Vanguard Multifactor fund VFMF has had a bad year

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
000
Posts: 2818
Joined: Thu Jul 23, 2020 12:04 am

Vanguard Multifactor fund VFMF has had a bad year

Post by 000 »

I was recently reading some old posts and found some discussion about VFMF (Vanguard U.S. Multifactor ETF) as a possible core portfolio holding.

As of 8/31/2020, VFMF has returned -8.11% YTD versus +9.39% by its benchmark (Russell 3000).

One year performance is +2.12% for VFMF and +21.44% for the benchmark.

VFMF's YTD and 1YR performances are close to those of the Vanguard Value ETF.

Questions
1. Why didn't the other factors (momentum, quality) help this year?
2. Is anyone thinking of reducing or increasing their VFMF allocation?
2b. For those holding VFMF, is it as a replacement for -- or complement to -- a raw beta fund, like a total market fund?
3. Is VFMF a fair representation of a multi-factor strategy? Or does it represent bad construction and/or tracking error?
Last edited by 000 on Fri Sep 18, 2020 3:41 pm, edited 1 time in total.
Tingting1013
Posts: 429
Joined: Mon Aug 24, 2020 5:44 pm

Re: VFMF has had a bad year

Post by Tingting1013 »

It hasn’t performed well as a Total Stock substitute that’s for sure.

On the other hand it has beaten Small Cap Value since inception.
User avatar
Forester
Posts: 1561
Joined: Sat Jan 19, 2019 2:50 pm
Location: UK

Re: VFMF has had a bad year

Post by Forester »

000 wrote: Fri Sep 18, 2020 12:49 am I was recently reading some old posts and found some discussion about VFMF (Vanguard U.S. Multifactor ETF) as a possible core portfolio holding.

As of 8/31/2020, VFMF has returned -8.11% YTD versus +9.39% by its benchmark (Russell 3000).

One year performance is +2.12% for VFMF and +21.44% for the benchmark.

VFMF's YTD and 1YR performances are close to those of the Vanguard Value ETF.

Questions
1. Why didn't the other factors (momentum, quality) help this year?
2. Is anyone thinking of reducing or increasing their VFMF allocation?
3. Is VFMF a fair representation of a multi-factor strategy? Or does it represent bad construction and/or tracking error?
Perhaps momentum gets watered down by all the other ingredients. Pick enough factors and it washes out to a market cap weighted fund.
User avatar
vineviz
Posts: 7843
Joined: Tue May 15, 2018 1:55 pm

Re: VFMF has had a bad year

Post by vineviz »

000 wrote: Fri Sep 18, 2020 12:49 am Questions
1. Why didn't the other factors (momentum, quality) help this year?
Momentum HAS helped so far this year, adding some performance relative to the total market. Quality has NOT helped because the quality factor has had negative returns year to date. Size helped a little bit, as did VFMF's slightly higher market beta, but value has had a VERY terrible year so far and there's no escaping that.
000 wrote: Fri Sep 18, 2020 12:49 am2. Is anyone thinking of reducing or increasing their VFMF allocation?
I'd hope not, but given the Boglehead affinity for performance-chasing I wouldn't be surprised.
000 wrote: Fri Sep 18, 2020 12:49 am 3. Is VFMF a fair representation of a multi-factor strategy? Or does it represent bad construction and/or tracking error?
I think most people would say VFMF is a relatively well-constructed multifactor fund.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
YRT70
Posts: 686
Joined: Sat Apr 27, 2019 8:51 am

Re: VFMF has had a bad year

Post by YRT70 »

vineviz wrote: Fri Sep 18, 2020 6:26 am
000 wrote: Fri Sep 18, 2020 12:49 am Questions
1. Why didn't the other factors (momentum, quality) help this year?
Momentum HAS helped so far this year, adding some performance relative to the total market. Quality has NOT helped because the quality factor has had negative returns year to date. Size helped a little bit, as did VFMF's slightly higher market beta, but value has had a VERY terrible year so far and there's no escaping that.
Has size really helped? YTD for IJR (small cap blend) is -13%, VTI is up 5%.
typical.investor
Posts: 2299
Joined: Mon Jun 11, 2018 3:17 am

Re: VFMF has had a bad year

Post by typical.investor »

vineviz wrote: Fri Sep 18, 2020 6:26 am
000 wrote: Fri Sep 18, 2020 12:49 am Questions
1. Why didn't the other factors (momentum, quality) help this year?
Momentum HAS helped so far this year, adding some performance relative to the total market. Quality has NOT helped because the quality factor has had negative returns year to date. Size helped a little bit, as did VFMF's slightly higher market beta, but value has had a VERY terrible year so far and there's no escaping that.
000 wrote: Fri Sep 18, 2020 12:49 am2. Is anyone thinking of reducing or increasing their VFMF allocation?
I'd hope not, but given the Boglehead affinity for performance-chasing I wouldn't be surprised.
000 wrote: Fri Sep 18, 2020 12:49 am 3. Is VFMF a fair representation of a multi-factor strategy? Or does it represent bad construction and/or tracking error?
I think most people would say VFMF is a relatively well-constructed multifactor fund.
I have no reason to doubt that, yet ...:

CAGR Mar 2018 - Aug 2020 (life of VFMF)
0.20% VFMF Vanguard US Multifactor ETF
7.29% USMF WisdomTree US Multifactor
6.08% MFUS PIMCO RAFI Dyn Multi-Factor US Eq ETF

Rm-Rf __ SMB __ HML __ MOM __ Alpha __ Annual Alpha
0.99 __ 0.37 __ 0.33 __ 0.11 __ -0.25% __ -3.04% __ VFMF Vanguard US Multifactor ETF
0.97 __ 0.13 __ 0.19 __ 0.18 __ 0.03% __ 0.41% __ USMF WisdomTree US Multifactor
0.95 __ -0.01 __ 0.21 __ 0.14 __ -0.12% __ -1.49% __ MFUS PIMCO RAFI Dyn Multi-Factor US Eq ETF

Obviously size and value hurt VFMF relative to the other funds, and it's lighter MOM exposure didn't help it as much as USMF and MFUS benefited. But what explains the negative alpha (trick question here as alpha means unexplained). Maybe only the idiosyncrasies of a short time period.

