Larry Swedroe: The Factors That Plague Factor Investing

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
Random Walker
Posts: 3955
Joined: Fri Feb 23, 2007 8:21 pm

Larry Swedroe: The Factors That Plague Factor Investing

Post by Random Walker » Tue Apr 16, 2019 10:56 pm

https://alphaarchitect.com/2019/04/16/t ... investing/

Larry references a recent paper by Arnott et.al Alice’s Adventures In Factorland: Three Blunders That Plague Factor Investing. This paper serves as the starting point for an essay by Larry discussing some concerns, misunderstandings, and unrealistic expectations people have about factor investing. He discusses overcrowding, unrealistic trading cost expectations, misunderstanding of downside shocks, misunderstanding of diversification benefits. Multiple excellent points are made and several very good papers referenced. Larry finishes up the essay with some real after expense data. One of my favorite points in the essay is “ so much of the criticism about factor investing ignores the fact that what is true for other factors can also be true of market beta”. References to Fama-French, Asness, Buffett, skew, kurtosis, correlations, Vanguard.

Dave

averagedude
Posts: 629
Joined: Sun May 13, 2018 3:41 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by averagedude » Tue Apr 16, 2019 11:48 pm

I am a skeptic of factor investing, even though i may tilt to value and size. If quality, size, momentum, and value was a no brainer, wouldn't every investor invest in these products. Time will tell, but the difference in returns will be closely correlated in my opinion. Perhaps there is no premium in these factors and smart beta is just a marketing strategy to obtain assets under management of the money you have to invest. Surely, active managers can outperform passive strategies, because they have way more skill and knowledge than the average investor? Why wouldn't they make you think otherwise?

DonIce
Posts: 601
Joined: Thu Feb 21, 2019 6:44 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by DonIce » Wed Apr 17, 2019 12:09 am

averagedude wrote:
Tue Apr 16, 2019 11:48 pm
Perhaps there is no premium in these factors and smart beta is just a marketing strategy to obtain assets under management of the money you have to invest. Surely, active managers can outperform passive strategies, because they have way more skill and knowledge than the average investor? Why wouldn't they make you think otherwise?
I'm skeptical of factor investing as well but this particular criticism is not a fair one. Factor investing can still be implemented in what can reasonably be termed a passive manner.

I think the evidence is fairly clear that "value" companies have delivered returns that exceed the stock market average over the long term. Benjamin Graham and Warren Buffet have clearly showed that value investing can be a source of above average returns and I would not argue with their investing philosophies, nor assume that they are obsolete today. That said, I don't know that it can be reasonably described as a "factor" in the current sense of the word, where it is an essentially orthogonal attribute of stocks unrelated to market beta, and where a company can be categorized based on its factors "i.e. 0.8 beta 0.4 size 0.7 value etc" and its future price movements relative to other stocks predicted on that basis. If big companies and small companies reliably moved in opposite directions once one factored out market beta, this relative movement could easily be arbitraged away by institutional traders until it didn't exist any more.

To me, there's only really two conceptually useful levels to think about companies on: they are part of the overall stock market and can all be averaged together to produce average returns (which is plenty adequate for most people), or they are individual companies whose individual fundamentals can be analyzed to try to select individual companies or groups of companies that one thinks may outperform. Balance sheets can be analyzed for value to try to select companies that fit the "value" criteria, but this should not be done in isolation but rather combined with an understanding of their financial stability, growth prospects, etc, to get an overall picture of whether one should invest in that particular company or not. Successful value investors don't just buy every company that's cheap; rather, they analyze the full breadth of information about the company to make a decision about whether it is worth adding to their portfolio or not.

User avatar
nisiprius
Advisory Board
Posts: 38469
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by nisiprius » Wed Apr 17, 2019 8:45 am

In the sense that they all have risk, and fractal-like chaotic ups and downs, perhaps market beta is like other factors. However, there are many ways in which it is qualitatively different from the others. If anyone were to actually say it is "just another factor" (Larry Swedroe does not say that in his article) it would be a false equivalence. It would be like saying that "fiat money has the same issues as cryptocurrency" or "the S&P 500 index has the same issues as active management."

Here are a number of ways in which market beta is fundamentally different from other factors.

1) It is possible to obtain near-perfect 1.0 loading on the market factor using a long-only portfolio.

2) Not only is it possible to do this, it is easy to do this, using literally dozens of mutual funds and ETFs--including many that antedate factor investing as we know it, and quite a few that do so at extremely low cost.

For example, Portfolio Visualizer is showing market factor loadings of:

1.00 (!) for Vanguard Total Stock Mkt Idx Inv VTSMX May 1992 Feb 2019
0.99 for Vanguard 500 Index Investor VFINX Sep 1976 Feb 2019
0.98 SPDR S&P 500 ETF SPY Feb 1993 Feb 2019

And even, for the very oldest mutual fund the world,

0.95 MFS Massachusetts Investors Tr A MITTX Jan 1971 Feb 2019

The market beta factor is far more investable than any other factor.

3) The market beta factor is the only factor that cannot possibly become overcrowded (short of pathological thought experiments). As many investors who want to, can have a portfolio with close to 1.00 loading on the market factor, without affecting the ability of any other investor to do something different, and without selectively driving up the prices of any group of stocks.

In contrast, micro-cap companies in CRSP deciles 9 and 10, only comprise 1.41% of the market cap of the total market. The Bridgeway Ultra-Small Company Market Fund, BRSIX, has an 0.71 loading on the size factor, but it has assets of only $290 million. Since it is drawing from 1.41% of a $28 trillion market, funds investing in these deciles mathematically cannot collectively grow past $400 billion. Mathematically, funds like BRSIX collectively cannot even reach the size of the $772 billion Vanguard Total Stock Market Index Fund, let alone the roughly $11 trillion mutual fund and ETF investors have invested in the stock market.*

It is impossible for investors have an average of more than 3-4% of their portfolio in funds providing high loading on size. Much less, of course, in small-cap value, or other more recently fashionable factors.

There is enough market beta for everyone who wants it. There is not enough of any other factor to go around.

4) Market beta has been studied for a long time. Indexes, as we understand them, were first described (or invented) by Irving Fisher in 1922, The Making of Index Numbers. The Cowles Commission report--a full-sized book--Common-Stock Indexes, 1871-1937 was an exercise in historical research and therefore has room for doubt on accuracy, but going forward people understood how to compile indexes, so we have reliable market indexes for at least eighty years.

Furthermore, we have real-world mutual fund performance for stock funds going back to 1924.

As far as I know, the oldest small-cap fund is the DFA US Micro Cap Fund, inception in 1981. Small-cap value funds? No fund could have deliberately targeted the value factor before the publication of the Fama-French paper in 1993 (and the DFA US Small-Cap Value Portfolio had its inception in that very year, 1993). But, since I'm counting MITTX as a "market beta" fund although it wasn't scientifically designed to be one, I guess we need to ask about any funds that had a small-cap value style. Does anybody know of older funds that intuitively steered themselves into the small-cap value style box?

5) The market factor is better defined than other factors, which to some extent are works in progress, or have different schools of thought as to how they should be defined. There is reasonable agreement on what constitutes a listed stock in the US market--and the areas of uncertainty affect only the smallest microcaps, which have negligible cap-weight on the index as a whole. In contrast, different authorities have different ideas on what constitutes "value" and where you should draw the line.


