factor based investing

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.

factor based investing

Postby larryswedroe » Tue Jul 16, 2013 5:32 pm

http://www.indexuniverse.com/hot-topics/19294-swedroe-factor-focused-investing-better.html

More and more you are seeing papers on this subject and the benefits of diversifying across factors that explain returms, including things like momentum, the carry trade, and now profitability

Hope you find it helpful

Best wishes
Larry
larryswedroe
 
Posts: 12188
Joined: Thu Feb 22, 2007 9:28 am
Location: St Louis MO

Re: factor based investing

Postby Random Walker » Wed Jul 17, 2013 9:38 am

For the individual investor, is the best way to incorporate momentum into a portfolio through patient buying and selling by the fund rather than a momentum fund? Likewise is a profitability screen to existing fund better than a specific fund based on profitability? Thanks,

Dave
Random Walker
 
Posts: 825
Joined: Fri Feb 23, 2007 9:21 pm

Re: factor based investing

Postby Gecko10x » Wed Jul 17, 2013 11:00 am

Larry, do you recommend including low-volatility in a factor-based portfolio?
Your other recent posts have focused on size, value, momentum, and profitability.
Gecko10x
 
Posts: 95
Joined: Thu Jun 06, 2013 9:10 pm

Re: factor based investing

Postby matonplayer » Wed Jul 17, 2013 12:59 pm

I too am wondering how best to use this information. If I have 5% of my portfolio in small cap value, should I reduce my total stock market allocation by 5% and use that money to buy a momentum-based fund?
matonplayer
 
Posts: 98
Joined: Wed May 22, 2013 11:02 pm

Re: factor based investing

Postby Oliver » Wed Jul 17, 2013 11:08 pm

The paper also gives support for etfs like Guggenheim S&P 500 Pure Value ETF (RPV) and Guggenheim S&P SmallCap 600 Pure Value ETF (RZV). These etfs allocate the weighting by value-ness not by capitalization.
Oliver
 
Posts: 73
Joined: Tue Sep 04, 2007 12:49 am

Re: factor based investing

Postby larryswedroe » Thu Jul 18, 2013 10:49 am

Re low volatility strategies suggest you read this post on the issue
http://www.cbsnews.com/8301-505123_162-57591323/can-you-win-with-low-risk-stock-strategies/
Larry
larryswedroe
 
Posts: 12188
Joined: Thu Feb 22, 2007 9:28 am
Location: St Louis MO

Re: factor based investing

Postby Oliver » Thu Jul 18, 2013 5:43 pm

larryswedroe wrote:Re low volatility strategies suggest you read this post on the issue
http://www.cbsnews.com/8301-505123_162-57591323/can-you-win-with-low-risk-stock-strategies/
Larry

Have you had a chance to review the paper listed below. They do address some of the issues raised in the paper you linked to. The authors are not academics. However, it has been published.

Second, our new sample enables us to address the main criticisms existing studies have received. For example, Bali and Cakici (2008) argue that the negative empirical relation between risk and return is driven by small-caps, especially the strong negative returns of high (idiosyncratic) volatility stocks. We address this concern by including only constituents of the S&P/IFC Investable Emerging Markets Index in our sample, and additionally by conducting a robustness test on the 50% largest stocks within this already liquid universe. Others, such as Scherer (2010), have argued that some of the effect may be due to exposure to the classic value premium. We therefore also adjust for such implicit factor loadings, using both parametric and non-parametric techniques. Yet another critique, by Amenc, Martellini, Goltz and Sahoo (2011), is that the relation between risk and return turns positive over longer holding periods. We therefore also analyze the performance characteristics of portfolios sorted on past risk over holding periods up to 5 years.


http://papers.ssrn.com/sol3/papers.cfm? ... id=2249660

Blitz, David, Pang, Juan and Van Vliet, Pim, The Volatility Effect in Emerging Markets (April 12, 2013). Emerging Markets Review, Vol. 16, pp. 31-45, 2013. Available at SSRN: http://ssrn.com/abstract=2249660
Oliver
 
Posts: 73
Joined: Tue Sep 04, 2007 12:49 am

Re: factor based investing

Postby boggler » Thu Jul 18, 2013 5:44 pm

How many factors are there now?

