1928 data, making it 14 years they way the quote counts years, but actually being 15 years long.)]

He's obviously not using actual fund data, since neither an S&P 500 Index fund or a SCV fund existed in 1928, so I used return data for 1928-2018 from the Dimensional data base for the "Market" [S&P 500 Index and CRSP (1-10 deciles) Index] and the Dimensional US SCV Index. I calculated the returns over the intervals Bogle quoted (and didn't quote). No cherry picking, I used his intervals.

Code: Select all

					Compound Annual Growth Rate (CAGR)							
Interval	Duration	S&P 500 Idx	CRSP (1-10) Idx	Dim SCV Idx		SCV Performance vs "Market'						
1928-1942	15 years	1.5%		0.8%		-1.9%			Underperformed	-2.6% to -3.4%		
1943		1 year		25.9%		28.4%		78.5%			Outperformed	50.1% to 52.6%
1944-1963	20 years	15.1%		14.5%		16.9%			Outperformed	1.7% to 2.4%		
1964-1976	13 years	6.4%		6.5%		13.6%			Outperformed	7.1% to 7.2%		
1977-1978	2 years		-0.5%		1.4%		22.7%			Outperformed	26.6% to 29.6%	
1979-1999	21 years	17.9%		17.6%		19.0%			Outperformed	1.1% to 1.4%		
2000-2018	19 years	4.9%		5.2%		11.4%			Outperformed	6.3% to 6.6%		
and for the entire 91 year interval:

Code: Select all

					Compound Annual Growth Rate (CAGR)					
Interval	Duration	S&P 500 Idx	CRSP (1-10) Idx	Dim SCV Idx		SCV Performance vs "Market'				
1928-2018	91 years	9.7%		9.5%		13.1%			Outperformed	3.4% to 3.6%
Below are the year by year comparisons for those who might care to see the detailed data I used:

Code: Select all

Year	S&P500 		CRSP(Mkt)	Dim US SCV 	
(J-D)	Index		Index		Index
1928	43.61%		38.39%		32.36%
1929	-8.41%		-15.19%		-37.13%
1930	-24.90%		-28.80%		-43.61%
1931	-43.35%		-43.52%		-55.45%
1932	-8.20%		-8.62%		-10.52%
1933	53.97%		56.65%		125.32%
1934	-1.43%		4.09%		-6.28%
1935	47.66%		44.43%		47.72%
1936	33.92%		32.32%		66.50%
1937	-35.02%		-34.74%		-50.56%
1938	31.14%		28.16%		32.55%
1939	-0.42%		2.83%		-3.91%
1940	-9.78%		-7.10%		-8.11%
1941	-11.58%		-10.08%		-0.20%
1942	20.33%		16.12%		34.05%
1943	25.91%		28.43%		78.55%
1944	19.73%		21.49%		52.56%
1945	36.41%		38.50%		65.42%
1946	-8.07%		-6.17%		-10.43%
1947	5.70%		3.59%		8.81%
1948	5.51%		2.11%		-4.94%
1949	18.79%		20.22%		19.71%
1950	31.74%		29.61%		63.41%
1951	24.02%		20.68%		9.93%
1952	18.35%		13.42%		8.95%
1953	-0.98%		0.67%		-10.60%
1954	52.62%		49.98%		64.40%
1955	31.54%		25.21%		23.78%
1956	6.56%		8.27%		1.66%
1957	-10.79%		-10.05%		-18.58%
1958	43.37%		45.02%		77.04%
1959	11.98%		12.67%		14.99%
1960	0.46%		1.16%		-10.74%
1961	26.89%		26.95%		29.21%
1962	-8.73%		-10.18%		-10.26%
1963	22.78%		20.98%		29.56%
1964	16.51%		16.13%		25.48%
1965	12.45%		14.46%		40.08%
1966	-10.05%		-8.74%		-9.64%
1967	23.99%		28.74%		69.32%
1968	11.08%		14.14%		48.98%
1969	-8.49%		-10.91%		-28.86%
1970	4.03%		0.00%		-1.25%
1971	14.32%		16.15%		15.13%
1972	18.98%		16.84%		7.77%
1973	-14.67%		-18.06%		-30.25%
1974	-26.46%		-27.04%		-17.84%
1975	37.21%		38.75%		65.81%
1976	23.85%		26.76%		58.55%
1977	-7.18%		-4.26%		22.39%
1978	6.57%		7.49%		22.96%
1979	18.42%		22.62%		35.39%
1980	32.41%		32.81%		24.44%
1981	-4.91%		-3.65%		20.34%
1982	21.41%		21.00%		36.94%
1983	22.51%		21.98%		48.90%
1984	6.27%		4.51%		1.74%
1985	32.17%		32.17%		29.18%
1986	18.47%		16.19%		8.57%
1987	5.23%		1.67%		-5.89%
1988	16.81%		18.03%		33.98%
1989	31.49%		28.86%		13.67%
1990	-3.10%		-5.96%		-24.20%
1991	30.47%		34.67%		47.08%
1992	7.63%		9.80%		34.80%
1993	10.07%		11.14%		25.96%
1994	1.32%		-0.06%		2.57%
1995	37.58%		36.79%		31.49%
1996	22.96%		21.35%		25.88%
1997	33.36%		31.38%		38.62%
1998	28.58%		24.30%		-5.51%
1999	21.04%		25.22%		8.15%
2000	-9.10%		-11.42%		19.99%
2001	-11.89%		-11.15%		28.59%
2002	-22.10%		-21.15%		-8.47%
2003	28.69%		31.62%		66.53%
2004	10.88%		11.97%		23.93%
2005	4.91%		6.16%		7.46%
2006	15.80%		15.48%		21.65%
2007	5.49%		5.81%		-11.81%
2008	-37.00%		-36.71%		-36.90%
2009	26.46%		28.82%		51.30%
2010	15.06%		17.73%		30.93%
2011	2.11%		0.77%		-6.56%
2012	16.00%		16.16%		16.65%
2013	32.39%		35.17%		41.95%
2014	13.69%		11.65%		3.46%
2015	1.38%		-0.45%		-8.43%
2016	11.96%		13.58%		37.20%
2017	21.83%		21.05%		7.38%
2018	-4.38%		-5.02%		-13.58%
ETA: Year by year return performance summary
For the 91 year period ...
SVC Index outperformed S&P 500 Index ....... 53 years to 38 years
SCV Index outperformed CRSP (1-10) Index ... 48 years to 43 years

