Is Low-Vol Anomaly really explained by Value factor?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Thu Apr 18, 2013 1:13 pm

My understanding is that DFA's conclusion about the low-volatility anomaly is that it is explained by the value factor. That is, a factor regression on a low-vol portfolio will reveal a high value loading. However, Falkenstein has this interesting blog post where he demonstrates that the value loading in low-vol is highly inconsistent. He says:

Secondly, when I look at the Minimum Variance and Low Beta portfolios I have created, I do not see a consistent value loading. Looking at the MVP, which I have data from 1998 to 2011, it estimates value beta of 0.34 using monthly data. Using daily data from July 2000 through July 2011, the value beta is 0.24. These are around what Shah gets. Yet it floats around like an incidental symptom, not something fundamental.

Followed by a chart showing that the time series of low-vol value factor is all over the place.

The blog post is here:
http://falkenblog.blogspot.com/2011/09/ ... -play.html

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by EDN » Thu Apr 18, 2013 1:20 pm

Yes, value and TERM. These guys hate it when they are trying to sell a new strategy and find out combos of old strategies work better. Happens all the time. They just hope there are enough uninformed to pay the bills.

Eric

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Thu Apr 18, 2013 1:40 pm

Eric, this is not a DFA bashing, so you can relax. Will you please back up your claims with evidence?

Did you look at the chart in the post showing the inconsistency of the value factor?

What exactly is the term loading on a low-vol portfolio such as SPLV?

Another poster has shown that a 5 factor regression of VPU explains 48% of the returns. Any ideas where the rest comes from?

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Thu Apr 18, 2013 1:47 pm


Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Thu Apr 18, 2013 2:03 pm

Thanks Larry. I'm familiar with your article and the assertion that it is merely value and term loading. I'm pointing to an analysis that disputes that assertion and would be grateful for any comments or analysis on it. e.g. Why do you suppose the value factor in low vol is so inconsistent? What is the term loading of low-vol indicies and does it fluctuate? I am not presently equipped to run a 5 factor regression on a low-vol index, but if anyone else were, it would be interesting to see the results, especially the consistency/inconsistency of the factor loadings. Thanks.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by EDN » Thu Apr 18, 2013 2:06 pm

Jebediah wrote:Eric, this is not a DFA bashing, so you can relax. Will you please back up your claims with evidence?

Did you look at the chart in the post showing the inconsistency of the value factor?

What exactly is the term loading on a low-vol portfolio such as SPLV?

Another poster has shown that a 5 factor regression of VPU explains 48% of the returns. Any ideas where the rest comes from?
No one said anything about DFA, that's a strawman.

On VPU, you must have missed my comments on that. Its a sector fund! Not a high dividend strategy (that holds all sectors). FF models work for diversified portfolios, not sectors with loads of unsystematic (diversifiable) risk.

And part of the low vol story is just that -- a huge overweight to Utilities, as much as about 40% in the 80s. So if Utilities do well, low vol looks good. If not, it doesn't. I don't know why I'd expect utilities (or any other sector) to have better risk adjusted returns than the market, so that is a non-starter for me.

What else? Yes, there is exposure to the value premium (of which some comes from Utility exposure). During years where growth beats value, low vol underperforms the market 75% of the time by about 5%. During years where value beats growth, low vol outperforms the market 70% of the time by over 3%. A 3F regression on low vol produces an R^2 of almost 80% (not higher likely because of extreme industry concentration), and a positive HmL coefficient that is statistically significant with a huge t-stat.

Finally, of course, low vol stragies have less exposure to the market factor (as do balanced stock & bond portfolios). The traditional way to lower risk is dilute stocks with bonds, not buy safer stocks. What low vol accomplishes is what we've been using for years -- stock allocations tilted to value with less than 100% in equities.

However, when you roll your own, you can choose the degree of stock exposure, and value/size exposure, as well as the right amount of interest rate sensitivity for each plan. Farming it out to a low vol ETF and not understanding where returns come from is a treacherous decision.

Low vol has gained traction from proponents who want to suspend asset pricing research that has taken place for the last 4 decades. If we go back to a CAPM, one factor world, low vol looks great, as does just about everything else!

Eric

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Thu Apr 18, 2013 2:30 pm

Thanks Eric.

I have two questions:

1. Why exactly would it be treacherous to invest in a low vol ETF whether or not you know about the value and term loadings? What is the hidden treachery in low-vol? I'm looking at, say,

http://www.bogleheads.org/wiki/Low_Vola ... ex_returns

and not following that logic.

2. Can you provide time series of the individual factor loadings that you cite as contributing to 80% of returns?

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Thu Apr 18, 2013 2:44 pm

Furthermore, look at the returns in the low-vol wiki page and compare it to the returns of a Large Value index. They are not similar, the Value index has a much higher SD (and worse returns). Why do you suppose that is?

User avatar
Ketawa
Posts: 1955
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Ketawa » Thu Apr 18, 2013 2:55 pm

I didn't look at the exact numbers, but here is my explanation. Low volatility is not only value, it's also lower beta and higher term. So, it would outperform over the last few decades compared to a plain large value index, depending on the time period. A better comparison would be low volatility vs (value + intermediate bonds). Some advantages of the latter choice are greater diversification and better tax location of assets.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by EDN » Thu Apr 18, 2013 3:07 pm

Why would it be dangerous to not know where your returns are coming from? Well, during a period where the strategy isn't working (because the return on beta is significant, or the value premium is negative), you might be fooled into thinking this anomaly no longer works. But if you understand the underlying source of returns, then you'll better understand why what is happening is happening, both good and bad.

The stats were from the time frame of 1968-2010, the DFA paper referenced in the link in your OP.

