We are pleased to present you with this collection of 13 articles on low-volatility investing. The articles included here share two things in common: they all dig into the low-volatility anomaly and they are all written by Robeco researchers. While some of these papers have been previously shared with our clients, this volume also includes our research published in the academic press, including the Journal of Portfolio Management and the Journal of Index Investing. And we’ve added some recent research on topics as diverse as applying the low-risk anomaly to emerging markets, smart-beta investing and how to exploit the low-risk anomaly in credit markets.
From the arc of these articles, it is clear to see that low-volatility investing has come a long way since the anomaly was first documented. And, from the vantage point of 2012, it appears that we have witnessed a quiet evolution.
We started our work on the subject of low volatility in 2005. This was also the year Pim joined the firm, after completing his doctoral thesis on downside risk and asset pricing in 2004. When we began, the low-risk anomaly was relatively unknown. This is curious, because the return potential of low-volatility stocks had been recognized in academic literature as early as the 1970s. No less than the renowned academics Fischer Black and Myron Scholes had acknowledged the unexpected high returns from low-risk stocks and how this finding conflicted with the capital asset pricing model and long-standing investment theory.
In one of our first joint research papers, we documented, for the first time, the existence of the low-volatility anomaly in European and Japanese markets. We also found that the anomaly increased over time and was also strongly present within the large and liquid segment of the stock market. Moreover, we connected the volatility anomaly to the rise of benchmark-driven investors. This research, published in the Journal of Portfolio Management in 2007, is included as the first article in this collection.
Later research, also included in Part 1, published under the title “Is the relation between volatility and expected stock returns positive, flat or negative?” provides a survey of the literature on the empirical relation between risk and return. We conclude that there is an overwhelming amount of evidence that the risk-return relationship is flat, or even negative. We also highlight the importance of compounding effects, which are typically ignored in academic research. If the price of a high-risk stock drops by 50% it needs a return of 100% to recoup the loss, while if the price of a low-risk stock drops by 10% it only needs to gain around 11% to break-even again. We conclude that the volatility effect is underestimated in most studies, because the effects of compounding are ignored.
The third article in this section provides background on Emerging Conservative Equities, a strategy that Robeco introduced in February 2011. To the best of our knowledge, our paper “The volatility effect in emerging markets” is the first to show that the low-volatility anomaly is not only present in developed equity markets, but also in emerging equity markets. In fact, the relation between risk and return in emerging markets is strikingly similar to that observed previously in developed markets. We also find that the empirical deviation from the theoretical risk-return relationship appears to be growing stronger over time, which might be related to the increasing participation of benchmark-driven investors, in line with the “limits to arbitrage” hypothesis. Finally, we observe low correlations between the volatility effects in emerging and developed equity markets, which argues against a common-factor explanation.
Part 1 concludes with an article which contends that the low-volatility anomaly implies that widely-used theoretical asset-pricing models, such as the capital asset pricing model (CAPM) and the Fama-French three-factor asset pricing model are flawed, because they still assume a positively linear relation between risk and return. The paper proposes an alternative asset-pricing model in which this relation is flat instead, and shows that this alternative model is much better able to explain the historical returns of a wide range of test portfolios. The model is motivated by the hierarchical way in which professional asset management is organized, the focus being first on the asset allocation decision, and next on finding fund managers who have been able to outperform their benchmarks. In our view, benchmark-driven investing lies at the root of the low-volatility anomaly. For more details, see “Agency-based asset pricing and the beta anomaly.”
Early on in the development of Robeco’s low-volatility strategies, we realized that bringing the low-volatility anomaly to the market required a new vehicle. We saw little means to incorporate low-risk investing into any of the existing portfolios Robeco was running. It required too big a shift from an information-ratio perspective to a Sharpe ratio perspective.
Almost without exception—and this was true for the entire asset management industry—tracking error was the standard measure of risk; and return was measured based on comparison with a benchmark. Subscribing to a low-volatility strategy, however, means a shift in focus from relative risk to absolute risk and from relative return to absolute return.
We introduced the low-volatility Global Conservative Equity strategy as an institutional fund in 2006. It would be nice to say it took off “like a rocket.” But that was not the case. Uptake was slow and Conservative Equity suffered from being unconventional. We were privileged, however, to find a few innovative institutional investors who trusted our research and shared our belief in the philosophy behind low-volatility investing. They understood the advantage of an equity strategy that could deliver equity market returns with a significant reduction in downside risk.
It was not until after the dramatic markets of 2008 and 2009, however, that potential clients realized what they may have missed. We finally began to hear, “I should have invested earlier,” at client meetings.