As Paul says ... "There is no risk in the past." So, I'm not sure Bogle's quote about the past means any more or any less than anyone else's. It's just data. However, accurate data should be used when make such claims.
"Re-verify our range to target ... one ping only."
Elysium
Posts: 4120
Joined: Mon Apr 02, 2007 6:22 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Elysium »

paul merriman wrote: Wed Oct 23, 2019 8:07 am How long is the long term? In ny case the long term is however long I live. When people ask me how my portfolio has done, I suggest they ask my kids after I've died as that is the point at which my investment decisions should be judged. If Vanguard would have been the better decision it suggests that large produced better returns than small and growth produced better returns than value. I won't be shocked if that's the outcome as no one knows if the future will be different than the past. At this point I am trusting the past but many people, much smarter than I, believe the future will be different. I don't envy the young but I do envy the fact that 50 years from now they will know the outcome. We have to keep in mind that no matter how loud we yell our beliefs and no matter how many facts we have to support our beliefs, it doesn't mean our beliefs are right. Good luck to you all. I wish you could all be right.
This is an excellent response, as good as it can get coming from someone taking a principled position in supporting deep value. It shows a level of maturity, wisdom, experience, and an rare humility that leaves out room for other possibilities. If everyone displayed these qualities, this topic would still continue to be discussed at length, but it would be much more cordial for all involved, and perhaps we would eventually get to a consensus.
marcopolo
Posts: 8445
Joined: Sat Dec 03, 2016 9:22 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by marcopolo »

