Larry Swedroe: Look Past Expense Ratios

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Wed Jul 11, 2018 9:08 am

http://www.etf.com/sections/index-inves ... nopaging=1

Good article showing that there is a lot more to consider than just the expense ratio when looking at funds. I think for most of us Bogleheads, the ER is easy to appreciate. It’s certainly a good starting point, but it can be overshadowed by other less obvious costs and fund characteristics. Larry reviews this issue in three different SV funds with different ERs, then he really emphasizes the point when he reviews a couple of alternatives.

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Wed Jul 11, 2018 9:14 am

Larry doesn’t mention it in the article, but the choice of a more expensive fund with deeper exposures to the desired factors can make good sense at the portfolio level. If one is targeting a certain factor exposure within the portfolio, it likely can be achieved with less of the more expensive fund that has deeper exposures. Thus cost differences are not as large as initially apparent. Moreover, by using less of a more expensive and more concentrated fund, one requires less exposure to market beta to gain the same desired exposure to the other factors. And it’s market beta most people are trying to diversify away from.

Dave

User avatar
nedsaid
Posts: 9697
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by nedsaid » Wed Jul 11, 2018 9:29 am

Larry raised a good point that DFA Funds give you better factor exposure compared to similar Vanguard Funds. So the somewhat higher fees charged by DFA in comparison to Vanguard might be worth the higher exposure to factors. But even then, when the Stock Market is in a Growth phase, Vanguard products will outperform as their funds are less valuey and have a higher market cap than the similar DFA funds. We have been in a Large Growth market since the 2008-2009 Financial Crisis.

However, Vanguard has introduced Factor related funds. I have read a thread here that discusses this but I have never done a comparison with these new funds compared to DFA products. Also, there are ETFs out there that factor tilt and these could be considered as DFA substitutes. There are new alternatives out there that compete with DFA so the pat answers we gave on Vanguard vs. DFA threads may no longer hold true.
A fool and his money are good for business.

mickens16
Posts: 324
Joined: Wed Apr 06, 2011 8:52 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by mickens16 » Wed Jul 11, 2018 9:31 am

Help me understand. Why does Larry write for etf.com but never writes about etf’s?

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Wed Jul 11, 2018 9:34 am

nedsaid wrote:
Wed Jul 11, 2018 9:29 am

However, Vanguard has introduced Factor related funds. I have read a thread here that discusses this but I have never done a comparison with these new funds compared to DFA products. Also, there are ETFs out there that factor tilt and these could be considered as DFA substitutes. There are new alternatives out there that compete with DFA so the pat answers we gave on Vanguard vs. DFA threads may no longer hold true.
Competition is a good thing! This will place pressure on all providers to reduce costs. On the flip side though, as these exposures become cheaper, more liquid, more investable, perhaps we need to expect lower future returns.

Dave

UniversityEmployee9
Posts: 71
Joined: Wed Mar 18, 2015 10:01 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by UniversityEmployee9 » Wed Jul 11, 2018 12:28 pm

nedsaid wrote:
Wed Jul 11, 2018 9:29 am
Larry raised a good point that DFA Funds give you better factor exposure compared to similar Vanguard Funds. So the somewhat higher fees charged by DFA in comparison to Vanguard might be worth the higher exposure to factors. But even then, when the Stock Market is in a Growth phase, Vanguard products will outperform as their funds are less valuey and have a higher market cap than the similar DFA funds. We have been in a Large Growth market since the 2008-2009 Financial Crisis.

However, Vanguard has introduced Factor related funds. I have read a thread here that discusses this but I have never done a comparison with these new funds compared to DFA products. Also, there are ETFs out there that factor tilt and these could be considered as DFA substitutes. There are new alternatives out there that compete with DFA so the pat answers we gave on Vanguard vs. DFA threads may no longer hold true.
DFA actually manages some John Hancock branded ETFs now as well: https://www.jhinvestments.com/etf

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Wed Jul 11, 2018 12:36 pm

When I was making the decision to go the advisor route, I created lots of spreadsheets enumerating the costs and potential benefits. This essay covers the “stuff I don’t even know that I don’t know” column of that spreadsheet.

Dave

User avatar
danielc
Posts: 366
Joined: Sun Dec 10, 2017 4:48 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by danielc » Wed Jul 11, 2018 12:37 pm

Tax efficiency are also and important cost that can easily overwhelm the expense ratio. Total Stock Market funds (incl ETFs) from Vanguard and iShares are extremely tax efficient. So, for example, you would be better off with a portfolio that has TSM at its core plus a tilt than an otherwise equivalent portfolio that slices across caps, sectors, or styles.

Even for tax-efficient TSM funds, tax drag can be 10x larger than the ER.

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Wed Jul 11, 2018 1:13 pm

danielc wrote:
Wed Jul 11, 2018 12:37 pm
Tax efficiency are also and important cost that can easily overwhelm the expense ratio. Total Stock Market funds (incl ETFs) from Vanguard and iShares are extremely tax efficient. So, for example, you would be better off with a portfolio that has TSM at its core plus a tilt than an otherwise equivalent portfolio that slices across caps, sectors, or styles.

Even for tax-efficient TSM funds, tax drag can be 10x larger than the ER.
That makes the argument for using a relatively more expensive and deeper exposure factor fund. Less of the more expensive factor fund is required to achieve desired tilt, and relatively more of the cheaper and more tax efficient TSM/core fund.

Dave

User avatar
danielc
Posts: 366
Joined: Sun Dec 10, 2017 4:48 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by danielc » Wed Jul 11, 2018 2:05 pm

Random Walker wrote:
Wed Jul 11, 2018 1:13 pm
That makes the argument for using a relatively more expensive and deeper exposure factor fund. Less of the more expensive factor fund is required to achieve desired tilt, and relatively more of the cheaper and more tax efficient TSM/core fund.
Without looking at the specifics... yes. But I also advocate trying to do some of the math. For example, iShares' IJS is an ETF with similar factor weights to the ones that Larry advocated but with less than half the ER (25bp versus 52bp and 60bp). I also strongly suspect that it is also more tax efficieent (e.g. zero cap gains distributions so far). Alternatively, RZV has much higher factor loadings than either of the funds that Larry mentioned and it still has a lower ER (35bp). Look:

Code: Select all

       |  ER     |  Rm-Rf    SMB     HML   |  SMB/ER    HML/ER 
DFSVX  |  52 bp  |  1.02     0.80    0.65  |  0.0154    0.0125
BOSVX  |  60 bp  |  1.04     0.90    0.68  |  0.0150    0.0113
IJS    |  25 bp  |  0.95     0.81    0.50  |  0.0324    0.0200
RZV    |  35 bp  |  1.19     1.18    1.06  |  0.0337    0.0303
So with both IJS and RZV you are buying twice as much SMB per dollar of ER relative to Larry's preferred funds. With IJS and RZV you are also buying, respectively, 2x and 3x the HML exposure per dollar of ER. If we consider the fact that TSM has an ER of around 3bp, the numbrers would look even better for IJS and RZV:

Code: Select all

       | SMB/(ER - 3)   HML/(ER - 3)
DFSVX  | 0.0163         0.0133
BOSVX  | 0.0158         0.0119
IJS    | 0.0368         0.0227
RZV    | 0.0369         0.0331

Tax efficiency is also likely to tilt the scales toward the ETFs; so again, IJS and RZV. Comparing IJS and RZV, we see that RZV buys you more HML per dollar. This is partially offset by the fact that RZV has a much lower liquidity which translates into a higher bid/ask spread (0.27% for RZV vs 0.08% for IJS). If you are at Fidelity you pay no commission for IJS. In the end, I think that it's a wash between RZV and IJS, and both of them soundly beat DFSVX and BOSVX.