Small cap
Value
Momentum
Profitability
what else?
boggler
 
Posts: 593
Joined: Thu Feb 07, 2013 3:29 am

Re: factor based investing

Postby larryswedroe » Thu Jul 18, 2013 6:43 pm

boggler
liquidity
term
default (credit)
Larry
larryswedroe
 
Posts: 12188
Joined: Thu Feb 22, 2007 9:28 am
Location: St Louis MO

Re: factor based investing

Postby nisiprius » Fri Jul 19, 2013 8:35 am

I'm sorry, Larry, I have a real problem with claims based on low correlations between synthetic factors, rather than low correlations between actual assets.

An actual asset that had true negative correlation with stocks--robust, persistent, and reliable, not just one that lasts for a few years and could be due to chance sampling error, as in long-term bonds and stocks (which have zero long-term correlation)--would be a wonderful thing. But I don't think there are any, if you restrict yourself to long positions.

You cannot invest in a factor. There are small-cap index funds and there are large-cap index funds but there are no SMB index funds.

Obviously, you can and do get negative correlations if you allow derivatives and short positions. Obviously, the BEARX fund has a strong, persistent negative correlation with stocks:

Image

In fact, Morningstar is showing a 15-year beta of -1.14.

Well then, would you or anyone else suggest that it would improve a portfolio that includes stocks to add BEARX to it? Of course not. Why not? Because the negative correlation is synthetic. It is a result of taking short positions, so it cancels out the return right along with the standard deviation--and adds cost into the bargain. Actually it speaks very well for BEARX that it lost less money than Total Stock made.

According to the 2010 SBBI Yearbook, the correlation coefficients from 1928-2009 of the actual asset classes themselves are:

Large growth vs. large value, 0.81
Small growth vs. small value, 0.87

and, the lowest correlation between any two members of the classic Fama-French factors,

Large growth vs. small value, 0.74

These are the actual assets you can use in a portfolio, so I think these must be the correlations that matter.

Anyone can create sums and differences of things that will have any correlations you like with other things.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
nisiprius
Advisory Board
 
Posts: 25756
Joined: Thu Jul 26, 2007 10:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: factor based investing

Postby scone » Fri Jul 19, 2013 9:11 am

"In only 17 percent of the months, both factors underperform the market simultaneously."

This is the part that really interests me. I'm thinking this feature should lower my risk overall, and lower the chance of rebalacning into a black hole, like Mrs. Watanabe in Japan. Thank you!
"Sometimes you eat the bear and sometimes the bear eats you." -- Preacher Roe
scone
 
Posts: 863
Joined: Wed Jul 11, 2012 5:46 pm

Re: factor based investing

Postby richard » Fri Jul 19, 2013 10:09 am

Once upon a time Fama & French came up with a model to the effect that you could analyze portfolio returns based on exposure to three factors. This model was supposed to do an excellent job of explaining portfolio returns.

Many suggested the FF model implied that results could be improved by diversifying by factors. In other words, given two portfolios with the same factor exposure, the one with a lower correlation among the factors would do better - various "barbell" type portfolios.

If so, the original FF work would seem wrong - it's not just factor exposure which determines returns. It seems hard to have it both ways - is it exposure or exposure plus portfolio construction? If the later, than FF should update their model.

The current crop of factor based investing just seems to push this further by adding more factors. It has the added issue of seeming to be data mining, over-fitting the data and other statistical problems.

