Academic Factor Exposure Versus Fund Factor Exposure

/Academic Factor Exposure Versus Fund Factor Exposure

Academic Factor Exposure Versus Fund Factor Exposure

By | 2017-10-06T12:42:05+00:00 April 26th, 2017|Factor Investing, Key Research, Tool Updates|28 Comments

Tomorrow I’ll be sitting with Pat O’Shaughnessy and Ben Johnson to discuss “Straight Talk About Smart Beta.” Here is a link to the big Morningstar event.

In preparation for our discussion we were spitballing ideas and Ben brought up the concept of helping investors understand the similarity/differences between live “factor” funds and the academic factor portfolios often used for research purposes.

As we will see from the case study below, there seems to be no resemblance.

In theory, if live funds roughly reflect the construction of the portfolios used to identify “alpha” in the academic research, one would expect to roughly perform in line with these theoretical portfolios. Of course, if live factor funds show no reflection whatsoever of the academic portfolios studied, then investors should not expect to achieve similar results to those that have been studied in the past.

For example, consider the generic book-to-market factor. The simplest way to express this “academically” is to sort all stocks from high to low on B/M, buy the top 10%, annual rebalance. Rinse, repeat. In fact, one can download the returns and study these returns by simply downloading the return series from Ken French’s website: http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/ftp/Portfolios_Formed_on_BE-ME_CSV.zip

If you run the stats on this portfolio relative to the passive index you’ll identify a long-term excess return of a few percent per annum. Doesn’t sound like much, but compound at 3% higher than the index over a 100 years and you start to seperate the millionaires from the billionaires. Of course, you’ll also identify that simple value portfolios are insanely volatile and go through long bouts of underperformance. Great, I understand the risk and reward. Sign me up, what’s the next step?

Let’s go buy a fund that does this strategy and capture this premium…right?

Uhh…not exactly…

We recently loaded “academic” factors into our visual active share database so users can explore how various funds actually line up against theoretical academic portfolios.

For example, start typing “academic” in the “add an ETF” search bar. You’ll notice that we have constructed decile portfolios for every factor characteristic in our database.

Let’s look at the cheapest stocks on P/E, P/B, and EV/EBIT:

 

Remarkably, there is currently not a ton of overlap between cheap P/B and cheap P/E. And mixed overlap for EV/EBIT and P/E and P/B. Or as Chris Meredith explains, “Factors aren’t commodities.

Let’s now look at the iShares VLUE value “factor” fund to see if they are exploiting the academic value factor. This fund is often used by advisors to exploit the so-called value premium. And how could it not? The name and ticker are pretty clear that this is a “value factor” fund. Awesome. But let’s dig into the weeds a bit.

First, let’s look at the VLUE holdings along the P/E and P/B dimension:

VLUE clearly doesn’t exploit cheap P/E or EV/EBIT characteristics.

How about we look along the P/E and P/B dimensions? The index construction file suggest they weight P/B fairly heavily (see index methodology here)

Nope, still not much value for the value factor fund.

How about we look at VLUE but substitute market cap on one dimension?

Bingo. Now we can visually identify what VLUE is all about. VLUE is not a value fund at all, it is simply a large-cap stock factor fund. If one were to sort on the top 10% of the stock universe on size they would essentially replicate the VLUE fund.

Fascinating.

Good luck finding live factor funds that actually have any resemblance to the academic portfolios studied and examined in the academic literature. If you find any, let us know.

Moreover, this basic analysis begs a simple question: If investors believe in the evidence presented in academic research, why are they buying factor funds with no resemblance to the portfolios that generated the evidence in the first place?

Weird.


  • The views and opinions expressed herein are those of the author and do not necessarily reflect the views of Alpha Architect, its affiliates or its employees. Our full disclosures are available here. Definitions of common statistics used in our analysis are available here (towards the bottom).
  • Join thousands of other readers and subscribe to our blog.
  • This site provides NO information on our value ETFs or our momentum ETFs. Please refer to this site.

Print Friendly, PDF & Email

About the Author:

After serving as a Captain in the United States Marine Corps, Dr. Gray earned a PhD, and worked as a finance professor at Drexel University. Dr. Gray’s interest in bridging the research gap between academia and industry led him to found Alpha Architect, an asset management that delivers affordable active exposures for tax-sensitive investors. Dr. Gray has published four books and a number of academic articles. Wes is a regular contributor to multiple industry outlets, to include the following: Wall Street Journal, Forbes, ETF.com, and the CFA Institute. Dr. Gray earned an MBA and a PhD in finance from the University of Chicago and graduated magna cum laude with a BS from The Wharton School of the University of Pennsylvania.