Then again VB (US small) had a 4.87% CAGR with a larger size loading (0.65) and similar MOM (.12) and HML (.27)

https://www.portfoliovisualizer.com/bac ... ion3_3=100

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

Since VFMF was heavy on quality and VB negative, I will conclude it was that and value.

https://www.portfoliovisualizer.com/bac ... ion3_3=100
rkhusky
Posts: 10213
Joined: Thu Aug 18, 2011 8:09 pm

Re: VFMF has had a bad year

Post by rkhusky »

vineviz wrote: Fri Sep 18, 2020 6:26 am I'd hope not, but given the Boglehead affinity for performance-chasing I wouldn't be surprised.
Yes, like so many that invest in small value stocks because they did well in the past. Or those that jump into momentum or quality or profitability or low volatility stocks because they performed well in the past.
MotoTrojan
Posts: 10659
Joined: Wed Feb 01, 2017 8:39 pm

Re: VFMF has had a bad year

Post by MotoTrojan »

rkhusky wrote: Fri Sep 18, 2020 7:10 am
vineviz wrote: Fri Sep 18, 2020 6:26 am I'd hope not, but given the Boglehead affinity for performance-chasing I wouldn't be surprised.
Yes, like so many that invest in small value stocks because they did well in the past. Or those that jump into momentum or quality or profitability or low volatility stocks because they performed well in the past.
I'd rather performance chase something that has a good long-run history but has done terrible in the last few years, than join this VGT/QQQ bandwagon of true performance-chasers :P.
AdrianC
Posts: 22
Joined: Mon May 23, 2016 6:31 am

Re: VFMF has had a bad year

Post by AdrianC »

I've noticed there's almost no difference in performance between VFMF and VSEQX (Strategic Equity Fund). They track very closely. Portfolios don't look the same, so probably coincidence, though they are both managed by the Vanguard Quantitative Equity Group. We own VSEQX in the kid's 529 plans.
rkhusky
Posts: 10213
Joined: Thu Aug 18, 2011 8:09 pm

Re: VFMF has had a bad year

Post by rkhusky »

MotoTrojan wrote: Fri Sep 18, 2020 7:17 am
rkhusky wrote: Fri Sep 18, 2020 7:10 am
vineviz wrote: Fri Sep 18, 2020 6:26 am I'd hope not, but given the Boglehead affinity for performance-chasing I wouldn't be surprised.
Yes, like so many that invest in small value stocks because they did well in the past. Or those that jump into momentum or quality or profitability or low volatility stocks because they performed well in the past.
I'd rather performance chase something that has a good long-run history but has done terrible in the last few years, than join this VGT/QQQ bandwagon of true performance-chasers :P.
Chasing after Large Growth performance is no different than chasing after Small Value.
MotoTrojan
Posts: 10659
Joined: Wed Feb 01, 2017 8:39 pm

Re: VFMF has had a bad year

Post by MotoTrojan »

rkhusky wrote: Fri Sep 18, 2020 7:53 am
MotoTrojan wrote: Fri Sep 18, 2020 7:17 am
rkhusky wrote: Fri Sep 18, 2020 7:10 am
vineviz wrote: Fri Sep 18, 2020 6:26 am I'd hope not, but given the Boglehead affinity for performance-chasing I wouldn't be surprised.
Yes, like so many that invest in small value stocks because they did well in the past. Or those that jump into momentum or quality or profitability or low volatility stocks because they performed well in the past.
I'd rather performance chase something that has a good long-run history but has done terrible in the last few years, than join this VGT/QQQ bandwagon of true performance-chasers :P.
Chasing after Large Growth performance is no different than chasing after Small Value.
One is chasing recent outperformance, one is chasing long-term outperformance during recent (significant) underperformance; in principal I get your point, but there is a difference IMHO. Beyond that, I personally believe there is evidence that one is expected to outperform while the other underperform based on long-term (many market cycle) returns and human behavior, but I know we won't agree on that one today.
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: VFMF has had a bad year

Post by Steve Reading »

000 wrote: Fri Sep 18, 2020 12:49 am 2. Is anyone thinking of reducing or increasing their VFMF allocation?
I'm buying it with every 401k paycheck.
000 wrote: Fri Sep 18, 2020 12:49 am 3. Is VFMF a fair representation of a multi-factor strategy? Or does it represent bad construction and/or tracking error?
No, the methodology is actually really good. I've slowly fallen in love with this fund.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
muffins14
Posts: 364
Joined: Wed Oct 26, 2016 4:14 am

Re: VFMF has had a bad year

Post by muffins14 »

000 wrote: Fri Sep 18, 2020 12:49 am I was recently reading some old posts and found some discussion about VFMF (Vanguard U.S. Multifactor ETF) as a possible core portfolio holding.

As of 8/31/2020, VFMF has returned -8.11% YTD versus +9.39% by its benchmark (Russell 3000).

One year performance is +2.12% for VFMF and +21.44% for the benchmark.

VFMF's YTD and 1YR performances are close to those of the Vanguard Value ETF.

Questions
1. Why didn't the other factors (momentum, quality) help this year?
2. Is anyone thinking of reducing or increasing their VFMF allocation?
3. Is VFMF a fair representation of a multi-factor strategy? Or does it represent bad construction and/or tracking error?
The YTD performance up to 9/17 for the SCV fund IJS is -20%, so it would seem that, yes, having momentum and less size helped VFMF over a pure SCV fund. One year performance for IJS is -7.24% as of 8/31.