Elysium wrote: Wed Oct 23, 2019 8:35 pm
paul merriman wrote: Wed Oct 23, 2019 8:07 am How long is the long term? In ny case the long term is however long I live. When people ask me how my portfolio has done, I suggest they ask my kids after I've died as that is the point at which my investment decisions should be judged. If Vanguard would have been the better decision it suggests that large produced better returns than small and growth produced better returns than value. I won't be shocked if that's the outcome as no one knows if the future will be different than the past. At this point I am trusting the past but many people, much smarter than I, believe the future will be different. I don't envy the young but I do envy the fact that 50 years from now they will know the outcome. We have to keep in mind that no matter how loud we yell our beliefs and no matter how many facts we have to support our beliefs, it doesn't mean our beliefs are right. Good luck to you all. I wish you could all be right.
This is an excellent response, as good as it can get coming from someone taking a principled position in supporting deep value. It shows a level of maturity, wisdom, experience, and an rare humility that leaves out room for other possibilities. If everyone displayed these qualities, this topic would still continue to be discussed at length, but it would be much more cordial for all involved, and perhaps we would eventually get to a consensus.
I agree. Its nice to see the argument for factor tilting with a dose of uncertainty and humility about the future. It is a nice contrast to the often smug and condescending tone I often sense from people who are so sure of their investing superiority, they assume anyone not doing the same is either doesn't know about the obvious benefits of factor tilting (ignorant), or are too stupid to take advantage of them. There have already been a few such post on this thread.
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
One Ping
Posts: 1087
Joined: Thu Sep 24, 2015 4:53 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by One Ping »

marcopolo wrote: Wed Oct 23, 2019 9:53 pm
Elysium wrote: Wed Oct 23, 2019 8:35 pm
paul merriman wrote: Wed Oct 23, 2019 8:07 am How long is the long term? In ny case the long term is however long I live. When people ask me how my portfolio has done, I suggest they ask my kids after I've died as that is the point at which my investment decisions should be judged. If Vanguard would have been the better decision it suggests that large produced better returns than small and growth produced better returns than value. I won't be shocked if that's the outcome as no one knows if the future will be different than the past. At this point I am trusting the past but many people, much smarter than I, believe the future will be different. I don't envy the young but I do envy the fact that 50 years from now they will know the outcome. We have to keep in mind that no matter how loud we yell our beliefs and no matter how many facts we have to support our beliefs, it doesn't mean our beliefs are right. Good luck to you all. I wish you could all be right.
This is an excellent response, as good as it can get coming from someone taking a principled position in supporting deep value. It shows a level of maturity, wisdom, experience, and an rare humility that leaves out room for other possibilities. If everyone displayed these qualities, this topic would still continue to be discussed at length, but it would be much more cordial for all involved, and perhaps we would eventually get to a consensus.
I agree. Its nice to see the argument for factor tilting with a dose of uncertainty and humility about the future. It is a nice contrast to the often smug and condescending tone I often sense from people who are so sure of their investing superiority, they assume anyone not doing the same is either doesn't know about the obvious benefits of factor tilting (ignorant), or are too stupid to take advantage of them. There have already been a few such post on this thread.
As someone once said, "There are many roads to Dublin." We don't all have to take the same one. Be it factor tilting, 3- or 4-funds, 2 funds, target date, ... Whichever approach you can stick to is probably better than the one you can't/won't be able to stick to ...
"Re-verify our range to target ... one ping only."
Longtermgrowth
Posts: 731
Joined: Thu Nov 26, 2015 12:59 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Longtermgrowth »

Well said, One Ping. I have a lot of respect for that someone. I hope views can meet in the middle again, allowing everyone to take their own route.

After all, we're all trying to keep costs low, diversify, take no more and no less risk than necessary (stock/fixed income allocation), and stay the course.
User avatar
Forester
Posts: 2400
Joined: Sat Jan 19, 2019 1:50 pm
Location: UK

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Forester »

If an investor over-weights SCV, they still benefit from the "wisdom of the crowd" which is essentially why indexing works. Over time 50% of the market cap of the S&P 500 rests with the largest 50 companies; so even TSM resembles the DJIA. It's nonsense to think that 100% TSM is as diversified as say, 70% TSM 30% SCV.

If one believes there's no SCV premium AND points to recent under-performance, that's all the more reason to own SCV since it's a differentiated return stream.
Amateur Self-Taught Senior Macro Strategist
User avatar
One Ping
Posts: 1087
Joined: Thu Sep 24, 2015 4:53 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by One Ping »

Longtermgrowth wrote: Wed Oct 23, 2019 11:13 pm Well said, One Ping. I have a lot of respect for that someone. I hope views can meet in the middle again, allowing everyone to take their own route.