*At year-end 2017 there was a total of $22 trillion in US mutual funds and ETFs. I'm guessing about half of that, $11 trillion--give or take a few trillions, you know--was in stocks.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

marcopolo
Posts: 2113
Joined: Sat Dec 03, 2016 10:22 am

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by marcopolo » Wed Apr 17, 2019 9:08 am

DonIce wrote:
Wed Apr 17, 2019 12:09 am
averagedude wrote:
Tue Apr 16, 2019 11:48 pm
Perhaps there is no premium in these factors and smart beta is just a marketing strategy to obtain assets under management of the money you have to invest. Surely, active managers can outperform passive strategies, because they have way more skill and knowledge than the average investor? Why wouldn't they make you think otherwise?
I'm skeptical of factor investing as well but this particular criticism is not a fair one. Factor investing can still be implemented in what can reasonably be termed a passive manner.

I think the evidence is fairly clear that "value" companies have delivered returns that exceed the stock market average over the long term. Benjamin Graham and Warren Buffet have clearly showed that value investing can be a source of above average returns and I would not argue with their investing philosophies, nor assume that they are obsolete today. That said, I don't know that it can be reasonably described as a "factor" in the current sense of the word, where it is an essentially orthogonal attribute of stocks unrelated to market beta, and where a company can be categorized based on its factors "i.e. 0.8 beta 0.4 size 0.7 value etc" and its future price movements relative to other stocks predicted on that basis. If big companies and small companies reliably moved in opposite directions once one factored out market beta, this relative movement could easily be arbitraged away by institutional traders until it didn't exist any more.

To me, there's only really two conceptually useful levels to think about companies on: they are part of the overall stock market and can all be averaged together to produce average returns (which is plenty adequate for most people), or they are individual companies whose individual fundamentals can be analyzed to try to select individual companies or groups of companies that one thinks may outperform. Balance sheets can be analyzed for value to try to select companies that fit the "value" criteria, but this should not be done in isolation but rather combined with an understanding of their financial stability, growth prospects, etc, to get an overall picture of whether one should invest in that particular company or not. Successful value investors don't just buy every company that's cheap; rather, they analyze the full breadth of information about the company to make a decision about whether it is worth adding to their portfolio or not.
Perhaps i misunderstood your point, but don't these two highlighted statements directly contradict each other?
Once in a while you get shown the light, in the strangest of places if you look at it right.

DonIce
Posts: 601
Joined: Thu Feb 21, 2019 6:44 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by DonIce » Wed Apr 17, 2019 9:17 am

marcopolo wrote:
Wed Apr 17, 2019 9:08 am
Perhaps i misunderstood your point, but don't these two highlighted statements directly contradict each other?
Sorry, my post was perhaps a bit rambling. I wasn't talking about factor investing any more at all by that bottom paragraph.

Topic Author
Random Walker
Posts: 3955
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by Random Walker » Wed Apr 17, 2019 9:24 am

Hi Nisi,
As you likely know, I’ve joined the “market beta is just another factor” crowd. Your point regarding ability to get a long only loading of 1 cheaply is a good one. Not sure about overcrowding; couldn’t overall valuation changes of the entire market reflect overcrowding? For individual portfolio construction purposes, most all of us have portfolios with the dominant majority of risk wrapped up in market beta. And I think moves towards more efficient portfolios involve diversifying away from market beta. So in a very practical sense market beta really is just another factor. Of course the most effective and cheapest start to diversifying away from market beta is high quality bonds.

Dave

User avatar
willthrill81
Posts: 10819
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by willthrill81 » Wed Apr 17, 2019 10:47 am

IMHO, the biggest issue related to factor investing is investors' impatience. The historic data are clear that any identified factor has underperformed the market at some point, and this can go on for years. As is evident by the posts of many on this forum, it's very difficult for many, probably most, investors to tolerate holding even a meaningful portion of an underperforming asset class for years. Most active investors certainly don't seem to be capable of doing so. David Stein of the "Money for the Rest of Us" podcast, who worked with endowment fund managers for years, says that most of those managers wouldn't tolerate underperformance for longer than about two years.

If you don't believe that you could hold an underperforming asset class for years on end, then you probably shouldn't have any factor exposure (i.e. own the total market).

It seems likely to me that this issue is a big part of the reason that many investors currently favor all U.S. portfolios.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

klaus14
Posts: 206
Joined: Sun Nov 25, 2018 7:43 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by klaus14 » Wed Apr 17, 2019 11:08 am

nisiprius wrote:
Wed Apr 17, 2019 8:45 am

3) The market beta factor is the only factor that cannot possibly become overcrowded (short of pathological thought experiments). As many investors who want to, can have a portfolio with close to 1.00 loading on the market factor, without affecting the ability of any other investor to do something different, and without selectively driving up the prices of any group of stocks.

In contrast, micro-cap companies in CRSP deciles 9 and 10, only comprise 1.41% of the market cap of the total market. The Bridgeway Ultra-Small Company Market Fund, BRSIX, has an 0.71 loading on the size factor, but it has assets of only $290 million. Since it is drawing from 1.41% of a $28 trillion market, funds investing in these deciles mathematically cannot collectively grow past $400 billion. Mathematically, funds like BRSIX collectively cannot even reach the size of the $772 billion Vanguard Total Stock Market Index Fund, let alone the roughly $11 trillion mutual fund and ETF investors have invested in the stock market.*

It is impossible for investors have an average of more than 3-4% of their portfolio in funds providing high loading on size. Much less, of course, in small-cap value, or other more recently fashionable factors.

There is enough market beta for everyone who wants it. There is not enough of any other factor to go around.
why should anyone care if SC will be overcrowded in the future or not? If it's not overcrowded yet, it's still investable.
if everyone tries to tilt towards SC in future, it's a very good thing for the investors that got in today.

Also, if too many people invest in a stock, that stock gets less valuey, so it's hard to overcrowd value factor.

Topic Author
Random Walker
Posts: 3955
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by Random Walker » Wed Apr 17, 2019 11:56 am

willthrill81 wrote:
Wed Apr 17, 2019 10:47 am
IMHO, the biggest issue related to factor investing is investors' impatience. The historic data are clear that any identified factor has underperformed the market at some point, and this can go on for years. As is evident by the posts of many on this forum, it's very difficult for many, probably most, investors to tolerate holding even a meaningful portion of an underperforming asset class for years. Most active investors certainly don't seem to be capable of doing so. David Stein of the "Money for the Rest of Us" podcast, who worked with endowment fund managers for years, says that most of those managers wouldn't tolerate underperformance for longer than about two years.

If you don't believe that you could hold an underperforming asset class for years on end, then you probably shouldn't have any factor exposure (i.e. own the total market).

It seems likely to me that this issue is a big part of the reason that many investors currently favor all U.S. portfolios.
I’m sure this is all true. For me though, investing is all about looking forward. That is why I place so much weight on Larry’s intuitive: risk based or behavioral based rationale to expect a premium to persist going forward. And when any factor, including market beta, can stink for a long time, makes sense to me to diversify across the factors broadly. From my perspective, with the low correlations between factors, over short periods one minimizes sequence of returns risk by diversifying. Especially true if you increase tilt to high expected return equity asset classes, decrease overall equity allocation, increase bond exposure.

Dave

User avatar
willthrill81
Posts: 10819
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by willthrill81 » Wed Apr 17, 2019 12:05 pm

Random Walker wrote:
Wed Apr 17, 2019 11:56 am
willthrill81 wrote:
Wed Apr 17, 2019 10:47 am
IMHO, the biggest issue related to factor investing is investors' impatience. The historic data are clear that any identified factor has underperformed the market at some point, and this can go on for years. As is evident by the posts of many on this forum, it's very difficult for many, probably most, investors to tolerate holding even a meaningful portion of an underperforming asset class for years. Most active investors certainly don't seem to be capable of doing so. David Stein of the "Money for the Rest of Us" podcast, who worked with endowment fund managers for years, says that most of those managers wouldn't tolerate underperformance for longer than about two years.

If you don't believe that you could hold an underperforming asset class for years on end, then you probably shouldn't have any factor exposure (i.e. own the total market).