I TLH between FNDA and VFMF, which have a somewhat high correlation, so I will increase my VFMF over time as I'm currently almost entirely in FNDA due to my last TLH
cheezit
Posts: 377
Joined: Sat Jul 14, 2018 7:28 pm

Re: VFMF has had a bad year

Post by cheezit »

Running factor regressions on PV for the trailing year, it looks like VFMF has a huge negative alpha (ca. -7%/year) for the trailing 1-year period that is absent from SCV funds like VFVA and IJS; this also holds if you add BaB to the factor model. Nb. the factor regression cuts off in June instead of August of 2020 for whatever reason. Maybe I screwed something up there. In any case, the stated ER for VFMF is low enough that the negative alpha has to be from something else. Trading frictions probably didn't help (the fund had 98% turnover last year!), especially if the model that Vanguard's quantitative equity group uses for this fund had them making trades during parts of the crash when liquidity was poor. Large (absolute) values of alpha sometimes indicate the model fit is poor, but R^2 is sky high so I don't think that was the problem. With 571 holdings, I don't think idiosyncratic risk was rearing its head.

Looking at the performance data for the same period, it looks like VFMF roughly tied the broad market until the SARS-CoV-2 crash, wherein it crashed harder and didn't recover as quickly.
columbia
Posts: 2933
Joined: Tue Aug 27, 2013 5:30 am

Re: VFMF has had a bad year

Post by columbia »

cheezit wrote: Fri Sep 18, 2020 8:51 am
Looking at the performance data for the same period, it looks like VFMF roughly tied the broad market until the SARS-CoV-2 crash, wherein it crashed harder and didn't recover as quickly.
If you look at where it's noticeably underweight and overweight (sector wise), no great surprise:
https://investor.vanguard.com/etf/profile/VFMF
muffins14
Posts: 364
Joined: Wed Oct 26, 2016 4:14 am

Re: VFMF has had a bad year

Post by muffins14 »

Couldn't the explanation for large alpha just be that the value methodology used by VFMF is different from that in the factor model, so some of the recent underperformance of "value" as implemented by VFMF shows up as alpha rather than HmL?
cheezit
Posts: 377
Joined: Sat Jul 14, 2018 7:28 pm

Re: VFMF has had a bad year

Post by cheezit »

muffins14 wrote: Fri Sep 18, 2020 9:13 am Couldn't the explanation for large alpha just be that the value methodology used by VFMF is different from that in the factor model, so some of the recent underperformance of "value" as implemented by VFMF shows up as alpha rather than HmL?
I'm sure VFMF defines every factor differently than F-F or AQR, but if a mismatch between Vanguard Quantitative Equity Group's factor definitions and the model's were the source of the discrepancy it should show up as a (relatively) low correlation of VFMF with the factor model. Sed contra, R^2 is 0.99x for the AQR model I used. If you're worried about fitting an elephant and making him wiggle his trunk with too many parameters, it's the same story with the basic F-F 3-factor model (R^2 also > 0.99 in that case for VFMF over the same period).
rkhusky
Posts: 10213
Joined: Thu Aug 18, 2011 8:09 pm

Re: VFMF has had a bad year

Post by rkhusky »

MotoTrojan wrote: Fri Sep 18, 2020 7:56 am
rkhusky wrote: Fri Sep 18, 2020 7:53 am
MotoTrojan wrote: Fri Sep 18, 2020 7:17 am
rkhusky wrote: Fri Sep 18, 2020 7:10 am
vineviz wrote: Fri Sep 18, 2020 6:26 am I'd hope not, but given the Boglehead affinity for performance-chasing I wouldn't be surprised.
Yes, like so many that invest in small value stocks because they did well in the past. Or those that jump into momentum or quality or profitability or low volatility stocks because they performed well in the past.
I'd rather performance chase something that has a good long-run history but has done terrible in the last few years, than join this VGT/QQQ bandwagon of true performance-chasers :P.
Chasing after Large Growth performance is no different than chasing after Small Value.
One is chasing recent outperformance, one is chasing long-term outperformance during recent (significant) underperformance; in principal I get your point, but there is a difference IMHO. Beyond that, I personally believe there is evidence that one is expected to outperform while the other underperform based on long-term (many market cycle) returns and human behavior, but I know we won't agree on that one today.
I think you can be successful with either as long as you stick with it. IMO, performance chasing suggests switching between strategies depending on what’s currently hot or popular. That usually doesn’t work out. You don’t have to worry about chasing sectors if you use Total Stock.
MotoTrojan
Posts: 10659
Joined: Wed Feb 01, 2017 8:39 pm

Re: VFMF has had a bad year

Post by MotoTrojan »

rkhusky wrote: Fri Sep 18, 2020 9:30 am
MotoTrojan wrote: Fri Sep 18, 2020 7:56 am
rkhusky wrote: Fri Sep 18, 2020 7:53 am
MotoTrojan wrote: Fri Sep 18, 2020 7:17 am
rkhusky wrote: Fri Sep 18, 2020 7:10 am
Yes, like so many that invest in small value stocks because they did well in the past. Or those that jump into momentum or quality or profitability or low volatility stocks because they performed well in the past.
I'd rather performance chase something that has a good long-run history but has done terrible in the last few years, than join this VGT/QQQ bandwagon of true performance-chasers :P.
Chasing after Large Growth performance is no different than chasing after Small Value.
One is chasing recent outperformance, one is chasing long-term outperformance during recent (significant) underperformance; in principal I get your point, but there is a difference IMHO. Beyond that, I personally believe there is evidence that one is expected to outperform while the other underperform based on long-term (many market cycle) returns and human behavior, but I know we won't agree on that one today.
I think you can be successful with either as long as you stick with it. IMO, performance chasing suggests switching between strategies depending on what’s currently hot or popular. That usually doesn’t work out. You don’t have to worry about chasing sectors if you use Total Stock.
Agreed. Flip-flopping is certainly the killer. Recently heard of someone moving their value portfolio entirely to large-growth, yikes.
User avatar
Robert T
Posts: 2740
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: VFMF has had a bad year

Post by Robert T »

.
So far VFMF looks okay (if wanting those types of tilts).

