By: Wesley R. Gray, Ph.D.
Unlike equity and bond investing, investing in commodities is a less familiar undertaking for many. Commodities behave differently than stocks and bonds, pose different risks, and can be taxed differently. And let’s be blunt: few retail and/or professional investors understand how to use commodity futures.
To address our own blind spots in the marketplace, we’ve spent the past few years investigating academic research and examining trading systems and strategies in the commodity futures space. At this stage I’d say our knowledge is fairly extensive, but we’re still learning and trying to get better. After acknowledging our own educational blind spots with respect to commodity futures, and subsequently studying and learning about the different ways they can be traded, used for hedging, and integrated into portfolios, we’re always curious to see how other “professionals” think about these contracts. Based our review of the product landscape and numerous discussions among “professional” asset allocators, the results don’t look promising.
For example, consider the discussion below from a popular robo-advisor’s website. The site compares equity-based ETFs, XLE and VDE, which buy the stock of commodity-related firms (e.g., Exxon or Chevron) to DJP, a futures-based ETN, which buys baskets of commodity futures based on the commodity’s “liquidity and economic significance.” Odd, to say the least.
First, the underlying assumption in the robo-advisors discussion is that these securities are somehow the same, which is implicit in their reference to tax-loss harvesting, which bounces investors across the 3 options so they can book losses, but maintain the same “exposure.” But the problem is that these exposures are not the same. The chart below highlights this fact without the use of numbers or equations or basic knowledge of commodity futures.
Oddly enough, the robo advisor highlights that XLE and VDE are “better” because they are “cheaper.” But how can one describe “cheapness” when one is comparing apples to oranges? Next, the commodity index chosen (DJP) doesn’t even consider the factors that matter in commodity investing (e.g., term structure and momentum). DJP is a classic case of products chasing performance.
To be clear, this isn’t a bash session on robo-advisors–we think robos are doing some great things in the business–to include better transparency, which is why we were able to actually get the chart above (most advisors/sites keep it “proprietary”). Nonetheless, the discussion is indicative of financial advisor knowledge of commodity futures.
So let’s try and get smart about commodity futures
First, a recap of what matters in commodity futures trading:
And here are some great papers on the subject by Erb/Harvey:
- Common misconceptions on Commodity Futures Investing
- The Tactical and Strategic Value of Commodity Futures
The lessons from Erb/Harvey are as follows:
- Commodity futures are NOT a play on commodity spot prices.
- Commodity futures should NOT be considered an asset class, but a trading strategy.
- Commodity futures do NOT deliver equity-like performance.
And here is the biggest common misconception we see (highlighted above):
Investing in commodity futures is not the same as investing in equities tied to commodities
For example, owning oil futures is not the same as owning Exxon stock.
If you don’t believe us, check out the lengthy post on this subject that we wrote recently.
What’s the bottomline?
Any good advisor should know how and why they invest a client’s wealth. A good advisor will certainly make bad calls and lose money at times, but the onus is on the advisor to make informed decisions.
By: Tian Yao
Do-It-Yourself tactical asset allocation weights are posted.
- No equity exposure!
- Hold cash and wait for the trend to become our friend
By: Wesley R. Gray, Ph.D.
Meb Faber tweeted an interesting question related to taxes, dividend payers, and the value premium the other day.
I’ll summarize the heart of the question:
How do dividend payers affect the value premium?
Instead of pontificating on the issue, we decided to directly address the question via empirical tests.
Portfolio construction details:
- Construct equally-weighted portfolio each year on June 30th. (use market cap on June 30th, as per the Asness/Frazzini paper).
- Focus on mid/large cap stocks. (Firms’ market cap below 40th percentile, excluding financial companies).
- Sort by BM and by firms which paid dividends the prior year.
- Rebalance annually.
Important to note is that the universe of non-dividend paying mid/large cap companies is actually fairly small. We start our analysis in 1990 because of this fact. The decile portfolios have ~20+ stocks at around that time period (i.e. 1990–>). Prior to 1990, the decile portfolios only held 5-10 stocks–sometimes fewer–so we do not include those results.
First, we dug into the performance of our book-to-market deciles since 1990 (see figure below) and compare them to the BM deciles obtained from the Ken French Data Library. The French BM deciles are value-weight across all market caps and use lagged market cap data, which is different than our construction. However, the empirical results below highlight that the distributions are approximately equal, with small differences.