In the wiki, I took a glance at the MSCI minimum volatility series and its annualized return from 2000-2012 of 4% per year. A 75% MSCI 750 Value, 25% Barclays Treasury Index mix earned +5.4%, and had less downside risk in 01-02 and 08. Returns will of course be a bit different from year to year because low vol is heavily concentrated in utilities and traditional value indexes have better sector diversification.

People are trying to avoid facing the reality of low bond yields and therefore these low vol strategies look like a better way to load up on stocks and get stock-like returns without the cost of low bond yields. But it ain't so. Returns are a function of industry concentration, and exposure to value premium and interest rate exposure that offset lower beta.

Eric

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Thu Apr 18, 2013 5:27 pm

Yes, the theory of term + value loadings makes sense. However, can we verify this explanation with data? Namely, what does the time series of value and term loadings for low-vol look like?

Also, I find it curious that EEMV (iShares EM low vol ETF) is, according to Morningstar, a large growth fund. Their growth:value score is 34:18, basically 2:1. Of course, MS could just have it wrong.

Also, Eric if you don't mind, for the sake of clarity, may we set aside all the stuff about everybody's bad motivations for buying and selling these ETFs? I'm really just interested in the pure analytics of the investment.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Thu Apr 18, 2013 6:37 pm

One more thing to remember
Vol strategies load heavily on certain industries, like utilities--which of course have benefited from falling rates--question is would they perform well in rising rates>
Larry

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Thu Apr 18, 2013 6:59 pm

EDN wrote:Why would it be dangerous to not know where your returns are coming from? Well, during a period where the strategy isn't working (because the return on beta is significant, or the value premium is negative), you might be fooled into thinking this anomaly no longer works. But if you understand the underlying source of returns, then you'll better understand why what is happening is happening, both good and bad.

The stats were from the time frame of 1968-2010, the DFA paper referenced in the link in your OP.

In the wiki, I took a glance at the MSCI minimum volatility series and its annualized return from 2000-2012 of 4% per year. A 75% MSCI 750 Value, 25% Barclays Treasury Index mix earned +5.4%, and had less downside risk in 01-02 and 08. Returns will of course be a bit different from year to year because low vol is heavily concentrated in utilities and traditional value indexes have better sector diversification.

People are trying to avoid facing the reality of low bond yields and therefore these low vol strategies look like a better way to load up on stocks and get stock-like returns without the cost of low bond yields. But it ain't so. Returns are a function of industry concentration, and exposure to value premium and interest rate exposure that offset lower beta.

Eric
Low volatility and minimum volatility are not quite the same. Low vol is very heavy in utilities and the like. Min vol needs to abide by pre set constraints ... the end result is more of a small tilt into utilities.

Data since 1988 http://www.msci.com/resources/factsheet ... -index.pdf
Maximize Diversification - Minimize Costs - Avoid Lotteries

stlutz
Posts: 4845
Joined: Fri Jan 02, 2009 1:08 am

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by stlutz » Thu Apr 18, 2013 8:30 pm

Low vol has gained traction from proponents who want to suspend asset pricing research that has taken place for the last 4 decades. If we go back to a CAPM, one factor world, low vol looks great, as does just about everything else!
Eric--actually if you read Falkenstein's stuff he bashes the CAPM more than anyone I've seen.

In terms of the history of the theory, CAPM was developed because of the hypothesis that higher risk=higher expected returns. The low vol. proponents and the value titling proponents both agree that CAPM has failed to explain returns however--more beta has not meant higher returns.. The question is where one goes from there.

Fama & Co. stuck to the view that higher returns can only come from higher risk, and went in search of other ways to define risk. The way this was done was to look for strategies that provided higher historical returns (value, size) and then develop ways to explain those as risk. Thus, the theory that more risk=more expected return held.

The low. vol. advocates have gone back and taken a different approach as result of the failure of CAPM. Beta and volatility are obvious and quite good way to define risk in the stock market, yet more risk by this definition does not lead to higher returns. As such, they have concluded that the higher risk=substantially higher expected returns assumption is false. If it is false, then why take a lot of risk?

stlutz
Posts: 4845
Joined: Fri Jan 02, 2009 1:08 am

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by stlutz » Thu Apr 18, 2013 8:42 pm

Also, I find it curious that EEMV (iShares EM low vol ETF) is, according to Morningstar, a large growth fund. Their growth:value score is 34:18, basically 2:1. Of course, MS could just have it wrong.
Morningstar is right. The ishares website shows the same thing. As you have noted, low vol/min. var fluctuates around between "value" and "not value." Right now it's in the "not value" area.

I have seen somewhere a study that showed that low. vol. outperforms when it has P/B ratios below the market and underperforms when the P/B is higher than the market, which makes logical sense to me. As such, based on current valuations I personally expect low. vol. to underperform the market over the next 5-10 years or so.
In the wiki, I took a glance at the MSCI minimum volatility series and its annualized return from 2000-2012 of 4% per year. A 75% MSCI 750 Value, 25% Barclays Treasury Index mix earned +5.4%, and had less downside risk in 01-02 and 08. Returns will of course be a bit different from year to year because low vol is heavily concentrated in utilities and traditional value indexes have better sector diversification.
First, as others have noted, most minimum variance strategies actually impose sector contraints to avoid the situation you've mentioned.

Second, the strategy you've outlined is not really consistent with a minimum variance approach either. For them, the goal is not to abandon asset class diversification. Instead, one should only take risks that have actually paid off historically with the assets classes one picks. So, you'd construct your minimum variance stock portfolio and then mix in a bond portfolio. One historically only gets paid for going out 2-3 years in duration and taking a limited amount of credit risk, so you probably go with a short-term investment grade bond fund.

