Skeptical of the hidden risk in Value - convince me

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
tadamsmar
Posts: 7325
Joined: Mon May 07, 2007 12:33 pm

Skeptical of the hidden risk in Value - convince me

Post by tadamsmar » Thu Jan 05, 2017 10:51 am

I was just thinking about the notion of hidden risk in Value. I know Fama seems to think it's the best explanation of Value's out-performance.

I do a good bit of analysis of money management in gambling which is basically the same analysis as for any other investment risk. I don't see that there is any hidden risk in gambling. Except when limited historical data is used and it underestimates the actual risk - is that really all that is being claimed in relation to Value?

I know that the notion that it's a persistent behavioral anomaly is not very compelling either.

PS: My main goal is to analyze the bracket pool and there is good evidence of a large behavioral anomaly there.

Quark
Posts: 1045
Joined: Sun Nov 01, 2015 5:32 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by Quark » Thu Jan 05, 2017 11:06 am

Fama's basic idea is that markets are efficient, so that you can't get more expected return without taking on more risk. He does not believe in behavioral economics.

If markets are efficient, either value's historic out-performance is just an anomaly and is not likely to persist (i.e., random luck) or there is something about value that is riskier. Fama finds widespread evidence of a value premium, so he believes it's a risk story, not a luck story. He has not identified the nature of the risk behind value.

I don't believe there's enough useful and reliable data to answer the question definitively, but people are prone to finding patterns and reasons (even when there are none), so there are lots of stories that purport to explain value.

Funds that were created to profit from behavioral anomalies do not have a great track record.

azanon
Posts: 1429
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Skeptical of the hidden risk in Value - convince me

Post by azanon » Thu Jan 05, 2017 11:12 am

For kicks, I went over to portfoliovisualizer, and entered in VG mid-cap value (VMVIX) 100%, vs. VG mid-cap growth (VMGRX). VMVIX was 18.24% SD vs. VMGRX at 17.28%. A higher SD, implies greater volatility, which implies greater risk.

I imagine you could pick any number of other similar comparisons, and you'll find the value one having a higher SD. But I don't want to take the time to do that.

Anyway, it was my understanding that value was mathematically more risky, so it transcends one's opinion, really. Nothing hidden, just crunch the numbers and calculate SD.
Last edited by azanon on Thu Jan 05, 2017 11:15 am, edited 2 times in total.

Valuethinker
Posts: 32630
Joined: Fri May 11, 2007 11:07 am

Re: Skeptical of the hidden risk in Value - convince me

Post by Valuethinker » Thu Jan 05, 2017 11:12 am

tadamsmar wrote:I was just thinking about the notion of hidden risk in Value. I know Fama seems to think it's the best explanation of Value's out-performance.

I do a good bit of analysis of money management in gambling which is basically the same analysis as for any other investment risk. I don't see that there is any hidden risk in gambling. Except when limited historical data is used and it underestimates the actual risk - is that really all that is being claimed in relation to Value?


The risk is of what happens in an extended situation like the 1930s I believe. Since we really only have good data for one such period, it's hard to make statistical assertions. But a tight money/ deflationary world is generally held to *not* be good for value stocks. Had the 2009 thing not been met with enormous monetary loosening (and some fiscal loosening) we would have seen value do even worse than it has since then.

I know that the notion that it's a persistent behavioral anomaly is not very compelling either.


Actually I do find that fairly compelling, particularly given the limits on arbitrage. The Quants may have been right in August 2007, but when there was a forced liquidation, value stocks crashed.

PS: My main goal is to analyze the bracket pool and there is good evidence of a large behavioral anomaly there.


Without looking it up, I can't comment on that.

Gambling to me is the "small cap growth" phenomenon writ large. You know that on average the house wins *but* you might hit the jackpot. I think that's called "right skew returns"?

rkhusky
Posts: 4197
Joined: Thu Aug 18, 2011 8:09 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by rkhusky » Thu Jan 05, 2017 11:28 am

azanon wrote:For kicks, I went over to portfoliovisualizer, and entered in VG mid-cap value (VMVIX) 100%, vs. VG mid-cap growth (VMGRX). VMVIX was 18.24% SD vs. VMGRX at 17.28%. A higher SD, implies greater volatility, which implies greater risk.

I imagine you could pick any number of other similar comparisons, and you'll find the value one having a higher SD. But I don't want to take the time to do that.

Anyway, it was my understanding that value was mathematically more risky, so it transcends one's opinion, really. Nothing hidden, just crunch the numbers and calculate SD.


Standard deviation (SD) does not equal risk. Unfortunately, true risk in investing is hard (impossible?) to compute. I view risk as the probability that my investments will not perform as well as some other choice over my investing future, along with the consequences if the negative side of my choices shows up.

User avatar
tadamsmar
Posts: 7325
Joined: Mon May 07, 2007 12:33 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by tadamsmar » Thu Jan 05, 2017 11:52 am

Valuethinker wrote:Gambling to me is the "small cap growth" phenomenon writ large. You know that on average the house wins *but* you might hit the jackpot. I think that's called "right skew returns"?


Money management in gambling involves exposing only a fraction of your bankroll, could be a small fraction.

Anyone holding the TSM is exposing a small fraction of their bankroll to small cap growth.

The math is essentially the same.

In the case of Blackjack, the house does loses on average against best play (depending on how the house arranges the game). See the paper "A Favorable Strategy for Twenty One", but the house does not face enough best play to matter or they take steps to limit best play. In the case of the bracket pool, there is typically no house.

User avatar
tadamsmar
Posts: 7325
Joined: Mon May 07, 2007 12:33 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by tadamsmar » Thu Jan 05, 2017 11:59 am

azanon wrote:For kicks, I went over to portfoliovisualizer, and entered in VG mid-cap value (VMVIX) 100%, vs. VG mid-cap growth (VMGRX). VMVIX was 18.24% SD vs. VMGRX at 17.28%. A higher SD, implies greater volatility, which implies greater risk.

