Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
bjr89
Posts: 151
Joined: Sat Apr 23, 2016 8:15 am

Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by bjr89 » Tue Oct 30, 2018 12:58 pm

I currently use Vanguard's small value index fund, but I've been thinking about switching to DFA lately (has nothing to do with past performance or differences in size or value measures).

Here's why. The three largest small value ETFs (VBR, IWN, and IJS) track CRSP, Russell, and S&P value indices. If you read the ranking methodology for these three indices you learn that what they deem to be value stocks are not just stocks with low price-to-fundamental ratios, but also those with poor earnings growth metrics. Consider the basic stock valuation model:

Expected Cash Flow / (Discount Rate Reflecting Risk - Growth Rate) = Present Value
This formula basically means stocks can trade at low price-to-fundamental ratios for two reasons (assuming efficient markets): 1. High risk 2. Low growth prospects.

Many, if not most, people who invest in these index funds do so with the belief that value stocks are higher risk, higher reward than the broad market. But since these funds hold the low-earnings-growth value stocks, it may just be that the stocks they hold are no riskier than average -- they just have lower earnings growth potential. This means they actually may not have higher expected returns (!).

DFA doesn't do this. A value stock is a value stock to DFA (basically) because it has a low price/book. They don't force it to also have low earnings growth rankings. This means their funds should do a better job capturing average differences in risk, not earnings growth. Although even better would be a fund that buys low price/fundamental stocks that also have high measures of profitability and earnings growth. This would isolate the discount rate as it would control for the growth rate.

Question is, does the team at Vanguard realize or understand this? Do their investors?

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by vineviz » Tue Oct 30, 2018 1:28 pm

bjr89 wrote:
Tue Oct 30, 2018 12:58 pm
Many, if not most, people who invest in these index funds do so with the belief that value stocks are higher risk, higher reward than the broad market. But since these funds hold the low-earnings-growth value stocks, it may just be that the stocks they hold are no riskier than average -- they just have lower earnings growth potential. This means they actually may not have higher expected returns (!).
I'm not sure what most people believe, but I do know that size, value, and quality are all well-established risk factors that account for a significant amount of the cross-section of stock returns. There is no such risk factor associated with earnings growth rates, so I don't know how you'd build either a theoretical or an empirical case that companies with high historical earnings growth should have higher expected returns.
bjr89 wrote:
Tue Oct 30, 2018 12:58 pm
DFA doesn't do this. A value stock is a value stock to DFA (basically) because it has a low price/book. They don't force it to also have low earnings growth rankings. This means their funds should do a better job capturing average differences in risk, not earnings growth.
DFA US Small Cap Value is, indeed a more "pure" value play than S&P Small-Cap 600 Value funds (though not, it's worth pointing out, the readily available S&P SmallCap 600 Pure Value ETF (RZV). Choosing DFA US Small Cap Value
over iShares S&P Small-Cap 600 Value ETF means, however, giving up much of the later's significant quality screen which is likely to cost at least as much return as the increased value exposure gets you.
bjr89 wrote:
Tue Oct 30, 2018 12:58 pm
Question is, does the team at Vanguard realize or understand this? Do their investors?
This you can bet on.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

bjr89
Posts: 151
Joined: Sat Apr 23, 2016 8:15 am

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by bjr89 » Tue Oct 30, 2018 2:16 pm

Right. Quality predicts return. And these indices select for low quality (eg, low ROA, low earnings growth) value stocks. Not good.

That wasn't my point about DFA. Yes, DFA selects smaller mkt cap and lower p/b...but that's beside the point. CRSP, Russell, and S&P all select low quality value stocks. DFA doesn't bias themselves toward low quality. Reason why quality works is because it controls for the part of the discount rate related to risk. That's why quality is more predictive in conjunction with value. Fama himself has said this. In fact, looking at quality alone doesn't work very well. You need quality and value to isolate the risk premium. Think about it this way. If a company looks good in the sense that it is expected to generate a lot of cash flow, but it still trades at a low valuation, something must REALLY be wrong with it (that is, there's a high expected variance around those high expected cash flows)... so it gets a high discount rate as a result of risk, not low growth.