Here are the factor loads since inception (using daily data to increase sample size):

0.02 = Annual alpha
0.98 = Market
0.45 = Size
0.47 = Value
0.18 = Momentum
0.19 = Profitability [quality]

97.7 = R^2
https://www.portfoliovisualizer.com/fac ... sion=false

Would prefer longer time period (backtested over the full market cycle) to better determine likely long-term characteristics.
.
Robot Monster
Posts: 1530
Joined: Sun May 05, 2019 11:23 am
Location: New York

Re: VFMF has had a bad year

Post by Robot Monster »

Steve Reading wrote: Fri Sep 18, 2020 8:15 am No, the methodology is actually really good. I've slowly fallen in love with this fund.
Would be interested in hearing more about your enthusiasm for the fund. I don't even understand the case for it, but the mere fact that both you, and vineviz, like the fund, that alone piques my interest.
Last edited by Robot Monster on Fri Sep 18, 2020 3:32 pm, edited 1 time in total.
"Happiness comes from being connected in the right ways to: other people, your work, something larger than yourself."
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: VFMF has had a bad year

Post by Steve Reading »

cheezit wrote: Fri Sep 18, 2020 8:51 am Running factor regressions on PV for the trailing year, it looks like VFMF has a huge negative alpha (ca. -7%/year) for the trailing 1-year period that is absent from SCV funds like VFVA and IJS; this also holds if you add BaB to the factor model. Nb. the factor regression cuts off in June instead of August of 2020 for whatever reason. Maybe I screwed something up there. In any case, the stated ER for VFMF is low enough that the negative alpha has to be from something else. Trading frictions probably didn't help (the fund had 98% turnover last year!), especially if the model that Vanguard's quantitative equity group uses for this fund had them making trades during parts of the crash when liquidity was poor. Large (absolute) values of alpha sometimes indicate the model fit is poor, but R^2 is sky high so I don't think that was the problem. With 571 holdings, I don't think idiosyncratic risk was rearing its head.

Looking at the performance data for the same period, it looks like VFMF roughly tied the broad market until the SARS-CoV-2 crash, wherein it crashed harder and didn't recover as quickly.
The reason is most likely too little data combined with methodology that is slightly different than the factors.
How can a fund achieve very high loadings, very high R^2, yet produce negative alpha? Like so:
1) Imagine FF momentum factor. They buy the stocks that outperformed in previous 12-2 months. These now are the "momentum stocks".
2) A momentum fund uses a different methodology. They buy the stocks that outperformed in the previous 6-2 months. So many of the same stocks as FF.
3) This means when FF momentum drops, so does the fund. And vice-versa. Because they share many of the same stocks, their correlation is very high and tight.
4) However, the momentum fund happened to pick the worst momentum stocks out of the FF momentum pot. Call it bad luck or whatever. So once you control for FF momentum, you'll see a residual negative alpha. And maybe even quite a high one!

So you might see a very high factor load (which just means the fund and FF are tightly correlated), you might see a high R^2 (which means FF momentum is doing a good job of explaining the variability of the momentum fund returns), and also see a very negative alpha (because, out of the FF momentum stocks, the momentum fund managed to pick the losers).

Solving it is simple:
1) Use much more data. If it truly was just bad luck then the law of large numbers should slowly rear its head and bring the alpha closer to zero. But it might not just be bad luck. It might be that VFMF's methodology actually systematically picks the worst factor stocks out of the factor bunch. In that case, you'll see a negative, and statistically significant alpha, over a long period.
2) Look for statistically significant alpha. Be careful here though since, given enough funds and different regressions, you're bound to hit many "statistically significant alphas" out of luck. FWIW, if you regress VFMF since inception using FF factors, the alpha isn't even statistically significant any more.
cheezit wrote: Fri Sep 18, 2020 9:28 am I'm sure VFMF defines every factor differently than F-F or AQR, but if a mismatch between Vanguard Quantitative Equity Group's factor definitions and the model's were the source of the discrepancy it should show up as a (relatively) low correlation of VFMF with the factor model.
No, I don't think that's true. A fund can achieve high (even very high) correlations to factors, and hence factors might be excellent at explaining the variability of returns, yet the returns of the fund itself could be much higher or much lower. It's hard to conceptualize but think of it as the factor loads creating a sinusodial and the fund returns are an identical sinusoidal, expect it's just -5% lower in the y-axis. Such a regression will show 1.0 loads, will show R^2 of 100%, yet the alpha is -5%.
So one probably could say VFMF has managed to choose the worst of the factor stocks thus far, most likely because its methodology resulted in that. The reader will decide whether it's just noise/bad luck, or VFMF's methodology will systematically keep picking the bad factor stocks.
muffins14 wrote: Fri Sep 18, 2020 9:13 am Couldn't the explanation for large alpha just be that the value methodology used by VFMF is different from that in the factor model, so some of the recent underperformance of "value" as implemented by VFMF shows up as alpha rather than HmL?
That's my bet as to what's happening.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: VFMF has had a bad year

Post by Steve Reading »

Robot Monster wrote: Fri Sep 18, 2020 11:00 am
Steve Reading wrote: Fri Sep 18, 2020 8:15 am No, the methodology is actually really good. I've slowly fallen in love with this fund.
Would be interested in hearing more about your enthusiasm for the fund. I don't even understand the case for it, but the mere fact that both you, and vineviz, like the fund, that alone piques my interest.
The case is that if you're an investor that wants exposure to quality, value and momentum, this fund will do it while also overweighing small and mid caps. The metrics it uses are reasonable. And the ER is very competitive for such a product. It's also all rules-based; there's no human stock-picking. It even has a screen to avoid highly volatile stocks. The thing has it all.