User avatar
grabiner
Advisory Board
Posts: 22191
Joined: Tue Feb 20, 2007 11:58 pm
Location: Columbia, MD

Re: Larry Swedroe: Look Past Expense Ratios

Post by grabiner » Wed Jul 11, 2018 9:14 pm

danielc wrote:
Wed Jul 11, 2018 2:05 pm
Tax efficiency is also likely to tilt the scales toward the ETFs; so again, IJS and RZV. Comparing IJS and RZV, we see that RZV buys you more HML per dollar. This is partially offset by the fact that RZV has a much lower liquidity which translates into a higher bid/ask spread (0.27% for RZV vs 0.08% for IJS). If you are at Fidelity you pay no commission for IJS. In the end, I think that it's a wash between RZV and IJS, and both of them soundly beat DFSVX and BOSVX.
And RZV, with its extremely low yield, is even more tax-efficient. Including both expense ratio and taxes, RZV is likely to be slightly less expensive than IJS, and comparable to VIOV, which tracks the same index.
Wiki David Grabiner

User avatar
tarheel
Posts: 359
Joined: Fri Sep 17, 2010 8:33 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by tarheel » Wed Jul 11, 2018 9:22 pm

danielc wrote:
Wed Jul 11, 2018 2:05 pm
Random Walker wrote:
Wed Jul 11, 2018 1:13 pm
That makes the argument for using a relatively more expensive and deeper exposure factor fund. Less of the more expensive factor fund is required to achieve desired tilt, and relatively more of the cheaper and more tax efficient TSM/core fund.
Without looking at the specifics... yes. But I also advocate trying to do some of the math. For example, iShares' IJS is an ETF with similar factor weights to the ones that Larry advocated but with less than half the ER (25bp versus 52bp and 60bp). I also strongly suspect that it is also more tax efficieent (e.g. zero cap gains distributions so far). Alternatively, RZV has much higher factor loadings than either of the funds that Larry mentioned and it still has a lower ER (35bp). Look:

Code: Select all

       |  ER     |  Rm-Rf    SMB     HML   |  SMB/ER    HML/ER 
DFSVX  |  52 bp  |  1.02     0.80    0.65  |  0.0154    0.0125
BOSVX  |  60 bp  |  1.04     0.90    0.68  |  0.0150    0.0113
IJS    |  25 bp  |  0.95     0.81    0.50  |  0.0324    0.0200
RZV    |  35 bp  |  1.19     1.18    1.06  |  0.0337    0.0303
So with both IJS and RZV you are buying twice as much SMB per dollar of ER relative to Larry's preferred funds. With IJS and RZV you are also buying, respectively, 2x and 3x the HML exposure per dollar of ER. If we consider the fact that TSM has an ER of around 3bp, the numbrers would look even better for IJS and RZV:

Code: Select all

       | SMB/(ER - 3)   HML/(ER - 3)
DFSVX  | 0.0163         0.0133
BOSVX  | 0.0158         0.0119
IJS    | 0.0368         0.0227
RZV    | 0.0369         0.0331

Tax efficiency is also likely to tilt the scales toward the ETFs; so again, IJS and RZV. Comparing IJS and RZV, we see that RZV buys you more HML per dollar. This is partially offset by the fact that RZV has a much lower liquidity which translates into a higher bid/ask spread (0.27% for RZV vs 0.08% for IJS). If you are at Fidelity you pay no commission for IJS. In the end, I think that it's a wash between RZV and IJS, and both of them soundly beat DFSVX and BOSVX.
Pretty impressive HML loading for RZV - what was weird was when I checked the loadings using a four-factor analysis it had a -0.50 momentum loading, whereas IJS did not. So performance could be very different with a big move down.

averagedude
Posts: 162
Joined: Sun May 13, 2018 3:41 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by averagedude » Wed Jul 11, 2018 9:30 pm

nedsaid wrote:
Wed Jul 11, 2018 9:29 am
Larry raised a good point that DFA Funds give you better factor exposure compared to similar Vanguard Funds. So the somewhat higher fees charged by DFA in comparison to Vanguard might be worth the higher exposure to factors. But even then, when the Stock Market is in a Growth phase, Vanguard products will outperform as their funds are less valuey and have a higher market cap than the similar DFA funds. We have been in a Large Growth market since the 2008-2009 Financial Crisis.

However, Vanguard has introduced Factor related funds. I have read a thread here that discusses this but I have never done a comparison with these new funds compared to DFA products. Also, there are ETFs out there that factor tilt and these could be considered as DFA substitutes. There are new alternatives out there that compete with DFA so the pat answers we gave on Vanguard vs. DFA threads may no longer hold true.
If you believe that DFA small cap value fund will outperform Vanguard's due to being more valuey and smaller, you should get the same results with a lower expense ratio by going 85 percent VSIAX and 15 percent IWC i shares micro cap etf. Better yet, you don't have to pay an advisor fee to obtain access to get these two funds.

User avatar
nedsaid
Posts: 9697
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by nedsaid » Thu Jul 12, 2018 12:03 pm

averagedude wrote:
Wed Jul 11, 2018 9:30 pm
nedsaid wrote:
Wed Jul 11, 2018 9:29 am
Larry raised a good point that DFA Funds give you better factor exposure compared to similar Vanguard Funds. So the somewhat higher fees charged by DFA in comparison to Vanguard might be worth the higher exposure to factors. But even then, when the Stock Market is in a Growth phase, Vanguard products will outperform as their funds are less valuey and have a higher market cap than the similar DFA funds. We have been in a Large Growth market since the 2008-2009 Financial Crisis.

However, Vanguard has introduced Factor related funds. I have read a thread here that discusses this but I have never done a comparison with these new funds compared to DFA products. Also, there are ETFs out there that factor tilt and these could be considered as DFA substitutes. There are new alternatives out there that compete with DFA so the pat answers we gave on Vanguard vs. DFA threads may no longer hold true.
If you believe that DFA small cap value fund will outperform Vanguard's due to being more valuey and smaller, you should get the same results with a lower expense ratio by going 85 percent VSIAX and 15 percent IWC i shares micro cap etf. Better yet, you don't have to pay an advisor fee to obtain access to get these two funds.
Well, actually I do own the Micro-Cap ETF that you mentioned alongside the Vanguard Small-Cap Value Index. I also own an ETF based on the S&P Small-Cap 600. Last I looked, Vanguard Small-Cap Value Index had been outperforming the DFA product as Vanguard is less valuey and has larger market cap, not surprising in a Large Growth market. The Micro-Cap ETF suffers from front-running but its returns have been acceptable. I have been very pleased with the S&P Small-Cap 600.