The figure below highlights that BM has been a shoddy way to access the value premium in recent history. This is not too surprising if you follow our blog and our research, which highlights that the king of valuation metrics are enterprise multiples, not BM. Regardless, we’ll stick with BM since that is an academic and practitioner favorite.
Dividend Payers vs. Non-Dividend Payers
Next, we look at how the BM deciles perform across dividend and non-dividend payers.
The decile results divided by dividend payers/non-payers are noisy, but it appears as though the BM version of the value anomaly is driven, at least in part, by dividend paying companies. The deciles are roughly equal across BM for non-payers, but we see a more pronounced compound annual growth rate difference between cheap dividend payers and expensive dividend payers.
Does the BM value measure drive dividend payer/non-payer results?
The analysis above suggests that dividend payers drive the value anomaly, when the value anomaly is measured via book-to-market. But as we highlighted at the outset, BM has never been an effective valuation metric. We understand that it is “famous” and used by huge firms such as DFA, but why use a sub-optimal approach to systematic value-investing in the first place?
The decile splits between dividend payers and non-dividend payers across price-to-earnings (PE) ratios is pictured below:
In contrast to the BM results, which clearly show that dividend payers drive the value anomaly, there doesn’t seem to be a clear cut relationship between the value anomaly and dividend payers when looking at PE ratios. Low PE dividend payers tend to outperform high PE dividend payers by a large amount, but low PE non-payers tend to beat high PE non-payers by a decent amount as well. From an empirical standpoint, it is harder to make the claim that the value anomaly–as measured by PE–is driven by dividend payers. One could say there is “weak” evidence for this claim.
But the next set of results muddy the water even further…
Consider the king of valuation metrics, EBIT/TEV, which is empirically the best-performing metric and one that we use in our own academic and practical work.
The results using enterprise multiples suggest there is weak evidence that non-dividend payers actually drive the value anomaly, which is in direct contrast to the evidence using BM as a valuation metric.
- BM is heavily influenced by payment classification and the evidence strongly suggests that dividend payers drive the value anomaly.
- PE is weakly influenced by payment classification and the evidence weakly suggests that dividend payers drive the value anomaly.
- EBIT/TEV is weakly influenced by payment classification and the evidence weakly suggests that dividend non-payers drive the value anomaly.
There is no robust evidence to suggest that dividend paying classification drives the value anomaly in one direction or the other.
By: Jack Vogel, Ph.D.
Active management has been out of favor for a while–high fees, high tax burdens, and poor long-term performance. But with the slow rise of actively managed ETFs, which have lower costs and more tax efficiency than traditional active mutual funds, the gateway to active management has potentially been reopened. This is certainly a positive move, but cheaper more tax-efficient active funds don’t answer the question of how should one use active exposures in a portfolio. We address this question in this post and propose several reasonable approaches one can take to incorporate active ETFs in to a diversified portfolio:
- Core-Satellite: The core of the portfolio is cheap index funds, the satellite funds are concentrated active ETFs.
- High-conviction: The core is active ETFs, combined with strategies and asset classes that tend to work well at different times.
Let’s dig into each of the approaches in more detail.
The Core-Satellite approach is fairly simple — for the “core” of the portfolio (let’s say 80%), invest in passive index funds. For the “satellite” of the portfolio (the other 20%), invest in highly active ETFs. Additional information can be found from the CFA institute and Vanguard.
Why would this be good for an advisor or a DIY investor?
One issue with going “all-in” on actively managed ETFs is that they tend to have a large deviations around an index (i.e., tracking error). For advisors who have to answer to short-horizon clients that review their accounts daily (or DIY investors who always compare themselves to an index), tracking error can create angry clients very quickly. The core-satellite approach may be optimal in this situation, because, by construction, a large part of the portfolio is allocated to passive index funds, which always keep the portfolio roughly inline with broad benchmarks. This core-satellite approach will lower tracking error of the overall portfolio, but give clients a shot at outperformance over time. How much is dedicated to passive and how much is dedicated to active really depends on the client-advisor relationship and the amount of time the advisor spends educating clients on thinking long-term when it comes to portfolio performance. The details of creating an effective core-satellite approach can get complex, but we outline some basic principles of concepts related to a core-satellite approach here.