And I think you're giving Vanguard and their investors way too much credit. Poorly designed products get made all the time in finance. For example, look at leveraged ETFs that suffer from volatility drag, or CCF funds that systematically lose money to contango. People sign up for the worst of things as a result of ignorance. Vanguard is obviously much better than most in the benevolence category, but that doesn't make them immune to error. And typical investors don't tend to read the fine print. I mean, I'm saying this as an investor. It's not hard to believe that most haven't even considered this.

alex_686
Posts: 3920
Joined: Mon Feb 09, 2015 2:39 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by alex_686 » Tue Oct 30, 2018 2:26 pm

bjr89 wrote:
Tue Oct 30, 2018 12:58 pm
DFA doesn't do this. A value stock is a value stock to DFA (basically) because it has a low price/book. They don't force it to also have low earnings growth rankings. This means their funds should do a better job capturing average differences in risk, not earnings growth. Although even better would be a fund that buys low price/fundamental stocks that also have high measures of profitability and earnings growth. This would isolate the discount rate as it would control for the growth rate.

Question is, does the team at Vanguard realize or understand this? Do their investors?
First, where did you come up with the low price/book thing. That is not my understanding of DFA - but then again I am not holding out myself as a expert - here to learn. That being said, I did attend a presentation by a DFA researcher and he stressed that the best factors was {small, not growth, not low profitability}. Which is a little different than {small, value, high profitability}. Plus all of the reading I have done suggests that the power of book value is falling expect for banks.

Second, I don't think investors realize this. Quality is an obscure factor. See all of the negative posts here on Bogleheads on Smart Beta. I don't see many investors clamoring for it. Vanguard tends to run index funds where there is a large market. No demand, no market. I don't see Vanguard blazing a trail here anytime soon.

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by vineviz » Tue Oct 30, 2018 2:28 pm

bjr89 wrote:
Tue Oct 30, 2018 2:16 pm
Right. Quality predicts return. And these indices select for low quality (eg, low ROA, low earnings growth) value stocks. Not good.
You’ve got that backwards, I’m afraid.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

MotoTrojan
Posts: 2461
Joined: Wed Feb 01, 2017 8:39 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by MotoTrojan » Tue Oct 30, 2018 2:32 pm

S&P has an earnings/quality screen. I am happy with my choice to hold S&P600 SV.

bjr89
Posts: 151
Joined: Sat Apr 23, 2016 8:15 am

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by bjr89 » Tue Oct 30, 2018 2:42 pm

Uh, no. S&P has a quality screen for their general universe. But within that universe they rank all stocks on their value factors (p/e, p/b, p/s) AND their growth factors (sales growth, the ratio of earnings change to price, and momentum). Stocks that rank high on value and low on growth get a higher ultimate value ranking than a stock that ranks high on value and high on growth. I'm saying they should only rank on the value factors, not both, since doing both biases you toward firms with low growth prospects that aren't necessarily riskier.

Another excerpt from Russell for illustration:

For each base index (the Russell 1000 and Russell 2000, and the smallest 1,000 in Russell Microcap),
stocks are ranked by their book-to-price ratio (B/P), their I/B/E/S forecast medium-term growth (2 year)
and sales per share historical growth (5 year). These rankings are converted to standardized units,
where the value variable represents 50% of the score and the two growth variables represent the
remaining 50%. They are then combined to produce a composite value score (CVS). Stocks are then
ranked by their CVS, and a probability algorithm is applied to the CVS distribution to assign growth and
value weights to each stock. In general, a stock with a lower CVS is considered growth, a stock with a
higher CVS is considered value and a stock with a CVS in the middle range is considered to have both
growth and value characteristics, and is weighted proportionately in the growth and value index. Stocks
are always fully represented by the combination of their growth and value weights; e.g., a stock that is
given a 20% weight in a Russell value index will have an 80% weight in the corresponding Russell
FTSE Russell | Russell U.S. Equity Indexes Construction and Methodology, v3.5, August 2018 27
growth index. Style index assignment for non-pricing vehicle share classes will be based on that of the
pricing vehicle and assigned consistently across all additional share classes.

https://us.spindices.com/documents/meth ... nload=true
http://www.crsp.com/files/Equity-Indexe ... uide_0.pdf
https://www.ftse.com/products/downloads ... ndexes.pdf

See for yourself, and be horrified.

