Updated Novy Marx Paper on Quality

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Updated Novy Marx Paper on Quality

Post by larryswedroe »

For those interested in this subject

http://rnm.simon.rochester.edu/research/QDoVI.pdf

Even addresses the small vs large stock issue here

Larry
User avatar
jeffyscott
Posts: 13438
Joined: Tue Feb 27, 2007 8:12 am

Re: Updated Novy Marx Paper on Quality

Post by jeffyscott »

Buying high quality assets without paying premium prices is just as
much value investing as buying average quality assets at discount
prices. Strategies that exploit the quality dimension of value are
profitable on their own, and accounting for both dimensions of value
by trading on combined quality and price signals yields dramatic
performance improvements over traditional value strategies.
Accounting for quality also yields significant performance
improvements for investors trading momentum as well as value.


Interesting...sounds like "growth-at-a-reasonable-price" wins :?:

IIRC, Robert Haugen had argued for that strategy long ago, for example in The New Finance: The Case Against Efficient Markets, published 1995.
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Updated Novy Marx Paper on Quality

Post by Robert T »

.
From the Novy-Marx paper - using data from July 1963 to Dec 2012

Small Value Strategy (sorted on average P/B and E/P ranks)
Net return = 10.9%
Turnover = 32.9
Max drawdown = -28.3

Joint Small Value and Profitability (sorted on average P/M and GP/A ranks)
Net return = 11.8%
Turnover = 25.8
Max drawdown = -18.2

This excluded financial stocks with account for about 25% of value strategies. So if we assume no difference in returns between financial stocks, when added to each portfolio this results in about a 0.68% higher return [(11.8-10.9)*0.75] with less historical draw-down and turnover.

We know that one of the sorts used by S&P is financial viability“Companies should have four consecutive
quarters of positive as-reported earnings, where as-reported earnings are defined as GAAP Net Income excluding
discontinued operations and extraordinary items.”
Russell uses a purely market cap split.
https://www.sp-indexdata.com/idpfiles/i ... ap_600.pdf

Does the S&P financial viability sort already capture some of the profitability premium? The answer is yes.

Feb 1995-December 2012 - “Profitability premium’ loads*

S&P600 = +0.06
Russell2000 = -0.09
Difference = +0.15

* These are the 'profitability premium' loads when added to the Fama-French three factor model. A 'profitability premium' series is available on Novy-Marx's website. The 0.15% difference is statistically significant. The standard deviation of returns, and portfolio turnover are also lower for S&P than Russell.

The monthly profitability premium over this period was 0.42%. From the above, the difference in return of the S&P600 and the Russell2000 due to ‘exposure’ to the profitability premium (financial viability sorts) = 0.76% in higher returns (0.15*42*42). The CRSP results are similar to Russell.

S&P seems to capture some of the ‘profitability premium” through its financial viability sort.

Robert
.
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Updated Novy Marx Paper on Quality

Post by Robert T »

.
The latest Equius Partners Asset Class has an article from David Booth of Dimensional which discusses the 'profitability dimension'

http://mobile.equiuspartners.com/newsle ... sional.pdf
  • From David Booth:

    "Expected returns = expected cash flow/price" ..."Expressing the relation this way highlights two of the dimensions of expected returns for equities—relative price and direct profitability. Higher expected returns are the result of having either higher expected cash flows or a lower price. The direct profitability dimension is tied to the numerator and the relative price dimension to the denominator. Stated another way, if two stocks sell at the same price, then the one with higher expected cash flows must have a higher expected return."
Just to note that there is an earlier article on DFA's website on "Is there still value in Book-to-Market? http://www.dfaus.com/2009/05/is-there-s ... ratio.html - which show sales/price and book-to-market to be better sorts than cash flow/price. Not sure how to reconcile the two.

Robert
.
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Updated Novy Marx Paper on Quality

Post by Robert T »

.
Here's a bit more analysis on the ‘profitability loads’ of various funds. Adding the profitability and momentum factors to the FF three factor model. The profitability fator data from Novy-Marx's website is use for the analysis throughout.

First, the GMO Quality fund. This indeed has a positive and significant profitability load of 0.23.