I imagine you could pick any number of other similar comparisons, and you'll find the value one having a higher SD. But I don't want to take the time to do that.

Anyway, it was my understanding that value was mathematically more risky, so it transcends one's opinion, really. Nothing hidden, just crunch the numbers and calculate SD.


Unless I am mistaken, the Fama data showed that value had a smaller SD at a given return, as in this example:

"The chart above shows the annual returns of the Russell 1000 Value and Growth Indexes from 1979 – 2003. The average annual return for the Russell 1000 Value Index was 15.42% and the average annual return for the Russell 1000 growth index was 14.46%. A portfolio’s volatility is typically measured in terms of standard deviation and as you can see from the chart above the volatility of returns is wider for the Russell 1000 growth index. The chart translates into a standard deviation for the Russell 1000 value of 14.20% and 20.65% for the Russell 1000 growth."

http://investorsolutions.com/knowledge- ... vesting-2/

psteinx
Posts: 2555
Joined: Tue Mar 13, 2007 2:24 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by psteinx » Thu Jan 05, 2017 12:05 pm

Count me in as a skeptic of the "value factor" being primarily a risk story.

It is pretty well accepted that general equity risk (beta) offers expected returns. Investors are cognizant of the risk return tradeoff. Investors who are risk averse are counseled to reduce equity exposure. Someone wanting high expected returns and who has a high risk tolerance will have a portfolio that is 100/0 equity/bond, or perhaps 80/20 or whatever. Someone who is risk averse will be perhaps 30/70 or even 0/100.

But you don't really see a corresponding phenomenon with value/growth. You don't see many investors saying, well, I'm only moderately risk tolerant, so within equities, I'll overweight growth.

Yes, I know some on this forum who take the opposite position - who believe they're super risk tolerant, so they're 100/0 equities, and heavily value tilted within equities. But IMO, for a risk story to be a major explanation for the "value factor", you need some folks on the other end of the tradeoff - ready to give up some return for lower risk, by overweighting growth vs. value. I don't see that to a large enough degree to explain the "value factor".

User avatar
grayfox
Posts: 4544
Joined: Sat Sep 15, 2007 4:30 am
Location: Earth

Re: Skeptical of the hidden risk in Value - convince me

Post by grayfox » Thu Jan 05, 2017 12:12 pm

Suppose there were two businesses that you could invest in, A and B.
ROE(A) = 10% and ROE(B) = 30%.
Simplifying assumption: ROE's are constant forever. No risk.
If you could buy each one at P/B=1, A would return 10% pa and B would return 30% pa.
Obviously you would choose to invest in B.

Now let the Market decide the price of A and B.
If PB(A) = 1, then the Market would drive PB(B) higher.
What should be the P/B of B?

:?: In other words, how much more should investors be willing to pay for a business with higher ROE?

I will give my answer below.
Last edited by grayfox on Thu Jan 05, 2017 12:51 pm, edited 2 times in total.
Gott mit uns.

User avatar
grayfox
Posts: 4544
Joined: Sat Sep 15, 2007 4:30 am
Location: Earth

Re: Skeptical of the hidden risk in Value - convince me

Post by grayfox » Thu Jan 05, 2017 12:49 pm

grayfox wrote:
:?: In other words, how much more should investors be willing to pay for a business with higher ROE?


Here's my answer. :arrow: It depends on how long you are willing to wait to break even.

If you pay P/B=1, your return will be ROE from the first year.
If you pay a premium, i.e. P/B > 1, you start in the hole, but eventually you break even, and in the very long run, your return approaches ROE.
The higher the premium you pay, i.e. the higher the P/B you are willing to pay, the longer it will take to break even.

:idea: Different investors with different investment horizons are willing to wait different amounts of time to see profits. So different investors will pay different P/B for Business B.
Gott mit uns.

lack_ey
Posts: 5542
Joined: Wed Nov 19, 2014 11:55 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by lack_ey » Thu Jan 05, 2017 1:09 pm

FWIW in the worst 5% of months for the market factor (well, market minus risk-free) in the US since 1927, value (HML) had an arithmetic average return of -0.88%. Over the full dataset, HML had an arithmetic average return of 0.40%. So the downside of value showed up to at least some small extent at the wrong times.

That said, it doesn't really seem to be the case in the smaller ex-US data.

In the Fama-French 2x3 size/value breakdown for US stocks since 1927, SCV does have higher standard deviation (and fatter tails) than SCB and SCG, and LCV does have higher standard deviation than LCB and LCG. Not by a lot. For monthly returns, for SCV/SCB/SCG it was 8.20%/7.02%/7.57% and for LCV/LCB/LCG it was 7.18%/5.70%/5.32%. A lot of these relationships vary by period and are not particularly robust. Since mid-1992 (Fama-French publication, just as a kind of arbitrary date), you're looking at for SCV/SCB/SCG 5.42%/5.11%/6.74% and for LCV/LCB/LCG 5.13%/4.22%/4.21%.

In any case I don't think the risk-based explanations for value convincingly explain 100% of the historical return differential.

azanon
Posts: 1429
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Skeptical of the hidden risk in Value - convince me

Post by azanon » Thu Jan 05, 2017 1:18 pm

rkhusky wrote:
azanon wrote:For kicks, I went over to portfoliovisualizer, and entered in VG mid-cap value (VMVIX) 100%, vs. VG mid-cap growth (VMGRX). VMVIX was 18.24% SD vs. VMGRX at 17.28%. A higher SD, implies greater volatility, which implies greater risk.

I imagine you could pick any number of other similar comparisons, and you'll find the value one having a higher SD. But I don't want to take the time to do that.

