Seeking Alpha

How Deep Is Your Value? A Critique Of 'Shallow Value' Factor ETFs

by: Tariq Dennison

"Value" is one of the oldest "smart beta" factors, and has been implemented by many funds in many ways over the past 25+ years.

Many of the largest value funds exhibit a "shallow value" filter, selecting roughly the cheapest half of a broad index like the S&P 500, and then market-cap weighting.

Dimensional Fund Advisors (DFA) and WisdomTree implement the value factor differently, and have so far outperformed both the Vanguard Value Fund and the S&P 500.

While there are scale reasons the "cheaper half" approach is as deep as many $10bio+ funds might go, we should expect to see more "deep value", "extreme value", and "value quintile" ETFs enter the market.

Beyond the simplest market cap weighted, price weighted, or equally-weighted index funds, a simple factor-based value fund should be one of the simplest and cheapest kind of equity index fund to both run and explain to clients. Many of us can easily teach an intern with a spreadsheet how to download a price-to-earnings, (P/E) or other value ratio for each of 50 - 5,000 stocks in a universe, and then create a weighted portfolio of stocks filtered by or tilted around these ratios. A few examples of "value index" rules applied to the S&P 500 universe (as an example) similar to those implemented in actual portfolios can be summarized in plain English as:

  1. Select the 250 stocks with the lowest P/E ratio, and weight those by market cap. (The "cheaper half" approach, weighted towards "the biggest of the cheap")
  2. Select the 100 stocks with the lowest price-to-book (P/B) ratio, and weight those equally (The "cheapest quintile" approach, plus a tilt away from large cap in weighting)
  3. Weight all 500 stocks by earnings instead of by market cap (The "WisdomTree" approach)

The idea of "value" as a factor on which to select or weight investments is probably as old as investing itself, but is widely said to have taken off in the form of "value funds" (as we know them today) in the early 1990s. In June 1992, Eugene Fama and Kenneth French published their famous paper highlighting evidence of persistent value and size factors in US stock returns from 1963 to 1990. In short, this paper (and numerous following papers) show that cheap stocks tend to outperform expensive stocks, and also that small cap stocks tend to outperform large cap stocks. Also in 1992, Morningstar Inc (MORN) introduced their well-known "style box", classifying stocks and stock funds by size / market cap into small vs medium vs large and by relative valuation ratio levels by value vs blend vs growth.

Vanguard Value vs Growth

Later that year, the Vanguard Group, best known for their low-cost, passive Vanguard 500 Index Fund (VFINX) tracking the S&P 500, launched the Vanguard Value Index Fund (VIVAX) along with its compliment, the Vanguard Growth Index Fund (VIGRX). Despite my tremendous respect for Vanguard, and the simple yet effective power of low cost indexing, I am less impressed with the significance of splitting Vanguard funds between growth and value given the close convergence in performance of these two funds over the past 26 years, as charted below. While there have been periods where growth has outperformed value (1998 - early 2000, and late 2015 - present), Vanguard's Value Index Fund has so far not really delivered significant value premium. Despite that, it has still become the largest US large cap value ETF listed below.

Vanguard Source: Yahoo! Finance

A few quick keystrokes on Bloomberg might show this lack of difference in performance is not really Vanguard's fault, as the broader Russell 1000 Value Index and Russell 1000 Value Index show very similar convergence. Going all the way back to 1980, the difference in annual performance between the growth and value versions of this US large cap benchmark is only about 31 basis points. By contrast, there has so far been a more significant difference in performance between the small cap Russell 2000 Value Index and the Russell 2000 Growth Index. This indicates that while a simple "cheaper half" strategy may not bring much edge to market cap weighted large cap portfolio, historical outperformance has been over 250 basis points in small caps, and so it is worth combining the value and size factors.

Russell 1000 vs 2000 Growth vs Value Source: Bloomberg,

Dimensional Fund Advisors (DFA) vs Vanguard in Value Fund Performance

Less than a year later, in early 1993, Dimensional Fund Advisors (DFA) launched a US Large Cap Value Portfolio (DFLVX), promoting the academic research underlying their value factor investment process, including by having Eugene Fama and Kenneth French themselves on the DFA board. Although DFA launched both large and small cap funds, they very notably did not launch any "growth" funds, showing a pure conviction towards the value factor as a driver of returns. My guess is that they launched both large and small cap funds to meet the requirements of larger institutional investors that may have more difficulty making large allocations to small cap stocks. The institutional requirement would explain why DFLVX has twice as much in assets ($32bio vs $16bio) as the DFA Small Cap Value Portfolio (DFSVX). So sticking with large cap value, the below chart shows that the DFA Large Value fund has significantly outperformed Vanguard's Value fund over the past 25 years.

