What are the "best" factors?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
sillysaver
Posts: 153
Joined: Thu Oct 08, 2015 5:24 pm

What are the "best" factors?

Post by sillysaver » Sun Sep 16, 2018 10:21 am

I was looking at moving from a portfolio of market-cap weighted indexes to factors and I find the choices available a little overwhelming. Which ones would you allocate capital to?

I frequently see references to Quality, Value, Size, Momentum and Yield. Do you need all these? How would you construct a portfolio that diversifies across multiple factors? I guess you could just evenly weight to each one. Is it better to use multi-factor funds? Is there such a thing as too many?

Allocating to a 3-fund portfolio is easy for DIY, but it seems like factor investing is less straightforward. I can see why this approach is being promoted today, as it necessitates all manner of funds and advisors, which is an opportunity for fees.

User avatar
vineviz
Posts: 2308
Joined: Tue May 15, 2018 1:55 pm

Re: What are the "best" factors?

Post by vineviz » Sun Sep 16, 2018 10:52 am

sillysaver wrote:
Sun Sep 16, 2018 10:21 am
Allocating to a 3-fund portfolio is easy for DIY, but it seems like factor investing is less straightforward. I can see why this approach is being promoted today, as it necessitates all manner of funds and advisors, which is an opportunity for fees.
Although I don't think factor investing is quite as complicated or mysterious as many members of this forum seem to think I DO believe that it is "less straightforward" than a market-cap index approach which is by definition the most simple approach.

In my mind the two biggest risks a factor investor faces are poor portfolio construction and lack of conviction.

Constructing a DIY factor-based portofolio, you MUST be diligent about ensuring that the different funds don't have opposing loadings on the same factor. If they do, you'll end up with a muddled portfolio that has no strong net loading on ANY factors. For this reason, I am a big believer in using a single well-designed multifactor fund like Vanguard U.S. Multifactor ETF (VFMF). This approach is likely to be more cost efficient and more factor-efficient.

Conviction comes into play in that some factors can outperform the market for a few years at a time. If you aren't SURE your portfolio is well-built, you run the risk of chasing the last successful factor instead of patiently waiting for the next one. This is another reason that multifactor fund is very useful.

Morningstar has described the most reliable factors as an "A-Team" of sorts: value, momentum, size, quality, and low volatility. Because you can very easily and cheaply add exposure to three of these with a simple small-cap value fund like iShares S&P Small-Cap 600 Value ETF (IJS), my opinion is that just replacing something between 50% to 100% of the US total stock market fund in your 3-fund portfolio with either IJS or VFMF is the easiest and best way to go.
Last edited by vineviz on Sun Sep 16, 2018 12:39 pm, edited 1 time in total.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

Random Walker
Posts: 3316
Joined: Fri Feb 23, 2007 8:21 pm

Re: What are the "best" factors?

Post by Random Walker » Sun Sep 16, 2018 11:55 am

I have the most confidence in value; it has both good risk based and behavioral based explanations behind it. People have gone back and forth about size, but intuitively it makes sense to me that smaller companies are riskier than big ones. MOREOVER, all the other factors (value, momentum, trend, profitability/quality) appear to be strongest in the smaller stocks. So hard to go wrong starting with a focus on tilting to SV. And if the funds can incorporate screens for momentum and quality, that would be good.

Dave

Random Walker
Posts: 3316
Joined: Fri Feb 23, 2007 8:21 pm

Re: What are the "best" factors?

Post by Random Walker » Sun Sep 16, 2018 12:00 pm

Trend (Time series momentum) is interesting. It’s purely behavioral, well known, and yet has some of the very strongest historical data supporting it. The size of the premium, it’s persistence, and its pervasiveness are strong. Many of us, like me, are much more inclined to have faith in risk based factors and less faith in behavioral based ones. But there are limits to arbitrage, and at some point the data is so strong it’s hard to ignore.

Dave

stlutz
Posts: 4839
Joined: Fri Jan 02, 2009 1:08 am

Re: What are the "best" factors?

Post by stlutz » Sun Sep 16, 2018 12:23 pm

I agree with vineviz that VFMF is a great choice. At the end of the day, "factor investing" is just making a [reasonable] bet that what has repeatedly backtested well will do well in actual portfolios going forward. If you're going to go with that approach, might as well make a few different bets as opposed to just picking one favorite. VFMF does this in a single fund.

User avatar
Taylor Larimore
Advisory Board
Posts: 27622
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: What are the "best" factors?