After all, we're all trying to keep costs low, diversify, take no more and no less risk than necessary (stock/fixed income allocation), and stay the course.
But, see ... the whole point of "there are many roads ..." is that our views shouldn't have to meet in the middle, or agree. That doesn't mean views different from our own are 'wrong' or that we shouldn't be civil and respectful when discussing them. :beer
"Re-verify our range to target ... one ping only."
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by YRT70 »

Elysium wrote: Tue Oct 22, 2019 12:38 pm
Leif wrote: Tue Oct 22, 2019 12:03 pm Having read the referenced article by Paul, why would anyone think that looking at 5 years of performance is more valuable than the last 80 years? Even more so, looking at segments along those 80 years.
Not just 5 years. SCV has trailed US TSM since 2004 for 15 years and counting by over 1% annualized.
In the 14 years from Jan 2004 to Dec 2017 TSM returned 8.94% while SCV returned 9.67% (PV data).
YRT70
Posts: 1289
Joined: Sat Apr 27, 2019 8:51 am

Re: Investing in the largest 10 companies?

Post by YRT70 »

Taylor Larimore wrote: Wed Oct 23, 2019 2:59 pm
Why do I want 20% of my money in the largest 10 companies?
rascott:

It makes me happy to know that at least 20% of my money is invested in the largest and most successful companies in the United States. :happy

Best wishes
Taylor
I think everybody would be happy with it when things are going as well as recently. The more interesting question I find is how happy would you have been with it in a period like 2000-2002, for example.
donaldfair71
Posts: 1241
Joined: Wed Mar 06, 2013 3:15 pm

Re: Investing in the largest 10 companies?

Post by donaldfair71 »

YRT70 wrote: Thu Oct 24, 2019 7:47 am
Taylor Larimore wrote: Wed Oct 23, 2019 2:59 pm
Why do I want 20% of my money in the largest 10 companies?
rascott:

It makes me happy to know that at least 20% of my money is invested in the largest and most successful companies in the United States. :happy

Best wishes
Taylor
I think everybody would be happy with it when things are going as well as recently. The more interesting question I find is how happy would you have been with it in a period like 2000-2002, for example.
One needs to use the strategy that allows one to stay the course when his/her strategy lags other strategies considered. This is really the primary consideration here between reasonable strategies.
Elysium
Posts: 4120
Joined: Mon Apr 02, 2007 6:22 pm

Re: Investing in the largest 10 companies?

Post by Elysium »

YRT70 wrote: Thu Oct 24, 2019 7:47 am
Taylor Larimore wrote: Wed Oct 23, 2019 2:59 pm
Why do I want 20% of my money in the largest 10 companies?
rascott:

It makes me happy to know that at least 20% of my money is invested in the largest and most successful companies in the United States. :happy

Best wishes
Taylor
I think everybody would be happy with it when things are going as well as recently. The more interesting question I find is how happy would you have been with it in a period like 2000-2002, for example.
You can see the results of all of those portfolios discussed on the forum since the beginning of 1999 listed by Madsinger monthly reports, here: https://www.bogleheads.org/wiki/Madsing ... ly_reports

The bottomline is saving and investing regularly and staying the course, you could get to your destination following any of these strategies listed by the sample portfolios, some like value tilted slice & dice portfolio comes with higher tracking error. The issue here is that, it appears many of the original slide & dicers have left or they are not very vocal anymore, either they have abandoned that strategy and instead perhaps chasing new factors, or just decided to overweight SCV by trimming things like REIT and LCV, where as the three fund portfolio remains the same.

I don't know this for sure, but we don't hear a lot from folks following the original 4 corners which is also what is recommended by Mr. Merriman in the article, instead we hear folks wanting to invest 50% in SCV and Intl SCV, which makes one wonder if this is a long term strategy or something that they will abandon after a few years in favor of some new found idea.
User avatar
Uncorrelated
Posts: 1237
Joined: Sun Oct 13, 2019 3:16 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Uncorrelated »