Whether you want to tilt to factor or not is a very important question. But IF you've decided you want to tilt, VFMF does basically everything I want it to.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
XacTactX
Posts: 66
Joined: Sat May 11, 2019 5:46 pm
Location: Southern California

Re: VFMF has had a bad year

Post by XacTactX »

I've thought about having this fund a few times but my biggest concern is negative alpha, in the past 2.5 years it generated around -5% alpha with a t-stat of 2.4, that's the only thing keeping me away.
SMLF | ISCF | EMGF | LendingClub | Cash
sycamore
Posts: 1225
Joined: Tue May 08, 2018 12:06 pm

Re: VFMF has had a bad year

Post by sycamore »

Robot Monster wrote: Fri Sep 18, 2020 11:00 am
Steve Reading wrote: Fri Sep 18, 2020 8:15 am No, the methodology is actually really good. I've slowly fallen in love with this fund.
Would be interested in hearing more about your enthusiasm for the fund. I don't even understand the case for it, but the mere fact that both you, and vineviz, like the fund, that alone piques my interest.

I do notice VFMF is tiny in total net assets, only $61 million. I also notice that the VFMF ETF page doesn't link to its fund version, VFMFX, and visa versa. Too small potatoes to get this right?
And only $25.9 million in assets for VFMFX.

VFMF and VFMFX are not simply different shares classes of the same fund (unlike say VTI and VTSAX). And looking at their portfolio composition, they're very similar but not quite the same (top 10 holdings are slightly different for example). No mutual fund -> ETF conversions allowed.
Robot Monster
Posts: 1530
Joined: Sun May 05, 2019 11:23 am
Location: New York

Re: VFMF has had a bad year

Post by Robot Monster »

Steve Reading wrote: Fri Sep 18, 2020 1:21 pm
Robot Monster wrote: Fri Sep 18, 2020 11:00 am
Steve Reading wrote: Fri Sep 18, 2020 8:15 am No, the methodology is actually really good. I've slowly fallen in love with this fund.
Would be interested in hearing more about your enthusiasm for the fund. I don't even understand the case for it, but the mere fact that both you, and vineviz, like the fund, that alone piques my interest.
The case is that if you're an investor that wants exposure to quality, value and momentum, this fund will do it while also overweighing small and mid caps. The metrics it uses are reasonable. And the ER is very competitive for such a product. It's also all rules-based; there's no human stock-picking. It even has a screen to avoid highly volatile stocks. The thing has it all.

Whether you want to tilt to factor or not is a very important question. But IF you've decided you want to tilt, VFMF does basically everything I want it to.
I am curious, in particular, about a tilt to quality, and see a Vanguard U.S. Quality Factor ETF exists. From BlackRock:

"We keep our strong overweight on quality. We see it as the most resilient exposure against a range of outcomes in terms of developments in the pandemic and economy."
Source

Would you dissuade someone, who was interested in such a tilt, to invest in this Quality ETF?
"Happiness comes from being connected in the right ways to: other people, your work, something larger than yourself."
MotoTrojan
Posts: 10659
Joined: Wed Feb 01, 2017 8:39 pm

Re: VFMF has had a bad year

Post by MotoTrojan »

Robert T wrote: Fri Sep 18, 2020 9:38 am .
So far VFMF looks okay (if wanting those types of tilts).

Here are the factor loads since inception (using daily data to increase sample size):

0.02 = Annual alpha
0.98 = Market
0.45 = Size
0.47 = Value
0.18 = Momentum
0.19 = Profitability [quality]

97.7 = R^2
https://www.portfoliovisualizer.com/fac ... sion=false

Would prefer longer time period (backtested over the full market cycle) to better determine likely long-term characteristics.
.
Woah didn't realize you could use custom regressions on PV, now i don't have to compromise between 4 (momentum) and 5 (quality) options! Slick!
User avatar
vineviz
Posts: 7843
Joined: Tue May 15, 2018 1:55 pm

Re: VFMF has had a bad year

Post by vineviz »

Robot Monster wrote: Fri Sep 18, 2020 1:28 pm
Steve Reading wrote: Fri Sep 18, 2020 1:21 pm
Robot Monster wrote: Fri Sep 18, 2020 11:00 am
Steve Reading wrote: Fri Sep 18, 2020 8:15 am No, the methodology is actually really good. I've slowly fallen in love with this fund.
Would be interested in hearing more about your enthusiasm for the fund. I don't even understand the case for it, but the mere fact that both you, and vineviz, like the fund, that alone piques my interest.
The case is that if you're an investor that wants exposure to quality, value and momentum, this fund will do it while also overweighing small and mid caps. The metrics it uses are reasonable. And the ER is very competitive for such a product. It's also all rules-based; there's no human stock-picking. It even has a screen to avoid highly volatile stocks. The thing has it all.

Whether you want to tilt to factor or not is a very important question. But IF you've decided you want to tilt, VFMF does basically everything I want it to.
I am curious, in particular, about a tilt to quality, and see a Vanguard U.S. Quality Factor ETF exists. From BlackRock:

"We keep our strong overweight on quality. We see it as the most resilient exposure against a range of outcomes in terms of developments in the pandemic and economy."
Source

Would you dissuade someone, who was interested in such a tilt, to invest in this Quality ETF?
I wouldn’t encourage a stand-alone tilt to quality, because it has a negative correlation with other desirable factor exposures (eg market, size, value).

My recommendation is to consider quality as a component of multifactor strategy.

If I WERE to buy a fund targeting quality specifically I’d be more attracted to Vanguard’s (VFQY) than IShares.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
hdas
Posts: 1327
Joined: Thu Jun 11, 2015 8:24 am

Re: VFMF has had a bad year

Post by hdas »

vineviz wrote: Fri Sep 18, 2020 6:26 am
I'd hope not, but given the Boglehead affinity for performance-chasing I wouldn't be surprised.
Would you mind sharing since when you've had strong value, size tilt in your portfolio?. thx. H
....
Elysium
Posts: 3223
Joined: Mon Apr 02, 2007 6:22 pm

Re: VFMF has had a bad year

Post by Elysium »

Vanguard QEG has a very poor track record in everything they have done so far, why would these factor funds be any different than their other long run factor funds. Let's see, long history of mediocre performance at best from Strategic Equity, Strategic Small Cap Equity, Explorer value, US Value (now merged to Value Index), and sleeves of many other underperforming active managed funds.