Of course, if you don't believe markets are reasonably efficient and don't believe in the usual relation between risk and return, then this may be less of a problem.
richard
 
Posts: 7474
Joined: Tue Feb 20, 2007 4:38 pm

Re: factor based investing

Postby Gecko10x » Fri Jul 19, 2013 10:48 am

Boy, the factors seem to get very little love around here 8-)
Gecko10x
 
Posts: 95
Joined: Thu Jun 06, 2013 9:10 pm

Re: factor based investing

Postby Browser » Fri Jul 19, 2013 10:59 am

Anything that gets you to an equity allocation that is NOT cap-weighted has done better than cap-weighting for quite awhile now. Factors, whatever...
If we have data, let’s look at data. If all we have are opinions, let’s go with mine. – Jim Barksdale
Browser
 
Posts: 3075
Joined: Wed Sep 05, 2012 5:54 pm

Re: factor based investing

Postby boggler » Fri Jul 19, 2013 11:16 am

There are so many factors these days that it seems like a slippery slope. How many funds would the optimal portfolio have to have?
boggler
 
Posts: 593
Joined: Thu Feb 07, 2013 3:29 am

Re: factor based investing

Postby Sammy_M » Fri Jul 19, 2013 11:19 am

As momentum and profitability are behavioral factors rather than risk factors, there should be skepticism about whether they will persist. Sooo...what is the upside versus downside of choosing to ignore these additional factors? From my standpoint...

Downside:
(D1) Missing the returns if they do persist.

Upside:
(U1) Saving maybe 30-50 bps in expenses and
(U2) Avoiding the frictional costs of higher turnover.
(U3) Not running the risk of abandoning the strategy at the wrong time.

Of these, I think U3 is the most significant (for me personally). I have far more confidence in my ability to stick with the 3F model than consistently rebalancing into, for example, a momentum fund. I therefore choose to only incorporate momentum/profitability when it is embedded within an existing fund (e.g., AQR Core). In using that fund, I'm betting D1 > U1 and U2. Who knows if that will be the case.
User avatar
Sammy_M
 
Posts: 1697
Joined: Sun Nov 25, 2007 9:30 am

Re: factor based investing

Postby vesalius » Fri Jul 19, 2013 3:23 pm

boggler wrote:There are so many factors these days that it seems like a slippery slope. How many funds would the optimal portfolio have to have?

Larry has on multiple occasions here stated he feels it best to find a core type fund that incorporates all of the factors as opposed to an individual fund for each factor. So far AQR and DFA are the only ones I know of, although they are out of reach of most DIYers. Bridgeway may be working on something, but again their new products in this theme have been advisor only. Rafi is probably the closest openly available and maybe Flexshares will add momentum and profitability screens on their Factor tilt line of ETFs in the future. iShares has the new iShares Enhanced U.S. small and large ETFs that are supposed to incorporate all of the factors, but these are listed as active ETFs not passive.
vesalius
 
Posts: 681
Joined: Tue Jul 13, 2010 8:00 pm
Location: Texas

Re: factor based investing

Postby Clearly_Irrational » Fri Jul 19, 2013 3:34 pm

nisiprius wrote:An actual asset that had true negative correlation with stocks--robust, persistent, and reliable, not just one that lasts for a few years and could be due to chance sampling error, as in long-term bonds and stocks (which have zero long-term correlation)--would be a wonderful thing. But I don't think there are any, if you restrict yourself to long positions.


Gold fills that requirement, but it has some drawbacks that make it a less than optimal choice in many cases, the largest of which are the lack of an intrinsic return and high volatility.
User avatar
Clearly_Irrational
 
Posts: 2608
Joined: Thu Oct 13, 2011 4:43 pm

Re: factor based investing

Postby nisiprius » Fri Jul 19, 2013 3:51 pm

Clearly_Irrational wrote:
nisiprius wrote:An actual asset that had true negative correlation with stocks--robust, persistent, and reliable, not just one that lasts for a few years and could be due to chance sampling error, as in long-term bonds and stocks (which have zero long-term correlation)--would be a wonderful thing. But I don't think there are any, if you restrict yourself to long positions.
Gold fills that requirement, but it has some drawbacks that make it a less than optimal choice in many cases, the largest of which are the lack of an intrinsic return and high volatility.
No, it does not have a robust, persistent, reliable negative correlation with stocks. What is accurate to say is that it has a near-zero correlation with stocks. -0.16 according to a 2010 essay by William J. Bernstein, so, OK, technically on the negative side of zero, but that's so small it's bound to depend on what endpoints you pick.