Code: Select all

GMO QUALITY FUND: Feb 2004 – December 2012

               Coefficient  (t-stat)

Aplha             -0.03      (-0.3)     
Mkt                0.84      (25.2) 
Size              -0.31      (-5.4) 
Value             -0.09      (-1.8) 
Profitability      0.23      ( 3.8)
Momentum           0.05      ( 2.0)

R^2                0.89
Second, the S&P600 Value index. A few points worth highlighting.
  • (1) the profitability load is positive at 0.09 and significant with a t-stat. of 2.1.
    (2) ‘profitability’ loads more in the S&P600 value index than the S&P600 growth index, for latter it is non-significant.
    (3) the three factor model on its own does a good job in explaining the variability in monthly return (R^2 = 0.94), the additional explanatory power of the profitability dimension is small. Personally I think the three factor model still provides a robust framework for structuring equity exposure/planning purposes.

Code: Select all

S&P600 VALUE: July 1995 – Dec. 2012

                  Coeff. (t-stat)  Coeff. (t-stat) Coeff. (t-stat)

Aplha             -0.06   (-0.6)   -0.11  (-1.1)   -0.09  (-1.0)  
Mkt                0.96   (45.6)    0.97  (45.4)    0.96  (42.4)
Size               0.68   (23.9)    0.69  (24.2)    0.69  (24.2)  
Value              0.56   (18.8)    0.57  (19.0)    0.57  (18.6)
Profitability                       0.09  ( 2.1)    0.09  ( 2.1)
Momentum                                           -0.01  (-0.8)    

R^2                0.94             0.94            0.94    
Third, the Bridgeway Ultra-Small Company Market fund. Interesting the ‘profitability’ load on this fund is similar to the GMO Quality fund at 0.21 (with t-stat of 2.8). So perhaps the stock screens to exclude those companies passing through to bankruptcy captures some of the profitability premium rather than the momentum premium as in the results below (significant profitability dimension effect, non-significant momentum effect).

Code: Select all

BRIDGEWAY ULTRA-SMALL COMPANY MARKET: Aug 1997 - Dec. 2012

               Coefficient  (t-stat)

Aplha              0.07      ( 0.4)     
Mkt                0.94      (22.3) 
Size               0.69      (13.0) 
Value              0.35      ( 6.3) 
Profitability      0.21      ( 2.8)
Momentum          -0.04      (-1.4)

R^2                0.84
Robert
.
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Updated Novy Marx Paper on Quality

Post by Robert T »

.
Some analysis and findings on the midcap series.

(1) The financial viability sorts of S&P doesn’t capture the profitability dimension in midcaps (unlike in small caps where is does). This may simply be reflective that the ‘profitability premium’ may be smaller in midcaps than smallcaps. From the Novy-Marx article in Journal of Finance - “more profitable growth firms tend to be larger than less profitable growth firms, more profitable value firms then to be smaller than less profitable value firms. So while little difference exists in size between unprofitable value and growth firms, profitable growth firms are large but highly profitable value firms are small” (my bold)

(2) The profitability loads across S&P, MSCI, and Russell midcap value series are non-significant. Perhaps reflective of the smallcap effect (quote above).

While the Novy-Marx paper (Table A9) also indicated that within small caps ‘controlling for profitability has little impact on the performance of value strategies’ there is some effect, and it seems larger in smallcaps if you sort first by profitability (which S&P does), then by value (BtM). If my reading of table A9 is correct if you control for value (BtM) and then sort by profitability the average difference between unprofitable and profitable portfolios is 0.54%, while if you control for profitability then sort by value the average difference between low BtM (growth) and high BtM (value) is 0.81%.