It seems likely to me that this issue is a big part of the reason that many investors currently favor all U.S. portfolios.
I’m sure this is all true. For me though, investing is all about looking forward. That is why I place so much weight on Larry’s intuitive: risk based or behavioral based rationale to expect a premium to persist going forward. And when any factor, including market beta, can stink for a long time, makes sense to me to diversify across the factors broadly. From my perspective, with the low correlations between factors, over short periods one minimizes sequence of returns risk by diversifying. Especially true if you increase tilt to high expected return equity asset classes, decrease overall equity allocation, increase bond exposure.

Dave
As long the 'low correlations' persist in the future, then this seems to be reasonable. But as many investors discovered the hard way last year, even a practically zero correlation between asset classes (i.e. stocks and bonds) does not at all preclude both going down at the same time. And the last recession demonstrated that correlations that were typically low can unexpectedly increase suddenly. That event shook up a lot of people's belief in the effectiveness of MPT.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

marcopolo
Posts: 2113
Joined: Sat Dec 03, 2016 10:22 am

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by marcopolo » Wed Apr 17, 2019 12:14 pm

willthrill81 wrote:
Wed Apr 17, 2019 10:47 am
IMHO, the biggest issue related to factor investing is investors' impatience. The historic data are clear that any identified factor has underperformed the market at some point, and this can go on for years. As is evident by the posts of many on this forum, it's very difficult for many, probably most, investors to tolerate holding even a meaningful portion of an underperforming asset class for years. Most active investors certainly don't seem to be capable of doing so. David Stein of the "Money for the Rest of Us" podcast, who worked with endowment fund managers for years, says that most of those managers wouldn't tolerate underperformance for longer than about two years.

If you don't believe that you could hold an underperforming asset class for years on end, then you probably shouldn't have any factor exposure (i.e. own the total market).

It seems likely to me that this issue is a big part of the reason that many investors currently favor all U.S. portfolios.
I think this is a very good point, and probably one of the reasons i have shied away from factors. Although, I do stay diversified internationally.

The question I have is regarding the required patience. In your mind, how long does one need to be patient? Is it 10, 20 years, or a life time commitment? Is there any duration of continued, consistent under performance, or even simple lack out performance that would/should cause one to question their assumptions about why a certain factor should out perform?
Once in a while you get shown the light, in the strangest of places if you look at it right.

User avatar
willthrill81
Posts: 10819
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by willthrill81 » Wed Apr 17, 2019 12:23 pm

marcopolo wrote:
Wed Apr 17, 2019 12:14 pm
The question I have is regarding the required patience. In your mind, how long does one need to be patient? Is it 10, 20 years, or a life time commitment? Is there any duration of continued, consistent under performance, or even simple lack out performance that would/should cause one to question their assumptions about why a certain factor should out perform?
That's a very good question, but there's unfortunately no definitive answer and there may not even be a good practical answer. Take a look at SCV. It's outperformed the S&P 500 in every 20 year period on record except one. But there have been many 10 year periods where it underperformed, and the one 20 year period cannot be ignored. How many factor investors could hold SCV if it underperformed the S&P 500 for a decade or more? More importantly, could you (it's a rhetorical question)?
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

Topic Author
Random Walker
Posts: 3955
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by Random Walker » Wed Apr 17, 2019 1:24 pm

willthrill81 wrote:
Wed Apr 17, 2019 12:05 pm
As long the 'low correlations' persist in the future, then this seems to be reasonable. But as many investors discovered the hard way last year, even a practically zero correlation between asset classes (i.e. stocks and bonds) does not at all preclude both going down at the same time. And the last recession demonstrated that correlations that were typically low can unexpectedly increase suddenly. That event shook up a lot of people's belief in the effectiveness of MPT.
Larry talks about this some in the article. Maybe us individual investors expect too much from diversification at times. Equity correlations will go up and down over time and tend towards 1 in bad times. But diversification across low correlated asset classes should work “eventually”.

Dave

Amadis_of_Gaul
Posts: 102
Joined: Tue Dec 18, 2018 5:57 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by Amadis_of_Gaul » Wed Apr 17, 2019 1:41 pm

nisiprius wrote:
Wed Apr 17, 2019 8:45 am


3) The market beta factor is the only factor that cannot possibly become overcrowded (short of pathological thought experiments). As many investors who want to, can have a portfolio with close to 1.00 loading on the market factor, without affecting the ability of any other investor to do something different, and without selectively driving up the prices of any group of stocks.

In contrast, micro-cap companies in CRSP deciles 9 and 10, only comprise 1.41% of the market cap of the total market. The Bridgeway Ultra-Small Company Market Fund, BRSIX, has an 0.71 loading on the size factor, but it has assets of only $290 million. Since it is drawing from 1.41% of a $28 trillion market, funds investing in these deciles mathematically cannot collectively grow past $400 billion. Mathematically, funds like BRSIX collectively cannot even reach the size of the $772 billion Vanguard Total Stock Market Index Fund, let alone the roughly $11 trillion mutual fund and ETF investors have invested in the stock market.*

It is impossible for investors have an average of more than 3-4% of their portfolio in funds providing high loading on size. Much less, of course, in small-cap value, or other more recently fashionable factors.

There is enough market beta for everyone who wants it. There is not enough of any other factor to go around.
A couple of thoughts on this.

First, although I admittedly have much to learn, it seems to me that the market beta factor _can_ become overcrowded. This doesn't happen because the price of some group of stocks is bid up relative to other stocks. It happens because the price of stocks, period, is bid up relative to other assets in a way not justified by the equity-risk premium. Isn't every stock bubble evidence of overcrowding on market beta?

Second, whenever discussion of factors turns to overcrowding, my mind always goes back to TIPS. Financial experts frequently advocate the inclusion of TIPS in portfolios, but TIPS only make up a small percentage of the total bond market. If everyone wanted TIPS, there would not be enough of them to go around either. Why is overcrowding a concern when it comes to various factors but not when it comes to TIPS?

international001
Posts: 899
Joined: Thu Feb 15, 2018 7:31 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by international001 » Wed Apr 17, 2019 2:04 pm

Excuse my ignorance.. SCV has a return premium because it's riskier. Isn't that using the beta factor?
Or is SCV return even higher that what the beta would predict?

Topic Author
Random Walker
Posts: 3955
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by Random Walker » Wed Apr 17, 2019 2:53 pm

international001 wrote:
Wed Apr 17, 2019 2:04 pm
Excuse my ignorance.. SCV has a return premium because it's riskier. Isn't that using the beta factor?
Or is SCV return even higher that what the beta would predict?
Yes, that’s my understanding. Small and value stocks have additional risks and expected premia beyond what can be explained by their higher beta.

Dave

Amadis_of_Gaul
Posts: 102
Joined: Tue Dec 18, 2018 5:57 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by Amadis_of_Gaul » Wed Apr 17, 2019 2:59 pm

international001 wrote:
Wed Apr 17, 2019 2:04 pm
Excuse my ignorance.. SCV has a return premium because it's riskier. Isn't that using the beta factor?
Or is SCV return even higher that what the beta would predict?
I'm far from an expert, but my understanding is that market beta explains most of the historically greater SCV return but not all of it. There are greater risks associated with small and value stocks that aren't present in the rest of the market. Investors don't like those risks, so the extra risk is rewarded with extra return.
Last edited by Amadis_of_Gaul on Wed Apr 17, 2019 6:09 pm, edited 1 time in total.