Another simple example. Two stocks trade at 15x P/E. One has higher projected earnings growth than the other. But it still trades at the same P/E as the other. Why? Because those earnings are being discounted at a higher rate that reflects higher risk. So, paradoxically, high "quality" (which we should maybe call quantitative earnings quality) implies higher risk, but only if you control for value. Again Fama has said this. And that's why DFA has the better designed product.

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by vineviz » Tue Oct 30, 2018 2:56 pm

bjr89 wrote:
Tue Oct 30, 2018 2:16 pm
Yes, DFA selects smaller mkt cap and lower p/b...but that's beside the point.
It's also not entirely true.

DFA US Small Cap Value holds stocks that are roughly the same size as S&P 600 Value index (for instance), only marginally more value-tilted (the difference is not statistically significant over the last ten years, and lower quality (as measured by the AQR quality factor).
bjr89 wrote:
Tue Oct 30, 2018 2:16 pm
And I think you're giving Vanguard and their investors way too much credit.
I think you're overreaching to make a point.

The folks at S&P, BlackRock, and Vanguard are way smarter than you seem to be willing to give them credit for. I'm not saying every product is a winner, but an accusation that they have no one in their quantitative finance or product management teams who understands the factor loadings of their indexes and funds can only be made from a position of misunderstanding about their intellect, training, and sophistication.

As for individual investors, I can't pretend to know what they know and don't know. Or even what they care about and don't care about.

My guess is that most investors who go to the trouble of purchasing a small cap value fund couldn't (and shouldn't) care less about any of this: as long as the fund is genuinely exposing them to the small size factor and the value factor, that's 98% of what matters.

I say 98% because the correlation between DFA US Small Cap Value and the three largest SCV ETFs (iShares Russell 2000 Value ETF, iShares S&P Small-Cap 600 Value ETF, & Vanguard Small-Cap Value ETF) is .99, .98, and .98 respectively. While it's entertaining to quibble over the nuances of how these funds and their indexes are built, I can't see an argument that the average Joe or Jane Investor needs to worry about it.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by vineviz » Tue Oct 30, 2018 3:00 pm

bjr89 wrote:
Tue Oct 30, 2018 2:42 pm
I'm saying they should only rank on the value factors, not both, since doing both biases you toward firms with low growth prospects that aren't necessarily riskier.
You do realize that S&P already has a SCV value index that does exactly what you want, right?

https://us.spindices.com/indices/equity ... pure-value

And that there is an ETF that tracks this index?

https://www.morningstar.com/etfs/arcx/rzv/quote.html
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

bjr89
Posts: 151
Joined: Sat Apr 23, 2016 8:15 am

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by bjr89 » Tue Oct 30, 2018 3:09 pm

No, "S&P Pure Value" is exactly what I don't want lol
"The companies at the bottom of the list have a higher Value Rank (and Value Score) and a lower Growth Rank (and Growth Score) and, therefore, exhibit pure value characteristics." -S&P

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by vineviz » Tue Oct 30, 2018 3:44 pm

bjr89 wrote:
Tue Oct 30, 2018 3:09 pm
No, "S&P Pure Value" is exactly what I don't want lol
"The companies at the bottom of the list have a higher Value Rank (and Value Score) and a lower Growth Rank (and Growth Score) and, therefore, exhibit pure value characteristics." -S&P
I’m not entirely sure you understand what you want, but I am sure you don’t understand how this index is constructed.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

bjr89
Posts: 151
Joined: Sat Apr 23, 2016 8:15 am

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by bjr89 » Tue Oct 30, 2018 4:31 pm

I just want low price to fundamentals. I don't want low price to fundamentals and an additional ranking that specifically identifies stocks with low earnings growth. The Pure indices do not solve this problem, they exacerbate it. I definitely understand how the indices are constructed.

stlutz
Posts: 4784
Joined: Fri Jan 02, 2009 1:08 am

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by stlutz » Tue Oct 30, 2018 8:34 pm

It's worth remembering that the standard value/growth indices that have been around a long time were created first as benchmarks and second as investment products.

If I'm going to split the market into a growth half and a value half, there is the question of what to do with stocks that exhibit both or neither characteristics. Value and growth investors can't both own 100% of the dream stocks that are both value and growth. And *somebody* owns the "junk" stocks that supposedly nobody would want. Therefore the indexes split these stocks between value and growth. If you are benchmarking active investors, this makes complete sense.