Post by Taylor Larimore » Sun Sep 16, 2018 12:28 pm

sillysaver wrote:
Sun Sep 16, 2018 10:21 am
I was looking at moving from a portfolio of market-cap weighted indexes to factors and I find the choices available a little overwhelming. Which ones would you allocate capital to?

I frequently see references to Quality, Value, Size, Momentum and Yield. Do you need all these? How would you construct a portfolio that diversifies across multiple factors? I guess you could just evenly weight to each one. Is it better to use multi-factor funds? Is there such a thing as too many?

Allocating to a 3-fund portfolio is easy for DIY, but it seems like factor investing is less straightforward. I can see why this approach is being promoted today, as it necessitates all manner of funds and advisors, which is an opportunity for fees.
sillysaver:

I am not a fan of "factor" investing (neither is Mr. Bogle). This Boglehead post explains why:

Factor Investing

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

sillysaver
Posts: 153
Joined: Thu Oct 08, 2015 5:24 pm

Re: What are the "best" factors?

Post by sillysaver » Sun Sep 16, 2018 12:29 pm

VFMF looks good. Based on the description it sounds like it's quality, value and momentum. With VFMF, maybe you could add IJS to get more small-cap exposure.

I have looked into Cambria's funds before. VAMO and GMOM seem like good ways to get exposure to value and momentum.

User avatar
nedsaid
Posts: 10665
Joined: Fri Nov 23, 2012 12:33 pm

Re: What are the "best" factors?

Post by nedsaid » Sun Sep 16, 2018 12:31 pm

Oddly enough, Momentum seems to be the best factor. Sometime ago, I saw an article, I think it was a Larry Swedroe article, that showed that Momentum had the best returns and was most persistent. I was pretty surprised. I would have thought that Value was the best.

My guess is that Quality/Profitability would be the second best factor. Quality seems to make value better as does setting momentum to neutral. Value has a great story as does Low Volatility. You want to buy Low Volatility when these stocks are in the Value area of the market.

Probably a multi-factor approach would be the best.

I have focused on Size, Value, and Momentum. I own individual stocks and I try to buy quality at a discount, not easy to do. My actual results have been a bit disappointing, particularly since Value has been on a 10 year vacation.
A fool and his money are good for business.

Theoretical
Posts: 1439
Joined: Tue Aug 19, 2014 10:09 pm

Re: What are the "best" factors?

Post by Theoretical » Sun Sep 16, 2018 1:01 pm

Value and Trend-following (Time-Series Momentum) are two of the best cases. Cross-sectional momentum has a lot of academic support and great results, but is nigh impossible to capture absent a lot of negative alpha.

ThrustVectoring
Posts: 583
Joined: Wed Jul 12, 2017 2:51 pm

Re: What are the "best" factors?

Post by ThrustVectoring » Sun Sep 16, 2018 2:37 pm

Bet Against Beta is really nice IMHO. There's really strong theoretical arguments for it (leveraged constrained investors bid up high risk / high return assets in lieu of levering up low risk / low return assets to an appropriate risk and reward), studies have shown it to pop up in many environments, and backtesting of Buffett's investing career has shown that most of his alpha is attributable to his use of reasonable amounts of leverage in this way.

Probably the best way of accessing this is through the treasury futures market, where there's extra headwind for the strategy from long-term bonds being required for regulatory compliance reasons in insurance companies and the like. But anyhow, implementation details and risk management is another whole post entirely. The tl;dr is to set aside something like $10k in savings to fund a single 2-year treasury futures contract (notional value $200k). The 20x leverage on a 1.8 year effective duration means you wind up with something with the volatility of a long-term bond but with significantly higher return.
Current portfolio: 60% VTI / 40% VXUS

hdas
Posts: 403
Joined: Thu Jun 11, 2015 8:24 am

Re: What are the "best" factors?

Post by hdas » Sun Sep 16, 2018 2:46 pm

ThrustVectoring wrote:
Sun Sep 16, 2018 2:37 pm
Bet Against Beta is really nice IMHO. There's really strong theoretical arguments for it (leveraged constrained investors bid up high risk / high return assets in lieu of levering up low risk / low return assets to an appropriate risk and reward), studies have shown it to pop up in many environments, and backtesting of Buffett's investing career has shown that most of his alpha is attributable to his use of reasonable amounts of leverage in this way.

Probably the best way of accessing this is through the treasury futures market, where there's extra headwind for the strategy from long-term bonds being required for regulatory compliance reasons in insurance companies and the like. But anyhow, implementation details and risk management is another whole post entirely. The tl;dr is to set aside something like $10k in savings to fund a single 2-year treasury futures contract (notional value $200k). The 20x leverage on a 1.8 year effective duration means you wind up with something with the volatility of a long-term bond but with significantly higher return.
Can you provide references for this strategy? Thanks
Stay the course and buy some more.