Finally, most advocates of minimum variance actually believe in value tilting, so it's not an either/or. They view value stocks as generally being less risky than growth stocks. As such, value tilting is a way to reduce risk, not increase it.
Last edited by stlutz on Thu Apr 18, 2013 8:43 pm, edited 1 time in total.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Thu Apr 18, 2013 8:43 pm

well we know high vol doesn't work, that is the lottery effect at work and also the inability to use margin or unwillingness leads those that want more beta to buy high beta stocks, thus explaining the failure of high volatility stocks like small growth. But low vol seems to be well explained by other issues,
Larry

stlutz
Posts: 4845
Joined: Fri Jan 02, 2009 1:08 am

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by stlutz » Thu Apr 18, 2013 8:46 pm

But low vol seems to be well explained by other issues
To get to the practical level, why should I care how much one return stream covaries with another one? If I can get the market return and take less risk than the market overall, who cares whether some academic or fund company selling a different strategy claims to "explain" the result?

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Fri Apr 19, 2013 7:03 am

stlutz wrote: Eric--actually if you read Falkenstein's stuff he bashes the CAPM more than anyone I've seen.

In terms of the history of the theory, CAPM was developed because of the hypothesis that higher risk=higher expected returns. The low vol. proponents and the value titling proponents both agree that CAPM has failed to explain returns however--more beta has not meant higher returns.. The question is where one goes from there.

Fama & Co. stuck to the view that higher returns can only come from higher risk, and went in search of other ways to define risk. The way this was done was to look for strategies that provided higher historical returns (value, size) and then develop ways to explain those as risk. Thus, the theory that more risk=more expected return held.

The low. vol. advocates have gone back and taken a different approach as result of the failure of CAPM. Beta and volatility are obvious and quite good way to define risk in the stock market, yet more risk by this definition does not lead to higher returns. As such, they have concluded that the higher risk=substantially higher expected returns assumption is false. If it is false, then why take a lot of risk?
Great summary :!:

I get how FF can suggest the size premium ... smaller stocks may return more because of more risk. But I do not get ... and hope to understand/learn ... the value premium ... do value stocks have greater risk? :?: Simba data (72-2012) on std. deviation suggests otherwise. IDK. I guess it matters how you define "risk"
Maximize Diversification - Minimize Costs - Avoid Lotteries

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Fri Apr 19, 2013 7:38 am

Image
http://www.msci.com/resources/factsheet ... ndices.pdf

If you are going to be globally diversified ... value weighting earned you a couple of percentage points greater return last 24 years and did have higher risk (question from my previous post)...

The Risk Weighted Index & Quality Index are interesting ... but one cannot invest in it w/ an ETF

The min vol point is what peaks my interest ...after reading a lot on this topic (I own a lot of ACWV combined with the 3 fund) - been happy with it ... but believe rethinking any strategy from time to time is a useful exercise.
stlutz wrote:
But low vol seems to be well explained by other issues
To get to the practical level, why should I care how much one return stream covaries with another one? If I can get the market return and take less risk than the market overall, who cares whether some academic or fund company selling a different strategy claims to "explain" the result?
+1 ... but I would like to know of flaws in my thinking, what happens if rates rise, am I really value tilting in disguise, etc.
Maximize Diversification - Minimize Costs - Avoid Lotteries

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Fri Apr 19, 2013 8:34 am

keep this in mind about value stocks
They typically have more leverage and thus benefit from inflation which reduces the real cost of debt,
However, if inflation becomes more than moderate and Fed steps in to drive real rates up to fight inflation then risks of value stocks shows up

As to risks of value stocks, here is a partial list I have posted many times
1.Baruch Lev and Theodore Sougiannis, “Penetrating the Book-to-Market Black Box,” Journal of Business Finance and Accounting (April/May 1999).
2.Robert F. Peterkort and James F. Neilsen, “Is the Book-to-Market Ratio a Measure of Risk?” Journal of Financial Research (Winter 2005).
3.Maria Vassalou and Yuhang Xing, “Default Risk in Equity Returns,” Journal of Finance (April 2004).
4.Xinting Fan and Ming Liu, “Understanding Size and the Book-to-Market Ratio: An Empirical Exploration of Berk’s Critique,” Journal of Financial Research (Winter 2005).
5.Howard W. Chan and Robert W. Faff, “Asset Pricing and the Illiquidity Premium,” The Financial Review (November 2005).
6.Charles Lee and Bhaskaran Swaminathan, “Price Momentum and Trading Volume,” Journal of Finance (October 2000).
7.Ralitsa Petkova, “Do the Fama-French Factors Proxy for Innovations in Predictive Variables?” Journal of Finance (April 2006).
8. Aydin Akgun and Rajna Gibson, “ Recovery Risk in Stock Returns,” Journal of Portfolio Management, Winter 2001.
9. Gerald R. Jensen and Jeffrey M. Mercer, “Monetary Policy and the Cross-Section of Expected Returns,” Journal of Financial Research, Spring 2002.
10.Gabriel Perez-Quiros and Allan Timmerman, “Firm Size and Cyclical Variations in Stock Returns,” July 1999.
11.Lu Zhang, “The Value Premium.” January 2002, http://papers.ssrn.com/sol3/papers.cfm? ... _id=351060
12.Joao Gomes, Leonid Kogan, and Lu Zhang, “ Equilibrium Cross-Section of Returns, March 2001. http://assets.wharton.upenn.edu/~zhanglu/
13. Moon K. Kim and David A Burnie, “The Firm Size Effect and the Economic Cycle,” Journal of Financial Research, Spring 2002.
14.Nai-fu Chen and Feng Zhang, Journal of Business, “Risk and Return of Value Stocks,” October 1998.
15. Clifford S. Asness, Tobias J. Moskowitz, and Lasse H. Pedersen∗Value and Momentum Everywhere, February 2009.
16. Joachim Grammig, “Creative Destruction and Asset Prices,” March 2011.
17. Jia Wang, Gulser Meric, Zugang Liu, and Ilhan Meric, “The Determinants of Stock Returns in the October 9, 2007-March 9, 2009 Bear Market,” The Journal of Investing (Fall 2011).
18. Nishad Kapadia, “Tracking Down Distress Risk,” May 2010
19. Angela J. Black, Bin Mao and David G. McMillan, “The Value Premium and Economic Activity: Long-run Evidence from the United States,” December, 2009.
20. Nicolae Garleanu, Leonid Koganz, and Stavros Panageas, “Displacement Risk and Asset Returns,” July 2008.

