Applied Quantitative Research (AQR) was founded in 1998 by Clifford Asness, David Kabiller, Robert Krial, and John Liew. Similar to Dimensional Fund Advisors, AQR's roots stem back to the University of Chicago where Asness, Liew, and Krial originally met as doctoral students. In fact, Cliff Asness was an assistant to Nobel Laureate Eugene Fama who is widely considered the “Father of Modern Finance.” All three would go on to work for Goldman Sachs on the newly-established quantitative research team.
In 1998, they left Goldman Sachs to start their own hedge fund whose strategy was based on the ideas they learned while at the University of Chicago, although taking a slightly different approach. In 2009, they decided to bring their alternative strategies into a mutual fund format. Today, AQR manages close to $175 billion across almost 40 different strategies.
What is unique about AQR, which is similar to Dimensional, is their close relationship with the academic community when it comes to formulating and implementing their investment ideas. Their strategies tilt towards the known dimensions of expected returns such as size and relative price (value). They also implement strategies based on a “quality” factor (QMJ) that is similar to the combination of profitability, investment, and relative-price factors that firms like Dimensional also screen for. Another major difference is AQR’s belief that Momentum (Jegadeesh & Titman, 1993) can be reliably captured in a separate strategy. When combined with a typical value strategy, momentum may provide a significant diversification benefit. It is important to note that Dimensional does screen for momentum in their strategies, but does not attempt to capture it in a separate strategy given the high amount of turnover (trading) that occurs in order to capture it.
What is important for investors to understand is that deciding to invest with AQR or Dimensional is really a difference in implementation on a similar philosophy. Both believe that a quantitative approach to investing with tilts towards known dimensions of expected return can yield better results for investors. Where they differ is how they interpret why these factors exist in the first place. While Dimensional believes markets are very efficient and these factors are associated with risk, AQR relaxes that assumption, believing there are arbitrage opportunities that can be reliably captured across many markets.
Today, we are going to examine the performance of AQR’s strategies to see how well they do in terms of their Morningstar assigned benchmarks as well as their Fama/French 3 Factor adjusted “alpha.”
It is also important to note that most AQR strategies have been around for less than a decade so any type of performance analysis will be extremely sensitive to inputs given the small sample size. We are not saying that our results are meaningless, but from a statistical point of view, results could vary substantially with the addition of just a couple more years' worth of data.
We have taken a deeper look at the performance of several other mutual fund companies and have come to one universal conclusion: They have failed to deliver on the value proposition they profess, which is to reliably outperform a risk comparable benchmark. You can review by clicking any of the links below:
Fees & Expenses
Our analysis begins with an examination of the costs associated with the strategies. It should go without saying that if investors are paying a premium for investment “skill,” then they should be receiving above average results consistently over time. The alternative would be to simply accept a market's return, less a significantly lower fee, via an index fund.
The costs we examine include expense ratios, front end (A), level (B) and deferred (C) loads, and 12b-1 fees. These are considered the “hard” costs that investors incur. Prospectuses, however, do not reflect the trading costs associated with mutual funds. Commissions and market impact costs are real costs associated with implementing a particular investment strategy and can vary depending on the frequency and size of the trades taken by portfolio managers. We can estimate the amount of costs associated with an investment strategy by looking at its annual turnover ratio. For example, a turnover ratio of 100% means the portfolio manager turns over the entire portfolio in 1 year. This is considered an active approach and investors holding these funds in taxable accounts will likely incur a higher exposure to tax liabilities to short term and long term capital gains distributions relative to those incurred by passively managed funds.
The table below details the hard costs as well as the turnover ratio for all 24 strategies offered by AQR that have at least 3 years of complete performance history. You can search this page for a symbol or name by using Control F in Windows or Command F on a Mac. Then, click the link to see the Alpha Chart. Also remember that this is what is considered an in-sample test. The next level of analysis is to do an out-of-sample test (for more information see here).
On average, an investor who utilized a strategy from AQR experienced a 0.93% expense ratio. This can have a substantial impact on an investor’s overall accumulated wealth if it is not backed by superior performance. The average turnover ratio for equity strategies offered by AQR was 100%. This implies an average holding period of about 12 months, on average. It is safe to say that AQR trades quite often, which is what we would expect in both alternative strategies and strategies attempting to capture momentum. Again, this is a cost that is not itemized to the investor, but is definitely embedded in the overall performance. In contrast, most index funds have very long holding periods -- decades, in fact, thus deafening themselves to the random noise that accompanies short-term market movements, and focusing instead on the long term.