The high-conviction approach is the approach we take with our personal wealth and most of our clients. Why we take this approach is described here and here. In this approach, the passive part of the portfolio does not exist because it is effectively captured in a long-only diversified portfolio already. There are many active strategies available, but we believe that Value and Momentum are the best long-term bets when it comes to active management.
Of course, the problem with high-conviction active portfolios is they aren’t the entire market, and can gyrate wildly around an index. If an advisor has short-term focused investors and the gyration is positive, you’re a hero, but if short-run performance is negative, you no longer have a career in asset management–yikes! We recommend that advisors building a high-conviction active portfolio combine a variety of top-shelf concepts so they help diversify their client’s exposures and also so they limit their own career risk (unless this isn’t a factor because of unique clients).
Sounds great, but if high conviction has a higher expected risk-adjusted return, why diworsify?
Consider high conviction value investing, which sounds so simple — buy the cheapest highest quality stocks you can find. The problem with these strategies is they can underperform for long stretches of time! After 6 years of underperformance, are you really going to stick with the strategy? For most advisors (and their clients) and DIY investors, the answer would be NO!
So diversifying across high-conviction active ideas is critical! Ideally we could find strategies that work well at different times, and then just allocate a portion to each of the strategies. For example, as shown here and here, Value and Momentum tend to work well at different times. So one might consider investing in BOTH value and momentum, as opposed to focusing on the absolute merit of one over the other.
Overall, we outline two reasonable approaches to using high conviction active ETFs: Core-satellite and high-conviction. For those advisors and investors who want to track an index and hope to beat the market by a small amount, the core-satellite approach may be the best route. For advisors and investors who are not as concerned with more informed clients and less short-run career risk, the high-conviction route may be a better approach.
By: Jack Vogel, Ph.D.
However, there is a new paper by Toby Moskowitz titled “Asset Pricing and Sports Betting” which examines how size, value and momentum affect sports betting contracts:
I use sports betting markets as a laboratory to test behavioral theories of cross-sectional asset pricing anomalies. Two unique features of these markets provide a distinguishing test of behavioral theories: 1) the bets are completely idiosyncratic and therefore not confounded by rational theories; 2) the contracts have a known and short termination date where uncertainty is resolved that allows any mispricing to be detected. Analyzing more than a hundred thousand contracts spanning two decades across four major professional sports (NBA, NFL, MLB, and NHL), I find momentum and value effects that move betting prices from the open to the close of betting, that are then completely reversed by the game outcome. These findings are consistent with delayed overreaction theories of asset pricing. In addition, a novel implication of overreaction uncovered in sports betting markets is shown to also predict momentum and value returns in financial markets. Finally, momentum and value effects in betting markets appear smaller than in financial markets and are not large enough to overcome trading costs, limiting the ability to arbitrage them away.
Some Interesting Points
The figure below explains the different price movements which are studied in the paper:
Here are the T-stats for the momentum betas in the figure below:
Analysis from the paper:
A consistent pattern emerges for the Spread and Over/under contracts in every sport, where the momentum betas exhibit a tent-like shape over the three horizons—near zero from open-to-end, significantly positive from open-to-close, and significantly negative from close-to-end, with the initial price movement from open-to-close related to momentum being fully reversed by the game outcome. The patterns for the Moneyline contracts exhibit the same tent-like shape, but are less pronounced, consistent with the Moneyline perhaps being less affected by “dumb” money and more dominated by “smart” money.
Then the paper shows the T-stats for the value betas in the figure below:
Analysis from the paper:
A consistent pattern is evident from the plots: a value contract’s betting line declines between the open and close and then rebounds between the close and game end, reaching the same level it started at the open. These patterns are consistent with an overreaction story for value, where value contracts, which measure “cheapness”, continue to get cheaper between the open and the close, becoming too cheap and thus rebounding positively when the game ends. This picture is the mirror image of momentum, where value or cheapness is negatively related to past performance, and hence the pictures for momentum and value tell the same story. (Though, recall the measures for value and momentum were only mildly negatively correlated.)