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by EDN » Fri Apr 19, 2013 10:16 am

stlutz wrote:
But low vol seems to be well explained by other issues
To get to the practical level, why should I care how much one return stream covaries with another one? If I can get the market return and take less risk than the market overall, who cares whether some academic or fund company selling a different strategy claims to "explain" the result?
I already explained this. It's not that one return stream covaries with another, it's that the strategy in question "low vol" is almost fully explained by exposure to priced sources of expected return. Knowing where returns come from help with expectations, and also allows one to decide for themselves if the bundled package is the best approach to achieving those results. For example, above, I showed why decomposing a portfolio into separate value stock and bond components (to get to a lower beta) and maintain same value exposure led to better results, would allow one to make allocation decisions per account (maybe bonds in IRAs, stocks in taxable), investment preferences (maybe fixed income should be tax-free bonds) and in the decumulation phase it is more efficient because you can explicitly target the sale of bond fund shares when stocks are down.

And you aren't taking "less risks", you are just taking less "market" risk, trading that for some value risk and interest rate risk.

Eric

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by EDN » Fri Apr 19, 2013 10:29 am

steve r wrote:
stlutz wrote: Eric--actually if you read Falkenstein's stuff he bashes the CAPM more than anyone I've seen.

In terms of the history of the theory, CAPM was developed because of the hypothesis that higher risk=higher expected returns. The low vol. proponents and the value titling proponents both agree that CAPM has failed to explain returns however--more beta has not meant higher returns.. The question is where one goes from there.

Fama & Co. stuck to the view that higher returns can only come from higher risk, and went in search of other ways to define risk. The way this was done was to look for strategies that provided higher historical returns (value, size) and then develop ways to explain those as risk. Thus, the theory that more risk=more expected return held.

The low. vol. advocates have gone back and taken a different approach as result of the failure of CAPM. Beta and volatility are obvious and quite good way to define risk in the stock market, yet more risk by this definition does not lead to higher returns. As such, they have concluded that the higher risk=substantially higher expected returns assumption is false. If it is false, then why take a lot of risk?
Great summary :!:

I get how FF can suggest the size premium ... smaller stocks may return more because of more risk. But I do not get ... and hope to understand/learn ... the value premium ... do value stocks have greater risk? :?: Simba data (72-2012) on std. deviation suggests otherwise. IDK. I guess it matters how you define "risk"
As I said before, the "low vol" camp wants us to live in the CAPM world, where anything with a higher standard deviation adjusted return is a free lunch and proof the market doesn't work. So either the market doesn't work, or your definition of how the market works (risk = standard deviation only) doesn't work. And it's probably the later. We knew the CAPM was flawed in the 70s. The non-behavioral camp of asset pricing researchers believes markets are in equilibrium, prices are mostly right, and expected returns differ from one security to the next in ways that we can quantify in very accurate ways, but that those expected returns are a function of exposure to some form of risk. It is true, the quantification of sources of expected return is a lot deeper than the quantification of the sources of those risks, but we do have a fairly good understanding of the risks, see Larry's 20 items above.

Very basically, it is easy to understand how standard deviation isn't a complete definition of risk. An investment that loses -2% every month is safer in terms of SD than one with greater variance but a +2% monthly return. Finally, I am always amazed when someone doubts where the risk in value stocks are. The purest, lowest priced companies obviously have the highest costs of capital, when they borrow money they are considered "junk" rated, certainly compared to growth companies like AAPL and GOOG. Some traditional value indexes with 50% value sorts do pick up some other types of companies -- regulated utilities, less distressed companies, but then they also have lower expected returns compared to a deeper value strategy. So it is all relative -- some "value" is lower risk, but also has lower returns.

Eric

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Fri Apr 19, 2013 11:42 am

EDN wrote: Very basically, it is easy to understand how standard deviation isn't a complete definition of risk. An investment that loses -2% every month is safer in terms of SD than one with greater variance but a +2% monthly return. Finally, I am always amazed when someone doubts where the risk in value stocks are ...
The returns of value and growth are ... in the long run ... similar .... with a slight higher performance for value (that is, it does not have -negative returns) ...
so value stocks have slightly higher returns and lower measured risk (std. dev.). FF suggests that the higher returns of value must come from greater risk ... I simply do not get this point. I guess we can agree to disagree.

On your larger point - value stocks have risks ... ABSOLUTELY. No one should think otherwise.

William Bernstein wrote of the risk of growth stocks versus value stocks a decade ago ... http://www.efficientfrontier.com/ef/902/vgr.htm. ... he basically concludes that both value and and growth stocks have their own unique set of risks ... clearly value investing is not risk free. In his view written just after the tech bubble burst, growth was riskier. I personally do not know enough to agree or disagree with this ... I read some of the papers Larry posted, I read some classic debates on value investing with him and another poster ... I just do not know one way or the other ...

-----

My interest in this thread has more to do with low vol investing ......

Scherer notes that some of the low vol premium can be explained by value. Actually, in his paper most (80+%) can be explained if and only if you add two additional factors ... Beta and Risk to the 3 factor model.

Clearly Beta and Risk are related to volatility ...

This post is about Low-Vol .... per M* ... low-vol investments like SPLV (least volatile S&P stocks) are tilted value stocks. So, I would expect yes ... low vol can be explained by value.

Personally, I own a minimum volatility index ACWV (and three fund portfolio w/ index funds). Other min vol investments include USMV (US) EEMV (Em Mkts.) These are not tilted value due to constraints. I wonder what Scherer would find if he tested it against min. vol. :?: IDK.