This isn't "negative correlation:"

Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
nisiprius
Advisory Board
 
Posts: 25756
Joined: Thu Jul 26, 2007 10:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: factor based investing

Postby Browser » Fri Jul 19, 2013 4:01 pm

Rick Ferri has suggested that assets with varying correlations are the very ones you want to hold for best diversification purposes. By that criterion it would seem that gold/stocks are a good pair.
If we have data, let’s look at data. If all we have are opinions, let’s go with mine. – Jim Barksdale
Browser
 
Posts: 3075
Joined: Wed Sep 05, 2012 5:54 pm

Re: factor based investing

Postby Gecko10x » Fri Jul 19, 2013 4:04 pm

Browser wrote:Rick Ferri has suggested that assets with varying correlations are the very ones you want to hold for best diversification purposes. By that criterion it would seem that gold/stocks are a good pair.


Rick does not suggest Gold, AFAIK, because it doesn't have an inherit positive return.

I agree re: varying/zero correlation.
Gecko10x
 
Posts: 95
Joined: Thu Jun 06, 2013 9:10 pm

Re: factor based investing

Postby Clearly_Irrational » Fri Jul 19, 2013 4:06 pm

nisiprius wrote:No, it does not have a robust, persistent, reliable negative correlation with stocks. What is accurate to say is that it has a near-zero correlation with stocks. -0.16 according to a 2010 essay by William J. Bernstein, so, OK, technically on the negative side of zero, but that's so small it's bound to depend on what endpoints you pick.


Well, if that's not good enough then bonds don't qualify either since intermediate total bond market runs at about +0.29 to total stock market. (Simba data or assetcorrelation.com)
User avatar
Clearly_Irrational
 
Posts: 2608
Joined: Thu Oct 13, 2011 4:43 pm

Re: factor based investing

Postby Oliver » Fri Jul 19, 2013 7:11 pm

vesalius wrote:
boggler wrote:There are so many factors these days that it seems like a slippery slope. How many funds would the optimal portfolio have to have?

Larry has on multiple occasions here stated he feels it best to find a core type fund that incorporates all of the factors as opposed to an individual fund for each factor. So far AQR and DFA are the only ones I know of, although they are out of reach of most DIYers. Bridgeway may be working on something, but again their new products in this theme have been advisor only. Rafi is probably the closing openly available and maybe Flexshares will add momentum and profitability screens on their Factor tilt line of ETFs in the future. iShares has the new iShares Enhanced U.S. small and large ETFs that are supposed to incorporate all of the factors, but these are listed as active ETFs not passive.


AQR is available at TD Ameritrade. I purchased AQR International Core Equity L QICLX. There was no minimum of but there is a fee ($50) to purchase the mutual fund.

Bridgeway small cap momentum fund is available without an advisor. As you mentioned, some of bridgeway funds are available only through an advisor. :( As a long-term investor with Bridgeway, I hope they will be available to all.
Oliver
 
Posts: 73
Joined: Tue Sep 04, 2007 12:49 am

Re: factor based investing

Postby nedsaid » Fri Jul 19, 2013 9:52 pm

I actually constructed my portfolio to try to take advantage of some of these factors, a couple of them unknowingly.

Another poster looked at my post and a description of my portfolio and commented that I was constructing a Total Stock Market Index with higher costs. Not quite correct, but a big grain of truth to what he said.

My concern is that as more and more factors get added, an investor that tries to invest in them all in effect reproduces an expensive total market index. I understand the small/value tilt and believe in it but it seems that the newer factors are contradictory. How does Value and Momentum go together in the same portfolio? There seems to be some contradiction there. But I actually am doing both but more value than momentum.

This is a very interesting discussion.
A fool and his money are good for business.
User avatar
nedsaid
 
Posts: 2579
Joined: Fri Nov 23, 2012 1:33 pm

Re: factor based investing

Postby Blue » Fri Jul 19, 2013 11:00 pm

Oliver wrote:AQR is available at TD Ameritrade. I purchased AQR International Core Equity L QICLX. There was no minimum of but there is a fee ($50) to purchase the mutual fund.



https://www.aqrfunds.com/OurFunds/Equit ... imums.aspx

The fund's web page (and M*) indicates a minimum investment of $5 million?