Elysium wrote: Wed Oct 23, 2019 4:17 pm
Uncorrelated wrote: Wed Oct 23, 2019 3:32 pm long quote
I have not tried to refute the existence of value premium in past data, or try to reject FF work, if earlier responses did not mention this, that is only because I was responding to specific claims of Value outperformance by other posters. When I say I am skeptical, I am in the camp of people like Bill Bernstein who has expressed doubts whether value premium has been diluted away, although he hasn't been going around saying that loud I did read it somewhere. My apologies to Dr.BB if this is not correct. Even Mr.Merriman points to the possibility of others disagreeing although he is sticking to the belief it will persist.
My second and third problems are with the deluge of factors lately. This makes the case even weaker imo and hard to implement together. Eugene Fama himself has rejected momentum for instance by itself. Then there are issues with having quality, low vol, value, all together. Some cancel others out. Low vol takes sector bets, so on.
All of this points to just one factor value that seems somewhat reliable, but you ask the question why at this point, because at the outset all of them appears to have merits. But when taken together falls apart imo. At least I have not seen a proper framework for inclusion of all in a portfolio.
That brings me back to the power of simplicity of Jack Bogle and his advice that the enemy of a good plan is pursuit of the perfect plan.
Fair.

The overflowing amount of factors being discovered is not a reason to reject good-old factors, just like the existence of poor passive indices (cloud computing ETF, cancer ETF) is not a reason to avoid total stock market.

I don't think that factors 'fall apart' when combined, just that it is difficult to create a long-only portfolio that has exposure to many different factors. This is not surprising, as each additional factor being discovered has dismissing explanatory effects.
Uncorrelated wrote: Wed Oct 23, 2019 3:32 pm I don't see the difference between ERP and factors. Do you happen to have some literature on hand that I can read?
Actually quite the opposite is true I think. I may not be able to find any literature showing how to calculate value premium going forward, but of course easy to find material on how to calculate ERP. You can calculate the ERP with CAPM by estimating future market returns minus risk free rate. How do you estimate future value premium? CAPM supports beta only for market risk, there is no HmL risk estimate in my understanding as it is only helpful to explain the returns by looking back. If there is a way Io do this systemically without individual security evaluation then would like to know. Therefore I think equity premium can be defined for future estimates but not value or size premiums. That said, there is actually no difference between finding equity premium of any set of securities, they will be whatever that comes out of based on current valuations and future forecasts minus the risk free rate. The larger point is, there is no discernible way to calculate a forward looking value premium, it can only be used as an explanatory factor.
It is easy to find literature to calculate the ERP, but it is also equally easy to find literature that such literature is flawed. For example:
Our article comprehensively reexamines the performance of variables that have been suggested by the academic literature to be good predictors of the equity premium. We find that by and large, these models have predicted poorly both in-sample (IS) and out-of-sample (OOS) for 30 years now; these models seem unstable, as diagnosed by their out-of-sample predictions and other statistics; and these models would not have helped an investor with access only to available information to profitably time the market.
From Welch and Goyal, A Comprehensive Look at The Empirical Performance of Equity Premium Prediction


If we assume that this literature is accurate, I don't see any reason to believe that predicting the value premium is more difficult than predicting the ERP. A simple historic average is likely to be a good guess.


CAPM is a different model than FF. Of course you can't explain factor returns with CAPM. CAPM is a tool to determine which bets to take, not a tool to predict future market premia.
Elysium
Posts: 4120
Joined: Mon Apr 02, 2007 6:22 pm

Re: [Paul Merriman: This strategy beats a total stock market fund]

Post by Elysium »

Uncorrelated wrote: Thu Oct 24, 2019 9:12 am I don't think that factors 'fall apart' when combined, just that it is difficult to create a long-only portfolio that has exposure to many different factors. This is not surprising, as each additional factor being discovered has dismissing explanatory effects.
This. Also, I believe the long-short strategies are proving to be equally difficult in practice, just looking at what AQR and some of the others are facing. I do think there is merit to the argument not to throw out the original factors, this is where FF and DFA are at, but I do wonder if this is because there are difficulties in including new factors, or a belief that old factors are still the best. My guess is that the direction is unclear, so when in doubt perhaps stay the course.
Uncorrelated wrote: Wed Oct 23, 2019 3:32 pm It is easy to find literature to calculate the ERP, but it is also equally easy to find literature that such literature is flawed. For example:
Our article comprehensively reexamines the performance of variables that have been suggested by the academic literature to be good predictors of the equity premium. We find that by and large, these models have predicted poorly both in-sample (IS) and out-of-sample (OOS) for 30 years now; these models seem unstable, as diagnosed by their out-of-sample predictions and other statistics; and these models would not have helped an investor with access only to available information to profitably time the market.
From Welch and Goyal, A Comprehensive Look at The Empirical Performance of Equity Premium Prediction


If we assume that this literature is accurate, I don't see any reason to believe that predicting the value premium is more difficult than predicting the ERP. A simple historic average is likely to be a good guess.