The next question we address is whether investors can expect superior performance in exchange for the higher costs associated with AQR’s “skill.” We compare each of the 3 strategies that have at least 3 years of performance history since inception and against its current Morningstar assigned benchmark to see just how well each has delivered on their perceived value proposition. We have included alpha charts for each strategy at the bottom of this article. Here is what we found:
- 67% (16 funds) have underperformed their respective benchmarks since inception, having delivered a NEGATIVE alpha
- 33% (8 funds) have outperformed their respective benchmarks since inception, having delivered a POSTIVE alpha
- 0% (0 funds) have outperformed their respective benchmarks consistently enough since inception to provide 95% confidence that such outperformance will persist as opposed to being based on random outcomes
Based on the historical performance of their strategies, it seems AQR has not provided superior returns for their investors. The vast majority of their funds have failed to outperform their Morningstar assigned benchmark. The inclusion of statistical significance is key to this exercise as it indicates which outcome is the most likely vs. random-chance outcomes.
How we define or choose or benchmark is extremely important, especially for funds where we know they are knowingly targeting dimensions of expected return. If we relied solely on commercial indices assigned by Morningstar, then we may lead to the false conclusion that AQR isn’t providing value for their investors. For example, both AQR and Dimensional target the size and value premiums. If large cap stocks and growth stocks have done well, then of course AQR and Dimensional will underperform. Because Morningstar is limited in terms of trying to fit the best commercial benchmark with each fund in existence, there is of course going to be some error in terms of matching up proper characteristics such as average market capitalization or average price-to-earnings ratio.
A better way of controlling for these possible discrepancies is to run multiple regressions where we account for the known dimensions (Betas) of expected return in the US (market, size, relative price, etc.). For example, if we were to look at all of the US based strategies from AQR, we could run multiple regressions to see what their alpha looks like once we control for the multiple Betas that we know are being systematically priced into the overall market. The chart below displays the average alpha and standard deviation of that alpha for the time period 8/1/2009 - 12/31/2016 (longest time period available). There were only 2 that fit these specifications.
As you can see, both AQR strategies had a statistically insignificant alpha once we controlled for risk exposure (Beta).
AQR is a very impressive firm whose roots stem back to the University of Chicago. Similar to Dimensional, they use a quantitative approach to finding known dimensions of expected return and build strategies around these factors in order to provide investors with higher expected returns. They are different from Dimensional in terms of providing alternative based strategies, their belief that there are arbitrage opportunities that can by systematically exploited, and their approach at attempting to capture the known dimensions of expected return (concentrated versus diversified). In terms of their attributes, AQR strategies are quite active and in alignment with most of the costs associated with the active management community. In terms of performance, they have not reliably outperformed their Morningstar assigned benchmarks nor their Fama/French 3 Factor adjusted benchmarks. Given the limited amount of years that AQR has been around, any statistical analysis should be taken with a grain of salt.
There will continue to be an academic debate in terms of why certain premiums exist (risk or behavioral) and firms like AQR and Dimensional will continue to implement ideas they believe will be profitable for investors. Because most of AQR’s and Dimensional’s ideas overlap in terms of strategy, it would be prudent to partner with the firm that provides broader diversification and lower overall costs. We will continue to monitor any developments made by both firms to help ensure that our IFA Index Portfolios represent the latest developments in academic research and innovative portfolio management. For now, we still recommend that investors buy, hold, and rebalance a globally diversified portfolio of index funds offered by our preferred fund partner, Dimensional Fund Advisors.
Here is a calculator to determine the t-stat. Don't trust an alpha or average return without one.
The Figure below shows the formula to calculate the number of years needed for a t-stat of 2. We first determine the excess return over a benchmark (the alpha) then determine the regularity of the excess returns by calculating the standard deviation of those returns. Based on these two numbers, we can then calculate how many years we need (sample size) to support the manager's claim of skill.
About the Authors
Tom Allen is an Accredited Investment Fiduciary (AIF®), Certified Cash Balance Consultant (CBC) and a Chartered Financial Analyst (CFA®) Level III Candidate. Tom received his Bachelor of Science in Management Science as well as his Bachelor of Art in Philosophy from the University of California, San Diego.
Mark Hebner - Founder, Index Fund Advisors, Inc.
Founder and President of Index Fund Advisors, Inc., and author of Index Funds: The 12-Step Recovery Program for Active Investors. He is a Wealth Advisor, with an MBA from the University of California at Irvine and a BS in Pharmacy from the University of New Mexico with a specialization in Nuclear Pharmacy.