----

On the post of 75% MSCI Value and 25% Treasuries since 2000 ... that finding does not move me. One could have been 100% treasuries and done even better!
Using Simba data for LCV and directly comparing it to low vol or min vol ... paints a different picture ... but, I one decade of data or so should be taken with a grain of salt.

I guess IF Low Vol investing lead to less bonds during the last decade or so ... that would have been a mistake.

----

I am really enjoying this post ...it is challenging my thinking ... thanks to all. :happy
Maximize Diversification - Minimize Costs - Avoid Lotteries

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Fri Apr 19, 2013 2:21 pm

steve r wrote: The returns of value and growth are ... in the long run ... similar .... with a slight higher performance for value (that is, it does not have -negative returns) ...
so value stocks have slightly higher returns and lower measured risk (std. dev.). FF suggests that the higher returns of value must come from greater risk ... I simply do not get this point.
Is it true that value has lower SD than growth? This is an interesting assertion. Larry or Eric care to comment?

For the last time I'll beg for someone to post 5 factor load time series of low vol or min var indicies.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Fri Apr 19, 2013 2:33 pm

Nope as I showed many times, at least if you go back as long as we have data for
The problem is people ignore the Great Depression when the risks of value stocks showed up, to show value has not been more volatile. But that IMO is wrong way to look at things

7/26-2/13 SDs annualized
LG 18.7 vs 26.2 for LV, so clearly value much more volatile and you don't have the data issues you might have with the small stocks
SG 28 vs 29.9 for SV

If look at 1950 onward get same result for large but smaller gap but reverse for small
LG 15.8 vs 18.3 for LV and SG 23 and SV 20.7

Which is why SG is called the black hole of investing

Larry

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Fri Apr 19, 2013 2:34 pm

Jebidiah

I think this is what you are looking for (see page 14):
http://www.northinfo.com/documents/391.pdf

As far as the standard deviation of value ... calculated with Simba data
http://www.bogleheads.org/forum/viewtopic.php?t=2520

A case is made that this is not an accurate measure of risk ....
Last edited by steve r on Fri Apr 19, 2013 2:36 pm, edited 1 time in total.
Maximize Diversification - Minimize Costs - Avoid Lotteries

stlutz
Posts: 4845
Joined: Fri Jan 02, 2009 1:08 am

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by stlutz » Fri Apr 19, 2013 2:35 pm

Eric: Just to make sure I understand you correctly, do you think that low volatility/minimum variance portfolio are more, equally, or less risky than that market overall?

EDN
Posts: 528
Joined: Sat Dec 29, 2012 7:55 pm
Location: Oklahoma City, OK

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by EDN » Fri Apr 19, 2013 2:40 pm

stlutz wrote:Eric: Just to make sure I understand you correctly, do you think that low volatility/minimum variance portfolio are more, equally, or less risky than that market overall?
Different risky. If you can afford to take value and term risk, but not as much volatility, it's safer. If you are OK with volatility, but business cycle or inflation risk (which would hurt value and term), it is riskier.

Jeb -- something like 0.7 beta, 0.0 size, 0.3 value, 0.5 term, and 0.1 credit. R^2 north of 80%

Eric

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Fri Apr 19, 2013 4:19 pm

steve r

Thanks for links. I was also looking at Scherer's 2010 ssrn paper and found it a bit hard to follow. He provides tables, but I still wonder about the factor time series. If these loadings wander significantly then it's highly misleading to peg them at their mean and call it case closed.

Regarding the SD of Value in simba data 1972 - present, you are right: value has lower SD than growth. Here are the numbers:

1972 - Present
LCV: SD = 17.97
TSM: SD = 18.18
LCG: SD = 20.68

Larry I'm not sure how you square this with your story above. 40 years is a long time for it to not hold true. If we say that value is positively correlated with volatility (has higher SD), then we are counting on the term premium (and it's unexpected outperformance) to explain the low vol anomaly. If value is negatively correlated with volatility, then that makes it easier to call the low vol anomaly a value story. But then why can't low vol (with whatever behavioral explanations behind it) be the thing that explains value as opposed to the other way around?

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Fri Apr 19, 2013 4:27 pm

Jebediah wrote:steve r

Thanks for links. I was also looking at Scherer's 2010 ssrn paper and found it a bit hard to follow. He provides tables, but I still wonder about the factor time series. If these loadings wander significantly then it's highly misleading to peg them at their mean and call it case closed.

Regarding the SD of Value in simba data 1972 - present, you are right: value has lower SD than growth. Here are the numbers:

1972 - Present
LCV: SD = 17.97
TSM: SD = 18.18
LCG: SD = 20.68

Larry I'm not sure how you square this with your story above. 40 years is a long time for it to not hold true. If we say that value is positively correlated with volatility (has higher SD), then we are counting on the term premium (and it's unexpected outperformance) to explain the low vol anomaly. If value is negatively correlated with volatility, then that makes it easier to call the low vol anomaly a value story. But then why can't low vol (with whatever behavioral explanations behind it) be the thing that explains value as opposed to the other way around?
In fairness to Larry ... William Berstein found likewise that during the Great Depression value lost more .. value down 89 percent verus 82 for growth. Berstein shows the trend I found goes back at least as far as 1963.
http://www.efficientfrontier.com/ef/902/vgr.htm
This is what Eric states by business cycle ... with the Great Depression being the ultimate risk ...or "fat tail" risk.

That said, I do wonder why can't low vol explain value as opposed to the other way around ...