Anyway, it was my understanding that value was mathematically more risky, so it transcends one's opinion, really. Nothing hidden, just crunch the numbers and calculate SD.


Standard deviation (SD) does not equal risk. Unfortunately, true risk in investing is hard (impossible?) to compute. I view risk as the probability that my investments will not perform as well as some other choice over my investing future, along with the consequences if the negative side of my choices shows up.


I'm told that's how Bridgewater (and many other firms) measure it. Enough do it that way, that I'm perfectly fine being demoted as merely the messenger here. Fortunately, even for a messenger though, it's certainly a logical conclusion to explain the SD.

Compare the SD of ST treasuries, and precious metal equities, just for an extreme example. One of those two will certainly be lower SD. And it doesn't take a rocket scientist to deduce why. I'm not being mean-spirited I promise, just trying to make a point here that I think is pretty self-evident.

ValueInvestor99
Posts: 46
Joined: Thu Dec 23, 2010 2:25 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by ValueInvestor99 » Thu Jan 05, 2017 1:32 pm

I don't see the distinction between value and growth. If a company has high earnings the company will spend it on growing the company.
If a company is growing, its earnings will seem depressed because accelerated depreciation.
If a company has a very high stock price, it can issue more stock to grow its book value.

User avatar
grayfox
Posts: 4544
Joined: Sat Sep 15, 2007 4:30 am
Location: Earth

Re: Skeptical of the hidden risk in Value - convince me

Post by grayfox » Thu Jan 05, 2017 2:25 pm

Assuming that you have thought about how much of premium investors might be willing to pay for stocks with higher Return-on-Equity ROE, what does that have to do with the price of eggs?

Everything. So-called growth stocks have higher ROE than so-called Value stocks. If Company A and B, with ROE 10% and 30%, retained all Earnings and re-invested them in the company at ROE, B would grow have higher Earnings Growth than A. Investors will pay a higher P/B for B than for A.

Below are the top ten holdings in Vanguard Value and Growth funds.

Vanguard Value Index Fund Investor Shares (VIVAX)
Month-End Holdings as of 11/30/2016

Code: Select all

Rank  Symbol  Holding             Percentage    ROE        P/B
1     MSFT    Microsoft Corp.         4.20%     22.44%     6.89
2     XOM     Exxon Mobil Corp.       3.40%      5.19%     2.19
3     JNJ     Johnson & Johnson       2.90%     22.09%     4.33
4     BRKA    Berkshire Hathaway Inc. 2.90%      9.02%     1.51
5     JPM     JPMorgan Chase & Co.    2.70%      0.38%     1.36
6     GE      General Electric Co.    2.60%      8.21%     3.43
7     WFC     Wells Fargo & Co.       2.30%     11.24%     1.56
8     T       AT&T Inc.               2.30%     12.05%     2.12
9     BAC     Bank of America Corp.   2.00%      6.27%     0.95
10    PG      Procter & Gamble Co.    2.00%     16.61%     3.95

                AVERAGE                         11.35%     2.83

Vanguard Growth Index Fund Investor Shares (VIGRX)
Month-End Holdings as of 11/30/2016

Code: Select all

Rank  Symbol Holding              Percentage       ROE     P/B
1     AAPL   Apple Inc.               6.20%     36.90%     4.83
2     GOOG   Alphabet Inc.            4.90%     15.23%     4.04
3     AMZN   Amazon.com Inc.          3.30%     13.93%    20.23
4     FB     Facebook Inc.            3.00%     15.71%     6.31
5     CMCS   Comcast Corp.            1.80%     15.74%     3.13
6     HD     Home Depot Inc.          1.80%    116.06%    29.41
7     KO     Coca-Cola Co.            1.70%     27.57%     6.89
8     V      Visa Inc.                1.60%     19.09%     6.90
9     DIS    Walt Disney Co.          1.60%     20.40%     3.97
10    PM     Philip Morris Int Inc.   1.50%

                 AVERAGE                        31.18%     9.52


For Value, the Average ROE and P/B is 11% and 2.8x.
For Growth, the Average ROE and P/B is 31% and 9.5x
Growth has about 3x the ROE and 3x the P/B.

That all makes sense, right? For companies that have higher ROE, pay higher P/B. It's all perfectly logical.

See where I'm going with this?
Gott mit uns.

User avatar
tadamsmar
Posts: 7325
Joined: Mon May 07, 2007 12:33 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by tadamsmar » Thu Jan 05, 2017 3:53 pm

grayfox wrote:Assuming that you have thought about how much of premium investors might be willing to pay for stocks with higher Return-on-Equity ROE, what does that have to do with the price of eggs?

Everything. So-called growth stocks have higher ROE than so-called Value stocks. If Company A and B, with ROE 10% and 30%, retained all Earnings and re-invested them in the company at ROE, B would grow have higher Earnings Growth than A. Investors will pay a higher P/B for B than for A.

Below are the top ten holdings in Vanguard Value and Growth funds.