ER of 1.01% beginning after ER cap expires in April 2015.
User avatar
Blue
 
Posts: 1017
Joined: Sat Jul 12, 2008 11:18 pm

Re: factor based investing

Postby boggler » Sat Jul 20, 2013 2:18 am

In an age of high-frequency trading and increasing market efficiency, do we expect beta to dominate more and more over time? Which of these factors have staying power?

In other words, is Total Stock Market going to get better or worse, from the perspective of investing in the various factors that explain returns?
boggler
 
Posts: 593
Joined: Thu Feb 07, 2013 3:29 am

Re: factor based investing

Postby Oliver » Sat Jul 20, 2013 4:12 am

Blue wrote:
Oliver wrote:AQR is available at TD Ameritrade. I purchased AQR International Core Equity L QICLX. There was no minimum of but there is a fee ($50) to purchase the mutual fund.



https://www.aqrfunds.com/OurFunds/Equit ... imums.aspx

The fund's web page (and M*) indicates a minimum investment of $5 million?

ER of 1.01% beginning after ER cap expires in April 2015.
I am aware of the that. However, td ameritrade offers it with no minimums. I have also purchased pcrix at TD Ameritrade. Pimco's site lists high minimums for that fund. The I in the fund name stands for institutional version of that fund.
Oliver
 
Posts: 73
Joined: Tue Sep 04, 2007 12:49 am

Re: factor based investing

Postby Oliver » Sat Jul 20, 2013 4:47 am

tarnation wrote:
tarnation wrote:
Sammy_M wrote:
tarnation wrote:It doesn't look to me that the bridgeway SCM fund purchases are restricted. The Omni funds are and they have that stated in the prospectus, but the SCM fund prospectus does not state that it is restricted.

Not available through Wellstrade. They "do not have a sales agreement" and I was guessing because they require an advisor.

I looked it up in TDAM and they give me a "buy" button and let me start an order. It shows $2k min. however, it also gave me a buy button for AQR ASMOX, but doesn't show any min info or anything. Will have to call them and see what the deal is.

Called TDAM and they said BRSMX was available with $2k initial investment. More interesting is they said the AQR SCM fund (ASMOX) was also available and the fund had waived minimums for TDAM. I didn't really know AQR was a choice for me.

I am not the first to purchase AQM funds from TD Ameritrade. This is from a thread on momentum funds. If I was going to move funds to TD Ameritrade to gain access, I would call and confirm that the funds would be available to me.
Oliver
 
Posts: 73
Joined: Tue Sep 04, 2007 12:49 am

Re: factor based investing

Postby Sammy_M » Sat Jul 20, 2013 5:42 am

Fidelity too. $75 to buy, $0 to sell.
User avatar
Sammy_M
 
Posts: 1697
Joined: Sun Nov 25, 2007 9:30 am

Re: factor based investing

Postby nisiprius » Sat Jul 20, 2013 8:56 am

Clearly_Irrational wrote:
nisiprius wrote:No, it does not have a robust, persistent, reliable negative correlation with stocks. What is accurate to say is that it has a near-zero correlation with stocks. -0.16 according to a 2010 essay by William J. Bernstein, so, OK, technically on the negative side of zero, but that's so small it's bound to depend on what endpoints you pick.


Well, if that's not good enough then bonds don't qualify either since intermediate total bond market runs at about +0.29 to total stock market. (Simba data or assetcorrelation.com)
I have a few time that bonds have near-zero correlation with stocks. There's an urban myth to the effect that bonds have negative correlation with stocks. I don't know where that comes from--recency, I think.

The distinction is important, because it is only a strong negative correlation that produces the almost-magic effects, like a zero-return asset improving a portfolio as a whole. A zero-return asset can improve a portfolio even if it has zero return if it has negative correlation. It can't do that if it merely has zero correlation. In order to improve a portfolio, a zero correlation asset needs to have a decent return of its own.