For my own investing, I stick with min vol (not low vol) which does not tilt. I do not get the value debate well enough to be comfortable with more value stocks. I may miss the value premium ... but am not comfortable investing in something I do not get.
Maximize Diversification - Minimize Costs - Avoid Lotteries

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Fri Apr 19, 2013 4:34 pm

I only pointed out Steve that to show value less volatile you have to EXCLUDE the very period when you would expect value to be the most risky.
And as I showed it is still there in LARGE value, which means likely overall that it's there for all value stocks in aggregate
So you might say, well value is perhaps not as risky as long as we don't get another depression. If only a cow had balls thinking
Assets that do badly in the worst of times should carry large risk premiums
Larry

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Fri Apr 19, 2013 5:00 pm

larryswedroe wrote:.... If only a cow had balls thinking Assets that do badly in the worst of times should carry large risk premiums
Laughing out loud at the way you explain things ..

Berstein points out your comment on the Great Depression ... but then adds:
"On the other hand, from 1973 until 1974, the reverse occurred, with large growth stocks losing 45%, versus only 26% for value. Similarly, from April 2000 to July 2002, large growth lost 44% versus only 27% for large value."

In 2008 LCV lost 36%, LCG lost 38, but more important for this thread on Low Vol ... both low vol and min vol (US) lost 28% :!: (telling but not enough data to be sure)
Maximize Diversification - Minimize Costs - Avoid Lotteries

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Fri Apr 19, 2013 5:14 pm

Larry

The data do not follow. At best the results are mixed.

Quote from the Bernstein post linked by steve:

The record in this regard is mixed. During the Great Depression, it was indeed the case: from September 1929 until June 1932, Ken French’s data show that large growth stocks lost "only" 82% of total return versus a loss of 89% for large value. Similarly, for the 12 months from October 1989 to September 1990, large growth and value lost 7% and 19%, respectively.

On the other hand, from 1973 until 1974, the reverse occurred, with large growth stocks losing 45%, versus only 26% for value. Similarly, from April 2000 to July 2002, large growth lost 44% versus only 27% for large value.



In the 08 downturn value lost a shade more than growth, but not much. Overall, considering 40 years of lower SD for value and the relative safety of value during the 73-74 and 2000-02 bear markets, I just don't see how you can claim value is conclusively riskier than growth, especially in bear markets.

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Fri Apr 19, 2013 5:24 pm

steve, beat me to it :)

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Fri Apr 19, 2013 5:26 pm

1930-39 the value premium was -1.65 percent a year with annual SD of almost 20 a year for the premium, which is about the SD of the stock market, not the ERP

The recession in 73-4 was not a deflationary recession, but an inflationary one where the higher leverage of value stocks is helped by the higher inflation

Check any financial crisis period and you'll find that there is a strong tendency for value to get hit, until crisis over, like summer 98 or during S&L crisis

It's just that we have not had many DEFLATIONARY crisis--doesn't mean risk not there, that's an issue of triumph of optimists

Note I do happen to be in the camp that the value premium is not ALL risk, it's part risk and big chunk of behavioral with SG anomaly and lottery effect a big part of it (it's why you see the much bigger value premium in small than large)

Larry

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Fri Apr 19, 2013 5:42 pm

Thanks for entertaining the discussion, Larry. Do you think that growth has an inflationary recession risk premium that is masked by the negative behavioral aspect?

Also, the higher one's value exposure, the more term risk they should take to hedge deflation risk?

stlutz
Posts: 4845
Joined: Fri Jan 02, 2009 1:08 am

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by stlutz » Fri Apr 19, 2013 7:26 pm

The record in this regard is mixed. During the Great Depression, it was indeed the case: from September 1929 until June 1932, Ken French’s data show that large growth stocks lost "only" 82% of total return versus a loss of 89% for large value
One big caveat on this data is that this is a comparision between a portfolio with a lot of companies vs one with very few. For example, in Dec. 1932, French assigns 154 companies to "large growth" and 32 companies to "large value". Other things being similar, one should always expect a portfolio with a lot of stocks to be less risky than a portfolio with very few. This really has nothing to do with the riskiness of value vs. not value.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Fri Apr 19, 2013 7:27 pm

Jebediah
First did bit more work on the value premium during Great Depression
From 1/30-5/32 it was -35%

Then from 9/33-3/35 it was -46%

Then when Fed tightened again in 37 sending us into another recession from 4/37 -5/40 it was -37.6%

Those are all cumulative not annualized figures. But to me that makes clear case of how serious risks value stocks have.

Yes if you own value stocks can take more term risk IF stick with the safest bonds which benefit from deflation, but you would not want to own longer term high yield

Sorry but don't understand the other question about growth

Best wishes
Larry

stlutz
Posts: 4845
Joined: Fri Jan 02, 2009 1:08 am

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by stlutz » Fri Apr 19, 2013 7:41 pm

This post is about Low-Vol .... per M* ... low-vol investments like SPLV (least volatile S&P stocks) are tilted value stocks
Note that if you look at the various "price-to-X" ratios, SPLV has higher ratios than the S&P 500. Right now, whether you do low vol or min. var., you are tilting toward growth, not value.

If the low volatility stocks fall more into the growth camp, then it probably is accurate to state that *right now*, value is riskier than growth.

stlutz
Posts: 4845
Joined: Fri Jan 02, 2009 1:08 am

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by stlutz » Fri Apr 19, 2013 8:24 pm

As I said before, the "low vol" camp wants us to live in the CAPM world
Wrong! The low vol camp believes that CAPM is false to a much larger extent than F/F do. It's simply not the case that there are only two ways to view how markets work, CAPM or F/F; just because you aren't in the F/F camp doesn't mean you believe that CAPM is true. CAPM is a very simple way to express that higher risk=higher expected return. The low vol. camp argues that this is false.

... where anything with a higher standard deviation adjusted return is a free lunch and proof the market doesn't work. So either the market doesn't work, or your definition of how the market works (risk = standard deviation only) doesn't work. And it's probably the later. We knew the CAPM was flawed in the 70s. The non-behavioral camp of asset pricing researchers believes markets are in equilibrium, prices are mostly right, and expected returns differ from one security to the next in ways that we can quantify in very accurate ways, but that those expected returns are a function of exposure to some form of risk.
The problem with terms like "free lunch", "the right price", "works vs. doesn't work" is that they are heavily value-laden as opposed to being purely empirical.