Vanguard Value Index Fund Investor Shares (VIVAX)
Month-End Holdings as of 11/30/2016

Code: Select all

Rank  Symbol  Holding             Percentage    ROE        P/B
1     MSFT    Microsoft Corp.         4.20%     22.44%     6.89
2     XOM     Exxon Mobil Corp.       3.40%      5.19%     2.19
3     JNJ     Johnson & Johnson       2.90%     22.09%     4.33
4     BRKA    Berkshire Hathaway Inc. 2.90%      9.02%     1.51
5     JPM     JPMorgan Chase & Co.    2.70%      0.38%     1.36
6     GE      General Electric Co.    2.60%      8.21%     3.43
7     WFC     Wells Fargo & Co.       2.30%     11.24%     1.56
8     T       AT&T Inc.               2.30%     12.05%     2.12
9     BAC     Bank of America Corp.   2.00%      6.27%     0.95
10    PG      Procter & Gamble Co.    2.00%     16.61%     3.95

                AVERAGE                         11.35%     2.83

Vanguard Growth Index Fund Investor Shares (VIGRX)
Month-End Holdings as of 11/30/2016

Code: Select all

Rank  Symbol Holding              Percentage       ROE     P/B
1     AAPL   Apple Inc.               6.20%     36.90%     4.83
2     GOOG   Alphabet Inc.            4.90%     15.23%     4.04
3     AMZN   Amazon.com Inc.          3.30%     13.93%    20.23
4     FB     Facebook Inc.            3.00%     15.71%     6.31
5     CMCS   Comcast Corp.            1.80%     15.74%     3.13
6     HD     Home Depot Inc.          1.80%    116.06%    29.41
7     KO     Coca-Cola Co.            1.70%     27.57%     6.89
8     V      Visa Inc.                1.60%     19.09%     6.90
9     DIS    Walt Disney Co.          1.60%     20.40%     3.97
10    PM     Philip Morris Int Inc.   1.50%

                 AVERAGE                        31.18%     9.52


For Value, the Average ROE and P/B is 11% and 2.8x.
For Growth, the Average ROE and P/B is 31% and 9.5x
Growth has about 3x the ROE and 3x the P/B.

That all makes sense, right? For companies that have higher ROE, pay higher P/B. It's all perfectly logical.

See where I'm going with this?


Those Average ROEs are a simple average that ignores the vastly different equity in the 10 firms. That's gotta overweight the heck out of Home Depot.

User avatar
tadamsmar
Posts: 7325
Joined: Mon May 07, 2007 12:33 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by tadamsmar » Thu Jan 05, 2017 3:58 pm

grayfox wrote:Assuming that you have thought about how much of premium investors might be willing to pay for stocks with higher Return-on-Equity ROE, what does that have to do with the price of eggs?

Everything. So-called growth stocks have higher ROE than so-called Value stocks. If Company A and B, with ROE 10% and 30%, retained all Earnings and re-invested them in the company at ROE, B would grow have higher Earnings Growth than A. Investors will pay a higher P/B for B than for A.

Below are the top ten holdings in Vanguard Value and Growth funds.

Vanguard Value Index Fund Investor Shares (VIVAX)
Month-End Holdings as of 11/30/2016

Code: Select all

Rank  Symbol  Holding             Percentage    ROE        P/B
1     MSFT    Microsoft Corp.         4.20%     22.44%     6.89
2     XOM     Exxon Mobil Corp.       3.40%      5.19%     2.19
3     JNJ     Johnson & Johnson       2.90%     22.09%     4.33
4     BRKA    Berkshire Hathaway Inc. 2.90%      9.02%     1.51
5     JPM     JPMorgan Chase & Co.    2.70%      0.38%     1.36
6     GE      General Electric Co.    2.60%      8.21%     3.43
7     WFC     Wells Fargo & Co.       2.30%     11.24%     1.56
8     T       AT&T Inc.               2.30%     12.05%     2.12
9     BAC     Bank of America Corp.   2.00%      6.27%     0.95
10    PG      Procter & Gamble Co.    2.00%     16.61%     3.95

                AVERAGE                         11.35%     2.83

Vanguard Growth Index Fund Investor Shares (VIGRX)
Month-End Holdings as of 11/30/2016

Code: Select all

Rank  Symbol Holding              Percentage       ROE     P/B
1     AAPL   Apple Inc.               6.20%     36.90%     4.83
2     GOOG   Alphabet Inc.            4.90%     15.23%     4.04
3     AMZN   Amazon.com Inc.          3.30%     13.93%    20.23
4     FB     Facebook Inc.            3.00%     15.71%     6.31
5     CMCS   Comcast Corp.            1.80%     15.74%     3.13
6     HD     Home Depot Inc.          1.80%    116.06%    29.41
7     KO     Coca-Cola Co.            1.70%     27.57%     6.89
8     V      Visa Inc.                1.60%     19.09%     6.90
9     DIS    Walt Disney Co.          1.60%     20.40%     3.97
10    PM     Philip Morris Int Inc.   1.50%

                 AVERAGE                        31.18%     9.52


For Value, the Average ROE and P/B is 11% and 2.8x.
For Growth, the Average ROE and P/B is 31% and 9.5x
Growth has about 3x the ROE and 3x the P/B.

That all makes sense, right? For companies that have higher ROE, pay higher P/B. It's all perfectly logical.

See where I'm going with this?


Those Average ROEs are a simple average that ignores the vastly different equity in the 10 firms. That's gotta overweight the heck out of Home Depot.

6 giants are 34 feet tall and 1 giant is 116 feet tall. What's the average height? Is it (34+116)/2 ??? What if the six are called Apple and the one is called Home Depot?

azanon
Posts: 1429
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Skeptical of the hidden risk in Value - convince me

Post by azanon » Thu Jan 05, 2017 3:59 pm

ValueInvestor99 wrote:I don't see the distinction between value and growth. If a company has high earnings the company will spend it on growing the company.
If a company is growing, its earnings will seem depressed because accelerated depreciation.
If a company has a very high stock price, it can issue more stock to grow its book value.


ValueInvestor99, with 23 posts, doesn't see the distinction between value and growth? Something's got to go.... the handle or this statement. :mrgreen:

User avatar
grayfox
Posts: 4544
Joined: Sat Sep 15, 2007 4:30 am
Location: Earth

Re: Skeptical of the hidden risk in Value - convince me

Post by grayfox » Thu Jan 05, 2017 4:29 pm

tadamsmar wrote:
Those Average ROEs are a simple average that ignores the vastly different equity in the 10 firms. That's gotta overweight the heck out of Home Depot.