DFA vs Vanguard Large Cap Value Funds vs WisdomTree Source: Yahoo! Finance

WisdomTree's Earnings Weighted Approach

Just as DFA earned the support of Fama and French, WisdomTree Investments Inc (WETF) took on a celebrity endorsement from Wharton professor and Stocks for the Long Run author Jeremy Siegel. When WisdomTree first launched dividend-weighted ETFs in 2006 and earnings-weighted ETFs in 2007, Siegel went on TV to promote them. The rationale was simple: when investing $100 into two companies earning the same amount of money, does it make sense to invest 9x as much in one trading at 90x earnings than the one trading at 10x earnings, or should the $100 be allocated evenly based on earnings? As the above chart shows, the WisdomTree Earnings 500 ETF (EPS), which basically takes the same 500 stocks in the S&P 500, but weights them according to earnings rather than size, has done well to track DFLVX and outperform both Vanguard Value and the S&P 500 since its launch 11 years ago.

Surprisingly to me at least, EPS is only WisdomTree's 28th largest ETF with barely over $200 million in assets, compared with billions each in its currency-hedged and dividend-weighted ETFs.

Why did the DFA and WisdomTree funds outperform Vanguard?

While one could go deep into the documentation explaining the different investment approaches of these three value funds, a quick look at the top holdings could give some clue as to why and how DFLVX and EPS outperformed VIVAX. For brevity, I only used the symbols instead of full company names in the tables below showing the top 10 holdings of each of the three funds. At a glance, it seems that:

  • Vanguard simply filtered out "growth" companies like Apple Inc (AAPL) and Alphabet Inc (GOOG) and cap-weighted the remaining "value" stocks.
  • EPS weighted companies by earnings, allowing them to hold high-performing, high-growth, yet still extremely profitable tech companies like AAPL and GOOG.
  • DFA seems to weight its portfolios using a combination of value and other factors, which are harder to guess without reading their methodology.
Holding in VIVAX Holding in DFLVX Holding in EPS
MSFT 6.6% XOM 3.9% AAPL 6.2%
BRK/B 3.1% T 3.8% MSFT 2.9%
JPM 3.0% PFE 3.7% JPM 2.4%
JNJ 2.8% INTC 3.5% GOOGL 2.2%
XOM 2.6% WFC 3.1% BRK/B 1.9%
BAC 2.3% CMCSA 2.5% BAC 1.8%
UNH 2.0% JPM 2.4% WFC 1.7%
WFC 2.0% CVX 2.3% VZ 1.6%
PFE 1.9% BAC 2.2% JNJ 1.5%
T 1.8% BRK/B 2.0% INTC 1.4%

Source: Vanguard, DFA, WisdomTree

DFA's Mutual Funds vs WisdomTree ETFs

While I respect DFA's research and execution, I personally dislike the fact that DFA remains committed to the mutual fund vehicle and has so far not launched their portfolios as openly accessible ETFs. I believe they do this largely to maintain control over their distribution and execution in ways that are more difficult to do with ETFs. I do believe not being an ETF has also eased the pressure on them to publish the holdings of DFLVX in a format other than PDF, which makes it slightly harder for me to process than most ETF providers' CSV or XLS files. That said, I do buy DFA funds in HSA Health Savings Accounts and 529 plans which are still on mostly mutual fund friendly rather than ETF friendly platforms. Given that DFA mutual funds and most smart beta ETFs have been quite cost competitive in keeping expense ratios below 50 basis points (often even below 30), I don't make too much of a fuss over this, though still personally and professional prefer the ETF instrument.

Leaderboard: The Top US Large Value ETFs as of 2018

Below are the six largest US-listed US large cap value ETFs as of 2018Q4:

ETF Symbol AUM Exp Ratio Launched P/E Past EPS Growth 3yr Return 10yr Return
Vanguard Value ETF VTV $44.1B 0.05 2004 17.0 5.1 16.6 10.2
iShares Russell 1000 Value ETF IWD $37.6B 0.20 2000 15.8 7.7 14.1 9.2
iShares S&P 500 Value ETF IVE $15.1B 0.18 2000 16.0 6.8 14.7 9.1
Schwab U.S. Large-Cap Value ETF SCHV $4.7B 0.04 2009 17.5 4.0 15.9
iShares Core S&P U.S. Value ETF IUSV $4.6B 0.04 2000 16.0 6.6 14.9 9.3
iShares Edge MSCI USA Value Factor ETF VLUE $3.7B 0.15 2013 13.3 16.1 17.3

Source: Fidelity

I read the above table as follows:

  • The Vanguard Value Fund (with the mutual fund "VIVAX" and ETF "VTV" simply being two different handles of the exact same fund) remains the largest despite the above highlighted under-performance in the US large value category. This can probably be attributed to its long lead as the low-cost provider.
  • The next two iShares funds, IWD and IVE, tracking the "cheaper half" of the Russell 1000 and S&P 500 respectively, straddle DFLVX's $32 billion AUM. These massive assets have been maintained despite underperforming even the Vanguard Fund with an expense ratio 3.5-4x as high as Vanguard's and almost as high as DFA's. The success of these funds can probably be attributed to the brand strength of iShares and the Russell and S&P indexes.
  • The next two funds are the cheapest US Large cap value ETFs on the table, and deserve a mention for two different reasons:
    • IUSV has a 93% overlap with IVE, and is basically a cheaper version of the S&P value product in line with the "iShares Core" series
    • SCHV also costs only 4 basis points, and has outperformed the iShares products, but still not performed as well as the Vanguard fund.
  • VLUE is from iShares's relatively new line of "Edge" ETFs implementing smart beta factors including value, as well as momentum, quality, and low volatility. At first, the Edge Value ETF looked like just a bit of marketing buzz to justify a slightly higher expense ratio, but then I noticed VLUE had a significantly lower P/E, significantly higher EPS growth, and noticeably better performance than other value ETFs.