When will financial industry learn to give up trying to beat market indexes using these failed quant strategies? AQR is failing big time, Vanguard QEG has a long mediocre record, DFA still sticks to FF value (perhaps the best out of all bad factor strategies). I predict VFMF will continue to be mediocre with flash in the pan brilliance every once, perhaps eventually to be closed/merged into plain indexes after 20 years.
HippoSir
Posts: 133
Joined: Tue Jul 03, 2018 2:56 pm

Re: VFMF has had a bad year

Post by HippoSir »

I find:

https://www.thinknewfound.com/style-dashboard

a useful resource at times like these. You can see that size/value/quality have all been quite negative this year. VFMF targets all of these, and is more concentrated than many other multifactor funds, so it's not surprising it's doing worse.

That said, it's still doing better than pure small value (or VFVA), so the factor blend is doing its job.
Angst
Posts: 2433
Joined: Sat Jun 09, 2007 11:31 am

Re: VFMF has had a bad year (So has VFMFX)

Post by Angst »

Robert T wrote: Fri Sep 18, 2020 9:38 am .
So far VFMF looks okay (if wanting those types of tilts).

Here are the factor loads since inception (using daily data to increase sample size):

0.02 = Annual alpha
0.98 = Market
0.45 = Size
0.47 = Value
0.18 = Momentum
0.19 = Profitability [quality]

97.7 = R^2
https://www.portfoliovisualizer.com/fac ... sion=false

Would prefer longer time period (backtested over the full market cycle) to better determine likely long-term characteristics.
.
Out of curiosity, I've edited Robert T's PV results to include (below) Vanguard's mutual fund version of VFMF, i.e. VFMFX. As others have already pointed out, these two funds are NOT direct subsets of the same overall fund like Vanguard index ETF/Mutual Fund pairs typically are. These funds however DO share identical "strategy and policy" statements on their portfolio page and fund managers. The "equity characteristics" and exact portfolio holdings have drifted apart somewhat. I don't know what's involved in trying to maintain their similarity over time, but one might imagine that the mutual fund version could gain at least a slight tracking and performance edge over the ETF, but apparently not so far. Perhaps there's something wrong with PV's calculation of that alpha:

Code: Select all

VFMF  VFMFX

0.02  -2.47  Annual alpha
0.98   0.99  Market
0.45   0.37  Size
0.47   0.39  Value
0.18   0.14  Momentum
0.19   0.19  Profitability [quality]

97.7   98.9  R^2 
https://www.portfoliovisualizer.com/fac ... sion=false
User avatar
Robert T
Posts: 2740
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: VFMF has had a bad year (So has VFMFX)

Post by Robert T »

Angst wrote: Fri Sep 18, 2020 2:49 pm Out of curiosity, I've edited Robert T's PV results to include (below) Vanguard's mutual fund version of VFMF, i.e. VFMFX. As others have already pointed out, these two funds are NOT direct subsets of the same overall fund like Vanguard index ETF/Mutual Fund pairs typically are. These funds however DO share identical "strategy and policy" statements on their portfolio page and fund managers. The "equity characteristics" and exact portfolio holdings have drifted apart somewhat. I don't know what's involved in trying to maintain their similarity over time, but one might imagine that the mutual fund version could gain at least a slight tracking and performance edge over the ETF, but apparently not so far. Perhaps there's something wrong with PV's calculation of that alpha:
Personally, I don't put much weight on the results for such a short-time periods (too much noise IMO), relative to longer-term characteristics. And prefer using monthly data over longer periods of time. With short-term data, you do get less robust results (as demonstrated) e.g. daily alpha on VFMFX is not statistically different from zero, even though large when scaled to annual (large standard error in the estimate).

While portfolio visualizer is an excellent tool, I worry users give too much weight to short-time period estimates (short life of funds). Personally I have never made a portfolio inclusion decision only on factor load estimate based on data availability for the life of a particular fund. For all, I needed to look at long-term characteristics of the underlying indexes for the longest time periods available for them to gain more confidence in the results. This exercise however is done outside of portfolio visualizer. As longer-term back-tested data aren't available for this exercise for VFMF - I simply skip the fund. Just my approach.
.
MotoTrojan
Posts: 10659
Joined: Wed Feb 01, 2017 8:39 pm

Re: VFMF has had a bad year

Post by MotoTrojan »

Elysium wrote: Fri Sep 18, 2020 2:17 pm Vanguard QEG has a very poor track record in everything they have done so far, why would these factor funds be any different than their other long run factor funds. Let's see, long history of mediocre performance at best from Strategic Equity, Strategic Small Cap Equity, Explorer value, US Value (now merged to Value Index), and sleeves of many other underperforming active managed funds.

When will financial industry learn to give up trying to beat market indexes using these failed quant strategies? AQR is failing big time, Vanguard QEG has a long mediocre record, DFA still sticks to FF value (perhaps the best out of all bad factor strategies). I predict VFMF will continue to be mediocre with flash in the pan brilliance every once, perhaps eventually to be closed/merged into plain indexes after 20 years.
:oops:
Topic Author
000
Posts: 2818
Joined: Thu Jul 23, 2020 12:04 am

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by 000 »

Thanks all for the discussion so far.