I did this because I didn't want to pay Merriman's organization 1% for Assets Under Management. The question is if the better factor exposure in DFA products can overcome the higher expense ratios and the advisor fee. I actually have been beating the Merriman Ultimate Buy and Hold Portfolio, in part because I have 66-67% stocks rather than the 60% in the Ultimate Buy and Hold. But even taking that into account, I have done well in comparison.
A fool and his money are good for business.

User avatar
nedsaid
Posts: 9697
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by nedsaid » Thu Jul 12, 2018 12:04 pm

UniversityEmployee9 wrote:
Wed Jul 11, 2018 12:28 pm
nedsaid wrote:
Wed Jul 11, 2018 9:29 am
Larry raised a good point that DFA Funds give you better factor exposure compared to similar Vanguard Funds. So the somewhat higher fees charged by DFA in comparison to Vanguard might be worth the higher exposure to factors. But even then, when the Stock Market is in a Growth phase, Vanguard products will outperform as their funds are less valuey and have a higher market cap than the similar DFA funds. We have been in a Large Growth market since the 2008-2009 Financial Crisis.

However, Vanguard has introduced Factor related funds. I have read a thread here that discusses this but I have never done a comparison with these new funds compared to DFA products. Also, there are ETFs out there that factor tilt and these could be considered as DFA substitutes. There are new alternatives out there that compete with DFA so the pat answers we gave on Vanguard vs. DFA threads may no longer hold true.
DFA actually manages some John Hancock branded ETFs now as well: https://www.jhinvestments.com/etf
Thanks, I am aware of this but have never taken time to analyze these products. They might be worth consideration. Hopefully others can chip in.
A fool and his money are good for business.

User avatar
nedsaid
Posts: 9697
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by nedsaid » Thu Jul 12, 2018 12:10 pm

Random Walker wrote:
Wed Jul 11, 2018 9:34 am
nedsaid wrote:
Wed Jul 11, 2018 9:29 am

However, Vanguard has introduced Factor related funds. I have read a thread here that discusses this but I have never done a comparison with these new funds compared to DFA products. Also, there are ETFs out there that factor tilt and these could be considered as DFA substitutes. There are new alternatives out there that compete with DFA so the pat answers we gave on Vanguard vs. DFA threads may no longer hold true.
Competition is a good thing! This will place pressure on all providers to reduce costs. On the flip side though, as these exposures become cheaper, more liquid, more investable, perhaps we need to expect lower future returns.

Dave
Hard to say, I am an advocate of buying good stuff and keeping it. I wouldn't dump DFA funds here because of increasing competition. Will all this competition lower future returns? Maybe. The thing is, we need to look at how much factor tilting is actually going on out there. Perhaps not as much as we think. Amazing numbers of people still have Ameriprise and Edward Jones brokerage accounts and are just beginning to understand the concept of diversification between stocks and bonds. Not sure how many folks even think about factors. Hedge funds, DFA, and AQR do a lot of factor investing but how big is this compared to the investment markets? Would be interesting if Larry would chip in with some data. Just because factors are known, it doesn't necessarily mean that they are being exploited enough to dampen future returns. Most folks out there, even here on Bogleheads, could care less.
A fool and his money are good for business.

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Thu Jul 12, 2018 1:23 pm

I wouldn’t sell an old fund and experience capital gains just to invest in a new fund with deeper factor exposures and lower costs, but I would strongly consider pointing new money towards the new fund.

Dave

User avatar
danielc
Posts: 366
Joined: Sun Dec 10, 2017 4:48 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by danielc » Thu Jul 12, 2018 2:04 pm

tarheel wrote:
Wed Jul 11, 2018 9:22 pm
Pretty impressive HML loading for RZV - what was weird was when I checked the loadings using a four-factor analysis it had a -0.50 momentum loading, whereas IJS did not. So performance could be very different with a big move down.
One thing I like about S&P 600 index is that stocks are screened for profitability, so they end up having a small RMW or QMJ loading. To be clear, the S&P 600 does not specifically aim for these factors, and the index itself predates those factors. What happens is that companies need to have shown profit over the last few quarters before they are allowed to enter the index (but they're not automatically removed when they become unprofitable). This rule naturally creates a "tilt" in the direction of RMW and QMJ. I don't know whether this has anything to do with their neutral MOM.

User avatar
danielc
Posts: 366
Joined: Sun Dec 10, 2017 4:48 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by danielc » Thu Jul 12, 2018 2:07 pm

nedsaid wrote:
Thu Jul 12, 2018 12:10 pm
The thing is, we need to look at how much factor tilting is actually going on out there. Perhaps not as much as we think. Amazing numbers of people still have Ameriprise and Edward Jones brokerage accounts and are just beginning to understand the concept of diversification between stocks and bonds. Not sure how many folks even think about factors.
But they don't have to. If they use an active fund from Edward Jones and the fund manager knows about factors, you still have all the people who have no idea what a factor is effectively chasing after them.

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Thu Jul 12, 2018 2:11 pm

I think Larry has written that the valuation spread between value and growth has not really decreased compared to the mid 1990’s, indicating that there is not significant chasing of the value factor.

Dave

User avatar
danielc
Posts: 366
Joined: Sun Dec 10, 2017 4:48 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by danielc » Thu Jul 12, 2018 2:15 pm

Random Walker wrote:
Thu Jul 12, 2018 2:11 pm
I think Larry has written that the valuation spread between value and growth has not really decreased compared to the mid 1990’s, indicating that there is not significant chasing of the value factor.

Dave
Bernstein expects the small and value premiums to both drop but not disappear. His advice is to assume that they'll be half of what they've been historically. Larry has made a case that at least a portion of the small and value premiums are real risks, but a lot of the arguments are behavioural and it is reasonable to worry that the behavioural portion of the premiums might not last.

User avatar
greg24
Posts: 3138
Joined: Tue Feb 20, 2007 10:34 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by greg24 » Thu Jul 12, 2018 2:15 pm

(Full disclosure: My firm, Buckingham Strategic Wealth, recommends AQR, Bridgeway and Dimensional funds in constructing client portfolios.)