The Vanguard/iShares etc. ETFs that follow these indexes are ways to get exposure to the universe of stocks that active managers are considering without the high expenses. Nothing more, nothing less. The index providers/Vanguard/iShares etc. make no claim that value stocks will beat growth stocks. That's a Boglehead thing.

So-called "smart-beta" indices are designed with the idea of beating the market. DFA funds are explicitly trying to beat the market as well. The question is of course whether those who are trying to beat the market actually pull it off.

If this form of active management is onto something, three factor regressions should show either significant positive alpha for these active funds and/or negative alpha for standard value/growth index funds. I've never been able to find this.

If one wants to do multi-factor investing, why not just narrow the list down to 50 individual stocks and invest that way? It's easy to get free trades nowadays, to the cost of doing this is trivial. That can't be any harder than studying all of the various indices and fund and selecting the best one.

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by vineviz » Tue Oct 30, 2018 9:03 pm

bjr89 wrote:
Tue Oct 30, 2018 4:31 pm
I just want low price to fundamentals. I don't want low price to fundamentals and an additional ranking that specifically identifies stocks with low earnings growth. The Pure indices do not solve this problem, they exacerbate it. I definitely understand how the indices are constructed.
Clearly you don't.

The Pure Value index includes only the "(c)onstituents of the S&P SmallCap 600 that exhibit the strongest value characteristics", which are defined by S&P as "ratios of book value, earnings, and sales to price".

In any case, Invesco S&P SmallCap 600 Pure Value ETF (RZV) has superior size, value, and quality factor loadings to DFA US Small Cap Value.

https://www.portfoliovisualizer.com/fac ... sion=false
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

stlutz
Posts: 4784
Joined: Fri Jan 02, 2009 1:08 am

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by stlutz » Tue Oct 30, 2018 9:52 pm

vineviz wrote:
Tue Oct 30, 2018 9:03 pm
bjr89 wrote:
Tue Oct 30, 2018 4:31 pm
I just want low price to fundamentals. I don't want low price to fundamentals and an additional ranking that specifically identifies stocks with low earnings growth. The Pure indices do not solve this problem, they exacerbate it. I definitely understand how the indices are constructed.
Clearly you don't.

The Pure Value index includes only the "(c)onstituents of the S&P SmallCap 600 that exhibit the strongest value characteristics", which are defined by S&P as "ratios of book value, earnings, and sales to price".
Regarding index construction, bjr89 is correct. The "pure" style indices exclude stocks that fall into both the growth value buckets. The pure value index does indeed only contain value stocks with low earnings growth.

The methodology doc is here: https://us.spindices.com/documents/meth ... nload=true

bjr89
Posts: 151
Joined: Sat Apr 23, 2016 8:15 am

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by bjr89 » Tue Oct 30, 2018 10:05 pm

stlutz, I think you hit on the key point. They were benchmarks first, products second.

Perhaps the Vanguard U.S. Value Factor ETF (VFVA) is what I'm looking for. But I don't really wanna invest until I know they aren't going to close it like a lot of other smart beta funds recently have. Only $38M in aum.

fennewaldaj
Posts: 272
Joined: Sun Oct 22, 2017 11:30 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by fennewaldaj » Tue Oct 30, 2018 10:24 pm

I have noticed this problem of sorts too. Basically in traditional value indexes a company is penalized (as in its more likely to appear in the growth index) for have good earnings growth and other growth metrics. It not so much that value investors don't like earnings growth its that they don't want to overpay for it. If you look through active value funds they tend to have better growth metrics than the indexes. For example if you look at IJS on morningstar it has some gross looking number in the growth metrics area like cash flow growth of -15%.

https://portfolios.morningstar.com/fund ... ctcode=COM

A fund like LSV small cap value has both better looking value numbers and better looking growth numbers.

https://portfolios.morningstar.com/fund ... ture=en-US

On the other hand active funds don't do better than indexes in the aggregate so maybe this isn't a problem.

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by vineviz » Tue Oct 30, 2018 11:27 pm

stlutz wrote:
Tue Oct 30, 2018 9:52 pm
Regarding index construction, bjr89 is correct. The "pure" style indices exclude stocks that fall into both the growth value buckets. The pure value index does indeed only contain value stocks with low earnings growth.
Sorry, but no. At least not completely: it's true that stocks in the "Pure Value" index are selected from the constituents of the base index that have 100% of their weight in the value basket, but it's not true that the "pure value index does indeed only contain value stocks with low earnings growth". The stocks in the Pure Value index are selected exclusively on their value score: there is no penalty for earnings growth.