Random Walker
Posts: 3316
Joined: Fri Feb 23, 2007 8:21 pm

Re: What are the "best" factors?

Post by Random Walker » Sun Sep 16, 2018 2:58 pm

hdas wrote:
Sun Sep 16, 2018 2:46 pm
ThrustVectoring wrote:
Sun Sep 16, 2018 2:37 pm
Bet Against Beta is really nice IMHO. There's really strong theoretical arguments for it (leveraged constrained investors bid up high risk / high return assets in lieu of levering up low risk / low return assets to an appropriate risk and reward), studies have shown it to pop up in many environments, and backtesting of Buffett's investing career has shown that most of his alpha is attributable to his use of reasonable amounts of leverage in this way.

Probably the best way of accessing this is through the treasury futures market, where there's extra headwind for the strategy from long-term bonds being required for regulatory compliance reasons in insurance companies and the like. But anyhow, implementation details and risk management is another whole post entirely. The tl;dr is to set aside something like $10k in savings to fund a single 2-year treasury futures contract (notional value $200k). The 20x leverage on a 1.8 year effective duration means you wind up with something with the volatility of a long-term bond but with significantly higher return.
Can you provide references for this strategy? Thanks
Larry Swedroe just wrote on this. Thought I would try to understand it better before I posted it separately.

https://alphaarchitect.com/2018/09/13/h ... sk-taking/

Dave

stlutz
Posts: 4839
Joined: Fri Jan 02, 2009 1:08 am

Re: What are the "best" factors?

Post by stlutz » Sun Sep 16, 2018 3:08 pm

Theoretical wrote:
Sun Sep 16, 2018 1:01 pm
Cross-sectional momentum has a lot of academic support and great results, but is nigh impossible to capture absent a lot of negative alpha.
MTUM has been doing it with positive alpha.

https://www.portfoliovisualizer.com/fac ... sion=false

stlutz
Posts: 4839
Joined: Fri Jan 02, 2009 1:08 am

Re: What are the "best" factors?

Post by stlutz » Sun Sep 16, 2018 3:24 pm

ThrustVectoring wrote:
Sun Sep 16, 2018 2:37 pm
Bet Against Beta is really nice IMHO. There's really strong theoretical arguments for it (leveraged constrained investors bid up high risk / high return assets in lieu of levering up low risk / low return assets to an appropriate risk and reward), studies have shown it to pop up in many environments, and backtesting of Buffett's investing career has shown that most of his alpha is attributable to his use of reasonable amounts of leverage in this way.

Probably the best way of accessing this is through the treasury futures market, where there's extra headwind for the strategy from long-term bonds being required for regulatory compliance reasons in insurance companies and the like. But anyhow, implementation details and risk management is another whole post entirely. The tl;dr is to set aside something like $10k in savings to fund a single 2-year treasury futures contract (notional value $200k). The 20x leverage on a 1.8 year effective duration means you wind up with something with the volatility of a long-term bond but with significantly higher return.
Long-Term Treasury bonds are generally about 2-3 times as volatile as short-term Treasuries. Levering up ST bonds by 20x is far more risky than using long-term bonds.

And that's perhaps where this anomaly can break down. That is, when the evidence indicates that taking a lot of risk generally doesn't work out leads to the conclusion that one should take highly leveraged bets on a Betting-Against-Beta "factor".

User avatar
nedsaid
Posts: 10665
Joined: Fri Nov 23, 2012 12:33 pm

Re: What are the "best" factors?

Post by nedsaid » Sun Sep 16, 2018 3:57 pm

vineviz wrote:
Sun Sep 16, 2018 10:52 am
sillysaver wrote:
Sun Sep 16, 2018 10:21 am
Allocating to a 3-fund portfolio is easy for DIY, but it seems like factor investing is less straightforward. I can see why this approach is being promoted today, as it necessitates all manner of funds and advisors, which is an opportunity for fees.
Although I don't think factor investing is quite as complicated or mysterious as many members of this forum seem to think I DO believe that it is "less straightforward" than a market-cap index approach which is by definition the most simple approach.

In my mind the two biggest risks a factor investor faces are poor portfolio construction and lack of conviction.