6 giants are 34 feet tall and 1 giant is 116 feet tall. What's the average height? Is it (34+116)/2 ??? What if the six are called Apple and the one is called Home Depot?


Yes, I don't know what's up with HD with the huge ROE. Those ROE and P/B numbers are from Yahoo Finance.

But even when I zero out HD's ROE, the average ROE for top ten in VIGRX is still 18.29%, which is 1.6x the Value ROE.

A point that I'm trying to make is that higher ROE stocks deserve higher P/B, and lower ROE deserves lower P/B. Value stocks having lower P/B doesn't mean that investors think that Value is riskier. Even if there was no risk, e.g. ROE for each stock was a known constant, the relationship would still hold.
Last edited by grayfox on Thu Jan 05, 2017 4:52 pm, edited 1 time in total.
Gott mit uns.

leonard
Posts: 5993
Joined: Wed Feb 21, 2007 11:56 am

Re: Skeptical of the hidden risk in Value - convince me

Post by leonard » Thu Jan 05, 2017 4:40 pm

tadamsmar wrote:I was just thinking about the notion of hidden risk in Value. I know Fama seems to think it's the best explanation of Value's out-performance.

I do a good bit of analysis of money management in gambling which is basically the same analysis as for any other investment risk. I don't see that there is any hidden risk in gambling. Except when limited historical data is used and it underestimates the actual risk - is that really all that is being claimed in relation to Value?

I know that the notion that it's a persistent behavioral anomaly is not very compelling either.

PS: My main goal is to analyze the bracket pool and there is good evidence of a large behavioral anomaly there.


So, are you asking us to convince you of the existence of value's outperformance?

Or, are you asking us to convince you why value outperforms?
Leonard | | Market Timing: Do you seriously think you can predict the future? What else do the voices tell you? | | If employees weren't taking jobs with bad 401k's, bad 401k's wouldn't exist.

stlutz
Posts: 3919
Joined: Fri Jan 02, 2009 1:08 am

Re: Skeptical of the hidden risk in Value - convince me

Post by stlutz » Thu Jan 05, 2017 7:07 pm

For kicks, I went over to portfoliovisualizer, and entered in VG mid-cap value (VMVIX) 100%, vs. VG mid-cap growth (VMGRX). VMVIX was 18.24% SD vs. VMGRX at 17.28%. A higher SD, implies greater volatility, which implies greater risk.


You compared an index value fund vs. an actively-managed growth fund. I think the growth index fund you want is VMGIX. If you compare like vs. like, value is less volatile than growth for this example (with not that much history). https://www.portfoliovisualizer.com/bac ... ion2_2=100

And in general, long-term comparisons of value vs. growth that do seek to compare like vs. like (i.e. not ones that define 20% of the market as value and 80% as growth) show value as being less volatile than growth.

stlutz
Posts: 3919
Joined: Fri Jan 02, 2009 1:08 am

Re: Skeptical of the hidden risk in Value - convince me

Post by stlutz » Thu Jan 05, 2017 7:20 pm

One of the things I've found from looking at the value/growth data from Ken French's website is that, on average, the value premium is larger in down markets than it is in up markets. That can lead in a number of possible directions:

a) Value outperforms not because it's more risky than growth but because it's less risky. That is, the theory is 180 degrees wrong.

b) Things like volatility and going down in price aren't really risk.

c) There are risky bear markets and non-risky bear markets. It's a risky bear market when value goes down more and a non-risky bear market when growth goes down more.

d) Value is riskier. However, this risk has not shown up in the historical periods we've been looking at. As such, the historical outperformance of value does not equate to the expected outperformance going forward, as the historical performance doesn't include the possible really bad outcomes for value relative to growth.

e) Simplistic explanations where I definitively explain historical market performance with no analysis beyond running a couple mathematical regressions is bunk. Instead of trying to prove that value is riskier than growth because it outperformed, I should start out by coming up with a list of ways to define and measure the risk of a stock, then analyze various buckets of stocks to see which is riskier, assess whether the relative riskiness of those buckets is permanent or time varying, and then determining whether excess return is in fact produced by the riskier buckets. Then you'd have a more unified theory of risk and return.

(Obviously option E is the avenue of exploration that I would find most useful).

User avatar
grayfox
Posts: 4544
Joined: Sat Sep 15, 2007 4:30 am
Location: Earth

Re: Skeptical of the hidden risk in Value - convince me

Post by grayfox » Thu Jan 05, 2017 7:24 pm

stlutz wrote:You compared an index value fund vs. an actively-managed growth fund. I think the growth index fund you want is VMGIX. If you compare like vs. like, value is less volatile than growth for this example (with not that much history). https://www.portfoliovisualizer.com/bac ... ion2_2=100

And in general, long-term comparisons of value vs. growth that do seek to compare like vs. like (i.e. not ones that define 20% of the market as value and 80% as growth) show value as being less volatile than growth.


Code: Select all

#   Initial Balance   Final Balance   CAGR   Std.Dev.   Best Year   Worst Year   Max. Drawdown   Sharpe Ratio   Sortino Ratio   US Mkt Correlation
1   $10,000   $20,671    7.53%    18.24%   37.61%   -36.64%   -56.46%    0.46   0.66   0.97
2   $10,000   $20,064    7.21%    19.03%   42.54%   -47.07%   -54.48%    0.43   0.61   0.94


So Midcap Value had both lower variance AND higher return. How about large cap value and growth, VIVAX vs VIGRX?

Code: Select all

#   Initial Balance   Final Balance   CAGR   Std.Dev.   Best Year   Worst Year   Max. Drawdown   Sharpe Ratio   Sortino Ratio   US Mkt Correlation
1   $10,000   $85,459    9.35%    14.86%   37.03%   -35.97%   -54.86%    0.52   0.74   0.94
2   $10,000   $76,849    8.87%    15.68%   42.21%   -38.32%   -53.58%    0.47   0.68   0.96


Large Value also had higher return AND lower variance. how about small cap value vs growth, VISVX vs VISGX?