In any case, the bottom line is that stocks with strong historical earnings growth aren't going to end up in a value factor fund in any case: the price/book, price/sales, price/cash flow ratios etc are too high for stocks with strong growth history any way. For example, DFA US Small Cap Value contains stocks with sales growth and book-value growth lower than (and similar forward EPS growth and historical cash-flow growth similar to) the S&P 600 Value index.

Furthermore, the S&P 600 SCV indexes have a higher loading on quality than the DFA funds do which is not surprising since the growth measures being criticized here have little bearing on the quality factor as defined in the literature by Frazzini et al. and which is associated with superior risk-adjusted returns in SCV stocks.

https://www.portfoliovisualizer.com/fac ... ssion=true
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

stlutz
Posts: 4784
Joined: Fri Jan 02, 2009 1:08 am

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by stlutz » Tue Oct 30, 2018 11:49 pm

Sorry, but no. At least not completely: it's true that stocks in the "Pure Value" index are selected from the constituents of the base index that have 100% of their weight in the value basket, but it's not true that the "pure value index does indeed only contain value stocks with low earnings growth". The stocks in the Pure Value index are selected exclusively on their value score: there is no penalty for earnings growth.
Read the document from S&P that is linked above.

Some stocks exhibit both growth and value characteristics. If a stock has enough growth characteristics that it can be partially included in the standard growth index, then it is not considered for inclusion in the pure value index (and vice versa for pure growth).

AlohaJoe
Posts: 3824
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by AlohaJoe » Wed Oct 31, 2018 2:35 am

bjr89 wrote:
Tue Oct 30, 2018 10:05 pm
But I don't really wanna invest until I know they aren't going to close it like a lot of other smart beta funds recently have. Only $38M in aum.
The only smart beta funds that have closed in 2018 are iShares sector funds that were shuttered in August. Things like "iShares Edge MSCI Multifactor Financials ETF".

Waiting is perfectly fine but if you look at ETF closures, basically few funds over $10 million closes.

Will that change in the future? Maybe? Maybe fund providers will start shutting down $30 million funds after 6 or 7 years after years of stalled or dwindling AUM? ETFs are relatively new so hard to say. But looking at fourth tier providers like VanEck and they seem happy to keep funds like that open for 5+ years so far.

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by vineviz » Wed Oct 31, 2018 4:04 am

stlutz wrote:
Tue Oct 30, 2018 11:49 pm
Sorry, but no. At least not completely: it's true that stocks in the "Pure Value" index are selected from the constituents of the base index that have 100% of their weight in the value basket, but it's not true that the "pure value index does indeed only contain value stocks with low earnings growth". The stocks in the Pure Value index are selected exclusively on their value score: there is no penalty for earnings growth.
Read the document from S&P that is linked above.

Some stocks exhibit both growth and value characteristics. If a stock has enough growth characteristics that it can be partially included in the standard growth index, then it is not considered for inclusion in the pure value index (and vice versa for pure growth).
I've read the methodology long ago. It doesn't contradict a thing that I wrote: these mythical "super high EPS growth but super cheap price/book" stocks do not exist in sufficient numbers or size to make a difference. If they did, the DFA fund would have stocks with significantly higher rates than the S&P 600 Value index funds. As we can see, it does not.

Instead, what happens is what you'd expect: the cheapest stocks have the lowest growth rates, so it really makes no difference on whether you perform a strict sort or a relative sort. You end up with the same basic factor loadings in any case.

All of which is a side show: the argument seems to be that the DFA US Small Cap Value fund is somehow arriving at a portfolio of higher quality stocks, and we can plainly observe (through factor analysis) that this is NOT the case.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

fennewaldaj
Posts: 272
Joined: Sun Oct 22, 2017 11:30 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by fennewaldaj » Wed Oct 31, 2018 5:24 am

vineviz wrote:
Wed Oct 31, 2018 4:04 am
stlutz wrote:
Tue Oct 30, 2018 11:49 pm
Sorry, but no. At least not completely: it's true that stocks in the "Pure Value" index are selected from the constituents of the base index that have 100% of their weight in the value basket, but it's not true that the "pure value index does indeed only contain value stocks with low earnings growth". The stocks in the Pure Value index are selected exclusively on their value score: there is no penalty for earnings growth.
Read the document from S&P that is linked above.