User avatar
willthrill81
Posts: 10819
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by willthrill81 » Wed Apr 17, 2019 3:04 pm

Random Walker wrote:
Wed Apr 17, 2019 1:24 pm
willthrill81 wrote:
Wed Apr 17, 2019 12:05 pm
As long the 'low correlations' persist in the future, then this seems to be reasonable. But as many investors discovered the hard way last year, even a practically zero correlation between asset classes (i.e. stocks and bonds) does not at all preclude both going down at the same time. And the last recession demonstrated that correlations that were typically low can unexpectedly increase suddenly. That event shook up a lot of people's belief in the effectiveness of MPT.
Larry talks about this some in the article. Maybe us individual investors expect too much from diversification at times. Equity correlations will go up and down over time and tend towards 1 in bad times. But diversification across low correlated asset classes should work “eventually”.

Dave
But that's just the problem. It's in bad times when diversification is most needed, but that also seems to be when it's most likely to fail. At least historically, it wasn't the long-term (i.e. "eventually") when most investors really needed diversification; stocks alone largely got investors where they wanted to go in the long-term. When people really want diversification is when stocks are down 30% or more. I'm personally not convinced that this ongoing search for 'lowly correlated assets' is going to be significantly more effective in terms of costs, returns, and diversification than a simple mix of equities and government-backed or guaranteed fixed income like Treasuries and CDs.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
patrick013
Posts: 2617
Joined: Mon Jul 13, 2015 7:49 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by patrick013 » Wed Apr 17, 2019 3:38 pm

I think I've learned to live with volatility and limited choices regarding correlation and returns and indexes that take turns being the better index this year or this decade. Isolating factors or intl equity would tend to increase this turn taking in the former and increase off shore diversification in the latter. So factors are running their cycle too.

Long story short when the market crashes all stocks basically crash. The choices then should be TRSY and CD's enough for several years even longer. An A++ annuity is more a luxury than a life preserver then IMO. Another convenience not unlike a corporate bond. Factors don't mean very much then when the market crashes because all stocks basically crash.



...
age in bonds, buy-and-hold, 10 year business cycle

pward
Posts: 391
Joined: Fri Dec 21, 2018 8:18 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by pward » Wed Apr 17, 2019 4:43 pm

willthrill81 wrote:
Wed Apr 17, 2019 3:04 pm
Random Walker wrote:
Wed Apr 17, 2019 1:24 pm
willthrill81 wrote:
Wed Apr 17, 2019 12:05 pm
As long the 'low correlations' persist in the future, then this seems to be reasonable. But as many investors discovered the hard way last year, even a practically zero correlation between asset classes (i.e. stocks and bonds) does not at all preclude both going down at the same time. And the last recession demonstrated that correlations that were typically low can unexpectedly increase suddenly. That event shook up a lot of people's belief in the effectiveness of MPT.
Larry talks about this some in the article. Maybe us individual investors expect too much from diversification at times. Equity correlations will go up and down over time and tend towards 1 in bad times. But diversification across low correlated asset classes should work “eventually”.

Dave
But that's just the problem. It's in bad times when diversification is most needed, but that also seems to be when it's most likely to fail. At least historically, it wasn't the long-term (i.e. "eventually") when most investors really needed diversification; stocks alone largely got investors where they wanted to go in the long-term. When people really want diversification is when stocks are down 30% or more. I'm personally not convinced that this ongoing search for 'lowly correlated assets' is going to be significantly more effective in terms of costs, returns, and diversification than a simple mix of equities and government-backed or guaranteed fixed income like Treasuries and CDs.
Diversification is a range, not an on or off switch. Different assets, factors, etc provide different levels of diversification. There's diversification that is completely uncorrelated (for instance the diversification that exists between stocks, bonds, and hard assets), and diversification that is of lowered correlation (this would be where "factors" fit in).

A couple things people misunderstand is that even though two assets are uncorrelated doesn't mean that both assets can't go in the same direction over time. I mean there are plenty of periods of time since the mid 80's where both stocks and bonds went up at the same time, and in the 2000-2007 period both stocks and gold went up in unison. Likewise in the 1970s both stocks and bonds went down at the same time. So if someone is looking for uncorrelated assets it's best to pick a combination where one asset has always historically been in a bull market. This is the form of diversification that saves your bacon when crisis' like 2008 hit. For instance, in 2008 someone that held long bonds and gold along side stocks did pretty well on the whole with a low drawdown and quick time to recover.

Then there is the lowered correlation form of diversification. This is where within an asset class you are trying to purposefully inject tracking error. The most popular version here being "factors" like SCV inside of a stock allocation. At the end of the day SCV stocks are still stocks, and they are still going to get bullied by market beta. If stocks are going down in general, SCV will also be going down, but it will inject tracking error that will cause it to go down at a different pace. Likewise, when stocks in general are going up, SCV will also go up but will inject tracking error into the portfolio that will cause it to go up at a different pace. While there are times this tracking error will work in your favor, and times it will work against you, over the long haul these "factors" tend to perform better than stocks as a whole. So when implementing them it's not for "diversification" in the traditionally understood sense, where one is counting on them to go up when stocks go down. It's more performance optimization and tracking error.

Now this performance optimization side effect can be leveraged like Dave mentioned, to allow someone to hold a lower amount of stocks so that they can hold a higher amount of the real uncorrelated diversifying assets like bonds and hard assets. This is the only way that these factors can be utilized to give "diversification" in the commonly understood sense of protecting your bacon in a crisis. I don't think people fully understand what "diversification" means, as it should not be shocking that when stocks are in a cyclical bear market that things like SCV, international stocks and corporate bonds are also in a cyclical bear market. This should be expected.

Topic Author
Random Walker
Posts: 3955
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by Random Walker » Wed Apr 17, 2019 7:13 pm

willthrill81 wrote:
Wed Apr 17, 2019 3:04 pm
I'm personally not convinced that this ongoing search for 'lowly correlated assets' is going to be significantly more effective in terms of costs, returns, and diversification than a simple mix of equities and government-backed or guaranteed fixed income like Treasuries and CDs.
I’m not convinced either, but very eager to see how it all works out over my investing career. I’ve got a heavily tilted equity allocation, big dose bonds, and substantial alternatives. Some of the data that has caused me to choose this path has shown that when the tilted portfolio lags, it tends to not lag by much. For example, many have commented that SV has underperformed this last decade. It has, but the returns have nonetheless been positive and generous. Sticking to a good plan likely more important than finding a better plan.

Dave

User avatar
willthrill81
Posts: 10819
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by willthrill81 » Wed Apr 17, 2019 7:15 pm

Random Walker wrote:
Wed Apr 17, 2019 7:13 pm
willthrill81 wrote:
Wed Apr 17, 2019 3:04 pm
I'm personally not convinced that this ongoing search for 'lowly correlated assets' is going to be significantly more effective in terms of costs, returns, and diversification than a simple mix of equities and government-backed or guaranteed fixed income like Treasuries and CDs.
I’m not convinced either, but very eager to see how it all works out over my investing career. I’ve got a heavily tilted equity allocation, big dose bonds, and substantial alternatives. Some of the data that has caused me to choose this path has shown that when the tilted portfolio lags, it tends to not lag by much. For example, many have commented that SV has underperformed this last decade. It has, but the returns have nonetheless been positive and generous. Sticking to a good plan likely more important than finding a better plan.

Dave
I can't argue with that! :beer
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

rkhusky
Posts: 6751
Joined: Thu Aug 18, 2011 8:09 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by rkhusky » Thu Apr 18, 2019 6:26 am

pward wrote:
Wed Apr 17, 2019 4:43 pm
A couple things people misunderstand is that even though two assets are uncorrelated doesn't mean that both assets can't go in the same direction over time.
In fact, if they are truly uncorrelated they should move in the same direction about half the time. If they always move in the opposite direction, they are anti-correlated.