Constructing a DIY factor-based portofolio, you MUST be diligent about ensuring that the different funds don't have opposing loadings on the same factor. If they do, you'll end up with a muddled portfolio that has no strong net loading on ANY factors. For this reason, I am a big believer in using a single well-designed multifactor fund like Vanguard U.S. Multifactor ETF (VFMF). This approach is likely to be more cost efficient and more factor-efficient.

Conviction comes into play in that some factors can outperform the market for a few years at a time. If you aren't SURE your portfolio is well-built, you run the risk of chasing the last successful factor instead of patiently waiting for the next one. This is another reason that multifactor fund is very useful.

Morningstar has described the most reliable factors as an "A-Team" of sorts: value, momentum, size, quality, and low volatility. Because you can very easily and cheaply add exposure to three of these with a simple small-cap value fund like iShares S&P Small-Cap 600 Value ETF (IJS), my opinion is that just replacing something between 50% to 100% of the US total stock market fund in your 3-fund portfolio with either IJS or VFMF is the easiest and best way to go.
Your comments about the Vanguard Multi-Factor ETF are very interesting. Just as Fidelity and Vanguard have become more like each other, Vanguard is now becoming more like DFA as well. For do-it-yourself investors, this is great news. Vanguard has a Market Neutral Fund out there, I wonder if they will go after the alternative investments and maybe go after AQR as well.
A fool and his money are good for business.

ThrustVectoring
Posts: 583
Joined: Wed Jul 12, 2017 2:51 pm

Re: What are the "best" factors?

Post by ThrustVectoring » Sun Sep 16, 2018 4:46 pm

stlutz wrote:
Sun Sep 16, 2018 3:24 pm

Long-Term Treasury bonds are generally about 2-3 times as volatile as short-term Treasuries. Levering up ST bonds by 20x is far more risky than using long-term bonds.
The two-year treasury futures contract currently has a DV01 of $35.77 per $200k contract. The 30-year ultra has a DV01 of $262.69 per $100k contract. If you have a fully-funded 30-year ultra treasury position, that $100k corresponds to 10 contracts at the risk tolerance level I specified, for an overall DV01 of $357.70. So overall, it's about 36% more sensitive to interest rates than buying 30-year treasuries outright.

So yeah, this is riskier than holding long-term bonds, which is intended. It's not "far" more risky, though, and it's still less risky than holding a 100% equities position (and many folks are fine with that as well!)
Current portfolio: 60% VTI / 40% VXUS

User avatar
JoMoney
Posts: 6284
Joined: Tue Jul 23, 2013 5:31 am

Re: What are the "best" factors?

Post by JoMoney » Sun Sep 16, 2018 5:03 pm

"Factor" investing is a pipe-dream that there is some persistent way to pick stocks that will outperform others.
It's an interesting, maybe useful, way to deconstruct and attribute performance in retrospect to some sort of common thematic, but the idea that stocks fitting that profile will somehow outperform is self-defeating when performance chasers actually try to do it.
Some of the popular factors are antithetical, like "value" and "momentum", or "small" and "quality".
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

KyleAAA
Posts: 6800
Joined: Wed Jul 01, 2009 5:35 pm
Contact:

Re: What are the "best" factors?

Post by KyleAAA » Sun Sep 16, 2018 6:04 pm

Factor investing is as simple as total market investing and doesn't require advisors or a mountain of funds. These days there are quite decent multifactor funds on the market but if you want to keep it ultra simple, exchange half of your total market index funds with small value index funds. That will get you 90% of the benefit. Quibbling over exact funds and implementation is a low value activity. And if you want to keep it really simple, don't bother looming for a good small value fund, just use IJS.

snailderby
Posts: 58
Joined: Thu Jul 26, 2018 11:30 am

Re: What are the "best" factors?

Post by snailderby » Sun Sep 16, 2018 6:42 pm

If the most popular factors are size, value, quality, momentum, and volatility, an S&P 600 value fund or ETF (like IJS or VIOV) should give you exposure to size, value, and to some extent quality (because the S&P 600 incorporates a basic profitability screen).

Vanguard's new multifactor fund should give you some level of exposure to all five factors. It tilts smaller than TSM because it's roughly 1/3 large, 1/3 mid, and 1/3 small. It screens out the most volatile stocks. And then it weights for value, quality, and momentum.

stlutz
Posts: 4839
Joined: Fri Jan 02, 2009 1:08 am

Re: What are the "best" factors?