Some stocks exhibit both growth and value characteristics. If a stock has enough growth characteristics that it can be partially included in the standard growth index, then it is not considered for inclusion in the pure value index (and vice versa for pure growth).
Instead, what happens is what you'd expect: the cheapest stocks have the lowest growth rates, so it really makes no difference on whether you perform a strict sort or a relative sort. You end up with the same basic factor loadings in any case.
But as I pointed out above there seem to be active funds that have high value loading and better than value index growth numbers.
https://www.morningstar.com/funds/XNAS/LSVQX/quote.html
Presumably there is some other perceived problem with these companies for them to have the valuations they do.

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by vineviz » Wed Oct 31, 2018 8:10 am

fennewaldaj wrote:
Wed Oct 31, 2018 5:24 am

But as I pointed out above there seem to be active funds that have high value loading and better than value index growth numbers.
https://www.morningstar.com/funds/XNAS/LSVQX/quote.html
Presumably there is some other perceived problem with these companies for them to have the valuations they do.
I suspect that part of what's going on is that, empirically, historical earnings growth rates aren't terribly predictive of future stock returns. This is why there is no one talks about an "earnings growth factor".

I don't know that much about the LSV fund or its process, but according to Morningstar even though its holdings have better historical earnings growth than the S&P 600 in general the LSV fund has lower expected earnings growth. Most likely they are catching struggling companies early in the process (according to the factor regressions the LSV stocks are more profitable but are still investing their assets more aggressively, which is a NEGATIVE predictor of future returns).

The LSV fund isn't very old, but from the performance attribution it's pretty obvious why its performance has been similar to iShares S&P Small-Cap 600 Value ETF (IJS): for whatever different PROCESS they are using to build their portfolios, the end RESULT in terms of factor exposure is strikingly similar.

https://www.portfoliovisualizer.com/fac ... ssion=true
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

User avatar
jhfenton
Posts: 3328
Joined: Sat Feb 07, 2015 11:17 am
Location: Ohio

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by jhfenton » Wed Oct 31, 2018 9:40 am

bjr89 wrote:
Tue Oct 30, 2018 10:05 pm
stlutz, I think you hit on the key point. They were benchmarks first, products second.

Perhaps the Vanguard U.S. Value Factor ETF (VFVA) is what I'm looking for. But I don't really wanna invest until I know they aren't going to close it like a lot of other smart beta funds recently have. Only $38M in aum.
I believe closure risk on VFVA is very low.

1. VFVA has been slowly but steadily accumulating assets.
2. They are only marketing VFVA and the factor funds to advisors, and I'm sure many advisors won't take the chance on brand new funds either. Once they get past their first anniversary (February 2019), I expect flows to pick up a bit.
3. Vanguard has never closed an ETF.

I own both VFVA and VFMFX/VFMF.

garlandwhizzer
Posts: 2016
Joined: Fri Aug 06, 2010 3:42 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by garlandwhizzer » Wed Oct 31, 2018 1:12 pm

vineviz wrote:

In any case, Invesco S&P SmallCap 600 Pure Value ETF (RZV) has superior size, value, and quality factor loadings to DFA US Small Cap Value.
True, and also true that RZV has been a disaster for shareholders since its inception 13+ years ago. Attractive factor loadings in the real world do not necessarily correlate with robust performance. In the case of RZV those juicy loadings produced consistent and substantial underperformance relative to SCB for 13 years running now. Its Morningstar rating has dropped to 2 stars and shareholders have voted with their feet--now RZV has only 166 mil in assets. Another example of the gulf between compelling academic factor theory and real world results. It's fine if you're a factor true believer and plan to stay on board forever. But it's important to realize that there are serious limits for predictability of market action regardless of how nice the story looks on academic cost-free backtesting much of which occurred before factors were described and markets were not dominated by professionals/algorithms. When you choose factor based investing, you're making a bet based on a convincing narrative. Like all bets, yours may or may not pay off over any time frame going forward.