User avatar
nisiprius
Advisory Board
Posts: 36033
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: Look Past Expense Ratios

Post by nisiprius » Thu Jul 12, 2018 2:38 pm

When has anyone recommending or selling a fund with a higher expense ratio ever said anything other than "yes, it costs a little more, but it's worth it?" And yet, Morningstar continues to say, most recently in 2016, that (my boldfacing)
Fund Fees Predict Future Success or Failure

The expense ratio is the most proven predictor of future fund returns. We find that it is a dependable predictor when we run the data. That's also what academics, fund companies, and, of course, Jack Bogle, find when they run the data.... There are many other things to consider, but investors should make expense ratios their first or second screen.
Factor loadings are theory. Expenses are dollars.

If we look at the performance of the three funds named by Larry Swedroe over the period of time they have existed, with BOSBX being Portfolio 1, blue; DFSVX, Portfolio 2, red; and VSIAX, Portfolio 3, yellow, we see:
Source
Image
I see three ways to interpret this.

1) The interpretation I like: it really didn't matter. After expenses, choosing the more expensive funds didn't do much good or much harm.

2) Moolah: Despite heavier factor loadings, the DFA fund happened to underperform the Vanguard fund. And the Bridgeway fund only outperformed the Vanguard fund by a hair.

3) Risk: The Bridgeway and DFA funds had higher standard deviation = a measure of volatility = an aspect of risk, and also had larger maximum drawdowns and worse worst years. And once you account for risk, you see that the Vanguard fund had Sharpe and Sortino ratios than the other two.

P.S. The time range is limited by VSIAX. Change it to VISVX (Investor shares), and the time range, now limited by BOSBX, goes back to September 2011, and, qualitatively, the results are the same as before. Throw out BOSBX and compare only DFSVX and VISVX, and DFSVX beats VISVX on return, but again has considerably higher risk and the difference in Sharpe ratio is tiny, 0.48 for DFSVX versus 0.46 for VSVIX.
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
Taylor Larimore
Advisory Board
Posts: 27106
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: Larry Swedroe: Look Past Expense Ratios

Post by Taylor Larimore » Thu Jul 12, 2018 2:49 pm

Larry Swedroe wrote:Using the regression tool available at Portfolio Visualizer enables us to see the factor loadings of the three funds (which shows how much exposure each has to the size and value factors). Data is from September 2011 (inception of BOSVX) through April 2018.

Fund----Size--Value
BOSVX--0.90--0.68
DFSVX--0.82--0.54
VSIAX--0.68---0.28

The table demonstrates that the three funds have very different exposures to these factors. Thus, investors should not just consider a fund’s expense ratio. Instead, investors should consider a fund’s cost relative to the factor exposure it provides. In other words, they should look at cost per unit of risk and return. Differences in factor loadings and fund construction and trading rules are why, while my firm has all three small value funds on its approved list, our recommended fund is BOSVX (despite its higher expenses).
Bogleheads:

According to Morningstar (6-30-2018), Larry might want to change his recommended high-expense factor fund (BOSVX) to the simple low-cost Vanguard index fund (VSIAX).

Fund---5-year return--Standard Deviation---Morningstar Risk--------Cost Projection per 10K

BOSVX*---11.73%--------------15.79------------ "Above Average"-----------------$335
VSIAX-----12.40%--------------12.49------------ "Below Average"------------------$40
*Longest period shown.
Morningstar: "If there's anything in the whole world of mutual funds that you can take to the bank, it's that expense ratios help you make a better decision. In every single time period and data point tested, low-cost funds beat high-cost funds."
Best wishes
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle

User avatar
danielc
Posts: 366
Joined: Sun Dec 10, 2017 4:48 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by danielc » Thu Jul 12, 2018 2:57 pm

nisiprius wrote:
Thu Jul 12, 2018 2:38 pm
If we look at the performance of the three funds named by Larry Swedroe over the period of time they have existed, with BOSBX being Portfolio 1, blue; DFSVX, Portfolio 2, red; and VSIAX, Portfolio 3, yellow, we see:
Source
Image
I see three ways to interpret this.

1) The interpretation I like: it really didn't matter. After expenses, choosing the more expensive funds didn't do much good or much harm.

2) Moolah: Despite heavier factor loadings, the DFA fund happened to underperform the Vanguard fund. And the Bridgeway fund only outperformed the Vanguard fund by a hair.

3) Risk: The Bridgeway and DFA funds had higher standard deviation = a measure of volatility = an aspect of risk, and also had larger maximum drawdowns and worse worst years. And once you account for risk, you see that the Vanguard fund had Sharpe and Sortino ratios than the other two.
That plot is only 6 years of data. You can't expect ANY strategy to show up on such a short period. The plot below shows 17 years of VSIAX (Portfolio 1) vs iShares IJS (Portfolio 2). This is still a short period, but it's much better. Now IJS looks like a winner despite the higher expense ratio. Yes, expenses matter. Risk tilts also matter a little.

Image

User avatar
danielc
Posts: 366
Joined: Sun Dec 10, 2017 4:48 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by danielc » Thu Jul 12, 2018 3:18 pm

Taylor Larimore wrote:
Thu Jul 12, 2018 2:49 pm
Bogleheads:

According to Morningstar (6-30-2018), Larry might want to change his recommended high-expense factor fund (BOSVX) to the simple low-cost Vanguard index fund (VSIAX).

Fund---5-year return--Standard Deviation---Morningstar Risk--------Cost Projection per 10K

BOSVX*---11.73%--------------15.79------------ "Above Average"-----------------$335
VSIAX-----12.40%--------------12.49------------ "Below Average"------------------$40
*Longest period shown.
Morningstar: "If there's anything in the whole world of mutual funds that you can take to the bank, it's that expense ratios help you make a better decision. In every single time period and data point tested, low-cost funds beat high-cost funds."
Best wishes
Taylor
Yeah. I doubt I would ever recommend BOSVX or a DFA fund because of the very high expense ratios. I will admit that I have bought IJS (iShares S&P 600 Value). That ETF has an ER higher than other funds I own (25bp vs ~ 8 bp), but lower than BOSVX and DFSVX (60 bp and 52 bp). One of the reasons I decided that IJS was acceptable is that it is very tax efficient. Using triceratop's spreadsheet for my own tax situation I found that IJS will cost me 54 bp including tax, while the Small Cap Value fund from Vanguard (VSS) will cost me 76 bp. So in my case, choosing the iShares fund with the higher ER actually reduces my total (ER + Tax) cost.

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Thu Jul 12, 2018 3:58 pm

danielc wrote:
Thu Jul 12, 2018 2:15 pm
Random Walker wrote:
Thu Jul 12, 2018 2:11 pm
I think Larry has written that the valuation spread between value and growth has not really decreased compared to the mid 1990’s, indicating that there is not significant chasing of the value factor.

Dave
Bernstein expects the small and value premiums to both drop but not disappear. His advice is to assume that they'll be half of what they've been historically. Larry has made a case that at least a portion of the small and value premiums are real risks, but a lot of the arguments are behavioural and it is reasonable to worry that the behavioural portion of the premiums might not last.
I believe the value, momentum, profitability/quality factor premia tend to be largest for the small equities. I think this is another reason to use small funds.