User avatar
packer16
Posts: 1213
Joined: Sat Jan 04, 2014 2:28 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by packer16 » Thu Apr 18, 2019 8:33 am

pward wrote:
Wed Apr 17, 2019 4:43 pm
willthrill81 wrote:
Wed Apr 17, 2019 3:04 pm
Random Walker wrote:
Wed Apr 17, 2019 1:24 pm
willthrill81 wrote:
Wed Apr 17, 2019 12:05 pm
As long the 'low correlations' persist in the future, then this seems to be reasonable. But as many investors discovered the hard way last year, even a practically zero correlation between asset classes (i.e. stocks and bonds) does not at all preclude both going down at the same time. And the last recession demonstrated that correlations that were typically low can unexpectedly increase suddenly. That event shook up a lot of people's belief in the effectiveness of MPT.
Larry talks about this some in the article. Maybe us individual investors expect too much from diversification at times. Equity correlations will go up and down over time and tend towards 1 in bad times. But diversification across low correlated asset classes should work “eventually”.

Dave
But that's just the problem. It's in bad times when diversification is most needed, but that also seems to be when it's most likely to fail. At least historically, it wasn't the long-term (i.e. "eventually") when most investors really needed diversification; stocks alone largely got investors where they wanted to go in the long-term. When people really want diversification is when stocks are down 30% or more. I'm personally not convinced that this ongoing search for 'lowly correlated assets' is going to be significantly more effective in terms of costs, returns, and diversification than a simple mix of equities and government-backed or guaranteed fixed income like Treasuries and CDs.
Diversification is a range, not an on or off switch. Different assets, factors, etc provide different levels of diversification. There's diversification that is completely uncorrelated (for instance the diversification that exists between stocks, bonds, and hard assets), and diversification that is of lowered correlation (this would be where "factors" fit in).

A couple things people misunderstand is that even though two assets are uncorrelated doesn't mean that both assets can't go in the same direction over time. I mean there are plenty of periods of time since the mid 80's where both stocks and bonds went up at the same time, and in the 2000-2007 period both stocks and gold went up in unison. Likewise in the 1970s both stocks and bonds went down at the same time. So if someone is looking for uncorrelated assets it's best to pick a combination where one asset has always historically been in a bull market. This is the form of diversification that saves your bacon when crisis' like 2008 hit. For instance, in 2008 someone that held long bonds and gold along side stocks did pretty well on the whole with a low drawdown and quick time to recover.

Then there is the lowered correlation form of diversification. This is where within an asset class you are trying to purposefully inject tracking error. The most popular version here being "factors" like SCV inside of a stock allocation. At the end of the day SCV stocks are still stocks, and they are still going to get bullied by market beta. If stocks are going down in general, SCV will also be going down, but it will inject tracking error that will cause it to go down at a different pace. Likewise, when stocks in general are going up, SCV will also go up but will inject tracking error into the portfolio that will cause it to go up at a different pace. While there are times this tracking error will work in your favor, and times it will work against you, over the long haul these "factors" tend to perform better than stocks as a whole. So when implementing them it's not for "diversification" in the traditionally understood sense, where one is counting on them to go up when stocks go down. It's more performance optimization and tracking error.

Now this performance optimization side effect can be leveraged like Dave mentioned, to allow someone to hold a lower amount of stocks so that they can hold a higher amount of the real uncorrelated diversifying assets like bonds and hard assets. This is the only way that these factors can be utilized to give "diversification" in the commonly understood sense of protecting your bacon in a crisis. I don't think people fully understand what "diversification" means, as it should not be shocking that when stocks are in a cyclical bear market that things like SCV, international stocks and corporate bonds are also in a cyclical bear market. This should be expected.
I agree with you in the practical implementation of factors (performance optimization) but IMO the additional risk you take on when you do this it timing risk. I have been running a trial on the portfolio of SCV/bond vs. SP500/bonds since Larry S's article in the NYT came out (late 2011). It is not a good result. Over that time period (Dec 2011 to Dec 2018) if you compare a 60/40 SP500/bonds to a 40/60 SCV/bonds the vol is roughly the same at 7% but the returns of the SCV/bonds is 6.0%/year while the S&P500/bonds is 8.6%/year. For someone relying on stock returns to reach a goal that is a large delta over a long period of time. So IMO this timing risk is big risk that is seldom mentioned & by some it is assumed that it will all come out in the end. As I have observed this over the past few years I become more skeptical of the later. Whether this turns around or not it is a risk of factor investing.

Also, the idea of factor diversification is good in theory but in practice, I have not seen it work. To the contrary, when I have seen folks mix factors it makes the decisions more complex & IMO subject to error. QSPIX, run by factor experts over one of the longest timeframes I am aware of, is a case in point.

Packer
Buy cheap and something good might happen

Topic Author
Random Walker
Posts: 3955
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by Random Walker » Thu Apr 18, 2019 9:04 am

Larry reminded me that although SV has underperformed in US over last decade, internationally SV has outperformed. Over last decade DISVX has outperformed MSCI ex-US AWI by over 2%. Strong argument for diversifying as broadly as possible across geographies, asset classes, styles, factors.

Dave

User avatar
House Blend
Posts: 4586
Joined: Fri May 04, 2007 1:02 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by House Blend » Thu Apr 18, 2019 9:32 am

rkhusky wrote:
Thu Apr 18, 2019 6:26 am
pward wrote:
Wed Apr 17, 2019 4:43 pm
A couple things people misunderstand is that even though two assets are uncorrelated doesn't mean that both assets can't go in the same direction over time.
In fact, if they are truly uncorrelated they should move in the same direction about half the time. If they always move in the opposite direction, they are anti-correlated.
Something even less well understood is that when measuring the correlation of X and Y, "directions" are measured relative to the averages of X and Y, not whether X and Y are up or down.

Stocks and bonds both have positive average returns, so both can deliver positive returns while making a negative contribution to their correlation.

User avatar
nisiprius
Advisory Board
Posts: 38469
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by nisiprius » Thu Apr 18, 2019 9:55 am

House Blend wrote:
Thu Apr 18, 2019 9:32 am
...Stocks and bonds both have positive average returns, so both can deliver positive returns while making a negative contribution to their correlation...
Or as someone said during 2008-2009 in this forum, "Low correlation means that during a crash, you get to watch all your asset classes fall at different speeds."

(For the record, though, that wasn't true during 2000-2002, when non-factor-based portfolios like the Vanguard LifeStrategy funds, and like my own, fell--but classic factor-based portfolios like the Bill Schultheis Coffeehouse portfolio--10% each large-cap blend, large-cap value, small-cap blend, small-cap value, REITs, international, and 40% Total Bond--eked out decent gains).
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

pward
Posts: 391
Joined: Fri Dec 21, 2018 8:18 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by pward » Thu Apr 18, 2019 12:06 pm

rkhusky wrote:
Thu Apr 18, 2019 6:26 am
pward wrote:
Wed Apr 17, 2019 4:43 pm
A couple things people misunderstand is that even though two assets are uncorrelated doesn't mean that both assets can't go in the same direction over time.
In fact, if they are truly uncorrelated they should move in the same direction about half the time. If they always move in the opposite direction, they are anti-correlated.
False, that would be a negative correlation. Uncorrelated means they all do their own thing, and they can go up together or down together or in separate directions entirely.

pward
Posts: 391
Joined: Fri Dec 21, 2018 8:18 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by pward » Thu Apr 18, 2019 12:10 pm

packer16 wrote:
Thu Apr 18, 2019 8:33 am

I agree with you in the practical implementation of factors (performance optimization) but IMO the additional risk you take on when you do this it timing risk. I have been running a trial on the portfolio of SCV/bond vs. SP500/bonds since Larry S's article in the NYT came out (late 2011). It is not a good result. Over that time period (Dec 2011 to Dec 2018) if you compare a 60/40 SP500/bonds to a 40/60 SCV/bonds the vol is roughly the same at 7% but the returns of the SCV/bonds is 6.0%/year while the S&P500/bonds is 8.6%/year. For someone relying on stock returns to reach a goal that is a large delta over a long period of time. So IMO this timing risk is big risk that is seldom mentioned & by some it is assumed that it will all come out in the end. As I have observed this over the past few years I become more skeptical of the later. Whether this turns around or not it is a risk of factor investing.