28 Comments

  1. John Hall April 26, 2017 at 9:53 am

    Is the label in the first chart mixing up Cheap P/E and Cheap P/B?

  2. Kevin April 26, 2017 at 10:28 am

    You’re right the visuals don’t show any value capture but VLUE stills shows a statistically significant loading to hml. I’m not sure what to make of it all.

  3. FIRV_TRADER April 26, 2017 at 11:11 am

    So, which ETF can I buy that best represents the EIT_BTM allocation?

    • Wesley R. Gray, PhD April 26, 2017 at 11:19 am

      No comment or I’ll get in trouble.
      Except to say the tool can quickly help you ascertain which funds do what they say they do vs those with strong sales departments.

      • FIRV_TRADER April 26, 2017 at 11:32 am

        Right. That’s a really annoying process; take ETF, plug into tool, asses output, repeat for next ETF. Someone needs to create a blog/website that provides a list of actionable ETFs/funds that best represent different factors. I recognize that licencors and ETF providers potentially have a conflict of interest and cannot provide such a list. If I didn’t have a real job, I’d create one myself.

        • Luke T April 26, 2017 at 9:52 pm

          Alpha Architect owns some ETFs. Maybe you should plug them into the tool?

      • dph April 27, 2017 at 5:37 pm

        Wes,

        I appreciate all your work. However how come you guys can’t talk about “other” guys ETFs. I know you have posed on how we can compare ETFs but why can’t you name a few outright that does the value well (other than your own)?

        Are there really formal rules against talking about (positive or negative)competitors or is this more of an unwritten rule type deal in your industry? I get the whole marketing thing and your own funds but not sure why you can’t be more specific about others.

        • Wesley R. Gray, PhD April 29, 2017 at 11:48 am

          Dph, happy to chat via email. Giving individual etf security recommendations exposes us to problems and our main objective is education not funding law firms. We ride the line by providing tools and education, but try and avoid specific recommendations to minimize perceived bias and compliciance issues.

          We’ll dig in more and see what is possible, given constraints…but not hopeful. Sorry!

  4. Govind April 26, 2017 at 11:32 am

    What you are showing is that there isn’t a lot of intersection with the the top deciles of these value factors. By concluding when that happens you have a closet index fund and not a value fund you are assuming that the second and third deciles have no factor power. They may have less factor impact than the top decile, but I’ve seen no evidence that they have no little or no impact. If that were the case, I doubt they would be used. HML sorts often include quartiles or terciles rather than deciles and still look valid.

    • Wesley R. Gray, PhD April 26, 2017 at 11:36 am

      The mojo is in the tails, not the middle of the roaders. So why focus on mediocre? The point is to extract excess risk premiums.

      • Wesley R. Gray, PhD April 26, 2017 at 11:37 am

        Agree there is some juice outside of extremes…but not near as much. Why bet on ace jack when u can take ace ace?

        • Govind April 26, 2017 at 11:54 am

          So what you saying without really saying it is “buy my ETFs”

          • Wesley R. Gray, PhD April 26, 2017 at 12:01 pm

            Nope. Saying investors should understand what they are buying, why they are buying it, and how much they are paying for it. Our products are arguably a horrific investment for the vast majority of consumers.

          • Gregory April 29, 2017 at 11:36 am

            VLUE is for main street investors who try to minimize tracking error but still take advantage of value factor. It is not a magical ETF but cheap and makes easier for a small investor to stick to it through thick and thin.

          • Wesley R. Gray, PhD April 29, 2017 at 11:44 am

            Totally agree.

  5. GRIMWEASEL April 26, 2017 at 3:57 pm

    This proves what some had opined all along. Most ‘smart-beta’ funds are marketing ploys. The big asset managers knew they were charging too much, but hey, make hay whilst the sun shines. Then Tommy Six-Pack cottoned onto the fee for lack of performance lark, and wandered over to Fool.com where they espoused Tracker Funds. Tommy bought trackers. Asset Managers’ AUM figs started to drop; they investigated; they saw Tommy loved passive. They gave him ‘active-passive’ in spades. Sacked the fund managers; hired algos and bots. Creamed off the economic rent until such time that the ETF vacuum sucked in all assets and crashed the market, hello Depression Redux.
    I think you need to play the factor zoo in the small cap arena. Hunt with the Minnows and not the Sharks. Tommy has the edge, because there is no career risk for Tommy and he is not limited by size or liquidity issues. He can happily sit in cash, use tactical approaches (dual momentum, moving averages etc) to sit out the bad times. Good drills Tommy.