Conclusion from the paper:
Examining momentum, value, and size characteristics of these contracts, analogous to those used to predict financial market security returns, I find that momentum exhibits significant predictability for returns, value exhibits significant but weaker predictability, and size exhibits no return predictability. The patterns of return predictability over the life of the betting contracts—from opening to closing prices to game outcomes—matches those from models of investor overreaction. The results suggest that at least part of the momentum and value patterns observed in capital markets could be related to similar investor behavior. The magnitude of return predictability in the sports betting market is about one-fifth that found in financial markets, where trading costs associated with sports betting contracts are too large to generate profitable trading strategies, possibly preventing arbitrage from eliminating the mispricing.
An interesting paper, showing that Value and Momentum work within the sports betting market, but the cost of trading on the signals is too large for profitable trades. This is probably why the “house always wins.”
It’s a good thing I watch countless hours of sports to form my own “expert” opinions!
By: Wesley R. Gray, Ph.D.
According to the latest Global Business Outlook survey jointly conducted by Duke University and CFO magazine, 55% of U.S. companies say they think the stock market is overvalued, while only 6% of them think the stock market is undervalued.
More than 1,200 CFOs, including 510 from the U.S., participated in this recent quarterly survey. The survey questions cover five main parts: Business Optimism, CFO top concerns, Employment and wages, stock market valuation, and risk management concerns. (Click to download: CFOsurveyOverview_2015Q3 Final)
CFOs are very bearish on the U.S. market,” said Fuqua professor Campbell R. Harvey, a founding director of the survey. “Our survey took place during a volatile time where there was a 10 percent market correction. Even after this drawdown, 55 percent of CFOs thought the market was overvalued. source.
Here are some stats on various metrics from 510 U.S. firms.
- CFOs’ expected earnings growth in next 12 months is 3.0%, down from 11.7% last year.
- CFOs’ expected revenue growth in the next 12 months is 3.5%, down from 7.3% last year.
Not a lot of bulls in CFO land…
By: Wesley R. Gray, Ph.D.
The Harvard Management Company report was recently released here. h.t. Tom for pointing out.
The biggest surprise was the spread in returns between the global 60/40 portfolio and the domestic 60/40 portfolio:
An 8.5% spread is quite large…
By: David Foulke
Always seek to simplify.
Occam’s razor teaches us we should cut away any extraneous factors that are unnecessary to explain something. Stated another way, we should avoid adding predictive elements unless they are absolutely necessary and strongly enhance our prediction.
But why should we believe this is necessarily true? Where’s the evidence? Forecasting is a subtle art. Perhaps more complex models can capture a wider range of class frequencies and offer more granular forecasts. Thus, maybe in the realm of predictions, we should favor complexity over simplicity?
Wharton Professor of Marketing J. Scott Armstrong has set about trying to measure the effectiveness of simple versus complex approaches to forecasting. He recently had a great post on the Wharton Blog Network relating to some research he’s done on this question.
In a new article in the Journal of Business Research, Armstrong discussed a recent meta study, or study of studies, he undertook involving 32 papers covering a range of forecasting methods (judgmental, causal, etc.), which included 97 independent comparisons of the accuracy of simple versus complex approaches.
What did the evidence say?
In a slam dunk, Armstrong found that in an extraordinary eighty-one percent of these independent comparisons, the simple forecasts beat complex ones. Moreover, the errors of from complex forecasts were 27 percent greater than for simple forecasts. So complexity reduced forecast accuracy and increased mistakes. In fact, in his survey of the research, Armstrong couldn’t find any papers that argued that complex predictive models beats simple ones.
Despite the seemingly conclusive nature of this finding, there’s also plenty of evidence that people still prefer complexity over simplicity. Why? It may be that there’s something persuasive about complexity itself.
For instance, Armstrong describes the “Dr. Fox effect,” from a famous experiment from 1970.
In the experiment, researchers had an actor, whom they called “Dr. Fox,” and described as a legitimate and esteemed “expert,” deliver a lecture that was intentionally engineered to consist of contradictions, meaningless references, non sequiturs and unintelligible mumbo jumbo. Yet despite the nonsensical nature of the talk, subjects gave Dr. Fox high satisfaction ratings.
What’s going on here?
Subjects were told Dr. Fox was a noted expert. He was lively and charismatic. He was even funny! He appeared to have a deep command of the material. He seemed to understand, and acted as if he understood. Yet…the material was very complicated. Very dense. And so, although it was difficult to understand him, the subjects still thought he was competent. They might have been saying to themselves, “well, if I can’t understand it, then it must be really high quality material indeed!” In a sense, it complexity itself that contributed to making Dr. Fox more credible.