Code: Select all

#   Initial Balance   Final Balance   CAGR   Std.Dev.   Best Year   Worst Year   Max. Drawdown   Sharpe Ratio   Sortino Ratio   US Mkt Correlation
1   $10,000   $54,958    9.93%    18.81%   37.19%   -32.05%   -56.13%    0.51   0.75   0.87
2   $10,000   $44,851    8.69%    20.39%   42.88%   -40.00%   -53.52%    0.43   0.63   0.87


It looks like Value has less variance than Growth AND has higher returns, no mater what the size.
Gott mit uns.

stlutz
Posts: 3919
Joined: Fri Jan 02, 2009 1:08 am

Re: Skeptical of the hidden risk in Value - convince me

Post by stlutz » Thu Jan 05, 2017 7:26 pm

It looks like Value has less variance than Growth AND has higher returns, no mater what the size.


Which is why the options (a) and (d) in my second post are legitimately worth considering.

User avatar
grayfox
Posts: 4544
Joined: Sat Sep 15, 2007 4:30 am
Location: Earth

Re: Skeptical of the hidden risk in Value - convince me

Post by grayfox » Thu Jan 05, 2017 7:46 pm

stlutz wrote:
It looks like Value has less variance than Growth AND has higher returns, no mater what the size.


Which is why the options (a) and (d) in my second post are legitimately worth considering.


I would go with a) Value outperforms not because it's more risky than growth but because it's less risky. That is, the theory is 180 degrees wrong.

Looking at the top ten holdings above, large Value stocks look less risky to me. It seems obvious to me that there is more certainty in Sales and Earnings with consumer staples like Procter & Gamble and Johnson & Johnson than with online retailer Amazon or social media Facebook.

Businesses have Life Cycles. When a business is new, it can have high growth rate because its starting from small, and can grow Sales and Earnings much faster than the overall economy or GDP. Once the company gets big, growth rate must come down. When its a mature, it can only have average growth. There is a lot of uncertainty about how long a young company can sustain the high growth period.

Many of the growth companies are relatively young: Apple 1976, Amazon 1994, Facebook 2004.
Many of the Large Value companies are very mature: P&G 1837, XOM 1870, J&J 1886, GE 1892.

There are probably some low P/B companies that are cheap because they are shaky. Maybe some of the SCV companies. There might be a risk explanation there.
Gott mit uns.

cheapskate
Posts: 705
Joined: Thu Apr 26, 2007 1:05 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by cheapskate » Thu Jan 05, 2017 8:05 pm

As someone who got his clock cleaned 2000-2003 (very large permanent loss of capital) investing in large cap tech (growth) stocks, I find the theory that Value is more risky than Growth completely ludicrous. In the top 10 stocks in the Large Cap Growth Index that grayfox posted, 6.60% is accounted for by Amazon and Facebook (neither of which can possibly be counted as "low risk"), that list also contains Apple, Comcast and Disney - all extremely vulnerable to sudden shifts in market/consumer sentiment. But if had to invest in 10 stocks equal weighted, I would have no hesitation investing in the top 10 stocks in the Vanguard Large Cap Value Index.

I invest in Value indexes solely to avoid holding Amazon, facebook, Comcast and so on. I also want to avoid holding the recent tech IPOs. To me, Risk is Ramen Noodles and No Healthcare in Retirement - and Aggressive Growth can possibly put me there, based on what I experienced in 2000-2003

User avatar
tadamsmar
Posts: 7325
Joined: Mon May 07, 2007 12:33 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by tadamsmar » Thu Jan 05, 2017 8:10 pm

leonard wrote:
tadamsmar wrote:I was just thinking about the notion of hidden risk in Value. I know Fama seems to think it's the best explanation of Value's out-performance.

I do a good bit of analysis of money management in gambling which is basically the same analysis as for any other investment risk. I don't see that there is any hidden risk in gambling. Except when limited historical data is used and it underestimates the actual risk - is that really all that is being claimed in relation to Value?

I know that the notion that it's a persistent behavioral anomaly is not very compelling either.

PS: My main goal is to analyze the bracket pool and there is good evidence of a large behavioral anomaly there.


So, are you asking us to convince you of the existence of value's outperformance?

Or, are you asking us to convince you why value outperforms?


The latter, why does it outperform.

bogglizer
Posts: 67
Joined: Tue Aug 16, 2016 8:56 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by bogglizer » Thu Jan 05, 2017 9:38 pm

I have always thought that it isn't that value stocks out-perform, but that growth stocks under-perform. The under-performance is because the evaluations are based on poorer quality estimates of future growth. In other words, growth stocks are not as close to the efficient frontier as value stocks, due to statistical over-evaluation (an error is more likely to increase the evaluation from log-normal statistics).

leonard
Posts: 5993
Joined: Wed Feb 21, 2007 11:56 am

Re: Skeptical of the hidden risk in Value - convince me

Post by leonard » Thu Jan 05, 2017 10:30 pm

tadamsmar wrote:
leonard wrote:
tadamsmar wrote:I was just thinking about the notion of hidden risk in Value. I know Fama seems to think it's the best explanation of Value's out-performance.

I do a good bit of analysis of money management in gambling which is basically the same analysis as for any other investment risk. I don't see that there is any hidden risk in gambling. Except when limited historical data is used and it underestimates the actual risk - is that really all that is being claimed in relation to Value?

I know that the notion that it's a persistent behavioral anomaly is not very compelling either.

PS: My main goal is to analyze the bracket pool and there is good evidence of a large behavioral anomaly there.


So, are you asking us to convince you of the existence of value's outperformance?