CAPM is a different model than FF. Of course you can't explain factor returns with CAPM. CAPM is a tool to determine which bets to take, not a tool to predict future market premia.
I have no disagreements with most of that, agree that estimating ERP is only a guess in direction and not likely to be perfect. The disagreement seems to be mostly about drawing a parallel between ERP and Value premium, that if one believes Equities have a premium compared to Fixed Income, then it also leads us to believe there is a Value premium compared to the broad market. I am not sure the literature on this is conclusive. I don't think even FF went in that direction, other than a suggestion that higher cost of capital and thus higher risk may lead to higher expected return. I think this is an idea taken over and promoted by the practitioners more than the academics themselves. This is why I am skeptical of that argument. That shouldn't prevent anyone from investing in SCV, because the arguments in favor of it is as good as it gets, but when people talk about investing 50% of their equities in a segment of market that represents only 2%, it gets very sketchy.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Investing in the largest 10 companies?

Post by willthrill81 »

YRT70 wrote: Thu Oct 24, 2019 7:47 am
Taylor Larimore wrote: Wed Oct 23, 2019 2:59 pm
Why do I want 20% of my money in the largest 10 companies?
rascott:

It makes me happy to know that at least 20% of my money is invested in the largest and most successful companies in the United States. :happy

Best wishes
Taylor
I think everybody would be happy with it when things are going as well as recently. The more interesting question I find is how happy would you have been with it in a period like 2000-2002, for example.
TSM investors were not happy with negative real returns from 2000-2009. Investors who diversified across factors though did well during the 'lost decade' though.

I find it interesting to hear some opine about TSM being 'maximally diversified' on one hand while they also seem to relish the idea of 20% of their U.S. stock being in ten companies. It seems to be a very clear example of very strong confirmation bias.
The Sensible Steward
User avatar
One Ping
Posts: 1087
Joined: Thu Sep 24, 2015 4:53 pm

Re: Investing in the largest 10 companies?

Post by One Ping »

willthrill81 wrote: Thu Oct 24, 2019 9:49 am I find it interesting to hear some opine about TSM being 'maximally diversified' on one hand while they also seem to relish the idea of 20% of their U.S. stock being in ten companies. It seems to be a very clear example of very strong confirmation bias.
With maybe a touch of cognitive dissonance thrown in for good measure ...
"Re-verify our range to target ... one ping only."
jdilla1107
Posts: 849
Joined: Sun Jun 24, 2012 8:31 pm

Re: Investing in the largest 10 companies?

Post by jdilla1107 »

One Ping wrote: Thu Oct 24, 2019 12:41 pm
willthrill81 wrote: Thu Oct 24, 2019 9:49 am I find it interesting to hear some opine about TSM being 'maximally diversified' on one hand while they also seem to relish the idea of 20% of their U.S. stock being in ten companies. It seems to be a very clear example of very strong confirmation bias.
With maybe a touch of cognitive dissonance thrown in for good measure ...
Would you feel better if each of these 10 companies decided to break into 10-20 companies each? Would that remove the risk you are concerned with?

Do you realize that big companies actually operate this way? Microsoft has 100s of business lines. You could say it's almost like 100s of small caps operating under 1 umbrella, because that's what it is.

Diversification should not be measured entirely by "corporate legal structure". Imagining that 10 small cap software companies are going to be more diversified than Microsoft is not thinking about the problem correctly. Management of large cap companies are constantly thinking about their risk concentrations across their business lines and optimizing for that. The market insists on this.
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Investing in the largest 10 companies?