Armstrong points out this effect may hold in academia, where the highly regarded journals tend to be more complex. Want to get published? Perhaps you should make your papers more dense and impenetrable to give them the best shot of getting published in the big name journals. If the writing is dense, then you must really know what you’re talking about. It’s harder to understand really smart people. By contrast, if your research is easy to read and accessible, then that might suggest you lack sophistication or your insights are too obvious. So as with Dr. Fox, again it is complexity itself that can contribute to credibility and positive assessments of competence.
Yet Armstrong’s meta study suggests this is the wrong intuition. We are falling victim to the siren song of complexity. We should do the opposite of what our gut tells us!
Instead, we should be suspicious of complexity, since it tends not to add predictive value, but rather to detract from the accuracy of forecasts. If we conclude the basis for a forecast is too complex, we should reject it on that basis alone. How might you go about doing this?
Armstrong suggests using a “simplicity checklist,” (a copy is here) as a way to assess whether an approach really is simple. The checklist focuses on prior knowledge used, relationship of model elements, and whether users can explain the forecasting process. If a model is not simple, and you can’t explain it, then you shouldn’t trust it.
I think this is useful thinking to apply to the asset management industry, where there is always a tug of war between simplicity and complexity. The Dr. Fox study reminds me of some industry dynamics I’ve witnessed.
In asset management, there are plenty of reasons to avoid simplicity and add complexity. After all, if it’s simple, then there’s nothing special about it. A client might say, “if it’s so easy for me to understand, and I can explain it to you, then maybe I know as much as you do, and maybe you’re not adding any value!”
Meanwhile, complexity sells. The “wow” effect is similar to the effect in academic journals. “If I can’t understand it, then it must be really good!”
Complexity is also a great substitute for a lack of substance. If a client wants a justification for something, you can devise a complex, heavily data mined solution that supports the decision, no matter what it is. Armstrong mentions the old saying, “if you can’t convince them, confuse them.”
So don’t be fooled by Dr. Fox! When in doubt, simplify, and avoid harmful complexity at all costs.
If you’d like to learn more about this in the context of finance, here is a post we wrote entitled, “Are you trying too hard?” The essence of the argument is to focus on simple robust processes and avoid complexity.
By: Wesley R. Gray, Ph.D.
Jack will be presenting at The World of ETF Investing Conference on 11/6 in NYC. Jack is an expert on all things finance and a wealth of knowledge on stock selection strategies.
Our CCO and COO Patrick Cleary will also be in attendance.
A link to the event can be found here.
Please let us know if you are around and would like to chat!
By: Jack Vogel, Ph.D.
There is an interesting discussion in the geeky world of academic finance literature between the intellectual muscle at AQR and academia.
The discussion revolves around the following question: “Does Active Share matter?” This is an important topic for active ETFs and Mutual Funds in the marketplace.
The original paper on this measure was written by Cremers and Petajisto and was published in the Review of Financial Studies in 2009 (top finance journal). Links to the paper can be found here and here. The abstract of the paper is the following:
We introduce a new measure of active portfolio management, Active Share, which represents the share of portfolio holdings that differ from the benchmark index holdings. We compute Active Share for domestic equity mutual funds from 1980 to 2003. We relate Active Share to fund characteristics such as size, expenses, and turnover in the cross-section, and we also examine its evolution over time. Active Share predicts fund performance: funds with the highest Active Share significantly outperform their benchmarks, both before and after expenses, and they exhibit strong performance persistence. Nonindex funds with the lowest Active Share underperform their benchmarks.
Main Finding of the paper: For non-index funds, the higher the active share, the better the performance. We tend to agree, as we have talked about diworsification in the past. However, just because a manager creates a more active portfolio (a necessary condition for outperformance), this doesn’t imply an active manager will actually have outperformance. The team at AQR (Frazzini, Friedman, and Pomorski), in a forthcoming article in the Financial Analyst Journal (link to the paper is here), address this question. The abstract is the following:
We investigate Active Share, a measure meant to determine the level of active management in investment portfolios. Using the same sample as Cremers and Petajisto (2009) and Petajisto (2013) we find that Active Share correlates with benchmark returns, but does not predict actual fund returns; within individual benchmarks, it is as likely to correlate positively with performance as it is to correlate negatively. Our findings do not support an emphasis on Active Share as a manager selection tool or an appropriate guideline for institutional portfolios.