One follow-up question:
2b. For those holding VFMF, is it as a replacement for -- or complement to -- a raw beta fund, like a total market fund?
Topic Author
000
Posts: 2818
Joined: Thu Jul 23, 2020 12:04 am

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by 000 »

Forester wrote: Fri Sep 18, 2020 6:08 am Perhaps momentum gets watered down by all the other ingredients. Pick enough factors and it washes out to a market cap weighted fund.
This has been my line of thinking too, except in this case it doesn't wash out to a market cap weighted fund. :(
User avatar
vineviz
Posts: 7843
Joined: Tue May 15, 2018 1:55 pm

Re: VFMF has had a bad year

Post by vineviz »

MotoTrojan wrote: Fri Sep 18, 2020 3:36 pm
Elysium wrote: Fri Sep 18, 2020 2:17 pm Vanguard QEG has a very poor track record in everything they have done so far, why would these factor funds be any different than their other long run factor funds. Let's see, long history of mediocre performance at best from Strategic Equity, Strategic Small Cap Equity, Explorer value, US Value (now merged to Value Index), and sleeves of many other underperforming active managed funds.

When will financial industry learn to give up trying to beat market indexes using these failed quant strategies? AQR is failing big time, Vanguard QEG has a long mediocre record, DFA still sticks to FF value (perhaps the best out of all bad factor strategies). I predict VFMF will continue to be mediocre with flash in the pan brilliance every once, perhaps eventually to be closed/merged into plain indexes after 20 years.
:oops:
Is there an "okay, Boomer" emoji yet?
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Topic Author
000
Posts: 2818
Joined: Thu Jul 23, 2020 12:04 am

Re: VFMF has had a bad year

Post by 000 »

vineviz wrote: Fri Sep 18, 2020 6:26 am
000 wrote: Fri Sep 18, 2020 12:49 am 1. Why didn't the other factors (momentum, quality) help this year?
Momentum HAS helped so far this year, adding some performance relative to the total market. Quality has NOT helped because the quality factor has had negative returns year to date. Size helped a little bit, as did VFMF's slightly higher market beta, but value has had a VERY terrible year so far and there's no escaping that.
It seems like VFMF should have performed somewhat better than just a Value ETF considering how well momentum strategies have done this year.

So perhaps value is outdimming all the other factors in this fund?
Jeff Albertson
Posts: 832
Joined: Sat Apr 06, 2013 7:11 pm
Location: Springfield

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by Jeff Albertson »

000 wrote: Fri Sep 18, 2020 12:49 am 1. Why didn't the other factors (momentum, quality) help this year?
“Rekenthaler’s Rule”: If the bozos know about it, it doesn’t work any more.
MotoTrojan
Posts: 10659
Joined: Wed Feb 01, 2017 8:39 pm

Re: VFMF has had a bad year

Post by MotoTrojan »

000 wrote: Fri Sep 18, 2020 3:49 pm
vineviz wrote: Fri Sep 18, 2020 6:26 am
000 wrote: Fri Sep 18, 2020 12:49 am 1. Why didn't the other factors (momentum, quality) help this year?
Momentum HAS helped so far this year, adding some performance relative to the total market. Quality has NOT helped because the quality factor has had negative returns year to date. Size helped a little bit, as did VFMF's slightly higher market beta, but value has had a VERY terrible year so far and there's no escaping that.
It seems like VFMF should have performed somewhat better than just a Value ETF considering how well momentum strategies have done this year.

So perhaps value is outdimming all the other factors in this fund?
Hard to get a good regression on such a young fund but it has a pretty healthy dose of value, and way less momentum than something pure like MTUM.

https://www.portfoliovisualizer.com/fac ... sion=false
Topic Author
000
Posts: 2818
Joined: Thu Jul 23, 2020 12:04 am

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by 000 »

Jeff Albertson wrote: Fri Sep 18, 2020 3:52 pm
000 wrote: Fri Sep 18, 2020 12:49 am 1. Why didn't the other factors (momentum, quality) help this year?
“Rekenthaler’s Rule”: If the bozos know about it, it doesn’t work any more.
Lol. I thought momentum was deliberately trying to follow the "bozos". :oops:

Who's the bigger bozo? The bozo or the bozo who follows him?
Last edited by 000 on Fri Sep 18, 2020 3:57 pm, edited 1 time in total.
muffins14
Posts: 364
Joined: Wed Oct 26, 2016 4:14 am

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by muffins14 »

000 wrote: Fri Sep 18, 2020 3:43 pm Thanks all for the discussion so far.

One follow-up question:
2b. For those holding VFMF, is it as a replacement for -- or complement to -- a raw beta fund, like a total market fund?
I use it as a complement, because the factor loading are more than my target. I am about 50/50 between VFMF and S&P 500 (via NTSX)
User avatar
vineviz
Posts: 7843
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by vineviz »

000 wrote: Fri Sep 18, 2020 3:46 pm
Forester wrote: Fri Sep 18, 2020 6:08 am Perhaps momentum gets watered down by all the other ingredients. Pick enough factors and it washes out to a market cap weighted fund.
This has been my line of thinking too, except in this case it doesn't wash out to a market cap weighted fund. :(
If the manager is willing to deviate substantially enough from the underlying universe of stocks then it's often possible to gain significant factor exposure to ALL the major factors without them "cancelling" out.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
Topic Author
000
Posts: 2818
Joined: Thu Jul 23, 2020 12:04 am

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by 000 »

vineviz wrote: Fri Sep 18, 2020 3:58 pm If the manager is willing to deviate substantially enough from the underlying universe of stocks then it's often possible to gain significant factor exposure to ALL the major factors without them "cancelling" out.
I'm not sure Vanguard "is willing to deviate substantially enough from the underlying universe of stocks" though.

Most of their active, style, and smart beta funds are fairly tame.

Do you have an example in mind? And can you please clarify for me what you consider all the major factors that can be complementarily targeted?

Thanks
muffins14
Posts: 364
Joined: Wed Oct 26, 2016 4:14 am

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by muffins14 »

000 wrote: Fri Sep 18, 2020 4:01 pm
vineviz wrote: Fri Sep 18, 2020 3:58 pm If the manager is willing to deviate substantially enough from the underlying universe of stocks then it's often possible to gain significant factor exposure to ALL the major factors without them "cancelling" out.
I'm not sure Vanguard "is willing to deviate substantially enough from the underlying universe of stocks" though.