Garland Whizzer

fennewaldaj
Posts: 272
Joined: Sun Oct 22, 2017 11:30 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by fennewaldaj » Thu Nov 01, 2018 12:52 am

garlandwhizzer wrote:
Wed Oct 31, 2018 1:12 pm
vineviz wrote:

In any case, Invesco S&P SmallCap 600 Pure Value ETF (RZV) has superior size, value, and quality factor loadings to DFA US Small Cap Value.
True, and also true that RZV has been a disaster for shareholders since its inception 13+ years ago. Attractive factor loadings in the real world do not necessarily correlate with robust performance. In the case of RZV those juicy loadings produced consistent and substantial underperformance relative to SCB for 13 years running now. Its Morningstar rating has dropped to 2 stars and shareholders have voted with their feet--now RZV has only 166 mil in assets. Another example of the gulf between compelling academic factor theory and real world results. It's fine if you're a factor true believer and plan to stay on board forever. But it's important to realize that there are serious limits for predictability of market action regardless of how nice the story looks on academic cost-free backtesting much of which occurred before factors were described and markets were not dominated by professionals/algorithms. When you choose factor based investing, you're making a bet based on a convincing narrative. Like all bets, yours may or may not pay off over any time frame going forward.

Garland Whizzer
Interestingly RPV the S+P 500 pure value fund has massively outperformed the vanguard large cap value index in the last 10 years. It does seem like the out performance came in the early part of the cycle so it might be some weird thing about when exactly it bought certain financials.

https://quotes.morningstar.com/chart/fu ... ture=en-US

fennewaldaj
Posts: 272
Joined: Sun Oct 22, 2017 11:30 pm

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by fennewaldaj » Thu Nov 01, 2018 1:16 am

vineviz wrote:
Wed Oct 31, 2018 8:10 am
fennewaldaj wrote:
Wed Oct 31, 2018 5:24 am

But as I pointed out above there seem to be active funds that have high value loading and better than value index growth numbers.
https://www.morningstar.com/funds/XNAS/LSVQX/quote.html
Presumably there is some other perceived problem with these companies for them to have the valuations they do.
I suspect that part of what's going on is that, empirically, historical earnings growth rates aren't terribly predictive of future stock returns. This is why there is no one talks about an "earnings growth factor".

I don't know that much about the LSV fund or its process, but according to Morningstar even though its holdings have better historical earnings growth than the S&P 600 in general the LSV fund has lower expected earnings growth. Most likely they are catching struggling companies early in the process (according to the factor regressions the LSV stocks are more profitable but are still investing their assets more aggressively, which is a NEGATIVE predictor of future returns).

The LSV fund isn't very old, but from the performance attribution it's pretty obvious why its performance has been similar to iShares S&P Small-Cap 600 Value ETF (IJS): for whatever different PROCESS they are using to build their portfolios, the end RESULT in terms of factor exposure is strikingly similar.

https://www.portfoliovisualizer.com/fac ... ssion=true
It looks like they sort by price to cash flow, price to earnings, and dividend yield and in general avoid companies with negative current earnings. Since they are a qaunt fund they don't go into exact details of there model. I have a brokerage link in my fidelity 403b/401a/457 that allows me to buy mutual funds but not etfs. My 403b/401a/457/wifes 401k space is much larger than accounts where I can buy what I want so I looked at a lot of active mutual funds deciding if I want to buy any or buy vanguard small cap value with a transaction fee. I ended up purchasing this fund due to its deep value orientation, low turnover, and highish name count (~300). LSV was founded by the writers of the other important academic paper about value investing from the early 90s "Contrarian Investment Extrapolation and Risk" They are mainly a institutional asset manager but they do have a few mutual funds. It is kind of interesting to compare their large value fund to DFAs (1999-2018) because despite having a fairly different investment process they have more or less mirrored each other.

https://quotes.morningstar.com/chart/fu ... ture=en-US

Anyway don't know what if anything to make of all that but it is interesting at least to me.


columbia
Posts: 979
Joined: Tue Aug 27, 2013 5:30 am

Re: Vanguard/Blackrock value index funds bias themselves to low growth, and not necessarily high risk, firms

Post by columbia » Thu Nov 08, 2018 4:49 pm

Mr. Clements discusses style cheating in the latest BH podcast.

Post Reply