Efficient markets theory assumes the following: a) People are rational, at least in aggregate. b) A rational investor demands a lot of additional return in exchange for taking a little bit of additional risk. c) therefore, riskier investments have higher expected returns.

If you come to a conclusion that is contrary to these axioms, then you must believe that markets are irrational, don't work, and give free lunches. Some people do believe all of these things. I don't. However, the value judgments are largely irrelevant--they tell you more about the views of the person opining than how investments perform. For this discussion, what matters is how markets actually function, not whether one likes the way that they do.

The low vol. argues that a look at the data tells you that many investors are willing to pay a premium for risk. There is no historical relationship between stock price volatility and long-term return (obviously there is a short-term one). The possibility of getting larger returns is very attractive to investors. As such, the price of such assets gets bid up. This means that a significant increase in the upside potential of an investment goes along with a significant increase in the downside potential as well. Potential returns are higher; expected returns are not.

I can argue either way whether this is rational or irrational, but who cares what I think about that?

Very basically, it is easy to understand how standard deviation isn't a complete definition of risk. An investment that loses -2% every month is safer in terms of SD than one with greater variance but a +2% monthly return.
Sure, but this doesn't exist. If you know an investment is going to go down 2% every month, then the price already would have dropped to reflect where investors think it will be in the future. The price then fluctuates because of changing opinions about what the future looks like.
Finally, I am always amazed when someone doubts where the risk in value stocks are. The purest, lowest priced companies obviously have the highest costs of capital, when they borrow money they are considered "junk" rated
Stocks in companies with junk-rated debt have, historically, performed poorly. If these are value stocks, they lower the value premium--take them out and the premium would be higher. Has there actually been analysis to show that companies with junk-rated debt usually fall in the value bucket? In a price-to-book world, such companies can actually either get discarded from the analysis because they have negative book value or they may be growth because their book value is very small.
So it is all relative -- some "value" is lower risk, but also has lower returns.
Okay, but this creates a problem.

-We know that lower vol does not mean lower returns.
-We know that value means higher returns
-If low vol. value stocks lowers returns, then low. vol. has to pick up the lost return elsewhere to get back to market-equaling returns.
-That can only mean that the source of the low vol "anomaly" is low vol. growth stocks! Just the opposite of what is being argued here.

Where this gets me to is this: If Low Vol=Value Tilting, then value stocks=low vol tilting. People can make up their own minds as to whether they think "volatility" has anything to do with "risk."

Jebediah
Posts: 515
Joined: Tue Aug 28, 2012 9:19 pm
Location: Denver, CO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by Jebediah » Fri Apr 19, 2013 11:41 pm

stlutz wrote: -That can only mean that the source of the low vol "anomaly" is low vol. growth stocks! Just the opposite of what is being argued here.

Where this gets me to is this: If Low Vol=Value Tilting, then value stocks=low vol tilting. People can make up their own minds as to whether they think "volatility" has anything to do with "risk."
I think the argument is that term risk is the missing piece of the puzzle. But it seems evident that low vol at least drifts between growth and value, and conversely growth and value have variable levels of volatility. Thus it is not true to say value is higher risk or higher vol, rather that it is sometimes higher risk, higher vol, and one of those sometimes is during a deflationary recession.

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Sat Apr 20, 2013 7:36 am

larryswedroe wrote: It's just that we have not had many DEFLATIONARY crisis--doesn't mean risk not there, that's an issue of triumph of optimists
This of course led to thoughts of what happened in Japan - who did experience deflation.
MSCI Japan Growth versus Value

July 1989 to July 1992 ... Growth lost 56 percent and Value lost 45.

Two years later ... Value was higher than in 1989, Growth still down 25%.

Fastfoward to March 2013

Value up a couple percent, Growth down 75. :oops:
Maximize Diversification - Minimize Costs - Avoid Lotteries

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Sat Apr 20, 2013 8:14 am

steve
What you had in Japan was a bubble in growth that burst, just as it did in US.
BTW- the research on financial distress shows that value stocks that are risky for financial distress reasons do outperform, but only to a point, the worst offenders, with the most financial distress risk actually underperform. My explanation is the lottery effect at work there at the deep end/
Best wishes
Larry

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Sat Apr 20, 2013 8:18 am

stlutz wrote: The low vol camp believes that CAPM is false to a much larger extent than F/F do.
I guess I am in a "camp" ... something for me to be aware of as any camp likely has its pitfalls.
stlutz wrote: The low vol. argues that a look at the data tells you that many investors are willing to pay a premium for risk. There is no historical relationship between stock price volatility and long-term return (obviously there is a short-term one). The possibility of getting larger returns is very attractive to investors. As such, the price of such assets gets bid up. This means that a significant increase in the upside potential of an investment goes along with a significant increase in the downside potential as well. Potential returns are higher; expected returns are not.
+1 ... but I have to be honest ... I am uncomortable with this behavioral explaination for the returns found in low. vol. investing. I tilt low vol because I see no upside to high vol. investing ... I choose to play this game as little as possible. I realize this lack of comfort runs counter to the data in front of me!

Where this leads me is (I am particularly curious what Stlutz thinks as you get low vol concepts) (I am in this space):

I am convinced low vol investing is less volatile. I am convinced that this will be delivered going foward.
I am not convinced that low vol investing will provide market like return going foward ... it may / may not ...
Going foward, it is unlikely that risk adjusted returns of low vol will be unfavorable in the long run.

My main cocerns are:

Staying the course if a repeat of the late 90s occur. (value investors have a similar concern). I am off to a spectacular start, so that helps.