Your Bernstein 50% haircut sounds reasonable to me. Larry has written about papers showing 1/3 haircut on premia after publication, and I think his firm assumes 25% reduction from historical averages.

Dave

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Thu Jul 12, 2018 4:02 pm

nisiprius wrote:
Thu Jul 12, 2018 2:38 pm
When has anyone recommending or selling a fund with a higher expense ratio ever said anything other than "yes, it costs a little more, but it's worth it?" And yet, Morningstar continues to say, most recently in 2016, that (my boldfacing)
Fund Fees Predict Future Success or Failure

The expense ratio is the most proven predictor of future fund returns. We find that it is a dependable predictor when we run the data. That's also what academics, fund companies, and, of course, Jack Bogle, find when they run the data.... There are many other things to consider, but investors should make expense ratios their first or second screen.
Factor loadings are theory. Expenses are dollars.

If we look at the performance of the three funds named by Larry Swedroe over the period of time they have existed, with BOSBX being Portfolio 1, blue; DFSVX, Portfolio 2, red; and VSIAX, Portfolio 3, yellow, we see:
Source
Image
I see three ways to interpret this.

1) The interpretation I like: it really didn't matter. After expenses, choosing the more expensive funds didn't do much good or much harm.

2) Moolah: Despite heavier factor loadings, the DFA fund happened to underperform the Vanguard fund. And the Bridgeway fund only outperformed the Vanguard fund by a hair.

3) Risk: The Bridgeway and DFA funds had higher standard deviation = a measure of volatility = an aspect of risk, and also had larger maximum drawdowns and worse worst years. And once you account for risk, you see that the Vanguard fund had Sharpe and Sortino ratios than the other two.

P.S. The time range is limited by VSIAX. Change it to VISVX (Investor shares), and the time range, now limited by BOSBX, goes back to September 2011, and, qualitatively, the results are the same as before. Throw out BOSBX and compare only DFSVX and VISVX, and DFSVX beats VISVX on return, but again has considerably higher risk and the difference in Sharpe ratio is tiny, 0.48 for DFSVX versus 0.46 for VSVIX.
Can’t argue with the above data. I’m sure you appreciate the timeframe is short. I would not put a lot of weight into the individual Sharpe and Sortino ratios. I think how an individual asset affects the Sharpe and Sortino of the portfolio as a whole is much more relevant.

Dave

Longtermgrowth
Posts: 535
Joined: Thu Nov 26, 2015 1:59 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Longtermgrowth » Thu Jul 12, 2018 4:18 pm

danielc wrote:
Thu Jul 12, 2018 3:18 pm
Using triceratop's spreadsheet for my own tax situation I found that IJS will cost me 54 bp including tax, while the Small Cap Value fund from Vanguard (VSS) will cost me 76 bp. So in my case, choosing the iShares fund with the higher ER actually reduces my total (ER + Tax) cost.
Did you mean VBR (Vanguard Small-Cap Value ETF)? VSS is an international small cap fund (Vanguard FTSE All-World ex-US Small-Cap ETF).

Also, Vanguard S&P Small-Cap 600 Value ETF (VIOV) costs 5bp less than IJS while tracking the same index.

User avatar
danielc
Posts: 366
Joined: Sun Dec 10, 2017 4:48 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by danielc » Thu Jul 12, 2018 4:50 pm

Longtermgrowth wrote:
Thu Jul 12, 2018 4:18 pm
Did you mean VBR (Vanguard Small-Cap Value ETF)? VSS is an international small cap fund (Vanguard FTSE All-World ex-US Small-Cap ETF).
Oh! I got the names mixed up. VBR is comparable to IJS after tax.
Longtermgrowth wrote:
Thu Jul 12, 2018 4:18 pm
Also, Vanguard S&P Small-Cap 600 Value ETF (VIOV) costs 5bp less than IJS while tracking the same index.
IJS has 20x more assets than VIOV and a slightly smaller spread. Perhaps more importantly, Vanguard is not very keen on people buying it. If you go to the list of "all" Vanguard ETFs, you will not see VIOV. You have to be very intentional about finding it. So I'm less confident about VIOV's future. I don't know how tax efficient VIOV is (it's not in triceratop's spreadsheet), but in general, iShares ETFs tend to be a little bit more tax efficient than Vanguard equivalents.

User avatar
nisiprius
Advisory Board
Posts: 36033
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: Look Past Expense Ratios

Post by nisiprius » Thu Jul 12, 2018 5:01 pm

danielc wrote:
Thu Jul 12, 2018 2:57 pm
That plot is only 6 years of data.
I know, but the three funds Larry Swedroe named have only existed for about six years. I commented later in my post on what happens if I first change VSIAX to VISVX, and then remove BOSBX. But I really wanted to include BOSBX as it had both the highest expenses and the highest loadings, and therefore was strongest test of "worth the expense."
The plot below shows 17 years of VSIAX (Portfolio 1) vs iShares IJS (Portfolio 2).
I can't reproduce your results. Could you copy and paste the link from PortfolioVisualizer? Look for the word "link" just after Portfolio Analysis Results. Usually you can copy that link, probably by right-clicking on it and choosing from a list of options. Then paste it into a posting.

Here's what I get when I try to reproduce your results using VSIAX and IJS. The long link will get garbled visually but it will work as a link. You can see if I did something wrong.

https://www.portfoliovisualizer.com/bac ... ion2_2=100

Here's what I get if I replace VSIAX with Investor shares, VISVX. Now I see your seventeen years of data and a growth to a final value of about $60,000, but I'm not seeing the big difference you're seeing.

https://www.portfoliovisualizer.com/bac ... ion2_2=100

I'm not seeing a big difference, and the Sharpe ratios (which I wanted to look at) are virtually identical.
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
HomerJ
Posts: 11318
Joined: Fri Jun 06, 2008 12:50 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by HomerJ » Thu Jul 12, 2018 5:59 pm

Random Walker wrote:
Wed Jul 11, 2018 1:13 pm
Less of the more expensive factor fund is required to achieve desired tilt
Then you are depending WAY too much on the factor fund returning exactly what you THINK it will.

You and Larry make this way too hard.

You calculate some expected return based on past results from some factor, then you put a very precise amount of your portfolio in that factor based on the expected return.

But what if the factor doesn't return what you THINK it will?
Last edited by HomerJ on Thu Jul 12, 2018 7:04 pm, edited 1 time in total.
The J stands for Jay

User avatar
danielc
Posts: 366
Joined: Sun Dec 10, 2017 4:48 am

Re: Larry Swedroe: Look Past Expense Ratios

Post by danielc » Thu Jul 12, 2018 6:44 pm

nisiprius wrote:
Thu Jul 12, 2018 5:01 pm
I can't reproduce your results. Could you copy and paste the link from PortfolioVisualizer? Look for the word "link" just after Portfolio Analysis Results. Usually you can copy that link, probably by right-clicking on it and choosing from a list of options. Then paste it into a posting.
That was entirely my fault. As noted earlier, I mixed up the ETF symbols. I strongly suspect that I wrote VSS instead of VSIAX. Sorry about that.