The Dirty Secret of "Core" Value and Growth Funds: Overlap

One big surprise when I compared the iShares S&P 500 Growth Fund (IVW) vs its Value counterpart IVE is that the two overlap significantly, currently with 27% of AUM overlapping between the funds. At the top of the overlap list are health care names Johnson & Johnson, Pfizer, Proctor & Gamble, and Merck, which one could argue are both growth companies and value priced, but such significant overlap reduces the edge of a value vs growth factor. This overlap is due to IVW and IVE each holding over 300 stocks from the S&P 500, so at best its holders would get the "cheapest 60%" of the S&P, without the value-tilt weighting of a fund like EPS.

Vanguard is much cleaner in avoiding overlaps; the Vanguard Growth ETF (VUG) "only" overlaps with VTV on 29 names, making up only 3% of each fund's AUM.

Being a "core" product probably implies the need to hold large assets that make up the median of the distribution, and this trade off between core and edge is arguably the new active vs passive debate.

There have been several studies on the performance of stocks in different value "quintiles" or "deciles". Many have shown that the more significant outperformance in growth vs value comes at the extreme deciles (e.g. the cheapest 10% of stocks vs the most expensive 10% of stocks in the universe), rather than owning broader slices that include the (often larger-cap) middle. A short sample list of these studies include:

It is probably a testament to how expensive it still is to launch and run an ETF that someone has still not come out with a "decile series" or even "quintile series" of ETFs on a major equity universe, tracking the top 10% vs 2nd 10%, ... to bottom 10% of stocks by some metric (P/E, P/TBV, etc.) in 10 easy to compare ETFs.

The "Dogs of the Dow" and "Foolish Four" strategies are two well-known low-tech and simple versions of a strategy based on selecting 10 stocks and 4 stocks, respectively, from the Dow Jones Industrial Average, closer to the "edge" than to its core of 30 stocks. These two strategies use so few stocks that they are easier to buy directly than through an ETF, but that also leaves them with much more single stock risk.

Why the iShares "Edge" seems worth the premium

As mentioned earlier, the first thing I noticed about VLUE that set it apart from the other value ETFs was the significant different in P/E and EPS growth metrics. Next, I noticed this seemed to be because VLUE holds only about 150 stocks, as opposed to around 300 in VTV and IVE. 17% of the assets in VLUE are in stocks with a P/E less than 10, versus only 11% of IVE and 7% of the S&P 500. VLUE also seems to have factored in quality, with almost 17% of the assets in VLUE in stocks with over 20% 5yr average ROE, vs 5% of the assets in IVE and 13% of the assets in the S&P 500. As another quality factor, VLUE has only 12% of its assets in stocks with accounting practices MSCI describes as "Very Aggressive", versus 18% of IVE and 21% of the S&P 500. VLUE components do lag slightly on ESG factors, but a whole other article could be written about that.

While VLUE would become more of an "Edge" product by moving further away from the core and reducing its holdings to 100 or fewer stocks (say as a "top quintile" product), these statistics show that VLUE already scores differently enough that it is likely to perform more like DFLVX or EPS than like IVE or VTV.

Further on the "edge", when I searched for "deep value", I did fund an ETF called, simply, the Deep Value ETF (DVP). This fund is even more on the "edge" with a concentrated portfolio of only 20 stocks. DVP has $156 million in assets, which is still small enough to run such a strategy, but its high expense ratio of 0.59% and standalone strategy may make it less appealing than a pure "decile" strategy. With respect to practitioners of deep value, I should say the term "deep value" often does deserve an amount of in-depth research into the quality and true intrinsic value of a business which is still mostly done manually until the AI catches up. So rather than calling a rules-based "cheapest decile" strategy "deep value", I'll call it "extreme value".


Of all the value funds above, VLUE and EPS remain my favorites, followed by DFLVX (mostly because the latter is not an ETF). The reasons I wrote 2 years ago about why I don't invest amounts over $100,000 into S&P 500 index funds may be getting challenged by the ever-falling prices of smart-beta ETFs, but there still do not seem to be many good, cheap ETF replacements for a direct basket of 20-50 "extreme value" stocks, equally weighted. In this age of abundant data and computing power and super-low-cost electronic trades, it will be a race to see whether decile strategies can thrive in ETF form, or whether investors trade these deciles as direct baskets. Meanwhile, the goal of investing in value ETFs should be to move away from the "core" and towards the "edge" until the right "extreme value" strategy is ready.

Tariq Dennison runs Hong Kong based asset management firm and US RIA GFM Asset Management, and is the author of the book "Invest Outside the Box: Understanding Different Asset Classes and Strategies"

Disclosure: I am/we are long SCHV, DFLVX, VLUE, EPS. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.