Main point of the paper: Active share should not be used as a manager selection tool. Basically, for a given index, they find that active share cannot be used as a reliable tool to identify out-performance.
So is Active Share a waste of time?
As Lee Corso says every Saturday morning during College Gameday, “Not so fast!”
The two authors of the original paper, Martijn Cremers and Antti Petajisto were quick to shoot down the AQR findings.
Here is the executive summary from Antti Petajisto:
All of the key claims of AQR’s paper were already addressed in the two cited Active Share papers: Petajisto (2013) and Cremers and Petajisto (2009).
1) The fact about the level of Active Share varying across benchmarks has been widely known for many years. Its performance impact was explicitly studied and discussed in the first drafts of Petajisto (2013) back in 2010, and the performance results remained broadly similar. The reason for the apparent discrepancy is AQR’s choice of summarizing results by benchmark, which effectively gives the same weight to the most popular index (S&P 500, assigned to 870 funds) and the least popular index (Russell 3000 Growth, assigned to 24 funds), which is not sensible as a statistical approach.
2) The issue about four-factor alphas varying across benchmark indices does nothing to change the fact that higher Active Share managers have been able to beat their benchmark indices. However, it does raise an interesting point about the four-factor approach to measuring performance, and in fact my coauthors and I wrote a long and detailed paper about this exact issue first in 2007 (published later as Cremers, Petajisto, and Zitzewitz (2013)).
3) AQR’s researchers argue that there is no theory behind Active Share and they remain mystified by the differences between Active Share and tracking error. It is unfortunate that they have entirely missed the lengthy sections of both Active Share papers that discuss this exact topic: pages 74-77 in Petajisto (2013) and sections 1.3, 3.1, and 4.1 in Cremers and Petajisto (2009). The short answer is that Active Share is more about stock selection, whereas tracking error is more about exposure to systematic risk factors. So clearly ignoring large and essential parts of the original Active Share papers is simply not the way to conduct impartial scientific inquiry.
If that executive summary wasn’t scathing enough, Martijn Cremers actually wrote a paper titeld “AQR in Wonderland: Down the Rabbit Hole of ‘Deactivating Active Share’ (and Back Out Again?)”
Here is the abstract:
The April 2015 paper “Deactivating Active Share”, released by AQR Capital Management, aims to debunk the claim that Active Share (a measure of active management) predicts investment performance. The claim of the AQR paper is that “neither theory nor data justify the expectation that Active Share might help investors improve their returns,” arguing that previous results are “entirely driven by the strong correlation between Active Share and the benchmark type.”
This paper’s first and main aim is to establish that the AQR paper should not be interpreted using typical academic standards. Instead, our conjecture is that this AQR paper falls into a wonderfully creative but altogether different genre, which we label the Wonderland Genre, as its main characteristic seems to be “Sentence First, Verdict Later.” For example, the results in the AQR-WP cannot be taken at face value, as the information that is not shared reverses their main conclusion.
Secondarily, we consider the plausible claim that benchmark styles matter and find that controlling for the main benchmark style, the predictability of Active Share is robust. While Active Share is only one tool among many to analyze investment funds and needs to be carefully interpreted for each fund individually, Active Share may indeed plausibly help investors improve their returns.
Thirdly and finally, we impolitely consider why AQR may not be a big fan of Active Share by taking a look at the AQR mutual funds offered to retail investors. We find that these tend to have relatively low Active Shares, have shown little outperformance to date (with performance data ending in 2014) and thus seem fairly expensive given the amount of differentiation they offer.
So who is the winner in the debate?
The answer is both are probably correct at some level. More concentration (less diworsification) probably has higher active share and in the past had higher returns. However, one cannot just take any random selection of stocks and expect to outperform (we show this here), the style of the investment matters, which was AQR’s argument (we prefer Value and Momentum).
Let us know what you think!
About Wesley R. Gray, PhD
After serving as a Captain in the United States Marine Corps, Dr. Gray received a PhD in finance. Dr. Gray served as a finance professor at Drexel University before starting the asset management firm Alpha Architect. Dr. Gray has published multiple...