The requirements for low correlation to be valuable are clear. It's not enough to have low correlation. The assets also need to be very roughly similar in both return and in volatility. A pair of assets that meet that criterion are stocks and long-term bonds. Stocks and cash don't: they have zero correlation, and cash has a positive return, but cash doesn't help because it has zero volatility. Stocks and gold don't: they have zero correlation, and both have high volatility, but gold has zero return.

The diversification/low correlation effect of Total Bond with Total Stock is muted because Total Bond has low volatility.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
nisiprius
Advisory Board
 
Posts: 25756
Joined: Thu Jul 26, 2007 10:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: factor based investing

Postby richard » Sat Jul 20, 2013 9:34 am

Sammy_M wrote:As momentum and profitability are behavioral factors rather than risk factors, there should be skepticism about whether they will persist. Sooo...what is the upside versus downside of choosing to ignore these additional factors? From my standpoint...

Profitability is an earnings yield variant. It's about as behavioral as p/e.
richard
 
Posts: 7474
Joined: Tue Feb 20, 2007 4:38 pm

Re: factor based investing

Postby Sammy_M » Sat Jul 20, 2013 9:37 am

nisi,
Please define what you mean when you say "doesn't help" and "improve a portfolio". Sharpe ratio?
User avatar
Sammy_M
 
Posts: 1697
Joined: Sun Nov 25, 2007 9:30 am

Re: factor based investing

Postby richard » Sat Jul 20, 2013 9:43 am

nisiprius wrote:The distinction is important, because it is only a strong negative correlation that produces the almost-magic effects, like a zero-return asset improving a portfolio as a whole. A zero-return asset can improve a portfolio even if it has zero return if it has negative correlation. It can't do that if it merely has zero correlation. In order to improve a portfolio, a zero correlation asset needs to have a decent return of its own.

The requirements for low correlation to be valuable are clear. It's not enough to have low correlation. The assets also need to be very roughly similar in both return and in volatility. A pair of assets that meet that criterion are stocks and long-term bonds. Stocks and cash don't: they have zero correlation, and cash has a positive return, but cash doesn't help because it has zero volatility. Stocks and gold don't: they have zero correlation, and both have high volatility, but gold has zero return.

Everyone seems to be searching for the assets your describe.

Consider the effects of those efforts. If I find two negatively correlated assets, what will I do? What effect will that (together with everyone else's similar efforts) have on correlations and other characteristics of the assets?
richard
 
Posts: 7474
Joined: Tue Feb 20, 2007 4:38 pm

Re: factor based investing

Postby Sammy_M » Sat Jul 20, 2013 9:53 am

richard wrote:
Sammy_M wrote:As momentum and profitability are behavioral factors rather than risk factors, there should be skepticism about whether they will persist. Sooo...what is the upside versus downside of choosing to ignore these additional factors? From my standpoint...

Profitability is an earnings yield variant. It's about as behavioral as p/e.

Not sure I get your point. You're agreeing that high profitability (i.e., free cash flow) is not a risk factor, correct? Is it that you're arguing that "value" (high BtM) is not a risk factor either?
User avatar
Sammy_M
 
Posts: 1697
Joined: Sun Nov 25, 2007 9:30 am

Re: factor based investing

Postby richard » Sat Jul 20, 2013 10:01 am

Sammy_M wrote:
richard wrote:
Sammy_M wrote:As momentum and profitability are behavioral factors rather than risk factors, there should be skepticism about whether they will persist. Sooo...what is the upside versus downside of choosing to ignore these additional factors? From my standpoint...

Profitability is an earnings yield variant. It's about as behavioral as p/e.

Not sure I get your point. You're agreeing that high profitability (i.e., free cash flow) is not a risk factor, correct? Is it that you're arguing that "value" (high BtM) is not a risk factor either?

I'm saying it is a risk factor, not behavior. p/e is about as standard a valuation metric as there is. Yield has an obvious economic meaning.