Post by Steve Reading »

jdilla1107 wrote: Thu Oct 24, 2019 4:13 pm
One Ping wrote: Thu Oct 24, 2019 12:41 pm
willthrill81 wrote: Thu Oct 24, 2019 9:49 am I find it interesting to hear some opine about TSM being 'maximally diversified' on one hand while they also seem to relish the idea of 20% of their U.S. stock being in ten companies. It seems to be a very clear example of very strong confirmation bias.
With maybe a touch of cognitive dissonance thrown in for good measure ...
Would you feel better if each of these 10 companies decided to break into 10-20 companies each? Would that remove the risk you are concerned with?

Do you realize that big companies actually operate this way? Microsoft has 100s of business lines. You could say it's almost like 100s of small caps operating under 1 umbrella, because that's what it is.

Diversification should not be measured entirely by "corporate legal structure"
Yes, if each of those companies broke into 10-20 different ones, that did completely different things, that would absolutely be better.

Evidently, we tilt because we don't see that being the case enough. I believe, for instance, that FB buying Whatsapp, IG, Oculus, etc is not at all appropriate enough diversification. It's still a similar space. If the large caps were truly conglomerates a la Berkshire Hathaway, I would honestly have no issues holding them at market cap. Like you say, they're like umbrellas for multiple smaller.

So it's just a matter of how diversified you really feel those Mega caps are.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
jdilla1107
Posts: 849
Joined: Sun Jun 24, 2012 8:31 pm

Re: Investing in the largest 10 companies?

Post by jdilla1107 »

305pelusa wrote: Thu Oct 24, 2019 4:34 pm
jdilla1107 wrote: Thu Oct 24, 2019 4:13 pm
One Ping wrote: Thu Oct 24, 2019 12:41 pm
willthrill81 wrote: Thu Oct 24, 2019 9:49 am I find it interesting to hear some opine about TSM being 'maximally diversified' on one hand while they also seem to relish the idea of 20% of their U.S. stock being in ten companies. It seems to be a very clear example of very strong confirmation bias.
With maybe a touch of cognitive dissonance thrown in for good measure ...
Would you feel better if each of these 10 companies decided to break into 10-20 companies each? Would that remove the risk you are concerned with?

Do you realize that big companies actually operate this way? Microsoft has 100s of business lines. You could say it's almost like 100s of small caps operating under 1 umbrella, because that's what it is.

Diversification should not be measured entirely by "corporate legal structure"
Yes, if each of those companies broke into 10-20 different ones, that did completely different things, that would absolutely be better.
So, If Microsoft announced they were breaking into separate companies:

- Home Entertainment
- Windows
- Business Solutions
- Cloud Computing
- Mobile and embedded devices
- A huge consulting organization

You would think to yourself, 'That's a lot more diversified now"?

That makes no sense.
User avatar
Steve Reading
Posts: 2959
Joined: Fri Nov 16, 2018 9:20 pm

Re: Investing in the largest 10 companies?

Post by Steve Reading »

jdilla1107 wrote: Thu Oct 24, 2019 4:41 pm
305pelusa wrote: Thu Oct 24, 2019 4:34 pm
jdilla1107 wrote: Thu Oct 24, 2019 4:13 pm
One Ping wrote: Thu Oct 24, 2019 12:41 pm
willthrill81 wrote: Thu Oct 24, 2019 9:49 am I find it interesting to hear some opine about TSM being 'maximally diversified' on one hand while they also seem to relish the idea of 20% of their U.S. stock being in ten companies. It seems to be a very clear example of very strong confirmation bias.
With maybe a touch of cognitive dissonance thrown in for good measure ...
Would you feel better if each of these 10 companies decided to break into 10-20 companies each? Would that remove the risk you are concerned with?

Do you realize that big companies actually operate this way? Microsoft has 100s of business lines. You could say it's almost like 100s of small caps operating under 1 umbrella, because that's what it is.

Diversification should not be measured entirely by "corporate legal structure"
Yes, if each of those companies broke into 10-20 different ones, that did completely different things, that would absolutely be better.
So, If Microsoft announced they were breaking into separate companies:

- Home Entertainment
- Windows
- Business Solutions
- Cloud Computing
- Mobile and embedded devices
- A huge consulting organization

You would think to yourself, 'That's a lot more diversified now"?

That makes no sense.
Huh? It has little to do with whether it's one or multiple entities.

Did you even read my post past what you quoted?
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
Post Reply