Also, the idea of factor diversification is good in theory but in practice, I have not seen it work. To the contrary, when I have seen folks mix factors it makes the decisions more complex & IMO subject to error. QSPIX, run by factor experts over one of the longest timeframes I am aware of, is a case in point.

Packer
What you are describing is exactly what I referred to as "tracking error". There are times the tracking error will work for you, and times it will work against you. And those times can be very long indeed because they are a byproduct of the underlying macro environment. These things play out over the course of decades. So I think that we are in agreement, just using different language to explain the same thing.

And to your final point there, the individual is the portfolios biggest risk. If someone cannot stick with a plan then they should not implement it. For someone that would be at risk of deviating if the tracking error is not going in their favor for many years, they are better off just holding the market and calling it a day. There's nothing wrong with holding the market, it's not like factor investing has proven to be substantially better than market returns over time, it just shows a minor (but still meaningful) increase in returns... and now that these factors are known will in all likelihood give a much lower increase in returns in the future.

garlandwhizzer
Posts: 2323
Joined: Fri Aug 06, 2010 3:42 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by garlandwhizzer » Thu Apr 18, 2019 1:08 pm

NISI wrote:

(For the record, though, that wasn't true during 2000-2002, when non-factor-based portfolios like the Vanguard LifeStrategy funds, and like my own, fell--but classic factor-based portfolios like the Bill Schultheis Coffeehouse portfolio--10% each large-cap blend, large-cap value, small-cap blend, small-cap value, REITs, international, and 40% Total Bond--eked out decent gains).
In the period above, 2000 - 2002, the bursting of the tech bubble, SCV rose from the grave and massively outperformed. Suddenly growth stocks, which had been driven higher by the investor euphoria in the 1990s, collapsed and investors started looking at long-ignored cheap assets like SCV and REITS whose valuations were at the time incredibly attractive. Hence size and value outperformed which is the usual case after a LCG bubble bursts (tech wreck 2000, nifty fifty LCG inflating in 1960s and 1970s, which likewise collapsed in the 1980s). Following the collapse of LCG bubbles, long ignored SCV has roared back historically, one reason why its strong outperformance comes in infrequent periodic multi-year spurts. The tendency for formation and inflation of bubbles in LCG driven by euphoria (this time it's different, the old rules don't apply) is certainty a factor driving SCV spurts of historical outperformance. Bubbles form in growth stocks, especially LCG, never in SCV. Neglected and struggling small stocks do not inspire euphoria. So one important question regarding SCV going forward near term is: do we currently have a LCG bubble that is destined to burst in the foreseeable future? Is SCV neglected and ignored, driving its valuations to very attractive levels? The answer to that in my opinion is probably not at present, certainly not to the degree of the tech bubble. Unlike the bull market of the late 1990s and the nifty fifty, investors as a whole have remained skeptical and never become euphoric during this entire current 10 year bull market. They have not yet created a massive LCG bubble as they are prone to do from time to time. Nor have they hugely neglected SCV with all the SCV funds and ETFs out there seeking to harvest the premiums. Severe imbalances between growth and value, between LC and SC, have not occurred this time. Such imbalances are the driving force for reversion to the mean where SCV shines. SCV has been struggling and underperforming for a long time and may continue to struggle for a while going forward. If and when investors get frustrated, bail out of SCV and head for where the action is, that may at last set the stage for a bright SCV future.

Garland Whizzer

pward
Posts: 391
Joined: Fri Dec 21, 2018 8:18 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by pward » Thu Apr 18, 2019 3:09 pm

garlandwhizzer wrote:
Thu Apr 18, 2019 1:08 pm
NISI wrote:

(For the record, though, that wasn't true during 2000-2002, when non-factor-based portfolios like the Vanguard LifeStrategy funds, and like my own, fell--but classic factor-based portfolios like the Bill Schultheis Coffeehouse portfolio--10% each large-cap blend, large-cap value, small-cap blend, small-cap value, REITs, international, and 40% Total Bond--eked out decent gains).
In the period above, 2000 - 2002, the bursting of the tech bubble, SCV rose from the grave and massively outperformed. Suddenly growth stocks, which had been driven higher by the investor euphoria in the 1990s, collapsed and investors started looking at long-ignored cheap assets like SCV and REITS whose valuations were at the time incredibly attractive. Hence size and value outperformed which is the usual case after a LCG bubble bursts (tech wreck 2000, nifty fifty LCG inflating in 1960s and 1970s, which likewise collapsed in the 1980s). Following the collapse of LCG bubbles, long ignored SCV has roared back historically, one reason why its strong outperformance comes in infrequent periodic multi-year spurts. The tendency for formation and inflation of bubbles in LCG driven by euphoria (this time it's different, the old rules don't apply) is certainty a factor driving SCV spurts of historical outperformance. Bubbles form in growth stocks, especially LCG, never in SCV. Neglected and struggling small stocks do not inspire euphoria. So one important question regarding SCV going forward near term is: do we currently have a LCG bubble that is destined to burst in the foreseeable future? Is SCV neglected and ignored, driving its valuations to very attractive levels? The answer to that in my opinion is probably not at present, certainly not to the degree of the tech bubble. Unlike the bull market of the late 1990s and the nifty fifty, investors as a whole have remained skeptical and never become euphoric during this entire current 10 year bull market. They have not yet created a massive LCG bubble as they are prone to do from time to time. Nor have they hugely neglected SCV with all the SCV funds and ETFs out there seeking to harvest the premiums. Severe imbalances between growth and value, between LC and SC, have not occurred this time. Such imbalances are the driving force for reversion to the mean where SCV shines. SCV has been struggling and underperforming for a long time and may continue to struggle for a while going forward. If and when investors get frustrated, bail out of SCV and head for where the action is, that may at last set the stage for a bright SCV future.

Garland Whizzer
I think the theory behind the SCV cycles has some merit, however I come to a slightly different conclusion on the current state of the market. I see many similarities between the "Nifty Fifty" and the current "FAANG" trade. One might even argue that the popularity of indexing itself has also contributed to a "Nifty Fifty" like large cap growth bubble, where since large cap growth stocks make up more of the index they get a higher amount of investment which becomes a self perpetuating bubble. That is part of why I personally am attracted to SCV for at least the next few years, as I think injecting some form of tracking error against LCG will be beneficial whenever the next recession does inevitably hit.

User avatar
willthrill81
Posts: 10819
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by willthrill81 » Thu Apr 18, 2019 3:21 pm

pward wrote:
Thu Apr 18, 2019 3:09 pm
garlandwhizzer wrote:
Thu Apr 18, 2019 1:08 pm
NISI wrote:

(For the record, though, that wasn't true during 2000-2002, when non-factor-based portfolios like the Vanguard LifeStrategy funds, and like my own, fell--but classic factor-based portfolios like the Bill Schultheis Coffeehouse portfolio--10% each large-cap blend, large-cap value, small-cap blend, small-cap value, REITs, international, and 40% Total Bond--eked out decent gains).
In the period above, 2000 - 2002, the bursting of the tech bubble, SCV rose from the grave and massively outperformed. Suddenly growth stocks, which had been driven higher by the investor euphoria in the 1990s, collapsed and investors started looking at long-ignored cheap assets like SCV and REITS whose valuations were at the time incredibly attractive. Hence size and value outperformed which is the usual case after a LCG bubble bursts (tech wreck 2000, nifty fifty LCG inflating in 1960s and 1970s, which likewise collapsed in the 1980s). Following the collapse of LCG bubbles, long ignored SCV has roared back historically, one reason why its strong outperformance comes in infrequent periodic multi-year spurts. The tendency for formation and inflation of bubbles in LCG driven by euphoria (this time it's different, the old rules don't apply) is certainty a factor driving SCV spurts of historical outperformance. Bubbles form in growth stocks, especially LCG, never in SCV. Neglected and struggling small stocks do not inspire euphoria. So one important question regarding SCV going forward near term is: do we currently have a LCG bubble that is destined to burst in the foreseeable future? Is SCV neglected and ignored, driving its valuations to very attractive levels? The answer to that in my opinion is probably not at present, certainly not to the degree of the tech bubble. Unlike the bull market of the late 1990s and the nifty fifty, investors as a whole have remained skeptical and never become euphoric during this entire current 10 year bull market. They have not yet created a massive LCG bubble as they are prone to do from time to time. Nor have they hugely neglected SCV with all the SCV funds and ETFs out there seeking to harvest the premiums. Severe imbalances between growth and value, between LC and SC, have not occurred this time. Such imbalances are the driving force for reversion to the mean where SCV shines. SCV has been struggling and underperforming for a long time and may continue to struggle for a while going forward. If and when investors get frustrated, bail out of SCV and head for where the action is, that may at last set the stage for a bright SCV future.

Garland Whizzer
I think the theory behind the SCV cycles has some merit, however I come to a slightly different conclusion on the current state of the market. I see many similarities between the "Nifty Fifty" and the current "FAANG" trade. One might even argue that the popularity of indexing itself has also contributed to a "Nifty Fifty" like large cap growth bubble, where since large cap growth stocks make up more of the index they get a higher amount of investment which becomes a self perpetuating bubble. That is part of why I personally am attracted to SCV for at least the next few years, as I think injecting some form of tracking error against LCG will be beneficial whenever the next recession does inevitably hit.
Of course, one could make the counterargument that all the heraldry regarding the historic SCV premium along with comparatively easy access to SCV may lead to overcrowding, resulting in the negation of the premium going forward. I'm not saying that I espouse that view, but I've heard it here multiple times.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

pward
Posts: 391
Joined: Fri Dec 21, 2018 8:18 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by pward » Thu Apr 18, 2019 3:26 pm

willthrill81 wrote:
Thu Apr 18, 2019 3:21 pm
Of course, one could make the counterargument that all the heraldry regarding the historic SCV premium along with comparatively easy access to SCV may lead to overcrowding, resulting in the negation of the premium going forward. I'm not saying that I espouse that view, but I've heard it here multiple times.
There is certainly a possibility that could happen. There is no such thing as a risk free return.

Of course the flip side is would the premium disappearing lead to it becoming unattractive again, which would in turn lead to the premium returning again? As popular as value funds are here, they are actually pretty unpopular on the whole right now.

rkhusky
Posts: 6751
Joined: Thu Aug 18, 2011 8:09 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by rkhusky » Thu Apr 18, 2019 5:47 pm

pward wrote:
Thu Apr 18, 2019 12:06 pm
rkhusky wrote:
Thu Apr 18, 2019 6:26 am
pward wrote:
Wed Apr 17, 2019 4:43 pm
A couple things people misunderstand is that even though two assets are uncorrelated doesn't mean that both assets can't go in the same direction over time.
In fact, if they are truly uncorrelated they should move in the same direction about half the time. If they always move in the opposite direction, they are anti-correlated.
False, that would be a negative correlation. Uncorrelated means they all do their own thing, and they can go up together or down together or in separate directions entirely.
Anti-correlation = negative correlation.

pward
Posts: 391
Joined: Fri Dec 21, 2018 8:18 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by pward » Thu Apr 18, 2019 5:54 pm

rkhusky wrote:
Thu Apr 18, 2019 5:47 pm
pward wrote:
Thu Apr 18, 2019 12:06 pm
rkhusky wrote:
Thu Apr 18, 2019 6:26 am
pward wrote:
Wed Apr 17, 2019 4:43 pm
A couple things people misunderstand is that even though two assets are uncorrelated doesn't mean that both assets can't go in the same direction over time.
In fact, if they are truly uncorrelated they should move in the same direction about half the time. If they always move in the opposite direction, they are anti-correlated.
False, that would be a negative correlation. Uncorrelated means they all do their own thing, and they can go up together or down together or in separate directions entirely.
Anti-correlation = negative correlation.
Oops, I read your post too quickly. Ignore my last comment.

DB2
Posts: 337
Joined: Thu Jan 17, 2019 10:07 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by DB2 » Thu Apr 18, 2019 9:47 pm

pward wrote:
Thu Apr 18, 2019 3:26 pm
willthrill81 wrote:
Thu Apr 18, 2019 3:21 pm
Of course, one could make the counterargument that all the heraldry regarding the historic SCV premium along with comparatively easy access to SCV may lead to overcrowding, resulting in the negation of the premium going forward. I'm not saying that I espouse that view, but I've heard it here multiple times.
There is certainly a possibility that could happen. There is no such thing as a risk free return.

Of course the flip side is would the premium disappearing lead to it becoming unattractive again, which would in turn lead to the premium returning again? As popular as value funds are here, they are actually pretty unpopular on the whole right now.
I was watching an interview with Bogle recently...to paraphrase, he said perhaps people switch to one or the other (Growth or Value) when one is priced lower than the other at some point...reversion to the mean ultimately. Quick look at the Vanguard small cap ETFs YTD, since at least 2004, Growth has beaten Value.

pward
Posts: 391
Joined: Fri Dec 21, 2018 8:18 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by pward » Fri Apr 19, 2019 10:30 am

DB2 wrote:
Thu Apr 18, 2019 9:47 pm
pward wrote:
Thu Apr 18, 2019 3:26 pm
willthrill81 wrote:
Thu Apr 18, 2019 3:21 pm
Of course, one could make the counterargument that all the heraldry regarding the historic SCV premium along with comparatively easy access to SCV may lead to overcrowding, resulting in the negation of the premium going forward. I'm not saying that I espouse that view, but I've heard it here multiple times.
There is certainly a possibility that could happen. There is no such thing as a risk free return.

Of course the flip side is would the premium disappearing lead to it becoming unattractive again, which would in turn lead to the premium returning again? As popular as value funds are here, they are actually pretty unpopular on the whole right now.
I was watching an interview with Bogle recently...to paraphrase, he said perhaps people switch to one or the other (Growth or Value) when one is priced lower than the other at some point...reversion to the mean ultimately. Quick look at the Vanguard small cap ETFs YTD, since at least 2004, Growth has beaten Value.
Yes there is definitely some reversion to the mean over time. There are definitely different economic climates that favor growth and value respectively. Being able to actually time them however is not an easy task. I mean as you mentioned growth has out performed for years and years now. At what point has the low "factor" been low long enough to warrant swapping? If you swap too early then you wind up under performing. One could use a momentum strategy, but those are prone to whipsaws, head fakes, and lots of short term cap gains. It would also require an understanding of technical and quantitative analysis (which the average Joe does not have). About the only non-technical way an average person could implement the strategy sensibly would be to have a separate growth and value allocation and rebalance between them so you're buying low and selling high over time. Whether that would actually perform better than just holding a blend index I'm not sure. Might be worth doing some backtesting on for the fun of it.