The Fama French model says value is a risk factor. Value may correlate with economic risks (for example, distress), but it's not obvious that it's an economic risk in itself.
richard
 
Posts: 7474
Joined: Tue Feb 20, 2007 4:38 pm

Re: factor based investing

Postby Sammy_M » Sat Jul 20, 2013 10:13 am

Value is associated with low p/e or low p/b. Not so for Profitability.

From Novy-Marx paper,
Profitability, measured by gross profits-to-assets, has roughly the same power as book-to-market predicting the cross-section of average returns. Profitable
firms generate significantly higher returns than unprofitable firms, despite having significantly higher valuation ratios.

So profitability would be closer to high e/b (since earnings are linked to profits, and assets are a component of book). I'm not sure how you're linking that to market price (p).
User avatar
Sammy_M
 
Posts: 1697
Joined: Sun Nov 25, 2007 9:30 am

Re: factor based investing

Postby Browser » Sat Jul 20, 2013 12:57 pm

As I understand it, the "factors" represent different kinds of risk; i.e., they are risk factors. You are presumably diversifying your risk exposure via factor investing. But it seems somewhat arbitrary and open-ended how "risk" is defined. It also seems circular to me to data-mine and then explain a particular dimension of total returns in terms of some sort of ex-post-facto "risk." Where does it begin and where does it end? I'm tending to agree with Nisiprius that it makes more sense to diversify across actual asset classes than invest in synthetic asset classes that drop out of a factor analysis study. Having spent a fair amount of time messing with factor analysis in my grad school days, you can kinda get the results to say whatever you want them to say; there's more than one solution. And for any particular solution it's somewhat subjective how the factors are defined based on factor loadings. I'm a little skeptical of this methodology as a way guide one's investment allocation strategy.
If we have data, let’s look at data. If all we have are opinions, let’s go with mine. – Jim Barksdale
Browser
 
Posts: 3075
Joined: Wed Sep 05, 2012 5:54 pm

Re: factor based investing

Postby Clearly_Irrational » Sat Jul 20, 2013 1:02 pm

nisiprius wrote:The requirements for low correlation to be valuable are clear. It's not enough to have low correlation. The assets also need to be very roughly similar in both return and in volatility.


I don't think that's true, after all the volatility of stocks and bonds are significantly different. To be fair, you mentioned afterwards that the effect is muted due to that.

nisiprius wrote:Stocks and gold don't: they have zero correlation, and both have high volatility, but gold has zero return.


Gold has zero intrinsic return, but it's return has not be zero over my lifetime. During times of economic stability it tends to actually have a negative return, while during periods of instability it has a strong positive return. Gold has several problems as an asset so I'm certainly not saying it's a clear & obvious addition. Other than Gold I haven't found any other assets that aren't strongly correlated with either stocks or bonds.

To get back on subject, I've found the small factor to be extremely reliable and well explained, value seems useful but it is on less solid ground justification wise, momentum is obviously behavioral but I think it's one that will persist though it's effect is fairly small. Many of the additional factors that have been developed beyond the four factor model seem to be of such low explanatory power I'm not sure it's worth trying to capture them.
User avatar
Clearly_Irrational
 
Posts: 2608
Joined: Thu Oct 13, 2011 4:43 pm

Re: factor based investing

Postby Random Walker » Sat Jul 20, 2013 1:15 pm

Clearly Irrational stated that some factors seem to be of such low explanatory power that maybe its not worthwhile to try to capture them. Obviously this depends tremendously on the costs involved in capturing: transaction, ER, taxes. I think it is quite possible that the costs are too great for some factors to be captured in their own funds, but perhaps much less costly to somewhat incorporate them by screens in already existing funds with patient buying and selling.

Dave
Random Walker
 
Posts: 825
Joined: Fri Feb 23, 2007 9:21 pm


Return to Investing - Theory, News & General

Who is online

Users browsing this forum: avalpert, baddriver, ClevrChico, exeunt, fourwedge, joe8d, John3754, NOVACPA, Rysto, SpaceCommander, SteveM, wilked and 94 guests