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Thu Jul 12, 2018 6:47 pm

HomerJ wrote:
Thu Jul 12, 2018 5:59 pm
Random Walker wrote:
Wed Jul 11, 2018 1:13 pm
Less of the more expensive factor fund is required to achieve desired tilt
Then you are depending WAY too much on the factor fund returning exactly what you THINK it will.

You and Larry make this way too hard.

You calculate some expected return based on past results from some factor, then you put 15% of your portfolio in that factor, which lets you be more conservative in the rest of your portfolio.

But what if the factor doesn't return what you THINK it will?
Good points. I have a couple of responses.
1. I put a lot of trust in Monte Carlo Simulation which displays a broad range of potential outcomes in addition to an expected mean outcome. An appropriate success threshold (e.g 85% of simulations) creates good margin of safety.
2. The MCS inputs include lower expected factor premia compared to historical averages: I think it’s 25% less.
3. Historically, a portfolio diversified across factors doesn’t underperform by much when it does underperform. When it outperforms, it’s more significant.
4. Past returns affect current wealth, current valuations, and future expected returns, so can always rerun MCS in future and make adjustments to AA (overall stock/bond split) as necessary
5. Always have Plan B: work longer, decrease spending, etc.

I bet you’ll say that with MCS I’m really making it way harder than I need to! But I gotta say I think the results are meaningful and likely surprising for many people. In my case, my AA is much less aggressive than I would probably have created on my own. Do you make assumptions about future returns for the market in formulating a plan? If so, are they based on valuations or historical returns data?

Dave

User avatar
HomerJ
Posts: 11318
Joined: Fri Jun 06, 2008 12:50 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by HomerJ » Thu Jul 12, 2018 7:06 pm

Random Walker wrote:
Thu Jul 12, 2018 6:47 pm
HomerJ wrote:
Thu Jul 12, 2018 5:59 pm
Random Walker wrote:
Wed Jul 11, 2018 1:13 pm
Less of the more expensive factor fund is required to achieve desired tilt
Then you are depending WAY too much on the factor fund returning exactly what you THINK it will.

You and Larry make this way too hard.

You calculate some expected return based on past results from some factor, then you put 15% of your portfolio in that factor, which lets you be more conservative in the rest of your portfolio.

But what if the factor doesn't return what you THINK it will?
Good points. I have a couple of responses.
1. I put a lot of trust in Monte Carlo Simulation which displays a broad range of potential outcomes in addition to an expected mean outcome. An appropriate success threshold (e.g 85% of simulations) creates good margin of safety.
2. The MCS inputs include lower expected factor premia compared to historical averages: I think it’s 25% less.
3. Historically, a portfolio diversified across factors doesn’t underperform by much when it does underperform. When it outperforms, it’s more significant.
4. Past returns affect current wealth, current valuations, and future expected returns, so can always rerun MCS in future and make adjustments to AA (overall stock/bond split) as necessary
5. Always have Plan B: work longer, decrease spending, etc.

I bet you’ll say that with MCS I’m really making it way harder than I need to! But I gotta say I think the results are meaningful and likely surprising for many people. In my case, my AA is much less aggressive than I would probably have created on my own. Do you make assumptions about future returns for the market in formulating a plan? If so, are they based on valuations or historical returns data?

Dave
I don't know how MCS allows you to allocate precise amounts to certain factors. What's the margin of error? Plus/minus 2%, 5%, 8%?

You allocate a small amount to factors, and you plan on getting the bottom of the MCS returns? Or the middle?

I mean you talk about THIS fund being more concentrated than THAT fund, so you can buy less of it. How much does that calculation depend on expected returns being correct?
The J stands for Jay

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Thu Jul 12, 2018 7:25 pm

Homer J,
I’m not expecting anywhere near the accuracy &/or precision you describe. I’m more setting an AA based on generalities. It’s because of that lack of numeric foresight that I diversify the portfolio in every way I can. Historically factors (including market) have provided premia over the long term. Each of them can underperform for long stretches. They are generally uncorrelated with each other. Portfolio volatility can be a big drag. Makes sense to diversify across uncorrelated factors that have provided premia and decrease overall equity allocation. We all have plenty of market beta in our portfolios. Typical 60/40 TSM portfolio has about 90% of its risk wrapped up in a single factor. Just trying to diversify away from that.

With regard to MCS, it’s hard to not anchor on the Mean result, but I try to be cognizant of the wide dispersion


Dave

User avatar
nisiprius
Advisory Board
Posts: 36033
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: Look Past Expense Ratios

Post by nisiprius » Thu Jul 12, 2018 8:09 pm

danielc wrote:
Thu Jul 12, 2018 6:44 pm
nisiprius wrote:
Thu Jul 12, 2018 5:01 pm
I can't reproduce your results. Could you copy and paste the link from PortfolioVisualizer? Look for the word "link" just after Portfolio Analysis Results. Usually you can copy that link, probably by right-clicking on it and choosing from a list of options. Then paste it into a posting.
That was entirely my fault. As noted earlier, I mixed up the ETF symbols. I strongly suspect that I wrote VSS instead of VSIAX. Sorry about that.
No problem. I've personally used the "facepalm" emoji three times this week. So far. Please keep posting!
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: 36033
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: Look Past Expense Ratios

Post by nisiprius » Thu Jul 12, 2018 8:14 pm

Larry Swedroe has PM-ed me with two comments on my posting. The first refers to my quotation from Russell Kinnel of Morningstar.
Larry Swedroe wrote:saying is that expense ratio is the most powerful predictor -- you misused the statement--as should say when loadings on factors are the same.
The second is
note also the six years had almost no size premium and negative value premium so should expect BOSVX to underperform, it has higher loadings on factors that had no premiums-- and did of course have higher ER. But should not have been expected. Note over its life BOSVX has outperformed DFSVX by quite a bit despite the higher ER
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
nedsaid
Posts: 9697
Joined: Fri Nov 23, 2012 12:33 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by nedsaid » Thu Jul 12, 2018 8:22 pm

Random Walker wrote:
Thu Jul 12, 2018 3:58 pm
danielc wrote:
Thu Jul 12, 2018 2:15 pm
Random Walker wrote:
Thu Jul 12, 2018 2:11 pm
I think Larry has written that the valuation spread between value and growth has not really decreased compared to the mid 1990’s, indicating that there is not significant chasing of the value factor.

Dave
Bernstein expects the small and value premiums to both drop but not disappear. His advice is to assume that they'll be half of what they've been historically. Larry has made a case that at least a portion of the small and value premiums are real risks, but a lot of the arguments are behavioural and it is reasonable to worry that the behavioural portion of the premiums might not last.
I believe the value, momentum, profitability/quality factor premia tend to be largest for the small equities. I think this is another reason to use small funds.