Or, are you asking us to convince you why value outperforms?


The latter, why does it outperform.


But, unless you are going to assert the market pricing mechanism is busted - by definition if an asset class out performs, it has more risk. Also, I don't understand the premise that the risk is "hidden". It's clearly evident - because the market on average prices it at a higher rate of return.

I know this is not the answer to "why" they out perform - but I don't even see the question of the risk being "hidden" as even being a question.
Leonard | | Market Timing: Do you seriously think you can predict the future? What else do the voices tell you? | | If employees weren't taking jobs with bad 401k's, bad 401k's wouldn't exist.

stlutz
Posts: 3919
Joined: Fri Jan 02, 2009 1:08 am

Re: Skeptical of the hidden risk in Value - convince me

Post by stlutz » Thu Jan 05, 2017 10:47 pm

But, unless you are going to assert the market pricing mechanism is busted - by definition if an asset class out performs, it has more risk.


That's not by definition at all. I think everyone agrees that riskier investments provide the potential for outperformance. That doesn't mean that the reverse is true--that outpeformance is always a product of risk.

And "busted" is value judgment about how markets should work which is not what the OP was asking. There are definitely some important people in finance and economics who believe that the way that markets should work is that riskier investments should provide superior returns the overwhelming majority of the time. That doesn't mean that markets do work that way nor does it mean that are not valid arguments that markets should not work that way.

jalbert
Posts: 2075
Joined: Fri Apr 10, 2015 12:29 am

Re: Skeptical of the hidden risk in Value - convince me

Post by jalbert » Thu Jan 05, 2017 11:00 pm

For large cap US stocks, I don't think any durable value premium has been observed for a long time, maybe not since 1984. (This of course would imply that the value and size factors are not independent, a basic defining requirement for linear models, which would imply that the Fama-French linear model is incorrect, but let's leave that aside).

In the great depression, large-cap US stocks were down 80% peak-to-trough, but SCV was down 90%. That may not sound so different until you consider that you have only half as much asset value remaining in the latter case.
Risk is not a guarantor of return.

User avatar
in_reality
Posts: 4180
Joined: Fri Jul 12, 2013 6:13 am

Re: Skeptical of the hidden risk in Value - convince me

Post by in_reality » Fri Jan 06, 2017 12:01 am

tadamsmar wrote:
azanon wrote:For kicks, I went over to portfoliovisualizer, and entered in VG mid-cap value (VMVIX) 100%, vs. VG mid-cap growth (VMGRX). VMVIX was 18.24% SD vs. VMGRX at 17.28%. A higher SD, implies greater volatility, which implies greater risk.

I imagine you could pick any number of other similar comparisons, and you'll find the value one having a higher SD. But I don't want to take the time to do that.

Anyway, it was my understanding that value was mathematically more risky, so it transcends one's opinion, really. Nothing hidden, just crunch the numbers and calculate SD.


Unless I am mistaken, the Fama data showed that value had a smaller SD at a given return, as in this example:

"The chart above shows the annual returns of the Russell 1000 Value and Growth Indexes from 1979 – 2003. The average annual return for the Russell 1000 Value Index was 15.42% and the average annual return for the Russell 1000 growth index was 14.46%. A portfolio’s volatility is typically measured in terms of standard deviation and as you can see from the chart above the volatility of returns is wider for the Russell 1000 growth index. The chart translates into a standard deviation for the Russell 1000 value of 14.20% and 20.65% for the Russell 1000 growth."

http://investorsolutions.com/knowledge- ... vesting-2/


Is volatility risk? I say no - not really.

What determines the price of growth stocks? To a large degree future expectations about what the company will be worth down the road. Obviously, circumstances will change (recession or boom) and so will valuations.

Value stocks are different by nature. Of course future expectation determine the price but it's a little bit of a different equation. I wouldn't expect value stocks to get as overpriced as growth during a boom since they aren't as well positioned to capture the growth.

Also, what happens in a downturn? Value stocks are more difficult to price at times of distress since price-to-book is much higher. The value of a company if it stays afloat can be very different than if it goes under and you are liquidating assets in a down economy which have been put together in many cases for a particular purpose. Research has shown that at times, prices can get separated from fundamentals since no-one truly knows whether the company will go under or stay afloat. Whichever view you have, your estimated price is going to be very different from those who have the other view. More so than for two different views on a growth stock. Even if it did go under, the valuations will be closer since there is relatively less to liquidate.

Anyway, value stocks have much higher risk of bankruptcy than growth stocks. Small stock have a higher risk of bankruptcy than large ones. Research shows that high-default-risk firms earn higher returns than low-default-risk firms to the extent that they are small in size and high in BM(book-to-market).**

In all other cases, it seems there is no difference between high and low default risk stocks.**

So yes you are right, there aren't hidden risks in value stocks. What there is though is default risk that is hard to price I think.

NOTE: You will find conflicting information(research) on whether default risk is rewarded by higher returns. Do note that when the market value of a firm's equity is used, it will show default risk is rewarded. This is as market value is a future looking measure. Using accounting models to measure default risk shows it is not rewarded by higher returns - obviously financial statements reflect the past and not expectations about future performance.

**M Vassalou, Y Xing, Default Risk in Equity Returns, The Journal of Finance Vol 59 Issue 2, April 2004, p831-868
[60% US _ 26% DEV _ 14% EM] | (-16% LC _ +8% MC _ +8% SC) | [47% FND/VAL _ 40% MKT _ 7% MOM _ 6% REIT] | (+/- 5% or *25% rebalancing bands)

azanon
Posts: 1429
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Skeptical of the hidden risk in Value - convince me

Post by azanon » Fri Jan 06, 2017 8:38 am

Is volatility a good measure of risk? I say so, and so have a vast array of investment professionals. I'll just have to agree to disagree.