Code: Select all

July 1995 - Dec. 2012

                S&P 400 Value      MSCI MCV         Russell MCV

                Coeff  (t-stat)   Coeff. (t-stat)   Coeff.  (t-stat)
Aplha            0.02   ( 0.1)     0.15  ( 1.4)      0.08   ( 0.9)
Mkt              0.97   (37.7)     0.95  (39.6)      0.97   (45.8)
Size             0.22   ( 7.0)     0.11  ( 3.5)      0.10    (3.9)
Value            0.49   (14.2)     0.61  (18.8)      0.60   (21.1)
Profitability   -0.03   (-0.6)    -0.02  (-0.4)     -0.02   (-0.5)
Momentum        -0.04   (-1.9)    -0.07  (-3.5)     -0.05   (-3.3)

R^2              0.91              0.92              0.93  
Robert
.
bs1
Posts: 23
Joined: Sun Jun 24, 2012 7:10 pm

Re: Updated Novy Marx Paper on Quality

Post by bs1 »

Robert,

your work seems to supports the idea that profitability is a growth strategy, similar to the way DFA markets it. (and LG may be best, since “more profitable growth firms tend to be larger than less profitable growth firms”.) it follows that it might be used to decrease your SV tilt (diversify risk factors) without sacrificing returns. do the “sorts” of any LG funds capture enough of the premium to make this approach worthwhile?

thank you.
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Re: Updated Novy Marx Paper on Quality

Post by richard »

Robert T wrote:.From David Booth:

"Expected returns = expected cash flow/price" ..."Expressing the relation this way highlights two of the dimensions of expected returns for equities—relative price and direct profitability. Higher expected returns are the result of having either higher expected cash flows or a lower price. The direct profitability dimension is tied to the numerator and the relative price dimension to the denominator. Stated another way, if two stocks sell at the same price, then the one with higher expected cash flows must have a higher expected return."
This sounds a whole lot like using p/e as a valuation metric.

Two problems with p/e are (1) using historic earnings rather than future earnings and (2) accounting earnings do not have a necessary relation to economic reality. Using an expected value is better than using a historic value, but only if we are good at predicting the future. Many regard cash flow as more accurate or useful than earnings, but cash flow has its own faults.
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Re: Updated Novy Marx Paper on Quality

Post by richard »

jeffyscott wrote:Buying high quality assets without paying premium prices is just as
much value investing as buying average quality assets at discount
prices. Strategies that exploit the quality dimension of value are
profitable on their own, and accounting for both dimensions of value
by trading on combined quality and price signals yields dramatic
performance improvements over traditional value strategies.
Accounting for quality also yields significant performance
improvements for investors trading momentum as well as value.


Interesting...sounds like "growth-at-a-reasonable-price" wins :?:

IIRC, Robert Haugen had argued for that strategy long ago, for example in The New Finance: The Case Against Efficient Markets, published 1995.
The paper includes "All these strategies thus fit comfortably under the general rubric of “growth at a reasonable price” (GARP) strategies" on page 3.
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Updated Novy Marx Paper on Quality

Post by larryswedroe »

Richard
Re earnings-the BIG contribution by Novy Marx was to move up the income statement, not looking at NET INCOME but gross profitability instead, avoiding many of the problems of manipulation of income--like the well known accrual problem

BTW-it turns out that highly profitability companies are growth companies and high profitability persists and so you get more growthy portfolio-so a profitability strategy will lower loading on value
Larry
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: Updated Novy Marx Paper on Quality

Post by Random Walker »

Larry, I remember Fama Junior writing something like buying the assets for a deep value portfolio involved buying dog stocks that were so run down that it hurt to look at them. That is why one needs autopilot formulaic passive screens. He said looking for the quality value stocks was just getting more growthy. Does this line of thinking fit with the profitability screens, contradict them, or irrelevant? Thanks,

Dave
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Updated Novy Marx Paper on Quality

Post by larryswedroe »

Dave
I don't think it contradicts him. Remember it's still a value stock, just not quite as valuey--so you get less of the value premium but pick up the profitability premium and the net impact is significant.
Same thing btw with MOM, add MOM and you get more growthy for obvious reasons--so you lose bit of the value premium but gain the MOM premium for another pickup.
Here you are getting two more small but not insignificant pickups by combining the three strategies.
Here is a rough example
Say value and MOM are both 4% premiums
By screening for MOM say you lose .2 on value loading, but add back 0.4 from negative MOM loading (value loads negatively on MOM). So that means you lose 80bp (4% x.2) from lower value loading but gain 1.6% from eliminating negative MOM screen. Again, just an example but that's the idea
Larry
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Re: Updated Novy Marx Paper on Quality

Post by richard »

Larry, people have been moving up the income statement for many years. As a random example, the early LBO players in the 1980s used cash flow as a rule of thumb valuation metric. As you say, the accounting problems with net income are well known, and go back much further.

