By Jack Forehand, CFA (@practicalquant)
Earlier this year, I wrote about the mechanics of value investing and the many decisions that go into building value portfolios. Many investors believe that all value-based strategies will produce similar performance. But in reality, the many portfolio construction decisions that value managers make, which I highlighted in the article, can lead to major differences both in their performance and in the ability of investors to stick with their strategies.
I wrote my previous article from the perspective of a practitioner developing value strategies, but after writing it, I received several questions about how the framework I discussed in the article can be used for investors on the other side of the equation who are evaluating value strategies. Since I think many of the same principles apply, I wanted to use this article to look at some popular value ETFs and how I might evaluate them using these criteria.
Just to be clear before I start, nothing I will write here should be interpreted as investment advice. My goal is to identify what I might look at when evaluating value ETFs, not to pick the best one. And as you will see when the analysis is done, there is no such thing as the optimal approach to building a value portfolio because a large part of the process of matching an ETF with an investor is a function of that investor’s personal goals and risk tolerance. Also, in the interest of full disclosure, I also want to note that I am the portfolio manager of an ETF that uses a value approach. I will not mention or discuss that product in this article, but I think it is important for anyone who reads my articles to understand any personal interest I have in what I am talking about.
With that out of the way, let’s move on to our framework.
What is Value?
When I talk about value investing, I am talking about a more traditional definition of it. That means buying cheap stocks relative to current fundamentals. That could mean earnings or cash flow or book value or sales or a variety of other metrics. There are some funds out there that use the word “value” in their name, but don’t fit into this traditional value mold. When value struggles for an extended period, this type of thing tends to happen more often. Managers come up with new ways to define value that fit better with what is happening. I have no problem with that, and those types of approaches have certainly worked better than traditional value recently, but I won’t analyze those types of funds here. So if your definition of value involves using discounted cash flow analysis to show that Tesla is a value stock based on what you think it will earn between now and 2032, this article likely isn’t for you.
The Different Approaches to Value
To illustrate how diverse the approaches that different funds take to value investing can be, I have picked six ETFs that utilize a value-based approach to analyze.
Here are the funds I will look at and some basic attributes of each one.
|Ticker||Name||Universe||Weighted Average Market Cap||Median Market Cap|
|QVAL||Alpha Architect US Quantitative Value ETF||Largest 40% of Stocks||$28 billion||$9.2 billion|
|DVP||Deep Value ETF||S&P 500||$28 billion||$14 billion|
|ZIG||The Acquirer’s ETF||Largest 25% of Stocks||$36 billion||$7.7 billion|
|VLUE||iShares Edge MSCI USA Value Factor ETF||Small and Mid-Cap Stocks||$101 billion||$15 billion|
|VFVA||Vanguard US Value Factor ETF||All Stocks||$35 billion||$2.8 billion|
|FOVL||iShares Focused Value Factor ETF||Small and Mid-Cap Stocks||$12 billion||$7.8 billion|
What is the Universe?
Following the same framework I used in my previous article, the first step in evaluating these ETFs is to look at the investment universe they use. This may seem like a pretty straight forward process, but there is actually a significant amount of nuance to it. At a high level, selecting from the largest universe possible offers a fund the best chance to find the stocks that meet the specific characteristics it is looking for. But on the other hand, small-cap stocks have higher indirect transaction costs, and those costs can eat away at the potential excess returns they can produce, especially as a fund gets larger. Funds typically deal with this problem by either excluding small-cap stocks or adding significantly to the number of positions in the fund to mitigate the negative effect of transaction costs. Of the funds in the table above, four of them use a universe that includes large- and mid-cap stocks, one uses a large-cap only universe, and one includes small-cap stocks. That fund also has significantly more holdings than the others to compensate for that.
Another way to look at the types of stocks that a fund owns is to look at either the weighted average or the median market cap of its holdings. I like to look at the median because it is less impacted by outliers (if I add Apple to my value portfolio, the average market cap would likely rise significantly, but the median would be much less affected), but both measures can have value. For these funds, the median market caps tend to show what you would expect based on the universes they use. Most of these funds are at the top of the mid-cap range or the bottom of the large-cap range, with the exception of the Vanguard fund, which has a much smaller median than all the others because it holds small-cap stocks. The iShares Edge Value ETF is an example of where the median can be misleading. It uses a modified market cap weighting system and therefore has a large portion of its assets in large- and mega-cap stocks. The weighted average market-cap picks that up (its average is much higher than all the other funds), but the median does not. So it is important to look at both measures.
Which Metric to Use
|Ticker||Primary Value Metrics||Value Composite Percentile|
|DVP||EV/EBITDA, Price/Free Cash Flow||10|
|VLUE||Price/Book, PE, Enterprise Value/Cash Flow||22|
|FOVL||Price/Book, Price/Dividend, Price/Earnings, Price/Cash Flow||17|
As you can see from the chart above, there is no standard way to look at value. Every one of these funds has a different way to analyze whether a stock is cheap. Some firms think that they have identified a metric or combination of metrics that is superior to the other options. Others don’t have a strong opinion and use a composite of factors. I personally tend to like composites because I think it is difficult to identify which value metric will do best in advance, but there is no right answer. I ran the holdings of each fund, which I pulled from their websites, through our system and looked at where they rank using a composite of value factors. This analysis confirmed that they all are doing what they say they are doing. All of them have a weighted average value composite score in the bottom quartile of the market, which indicates they are buying cheap stocks.