Not really tested under multiple business cycle environments. In particular a high interest rate environment. My guess is people are moving to this space because bond returns are perceived as too low. Will this space be attractive if you can reduce volatility with high yielding bonds? IDK.

Others?
Maximize Diversification - Minimize Costs - Avoid Lotteries

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Sat Apr 20, 2013 8:29 am

larryswedroe wrote:steve
BTW- the research on financial distress shows that value stocks that are risky for financial distress reasons do outperform, but only to a point, the worst offenders, with the most financial distress risk actually underperform. My explanation is the lottery effect at work there at the deep end/
Best wishes
Larry
Interesting ... and it makes sense. It also fits the idea that during a severve recession such companies are more likely to underperform ... thus term risk. If memory serves me correctly, the government in Japan did everything it could to not allow companies to fail. That may be another reason why their expereince was different. Also, Japan's deflation was not global like the deflation of the 1930s. As long as distressed companies could export ... they could survive.

I particulary like the observation of the lottery effect at the low priced end as well ... the low price of "penny" stocks make them "value"

I would imagine such stocks would be volatile ... and not picked up in low vol indexes once they reset (which is twice a year).
Maximize Diversification - Minimize Costs - Avoid Lotteries

stlutz
Posts: 4845
Joined: Fri Jan 02, 2009 1:08 am

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by stlutz » Sat Apr 20, 2013 10:56 am

I guess I am in a "camp" ... something for me to be aware of as any camp likely has its pitfalls.
Ha ha! And what I meant by that were the people who've written papers on the topic and who tend to cite one another. Off the top of my head, names to google to read more papers would be Falkenstein, Haugen & Baker, and Clarke & De Silva.
+1 ... but I have to be honest ... I am uncomfortable with this behavioral explanation for the returns found in low. vol. investing. I tilt low vol because I see no upside to high vol. investing ... I choose to play this game as little as possible. I realize this lack of comfort runs counter to the data in front of me!
I think we're kind of in the same place on this. My personal view is that the future is generally not much like the past, and therefore the past is not that great of a guide for which stocks will do best in the future. That said, we really don't have anything better than the past to go off of. When people on message boards like this debate "explanations" for data, what is really going on is trying to find something of a guarantee that the past will be like the future, thus turning what they should view as a 60/40 probability into a 90/10 one. In the end, I am 100% sure that the market couldn't care less about how confident an investor is in his expectations, however. :)
My main concerns are:

Staying the course if a repeat of the late 90s occur. (value investors have a similar concern). I am off to a spectacular start, so that helps.

Not really tested under multiple business cycle environments. In particular a high interest rate environment. My guess is people are moving to this space because bond returns are perceived as too low. Will this space be attractive if you can reduce volatility with high yielding bonds? IDK.
Going in reverse order, for what it's worth, the historical data on low vol. is robust across markets and business cycle environments. As has been noted, people have known that CAPM didn't explain things very well decades ago; it's more recent, however, for the investment community to seize on this and start offering products based upon it.

Low vol. stocks and bonds do in some ways compete for the same money. How low vol. would do in a rising rate environment would depend on the reason for the rising rates--such an increase has often been bad for stocks overall, not just one type of stock.

On staying the course, this is an issue for all investors, not just those "tilt" in a particular direction. Investor behavior is the risk that kills most of us more than the particular type of stocks we own. That is why I tend to advocate 60/40 portfolios for young investors even though I do think that stocks are highly likely to beat bonds over time--I think people are more likely to stick with this type of portfolio than a 100% stocks one.

On expectations going forward, I predict that low vol. will underperform the market over the intermediate term. This space has become quite popular--the low vol. ETFs have been a huge success in terms of bringing in assets. Right now I think the sweet spot is in mid vol.--cutting off the highest and lowest risk names.

What have I done to my personal portfolio based on this expectation? Nothing. I believe in staying the course more than my forecasts. :D

Note: I have a token (~.5% of my portfolio) position in USMV, so I'm not a big minimum variance investor in practice.

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Sat Apr 20, 2013 1:02 pm

Thanks Stultz
Maximize Diversification - Minimize Costs - Avoid Lotteries

winguy
Posts: 41
Joined: Sun Nov 06, 2011 11:55 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by winguy » Mon Apr 29, 2013 10:13 am

Hi,

Since low volatility is explained by value and term risk, does that mean it'll not do well when interest rates rise?

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by larryswedroe » Mon Apr 29, 2013 3:17 pm

winguy
That is what it would mean--if term premium becomes negative. But have to look at the total exposure to all the factors, so beta, and size, value and term and default
Larry

steve r
Posts: 506
Joined: Mon Feb 13, 2012 8:34 pm

Re: Is Low-Vol Anomaly really explained by Value factor?

Post by steve r » Mon Apr 29, 2013 3:58 pm

winguy wrote:Hi,

Since low volatility is explained by value and term risk, does that mean it'll not do well when interest rates rise?


Well ... it matters what you think the answer to the question in the thread is. Many think the answer is yes ... however...

As STLUTZ points out the holdings on M* do not suggest this is the case.
Morningstar is right. The ishares website shows the same thing. As you have noted, low vol/min. var fluctuates around between "value" and "not value." Right now it's in the "not value" area.
As Falkenstein notes (linked paper in OPs opening)
the value beta may be around 0.3, it doesn't seem consistent enough to be fundamental.
I might add, correlations do not always mean causation ... swimming pool accidents are correlated with ice cream truck sales ... the events are both caused by high temps. I wonder, but have not researched enough to know for sure, that the positive .3 findings are because both are doing well at the same time, not because one causes the other ... IDK ...

That said, I play this space with the min variance products (ACWV in my case) because it limits exposure to utilities (to a little more than cap weighted). USMV does likewise. SPLV (low vol) has heavy exposure to utilities and is much and has more of a value tilt in holdings. BH WIKI has more details.
Maximize Diversification - Minimize Costs - Avoid Lotteries

Post Reply