Your Bernstein 50% haircut sounds reasonable to me. Larry has written about papers showing 1/3 haircut on premia after publication, and I think his firm assumes 25% reduction from historical averages.

Dave
Yes, the academic publishing effect. That is observed premiums are cut after the academics publish papers about them. Also hard to believe that the behavioral part of the factor premiums will go away, have the quantitative investors repealed the cycles of greed and fear? I think not.
A fool and his money are good for business.

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Thu Jul 12, 2018 9:05 pm

nedsaid wrote:
Thu Jul 12, 2018 8:22 pm

Yes, the academic publishing effect. That is observed premiums are cut after the academics publish papers about them. Also hard to believe that the behavioral part of the factor premiums will go away, have the quantitative investors repealed the cycles of greed and fear? I think not.
The more I learn, the more I believe in the behavioral explanations. Value has both risk and behavioral stories, and that strengthens my belief tremendously.

Dave

User avatar
HomerJ
Posts: 11318
Joined: Fri Jun 06, 2008 12:50 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by HomerJ » Thu Jul 12, 2018 9:43 pm

Random Walker wrote:
Thu Jul 12, 2018 7:25 pm
Homer J,
I’m not expecting anywhere near the accuracy &/or precision you describe.
Then I don't understand how you can "tune" your portfolio to be super efficient using factors. Larry has said many times that one can lower their equity exposure, if they use factors to give themselves higher expected returns on a smaller percentage of a portfolio.

But that's just adding another risk. What if factors don't deliver in the future like they did in the past?
The J stands for Jay

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Thu Jul 12, 2018 10:12 pm

HomerJ wrote:
Thu Jul 12, 2018 9:43 pm

But that's just adding another risk. What if factors don't deliver in the future like they did in the past?
Exactly! The point is that market beta is just another factor, no more and no less. It also may not deliver over any given investing horizon. Sounds like a strong reason to diversify across independent sources of return. Would be fascinating to see how a home biased Japan TSM investor has done compared to one diversified across factors over the last 40 years.

Dave

User avatar
FIREchief
Posts: 2234
Joined: Fri Aug 19, 2016 6:40 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by FIREchief » Thu Jul 12, 2018 10:32 pm

Every time I see this thread title I am reminded of one saying. "Pay no attention to the man behind the curtain." :annoyed
.
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.

fennewaldaj
Posts: 139
Joined: Sun Oct 22, 2017 11:30 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by fennewaldaj » Fri Jul 13, 2018 3:27 am

Random Walker wrote:
Thu Jul 12, 2018 10:12 pm
HomerJ wrote:
Thu Jul 12, 2018 9:43 pm

But that's just adding another risk. What if factors don't deliver in the future like they did in the past?
Exactly! The point is that market beta is just another factor, no more and no less. It also may not deliver over any given investing horizon. Sounds like a strong reason to diversify across independent sources of return. Would be fascinating to see how a home biased Japan TSM investor has done compared to one diversified across factors over the last 40 years.

Dave
Although I have elected to tilt my portfolio I just don't buy the argument that smb and hml are on as strong a footing as market beta. I realize the prior data would lead one to that conclusion however prior data can lead one to a lot of conclusions. So as has been discussed many times you should have a strong reason to believe that certain trends will persist. Stocks are structurally very different from bonds. I have less confidence in the inherent difference of small vs large company stocks or cheap stocks vs expensive stocks. This causes me to give a different level of believe to persistence of beta vs smb and hml. That said I still find the evidence strong enough to tilt to these factors but not enough to say only buy a small value fund or to buy long/short funds. I think of this in a Bayesian way with prior probabilities. My prior probability that stocks will out perform bonds is simply higher than the other two things. The data that we have pushes my view stocks will outperform bonds over long periods to pretty high levels (like 95%) but it has not pushed my confidence in the other conclusions as high as my priors for them were lower in the first place. So i might say my confidence of a positive smb is (~55%) and my confidence of a positive hml is (~70%). My priors for those two things were about the same but the hml data is stronger than the smb data.

Random Walker
Posts: 2787
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by Random Walker » Fri Jul 13, 2018 6:13 am

Fennewaldaj,
You and I think very similarly. We weigh more heavily the factors we have the greatest belief in, perhaps we only differ in one respect. Even the SV fund with the deepest exposure to SmB and HmL still has a huge dose of the market factor, so there’s plenty of that in a portfolio without a TSM fund.

Dave

User avatar
nisiprius
Advisory Board
Posts: 36033
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: Look Past Expense Ratios

Post by nisiprius » Fri Jul 13, 2018 6:44 am

Random Walker wrote:
Thu Jul 12, 2018 10:12 pm
...The point is that market beta is just another factor, no more and no less...
I think it is very different from other factors, in two aspects: ease of access, and coexistence with the efficient market hypothesis.

1) You get a good loading on market beta at no additional cost once you decide to invest in a stock mutual fund. PortfolioVisualizer is showing me loadings of 0.99 for Vanguard 500 Index Fund, 1.00 for Total Stock, 0.95 for American Growth Fund of America, 0.97 for DFA US Micro Cap.

Furthermore, you can get it with a long-only position. Increasingly, though, we are being told that factors "are" long-short portfolios and that you cannot get really effective loadings on them solely through tightly focussed stock screening--you really ought to have a long-short portfolio--and, usually, leverage. The need for short positions is another thing that makes the other factors different from market beta.

2) In order to believe that market beta has a persistent and worthwhile premium, one needs to believe only that businesses are capable of making money and that (in a loosely-coupled general sort of long-term way) stocks of money-making companies will be money-making stocks. That's reasonably easy to believe. More important, you can believe in it without assuming an inefficient market.

In order to believe that any other factor has superior risk-adjusted return, it is necessary to believe three things. First, that the market is inefficient. Second, that you can profit from that inefficiency and take money away from other investors without psyching them out individually; a simple screening formula does it. Third, that the inefficiency cannot disappear no matter how many people have heard about it.

Factor advocates do have arguments, some backed by data, supporting these premises, but I don't see how anyone can have the same strength of conviction in them. I don't know how true the EMH is, but there are hundreds of people out there saying, in various ways, "I can take money away from all the other investors who are trying to take money away from me." To have a strong conviction in factors, you have to judge their "I can take other investors' money" claims, against those of other contenders claiming the same thing.