    • Wesley Gray, PhD May 12, 2017 at 7:23 am

      well, now you can visualize what is actually going on. So hopefully the tool is useful for decision-making and understanding

  6. David Foulke April 27, 2017 at 10:03 am

    Wes, help me out with something. You show that VLUE is literally all over the decile map on P/B, P/E and EV/EBIT. But VLUE’s avg P/E of ~15x and avg P/B of ~2x are still cheaper than the S&P 500’s P/E of ~25x and P/B ~3.1x (I couldn’t find EV/EBITs). And VLUE’s 15 bps is a very cheap expense ratio. So intuitively, it seems like a better long run bet than the S&P 500, which itself costs maybe 5 bps. How might you assess, in expectation, how good of a deal this is for investors?

    • Wesley Gray, PhD May 12, 2017 at 7:23 am

      Hey David,
      I think that might be the simple average, but the weighted avg off factset is 22.1. And because VLUE obviously has a mega-cap bias, the weighted avg, is more comparable to the SP 500 P/E. So yes, it is marginally cheaper, but not by much. You would arguably need to be in the decile/quintile of cheap stuff to really capture any mojo. You’d also want to avoid mega-caps, where we already know the value premium is marginal, at best.
      So I think VLUE won’t hurt you that much because it is cheap and tax-efficient, but I’m not sure it will be materially different than the generic index fund over a 10yr horizon.

      • David Foulke May 12, 2017 at 8:58 am

        Aha! Of course! It didn’t even occur to me they would provide simple averages, rather than value weighting them. Makes perfect sense. I was suckered by Wall Street sales tactics yet again. So yes, I see now that an incremental 10 bps for a little bit of extra value weighting in large caps may be worth it, but only marginally. This is a great poster child for closet indexing.

        • Wesley Gray, PhD May 12, 2017 at 2:06 pm

          probably best to just plot out the holding on visual active share and you’ll get a sense for the real portfolio characteristic exposures. People can fudge statistics, but pictures are hard to deny.

          I think VLUE is a perfectly reasonable product at some level. Definitely a closet indexer, but also not charging that much, at the margin. Of course, if your intent is to access the value premium, you really aren’t gonna capture it over time — you’re gonna capture mega-cap beta over time. As long as the consumer is aware of this, what the heck…not a bad option!

  7. Mitesh Patel April 27, 2017 at 11:53 am

    Alternative ETF sponsors:

    AUMShares – built for capacity and not much else

    ClosetShares – let’s be HonestShares

    GeniusShares – the very smartiest smarty smart beta

  8. Jim Worden April 27, 2017 at 3:37 pm

    Doesn’t this have to do more with the different ways academia looks at factors? BlackRock is using MSCI and MSCI’s definition of “Value” is different than Ken French or how S&P or Russell or Analytic Investors or AQR or Research Affiliates look at “Value.” Just curious how much of this has to do with definition of value and whether multiple value metrics are used and the ranking is different.

    Wes, I’m curious also about your thoughts on isolating the “Value” factor by removing Size, Quality, Low Volatility, and Momentum or whether you believe it is okay to have overlap. I’m also interested to hear your thoughts on how “Value” may have drifted from its historical valuation (e.g. the value of value) and whether this should be factored in or not. “Value” companies may also exhibit different characteristics during different periods of the economic cycle and factors such as interest rates may also impact these companies. Thanks!

    • Wesley Gray, PhD May 12, 2017 at 7:12 am

      Yes, a lot of folks blend the ratios, but this won’t change the story much. The primary driver in product manufacturing is tracking error minimization algos, which means the portfolios will be driven by generic beta more than anything. They won’t load on the characteristics that drive the historical factor premiums, almost mechanically.

      Well, if you are long-only (and even long-short) this is tough to do and arguably getting too cutesy.

      With respect to factor timing and the RAFI type arguments, I think Asness has addressed this very well. Basically, our conclusion is similar to Cliff’s — factor timing is difficult, if not impossible. That said, the narrative is compelling, but the underlying evidence and economic theories are muddy at best.

      Great questions and I wish we had all the answers. We’ll keep investigating and having fun along the way!

Leave A Comment