I regard using historical income statement data as a serious problem. With a high quality bond, you know future yield. With stocks, you don't know future earnings, cash flow or whatever income statement line item you favor. Past income statement data may or may not be a good estimator for future data.

Using income statement numbers has more theoretical justification than using balance sheet numbers. b/p may have empirical justification, but it's lacking in a solid theory. In addition, balance sheet numbers are not without serious accounting issues.

Interesting that this new focus moves away from the traditional b/p version of value.
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Updated Novy Marx Paper on Quality

Post by larryswedroe »

Richard
I understand the skepticism which is always healthy but FF looked at the data and found strong statistical significance that profitability persists out for about 7 years if memory serves.
Larry
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Updated Novy Marx Paper on Quality

Post by Robert T »

bs1 wrote:Robert,

your work seems to supports the idea that profitability is a growth strategy, similar to the way DFA markets it. (and LG may be best, since “more profitable growth firms tend to be larger than less profitable growth firms”.) it follows that it might be used to decrease your SV tilt (diversify risk factors) without sacrificing returns. do the “sorts” of any LG funds capture enough of the premium to make this approach worthwhile?

thank you.
My view is:

(1) The profitability premium seems more difficult to capture with available options than the market, size or value premiums re: the profitability load on the GMO Quality fund is only 0.23. On other growth funds (at least the ones I have briefly looked at) it is smaller.

(2) The profitability premium is smaller than the value premium (as in the Novy-Marx paper), and to me the risk based story for the value premium is more convincing i.e. more likely to persist - in my view (same for market and size premiums). This makes matching factor exposure to 'needed' expected return more straight forward with market, size, and value than with the profitability dimension.

(3) There are many available ways to capture a fairly large share of any size and value premiums.

(4) I already capture some of the profitability premium through the S&P 600 value and Bridgeway Ultra-small company fund (as per the above analysis).

It will be interesting to see what other options become available. But at this stage will just stay-the-course with my current factor targets, has served very well over last decade.

Best,

Robert
.
james22
Posts: 1992
Joined: Tue Aug 21, 2007 2:22 pm

Re: Updated Novy Marx Paper on Quality

Post by james22 »

jeffyscott wrote:Buying high quality assets without paying premium prices is just as
much value investing as buying average quality assets at discount
prices. Strategies that exploit the quality dimension of value are
profitable on their own, and accounting for both dimensions of value
by trading on combined quality and price signals yields dramatic
performance improvements over traditional value strategies.
Accounting for quality also yields significant performance
improvements for investors trading momentum as well as value.


Interesting...sounds like "growth-at-a-reasonable-price" wins :?:

IIRC, Robert Haugen had argued for that strategy long ago, for example in The New Finance: The Case Against Efficient Markets, published 1995.
And Buffett began applying the strategy 30 years before then.
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Updated Novy Marx Paper on Quality

Post by larryswedroe »

James
That's right, we now know that it was his genius--he did not have stock picking skills, but he identified this factor 40 years before academics were able to do so.
See this piece I wrote on the subject


http://www.cbsnews.com/8301-505123_162- ... he-market/

Best wishes
Larry
james22
Posts: 1992
Joined: Tue Aug 21, 2007 2:22 pm

Re: Updated Novy Marx Paper on Quality

Post by james22 »

Great piece, thanks.
jhd
Posts: 99
Joined: Fri Jan 18, 2008 5:18 pm
Location: San Francisco, CA

Re: Updated Novy Marx Paper on Quality

Post by jhd »

Robert T wrote: (1) The profitability premium seems more difficult to capture with available options than the market, size or value premiums re: the profitability load on the GMO Quality fund is only 0.23. On other growth funds (at least the ones I have briefly looked at) it is smaller.
Robert, have you or anyone else analyzed the DFA Growth funds for their profitability load? E.g. DUSLX and DSCGX. (Apologies if you have and I just can't find it.)
Post Reply