Having high volatility means whatever asset class you own can drop sharply or rise sharply. If you own an asset class that can rise sharply or drop sharply, then you own a risky asset class. It's really that simple.

And as for portfolios, it's risk (if one wants to estimate it) is also best measured by estimating the portfolio's expected volatility. It's how its done at portfolicharts. It's how its done at a lot of places.

User avatar
tadamsmar
Posts: 7325
Joined: Mon May 07, 2007 12:33 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by tadamsmar » Fri Jan 06, 2017 10:16 am

Risk is the risk of losing money. Not sure if there is a formula for relating volatility to the risk of losing money, and if there is then the consumption schedule would have to be a factor.

Volatility also impacts investment growth. For a given average yearly return, higher year-wise volatility means lower longer term growth.

User avatar
in_reality
Posts: 4180
Joined: Fri Jul 12, 2013 6:13 am

Re: Skeptical of the hidden risk in Value - convince me

Post by in_reality » Fri Jan 06, 2017 10:40 am

tadamsmar wrote:Risk is the risk of losing money. Not sure if there is a formula for relating volatility to the risk of losing money, and if there is then the consumption schedule would have to be a factor.


I do agree. Volatility can cost you dearly if you have to sell at a low. Sequence of returns risk and the impact of starting retirement in a downturn is worrisome. (though relatives who took early retirement in 2007 made it fine by adjusting their spending). Still we can mitigate it by structuring things so we don't get forced into selling at a low.

tadamsmar wrote:Volatility also impacts investment growth. For a given average yearly return, higher year-wise volatility means lower longer term growth.


Oh dear, is that why value outperforms -- because growth is, as you say, more volatile? In a way it's like bungee jumping though, you know the growth will return after the recession ends and you'll get yanked up after the bottom. If your cord breaks (in a bankruptcy) though, game over.

I guess I think there should be style points or some kind of reward for bungee jumping with a frayed cord (which kinda seems like a distressed company to me in some ways), otherwise who would ever do it? I think it's definitely true that a properly working cord is going to have more volatility than a frayed one that breaks(up down, up down, up down VERSUS down splat).

So to me, yeah if I don't have to sell, what is the problem with volatility? It's the plummet of a value trap that worries me. Why buy a cord that doesn't bounce as well and might break? I would only buy such a cord at a discount (i.e. lower P/E).
[60% US _ 26% DEV _ 14% EM] | (-16% LC _ +8% MC _ +8% SC) | [47% FND/VAL _ 40% MKT _ 7% MOM _ 6% REIT] | (+/- 5% or *25% rebalancing bands)

rkhusky
Posts: 4197
Joined: Thu Aug 18, 2011 8:09 pm

Re: Skeptical of the hidden risk in Value - convince me

Post by rkhusky » Fri Jan 06, 2017 11:17 am

azanon wrote:
rkhusky wrote:Standard deviation (SD) does not equal risk. Unfortunately, true risk in investing is hard (impossible?) to compute. I view risk as the probability that my investments will not perform as well as some other choice over my investing future, along with the consequences if the negative side of my choices shows up.


I'm told that's how Bridgewater (and many other firms) measure it. Enough do it that way, that I'm perfectly fine being demoted as merely the messenger here. Fortunately, even for a messenger though, it's certainly a logical conclusion to explain the SD.

Compare the SD of ST treasuries, and precious metal equities, just for an extreme example. One of those two will certainly be lower SD. And it doesn't take a rocket scientist to deduce why. I'm not being mean-spirited I promise, just trying to make a point here that I think is pretty self-evident.


SD is used because it is easy to compute, it has nice mathematical properties, and, as a first cut, it does provide a rough separation between safer and riskier investments. It just doesn't tell the whole story. One can generate investment returns that look completely different, but have the same SD.

jalbert
Posts: 2075
Joined: Fri Apr 10, 2015 12:29 am

Re: Skeptical of the hidden risk in Value - convince me

Post by jalbert » Mon Jan 09, 2017 2:45 pm

Standard deviation of future return over your holding period is an excellent measure of your risk in holding an investment. This can only be computed if you know the probability distribution of future returns.
Risk is not a guarantor of return.

leonard
Posts: 5993
Joined: Wed Feb 21, 2007 11:56 am

Re: Skeptical of the hidden risk in Value - convince me

Post by leonard » Tue Jan 10, 2017 4:29 pm

stlutz wrote:
But, unless you are going to assert the market pricing mechanism is busted - by definition if an asset class out performs, it has more risk.


That's not by definition at all. I think everyone agrees that riskier investments provide the potential for outperformance. That doesn't mean that the reverse is true--that outpeformance is always a product of risk.

And "busted" is value judgment about how markets should work which is not what the OP was asking. There are definitely some important people in finance and economics who believe that the way that markets should work is that riskier investments should provide superior returns the overwhelming majority of the time. That doesn't mean that markets do work that way nor does it mean that are not valid arguments that markets should not work that way.


Why would the market be efficient in only one direction? that doesn't make sense. The asset class can outperform - but for some reason it's not tied to risk and is some sort of free lunch? No way. Too many arbitrageurs out there.
Leonard | | Market Timing: Do you seriously think you can predict the future? What else do the voices tell you? | | If employees weren't taking jobs with bad 401k's, bad 401k's wouldn't exist.

jalbert
Posts: 2075
Joined: Fri Apr 10, 2015 12:29 am

Re: Skeptical of the hidden risk in Value - convince me

Post by jalbert » Wed Jan 11, 2017 11:10 pm

There are definitely some important people in finance and economics who believe that the way that markets should work is that riskier investments should provide superior returns the overwhelming majority of the time.

I think the correct statement is that achieving superior return requires taking substantial risk the overwhelming majority of the time.
Risk is not a guarantor of return.

Post Reply