How Many Stocks to Hold
|Ticker||Number of Stocks||Active Share|
There is wide variation in the number of stocks held by each of these funds. The most focused ETF, the Deep Value ETF, holds only 20 stocks, while the ETF with the most stocks, the Vanguard Value Factor ETF, holds 852. Those two portfolios will obviously produce very different return paths for their investors. As a general rule, the more concentrated funds tend to produce more exposure to the value factor, which if history is any guide should lead to more excess return over the market. But with that potential excess return comes increased volatility and typically longer and deeper periods of underperformance, which makes them harder to stick with.
Active share is another good measure that shows you how different a portfolio is than the market. It typically ranges from 0% (the portfolio is the market) to 100% (the portfolio has no holdings in common with the market). Active share doesn’t tell you that a fund’s portfolio is superior to the market. It just shows you that it is different, both in terms of the underlying holdings and the weights of those holdings relative to the index. I could build a portfolio of expensive, high debt companies with negative cash flow and it would have high active share, but it wouldn’t be something anyone should invest in, so it is important to not put too much weight in the measure. In my opinion, active share is best used as a guide to help understand the potential of a fund for excess returns. A fund with high active share has the potential to produce substantial outperformance, but also will likely deviate significantly from the market and have extended periods of underperformance on its path to get there. It should be no surprise that the more focused funds above have the highest active share. ZIG has an active share that exceeds 100% because it also shorts expensive stocks in addition to buying cheap ones.
How Often to Rebalance
Typically, value portfolios require less frequent rebalancing than other factors like momentum. So it isn’t a surprise that most of these funds are rebalanced quarterly or less often than that. FOVL is rebalanced monthly, but it has a specific set of thresholds that need to be met for a rebalance to occur, so it doesn’t trade every month. VFVA is the one active ETF in the group and it uses its own custom rebalancing approach. With the value strategies we run on Validea, we have found that the optimal rebalancing period can vary, and some of them have performed better when rebalanced less frequently, so I don’t think there are many conclusions that can be drawn based on how often these funds rebalance. If we were looking at momentum funds, I would say that more frequent rebalancing may be an advantage, but with value, the evidence is mixed.
|Ticker||Sector Overweight||Sector Underweight|
The chart above looks at each fund’s biggest sector overweight and underweight relative to its peer group. What might be surprising to some is how different these are given that all of these are value strategies. Some funds like VLUE have systems in place to limit deviation from the market’s sector weights. Given that most value funds are underweight technology, it makes sense that a system like that would lead to a Technology overweight relative to the value fund category. The primary metric each fund uses also plays a big role here. QVAL uses EV/EBITDA, which isn’t a good way to evaluate Financials, as its primary metric, so it has no Financials in its portfolio. FOVL, on the other hand, which uses more traditional value metrics, has over 40% of its portfolio in Financials. Since different metrics can lead to very different results, it is important to understand the implicit sector bets you are making when choosing a value ETF.
How to Weight Individual Positions
|Ticker||Position Weight Standard Deviation|
The final value strategy evaluation criteria I looked at in my article was position weighting. Many quant-based strategies tend to equal weight positions, but you will see other approaches as well such as conviction weighting, market cap weighting or even inverse volatility weighting. Position weighting may not play as important a role in an evaluation of a value strategy as some of the other criteria above, but it is still important to pay attention to, especially if the fund has a large portion of its assets in a small number of positions. Since ETFs disclose their holdings every day, figuring out how they weight their positions just requires going to their website. A more advanced method I also like to use is taking the standard deviation of the fund’s individual position weights. This gives you an idea how much dispersion there is relative to equal weight. Funds that use equal weighting will have low standard deviations and fund’s that use other approaches will have higher values. Based on this data, it is pretty clear that DVP and VLUE are both using alternative approaches, while the others look like they are equal weighting their positions.
I didn’t include this in my original article, but the final thing I like to look for is anything out of the ordinary employed by a fund. For example, ZIG is a 120/20 fund, which means that it invests 120% of its assets in value stocks and has short positions equal to 20% of its assets in expensive stocks. This type of approach should serve to magnify its returns if value stocks outperform growth stocks, but it will hurt returns if the opposite occurs. It is important to understand things like this that are unique to each fund before you invest.
I also haven’t mentioned fees anywhere in my analysis since I was focusing on portfolio characteristics, but they are obviously a very important part of the process of choosing any ETF. My view on fees is that you need to make sure you pay for what you get. Well-constructed, concentrated ETFs can justify higher fees by getting more exposure to the value factor. Products that look much more like the index will typically have lower fees, but also less factor exposure. The one type of product to watch out for is the closet indexing fund that looks very similar to the index, but charges the high fees of an active product. None of the ETFs I have analyzed here fit that bill, though.
So where does all of this analysis leave us in our search for the best value ETF? Ultimately, I think it leaves us with the conclusion that there isn’t one. Which value ETF to invest in, or whether to invest in one at all, is a decision that should be primarily determined by factors that are specific to each individual. For an investor who is a believer in value that can deal with the fact that their portfolio will have major deviations from the market, the focused ETFs could be worth looking at. For others who want exposure to value, but want a much smoother ride, the ETFs that look much more like the market could make sense. And for many investors, investing in a market index fund is the way to go because it avoids all the behavioral issues that come with looking different than the market. For investors who do choose to invest in value ETFs, it is important to look under the hood and understand what you are getting. I think the framework presented above can be a good way to do that.
Photo: Copyright: 123rf.com / fotogestoeber