No less a person than Eugene Fama has stated
Basically this a risk story the way we tell it, so there is no optimal portfolio.... in every asset pricing model, the market portfolio is always an efficient portfolio.... And investors can decide to tilt away from that based on their personal tastes. But that's what it amounts to. You can decide to tilt toward more value or smaller size based on your tastes for these dimensions of risk. But you needn't do it. You could also decide to go the other way. You could look at the premiums and say, no, I think I like the growth stocks better. Then, as long as you get a diversified portfolio of them, I can't argue with that either. So there's a whole multi-dimensional continuum here of efficient portfolios that anybody can decide to buy that I can't quarrel with. And I have no recommendations about because I think it's totally a matter of taste. If you eat oranges and I eat apples I can't really quarrel very much with that.
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
vineviz
Posts: 543
Joined: Tue May 15, 2018 1:55 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by vineviz » Fri Jul 13, 2018 8:34 am

nisiprius wrote:
Fri Jul 13, 2018 6:44 am
Increasingly, though, we are being told that factors "are" long-short portfolios and that you cannot get really effective loadings on them solely through tightly focussed stock screening--you really ought to have a long-short portfolio--and, usually, leverage. The need for short positions is another thing that makes the other factors different from market beta.
You seem to be mixing up the way that factor returns are computed (by constructing differential, or long-short) portfolios with the way that factor portfolios are invested. The size premium is measured by subtracting the returns of the largest stocks from the returns of the smallest stocks, but that doesn't mean we are actually shorting anything when we invest. All the factor based funds we discuss here, and virtually every one available to individual investors, are long-only portfolios.
nisiprius wrote:
Fri Jul 13, 2018 6:44 am
In order to believe that any other factor has superior risk-adjusted return, it is necessary to believe three things. First, that the market is inefficient.
Actually, no.

You have to be careful with your mental model here, because if we are talking about factors besides market beta then we are talking about a world in which there isn't just one measure of risk. In such a world, a measure (like Sharpe) that includes only one dimension of risk is not sufficient to describe risk-adjusted returns.

If volatility is the only measure of risk, then the efficiency frontier is a hyperbolic line and there is a single tangency point. But if there are two dimensions of risk (aka two factors), then the efficient frontier becomes three dimensional (imagine that hyperbola being rotated on an axis) and you get a tangency line. In short, the multifactor efficient frontier is not identical to the mean-variance efficient frontier. They are related, but the former exists in more dimensions than the latter.

So you don't need to assume market inefficiency. It is sufficient to assume merely that different investors care more or less about their exposure to different sources (or dimensions) of risk.

User avatar
nisiprius
Advisory Board
Posts: 36033
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: Look Past Expense Ratios

Post by nisiprius » Fri Jul 13, 2018 8:49 am

vineviz wrote:
Fri Jul 13, 2018 8:34 am
...So you don't need to assume market inefficiency. It is sufficient to assume merely that different investors care more or less about their exposure to different sources (or dimensions) of risk...
On the one hand, you're correct. And it meshes in well with the quotation from Fama that I cited.

On the other hand, it just pushes the debate into a different area. In order to show that factor-based investing is superior to total-market investing for me, you need to convince me that my personal taste for risk is different from the market's, and that the factor-based portfolio is more to my taste than a total market portfolio.

For example, say that I happen to like a high probability of a small downside and a low probability of a big upside--positive skew, "lottery ticket." Then, if what I think I've read is correct, I ought to like small growth stocks. On the other hand, if I want a high probability of a slow and steady upside coupled with a low probability of a severe downside, I ought to like small value stocks. That would be personal preference. However, I think I hear factor advocates suggesting that everybody ought to prefer small value, perhaps because widespread taste for small growth has driven its price up. That would make sense if I do not care about the skewness of returns, but if I like a positive skew, then I might still prefer small growth if shaping the distribution that way is worth the extra price.

I don't even know a good way to assess risk tolerance itself. I've never seen anything that purported to help you match at a finer level of detail.

Do you know of a risk tolerance questionnaire that would help an investor assess whether, based on their personal tastes, they ought to prefer total market investing or factor-based investing?
Last edited by nisiprius on Fri Jul 13, 2018 8:50 am, edited 1 time in total.
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: 36033
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: Look Past Expense Ratios

Post by nisiprius » Fri Jul 13, 2018 8:50 am

Larry Swedroe has PMed me to say:
IMO big and common mistake

you don't need as you say to believe that another factor has higher risk adjusted return, It can even have lower risk-adjusted return in isolation and add value because it adds a unique/independent and thus low correlating risk which might mix well with the other risks in the portfolio. That's huge error to make and leads to bad conclusions, all one needs is a risk premium and unique risks.

Also you can easily access factors in long only portfolios as DFA and Bridgeway and others do and one way to do it is to avoid the stocks in the short side of the factor

Best wishes
Larry
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
vineviz
Posts: 543
Joined: Tue May 15, 2018 1:55 pm

Re: Larry Swedroe: Look Past Expense Ratios

Post by vineviz » Fri Jul 13, 2018 9:22 am

nisiprius wrote:
Fri Jul 13, 2018 8:49 am
For example, say that I happen to like a high probability of a small downside and a low probability of a big upside--positive skew, "lottery ticket." Then, if what I think I've read is correct, I ought to like small growth stocks. On the other hand, if I want a high probability of a slow and steady upside coupled with a low probability of a severe downside, I ought to like small value stocks. That would be personal preference. However, I think I hear factor advocates suggesting that everybody ought to prefer small value, perhaps because widespread taste for small growth has driven its price up. That would make sense if I do not care about the skewness of returns, but if I like a positive skew, then I might still prefer small growth if shaping the distribution that way is worth the extra price.
Given the diversity of factor investors, I wouldn't be surprised to find that one of them said just about anything.

But my take on this is probably similar to yours, though I might frame it slightly differently.

Some people derive utility from being (or at least feeling) popular and "in the know". Such people might prefer to dine at the newest trendy restaurants, always drive a latest car, wear the current fashions, or have the highest tech phone. These people might also prefer to have in their portfolios the stocks or funds that have performed best in the recent past, as much as for their cocktail conversational value as for their actual long-term returns. Also, these people are probably not spending a lot of time online at www.bogleheads.org, I'd wager.

These people, quite rationally given their utility curve, will disproportionately own large cap growth stocks. For these people the "risk" of not owning stocks like Amazon, salesforce.com, or Netflix is great because (for them) the risk of potentially missing out is more painful than the risk of low expected returns.

Would I say that such a person could earn higher returns by investing in Vanguard U.S. Multifactor ETF (VFMF) than owning a portfolio of individual stocks like AMZN, CRM, and NFLX? Yeah, probably. Would I say that such a person SHOULD own VFMF? Maybe, but not if I had an honest appraisal of what they really value and don't value.
nisiprius wrote:
Fri Jul 13, 2018 8:49 am
I don't even know a good way to assess risk tolerance itself. I've never seen anything that purported to help you match at a finer level of detail.
Yeah, me either, but I think people tend to sort themselves out. Right?

I mean, I look at the academic studies on factor investing and find the evidence compelling. I've spent time investigating it and thinking about it, and got there. I certainly don't care about owning the latest and greatest of ANYTHING merely because it is flashy and new, and don't spend any more time at cocktail parties than I need to. So my portfolio works for me, and I don't mind arguing its merits because I guess I have an implicit trust that IF you (for instance) find my arguments compelling that it will work for you too. And if you DON'T find my arguments compelling, they presumably you'll some approach that DOES work for you more compelling.

Post Reply