Post by stlutz » Sun Sep 16, 2018 7:16 pm

The two-year treasury futures contract currently has a DV01 of $35.77 per $200k contract. The 30-year ultra has a DV01 of $262.69 per $100k contract. If you have a fully-funded 30-year ultra treasury position, that $100k corresponds to 10 contracts at the risk tolerance level I specified, for an overall DV01 of $357.70. So overall, it's about 36% more sensitive to interest rates than buying 30-year treasuries outright.
However, long-term rates are less volatile than short-term rates. So, 10x more duration risk for a long-term bond doesn't translate into 10x more realized volatility. So, you probably have about 75% more interest rate sensitivity.

Apologies to the OP for creating a tangent, but to circle back to the OP's original question--this is the problem with asking which is the "best" factor. Everyone has their favorite and everyone's favorite has problems with it. Doesn't mean you still can't "beat the market" with factor investing, but if it was obviously what the best investments are going to be, they would have already been bid up in price.

sillysaver
Posts: 153
Joined: Thu Oct 08, 2015 5:24 pm

Re: What are the "best" factors?

Post by sillysaver » Sat Sep 22, 2018 3:50 pm

I think I will go with an incremental approach that minimizes costs and taxable events, and optimizes tax placement. Besides the typical broad-based funds in my 401k, my IRA's and taxable accounts consist of:

- Schwab Total Stock Market Index Fund (SWTSX)
- Schwab International Index Fund (SWISX)
- Schwab Emerging Markets Equity ETF (SCHE)
- Schwab Total Bond Market Fund
- Van Eck Gold Miners ETF (GDX)
- Cambria Global Value (GVAL)

I am thinking of making the following changes:
- Highest expected returns in the Roth. Here I am thinking of emerging market yield (EYLD - Cambria Emerging Shareholder Yield or DGS - Wisdom Tree Small Cap Dividend) and/or a small cap value fund like IJS. My wife has another Roth that owns GVAL (Cambria Global Value) and GDX, which I will probably keep for value tilt and for exposure to "alts".

I've been thinking about increasing the overall allocation to emerging markets and small caps. We have a lot of time ahead of us and are still in the accumulation stages. I would probably completely fill up the Roth's with emerging/small/value and sell the broad-based index funds in these accounts

- Replace passive broad market funds in the traditional IRA with factor funds. I like VFMF (Vanguard Multifactor ETF) for the US portion and the Schwab Fundamental International Large Company Index ETF (FNDF) for international. Schwab's ETF's are extremely low cost and the fundamental index funds are good, too, at about 0.25%.
- Keep the bond index fund (already in the traditional IRA)
- Will probably leave my taxable accounts alone, or if the opportunity arises to harvest tax losses, sell and replace with the aforementioned fundamental and factor funds from Vanguard and Schwab (VFMF, FNDF).
- Keep market cap weighted index funds and target date funds in the 401k's, keeping costs to a minimum. We will want some "market" exposure to minimize tracking error regret anyways, and in these accounts with limited choices is all about managing costs.

selters
Posts: 619
Joined: Thu Feb 27, 2014 9:26 am

Re: What are the "best" factors?

Post by selters » Mon Sep 24, 2018 4:26 am

Small cap minus junk seems pretty consistent, even for short time periods.

User avatar
patrick013
Posts: 2425
Joined: Mon Jul 13, 2015 7:49 pm

Re: What are the "best" factors?

Post by patrick013 » Mon Sep 24, 2018 1:23 pm

Since 2008, factors that have outperformed large-cap stocks are
quality and small size. This year momentum is reported to be the
best. Usually value and size are considered best from long term
observations.
age in bonds, buy-and-hold, 10 year business cycle

gtwhitegold
Posts: 266
Joined: Fri Sep 21, 2012 1:55 pm

Re: What are the "best" factors?

Post by gtwhitegold » Tue Sep 25, 2018 1:36 pm

Outside of my TSP, I tilt all of my holdings. I personally have 60/40 between US and International. In International, I am split 60/40 Emerging and Developed with the overweight in Emerging Markets due to lower correlations with US Equities. To keep it simple, I have the same allocation between small and large cap stocks overweighting small caps. I am also using multi-factor funds where possible.

I am about to simplify things by holding Vanguard US funds (VSTCX and VFMF). I haven't decided on International, but I'm currently using AQR Multi-Style Funds for large cap developed and emerging markets large caps (QICNX and QEELX respectively).

International Developed Small Caps I'm moving to ISCF (iShares Multifactor) and a split between FEMS (First Trust AlphaDEX Emerging Markets Small Caps) and DGS (Wisdomtree Emerging Markets Small Cap Dividend) due to FEMS not including Korea.

If I had access to DFA, I would definitely use DEMSX (EM small caps) and would consider others.

Post Reply