Most of their active, style, and smart beta funds are fairly tame.

Do you have an example in mind? And can you please clarify for me what you consider all the major factors that can be complementarily targeted?

Thanks
VFMF is the example
User avatar
Steve Reading
Posts: 2460
Joined: Fri Nov 16, 2018 10:20 pm

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by Steve Reading »

000 wrote: Fri Sep 18, 2020 4:01 pm Most of their active, style, and smart beta funds are fairly tame.
On the contrary, Vanguard's factor funds achieve some truly remarkable factor loads. The "trick" is in their methodology. By weighing stocks purely by their factor score, you basically end up with only the strongest factor stocks, weighted by how strong their score is on top of it.

Which means, IMO, these funds are meant to be added to portfolios that are mostly market-cap weighted already. They're like very concentrated doses of factors.
"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson
Elysium
Posts: 3223
Joined: Mon Apr 02, 2007 6:22 pm

Re: VFMF has had a bad year

Post by Elysium »

vineviz wrote: Fri Sep 18, 2020 3:48 pm
MotoTrojan wrote: Fri Sep 18, 2020 3:36 pm
Elysium wrote: Fri Sep 18, 2020 2:17 pm Vanguard QEG has a very poor track record in everything they have done so far, why would these factor funds be any different than their other long run factor funds. Let's see, long history of mediocre performance at best from Strategic Equity, Strategic Small Cap Equity, Explorer value, US Value (now merged to Value Index), and sleeves of many other underperforming active managed funds.

When will financial industry learn to give up trying to beat market indexes using these failed quant strategies? AQR is failing big time, Vanguard QEG has a long mediocre record, DFA still sticks to FF value (perhaps the best out of all bad factor strategies). I predict VFMF will continue to be mediocre with flash in the pan brilliance every once, perhaps eventually to be closed/merged into plain indexes after 20 years.
:oops:
Is there an "okay, Boomer" emoji yet?
Not a boomer, a solid Gen X. I can spot a few millennials here though, think we need an emoji for that :wink:
typical.investor
Posts: 2299
Joined: Mon Jun 11, 2018 3:17 am

Re: Vanguard Multifactor fund VFMF has had a bad year

Post by typical.investor »

Steve Reading wrote: Fri Sep 18, 2020 4:28 pm
000 wrote: Fri Sep 18, 2020 4:01 pm Most of their active, style, and smart beta funds are fairly tame.
On the contrary, Vanguard's factor funds achieve some truly remarkable factor loads. The "trick" is in their methodology. By weighing stocks purely by their factor score, you basically end up with only the strongest factor stocks, weighted by how strong their score is on top of it.

Which means, IMO, these funds are meant to be added to portfolios that are mostly market-cap weighted already. They're like very concentrated doses of factors.
Not for the momentum loading though...
YRT70
Posts: 686
Joined: Sat Apr 27, 2019 8:51 am

Re: VFMF has had a bad year

Post by YRT70 »

000 wrote: Fri Sep 18, 2020 3:49 pm
vineviz wrote: Fri Sep 18, 2020 6:26 am
000 wrote: Fri Sep 18, 2020 12:49 am 1. Why didn't the other factors (momentum, quality) help this year?
Momentum HAS helped so far this year, adding some performance relative to the total market. Quality has NOT helped because the quality factor has had negative returns year to date. Size helped a little bit, as did VFMF's slightly higher market beta, but value has had a VERY terrible year so far and there's no escaping that.
It seems like VFMF should have performed somewhat better than just a Value ETF considering how well momentum strategies have done this year.

So perhaps value is outdimming all the other factors in this fund?
Small has had a bad year too so far.
Topic Author
000
Posts: 2818
Joined: Thu Jul 23, 2020 12:04 am

Re: VFMF has had a bad year

Post by 000 »

YRT70 wrote: Sat Sep 19, 2020 12:53 am Small has had a bad year too so far.
True, but I think it's worth pointing out that the benchmark for this fund is the Russell 3000, which is a cap-weighted near-total market index.

The Vanguard Product Detail page (linked in my OP) has this to say about the fund's Objective:
Vanguard U.S. Multifactor ETF seeks to provide long-term capital appreciation by investing in U.S. stocks which target three factors, value, momentum, and quality, after an initial volatility screen. Such a strategy offers potential diversification benefits that can help reduce the active risk associated with exposure to a single factor.
And it says this about the Investment Approach (bold emphases mine):
  • Advisor uses a rules-based quantitative model to evaluate U.S. common stocks and construct a U.S. equity portfolio that seeks to achieve exposure to multiple factors.
  • After applying an initial screen to remove the most volatile stocks in the universe, stocks are then selected according to their equally weighted ranking across three targeted factors; momentum-stocks that exhibit strong recent performance, quality-stocks that exhibit strong fundamentals, and value-stocks with low prices relative to fundamentals.
  • Portfolio includes a diverse mix of stocks representing many different market capitalizations (large, mid, and small), market sectors, and industry groups, and holds hundreds of names to diversify idiosyncratic stock risk.
  • Portfolio is rebalanced as needed to maintain consistent exposure to the targeted factors.
  • Seeks long-term capital appreciation.
  • Typically, at least 80% of the fund’s assets will be invested in securities issued by U.S. companies.
Seeing as small caps are some of "the most volatile stocks in the universe" and the fund seeks to invest in a "diverse mix of stocks representing many different market capitalizations (large, mid, and small)", it's not obvious from the high-level fund description that a small cap tilt is included. Indeed, size is not one of the three factors mentioned as being targeted by the fund.

I suppose someone might be expected to read through the lines and observe that small is more valuely than large right now, but conversely small does not currently have more momentum or quality than large.
Post Reply