User avatar
patrick013
Posts: 2617
Joined: Mon Jul 13, 2015 7:49 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by patrick013 » Fri Apr 19, 2019 1:09 pm

pward wrote:
Fri Apr 19, 2019 10:30 am

Yes there is definitely some reversion to the mean over time. There are definitely different economic climates that favor growth and value respectively. Being able to actually time them however is not an easy task. I mean as you mentioned growth has out performed for years and years now. At what point has the low "factor" been low long enough to warrant swapping? If you swap too early then you wind up under performing. One could use a momentum strategy, but those are prone to whipsaws, head fakes, and lots of short term cap gains. It would also require an understanding of technical and quantitative analysis (which the average Joe does not have). About the only non-technical way an average person could implement the strategy sensibly would be to have a separate growth and value allocation and rebalance between them so you're buying low and selling high over time. Whether that would actually perform better than just holding a blend index I'm not sure. Might be worth doing some backtesting on for the fun of it.
https://kz0fbw.bn.files.1drv.com/y4mPMs ... pmode=none

The above file doesn't show growth vs. value but rather LC vs MC vs SC. So thru the years the returns vary with each year. Lengthy histories would probably show that for any factor clearly. Timing is just spending or rebalancing from the factor with the highest return this year.
Last edited by patrick013 on Thu Apr 25, 2019 11:48 am, edited 3 times in total.
age in bonds, buy-and-hold, 10 year business cycle

User avatar
Taylor Larimore
Advisory Board
Posts: 28392
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by Taylor Larimore » Fri Apr 19, 2019 1:36 pm

Random Walker wrote:
Tue Apr 16, 2019 10:56 pm
https://alphaarchitect.com/2019/04/16/t ... investing/

Larry references a recent paper by Arnott et.al Alice’s Adventures In Factorland: Three Blunders That Plague Factor Investing. This paper serves as the starting point for an essay by Larry discussing some concerns, misunderstandings, and unrealistic expectations people have about factor investing. He discusses overcrowding, unrealistic trading cost expectations, misunderstanding of downside shocks, misunderstanding of diversification benefits. Multiple excellent points are made and several very good papers referenced. Larry finishes up the essay with some real after expense data. One of my favorite points in the essay is “ so much of the criticism about factor investing ignores the fact that what is true for other factors can also be true of market beta”. References to Fama-French, Asness, Buffett, skew, kurtosis, correlations, Vanguard.

Dave
Bogleheads:

You can read my thoughts about "factor investing" here":

Factor Investing

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle

pward
Posts: 391
Joined: Fri Dec 21, 2018 8:18 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by pward » Fri Apr 19, 2019 1:38 pm

patrick013 wrote:
Fri Apr 19, 2019 1:09 pm
pward wrote:
Fri Apr 19, 2019 10:30 am

Yes there is definitely some reversion to the mean over time. There are definitely different economic climates that favor growth and value respectively. Being able to actually time them however is not an easy task. I mean as you mentioned growth has out performed for years and years now. At what point has the low "factor" been low long enough to warrant swapping? If you swap too early then you wind up under performing. One could use a momentum strategy, but those are prone to whipsaws, head fakes, and lots of short term cap gains. It would also require an understanding of technical and quantitative analysis (which the average Joe does not have). About the only non-technical way an average person could implement the strategy sensibly would be to have a separate growth and value allocation and rebalance between them so you're buying low and selling high over time. Whether that would actually perform better than just holding a blend index I'm not sure. Might be worth doing some backtesting on for the fun of it.
https://kz1cba.bn.files.1drv.com/y4pl4z ... png?psid=1

The above file doesn't show growth vs. value but rather LC vs MC vs SC. So thru the years the returns vary with each year. Lengthy histories would probably show that for any factor clearly. Timing is just spending or rebalancing from the factor with the highest return this year.
I just did a rough backtest on portfolio visualizer, using large cap blend vs 50/50 large cap growth/value with 5% rebalance bands (to give momentum a bit more room to run than yearly rebalancing) and since 1972 that ended up with a .5% extra CAGR. Now that's only one country and one set of start/end date data points so it would definitely need a bit more research before saying the probability is that the strategy would continue to slightly out perform blend. But not too shabby on first glance. https://www.portfoliovisualizer.com/bac ... 0&total3=0

DonIce
Posts: 601
Joined: Thu Feb 21, 2019 6:44 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by DonIce » Fri Apr 19, 2019 1:54 pm

pward wrote:
Fri Apr 19, 2019 1:38 pm
I just did a rough backtest on portfolio visualizer, using large cap blend vs 50/50 large cap growth/value with 5% rebalance bands (to give momentum a bit more room to run than yearly rebalancing) and since 1972 that ended up with a .5% extra CAGR. Now that's only one country and one set of start/end date data points so it would definitely need a bit more research before saying the probability is that the strategy would continue to slightly out perform blend. But not too shabby on first glance. https://www.portfoliovisualizer.com/bac ... 0&total3=0
Don't think the result is significant, because if you just plug in 100% large cap value, you get about double the outperformance. Value simply did better since 1972. So the extra performance isn't from the 50/50 split and rebalancing alpha, but just from taking a bet on value (you can also verify this by turning off rebalancing and seeing that you get an identical result to within a few bps).

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

User avatar
nedsaid
Posts: 11817
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by nedsaid » Fri Apr 19, 2019 3:44 pm

Larry's analysis comparing Vanguard Indexes to DFA Funds implies that investors can capture these premiums:

Size factor: 1.10% for Small US Stocks, 2% for US Micro-Caps, 4.1% for International Developed Markets, and 4.0% for Emerging Markets.

Value factor: 0.6% for Large US Stocks, 0.8% for Small US Stocks, 1.8% for Large Developed Market Stocks, 0.7% for Small International Stocks, and 3.3% for Large Emerging Markets Stocks.

Not too surprising but if you pay 1.0% Assets Under Management Fee to an advisor to get access to DFA funds, your premiums will be pretty small. There are good ETF products that investors can buy that are very competitive with DFA in both price and performance. An informed do-it-yourself investor could take a crack at factor investing and have a good chance to succeed. Don't think I would pay an advisor for access to DFA.

Size and Value premiums are better with International which might imply US markets are more efficient.
A fool and his money are good for business.

User avatar
2pedals
Posts: 832
Joined: Wed Dec 31, 2014 12:31 pm

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by 2pedals » Fri Apr 19, 2019 4:08 pm

I don't have the willingness to concentrate my investments in market factors. If I did it could make me regret picking that factor or factor(s) over the next 10 to 20 years.

A big factor for my unwillingness is that I may not live long enough to see results perform better than the market.

User avatar
nisiprius
Advisory Board
Posts: 38469
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: The Factors That Plague Factor Investing

Post by nisiprius » Fri Apr 19, 2019 4:23 pm

In Stay the Course (2019), pp. 93-4, John C. Bogle notes that
In 1989, I gave a speech... promising that the creation of separate growth and value index funds "awaits only the development of soundly-constructed... index.

Standard & Poor's Corporation introduced both indexes in 1992. We wasted little time. On November 2, 1992, we created two new index funds, separating the stocks of the S&P 500 into growth and value portfolios.

My concept for how investors should approach these two funds was simple. Investors who were accumulating assets should consider Vanguard Growth Index Fund.... When they reached retirement and began the distribution phase of their investment lives, investors would seek higher dividend income and lower volatility. Voila! Vanguard Value Index Fund....

In my annual reports to shareholders, I repeatedly warned that switching between the two funds based on expectations about short-term returns of these two market segments was likely to be counterproductive....

My warnings fell largely on deaf ears.... over the quarter century that followed their creation... investors in both funds too often switched between the two series and earned far lower returns than the funds themselves.
From this, I observe that
  • Vanguard actually beat DFA to market with a value factor fund.
  • John C. Bogle's views on the uses of factors correspond to pre-Fama-French understandings of "growth" and "value."
  • His explanation of the use of "value" is closer to "dividend stock investing" than "factor-based investing."
  • He specifically warned against trying to market-time value and growth.
  • Despite his comments about lower volatility in value, the two funds have been extremely close in return, volatility, maximum drawdown, and risk-adjusted return. This is particular interesting in that the biggest ups and downs occurred at different time so the two curves are fairly different from each other.
Portfolio 1 is 100% Value Index.
Source

Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

Post Reply