Global Conservative Equities celebrated its fifth anniversary last year. At that point (and subsequently), it has achieved positive excess returns and lower volatility than either the MSCI World or MSCI Minimum Volatility Indexes. Robeco now manages more than EUR 2 billion in low-volatility equities.
Over the years, we have published a stream of research on low-volatility topics. Many of them were written in answer to client questions. In Part Two, for example, we address issues related to the long-term performance of low-volatility portfolios and strategic allocation to premiums in the equity market.
In “Low volatility investing: a long-term perspective” eight decades of US market returns were examined to compare the performance of low-volatility versus high-volatility portfolios. One of the main takeaways of this research is that investors in low-volatility stocks would have never experienced a “lost decade,” as opposed to investors in high-risk stocks, who would have had negative returns in the 1930s, 1970s and the 2000s. The main drawback of low-volatility stocks is that they tend to lag in strong thematic bull markets, although this is mainly a relative-return issue. Furthermore, examination of decades such as the 1940s, 1950s and 1960s shows that additional valuation and sentiment factors are very helpful in enhancing generic low-volatility strategies. This article, as well as “Low-risk stocks highly suitable for long-term investors,” were written as we saw institutional investors interested in adding a low-volatility strategy to their portfolios, but still grappling with some aspect of it, such as high-tracking error or the focus on absolute, not relative, risk and return. These papers also acknowledge a growing dissatisfaction among some investors with market-capitalization-weighted indexes.
We addressed this issue head-on in “Strategic allocation to premiums in the equity market” and “Smart-beta investing”. In the latter, we examined and compared the long-term returns, portfolio characteristics and qualitative factors of a number of different alternative betas. Our analysis included strategies based on minimum volatility, risk and value; and we compared these with a market-capitalization-weighted index and Robeco’s Conservative Equity strategy as well.
Our conclusion was that investors need to strategically consider their exposures to specific equity factor premiums. And, as we recommend in the conclusion to our article “Strategic allocation to premiums in the equity market,” a variety of equity premiums should be considered. Even using conservative assumptions, we believe the opportunity available from well-documented equity factor premiums, such as value, momentum and low volatility, will continue to be sizable in the future.
In “Combining low-volatility and fundamental investing,” we zoom in on the relation between low-volatility investing and fundamental indexation. We find that fundamental investing is essentially a value strategy, and while it shares a common goal with low-volatility investing to improve on the capitalization-weighted market portfolio, the two strategies are very different with dissimilar performance characteristics. With low-volatility, the first step toward improving the Sharpe ratio comes from reducing volatility, while fundamental investing is all about enhancing return. We recommend exposure to both effects.
Some more practical aspects of low-volatility investing are included in Part Three. In particular, we present “How to benchmark low-volatility strategies.” It is important to remember that the MSCI Minimum Volatility Indices were not introduced until 2008, two years after the launch of Conservative Equity. What the article also highlights, however, is how many “passive” low-volatility strategies are actually active strategies in disguise, requiring numerous subjective decisions.
In Part Three, you will also find an article based on our more than five-years of experience running Conservative Equity. “How distress risk can improve a low-volatility strategy,” presents our research that distress risk is unrewarded. Including a distress-risk model in a low-volatility equity strategy can avoid a number of common pitfalls, such as underperformance in bear markets, high turnover in illiquid stocks and concentrated industry exposures. Part Three concludes with a concise and useful summary, “Ten things you should know about low-volatility investing.”
The distress-risk model we developed for Conservative Equities proved to be a bridge to a new strategy. In March 2012, we introduced Conservative Credits. Basically, we found that just as in equities, the low-risk anomaly exists in bond markets. And, just as in equities, it is driven by benchmark-driven investing, where portfolio managers are evaluated based on returns relative to an index. This causes them to overpay for higher risk assets and to neglect lower-risk assets. The distress-risk model is a key factor in selecting low-risk credits in the Conservative Credits strategy. You can read more about it in “The low-risk anomaly in credits,” which comprises Part Four.
Less than ten years ago, we began with identifying and testing an investment anomaly: low-risk stocks have better risk-adjusted returns than high-risk stocks. Empirical evidence documented its existence and, after a few years, the focus shifted from whether it was real to the best way to implement it in portfolios. Gradually, other alternatives to the traditional capitalization-weighted index were developed using systematic factor strategies to capture beta in a “smarter” way. Today, we find a growing number of institutional investors believing that the long-term game is not to beat a benchmark, but to harvest factor returns. The discussion is increasingly centered on the strategic selection of specific active strategies to include in an equity portfolio.
We are pleased to have participated in this important debate and trust that you will find the results of our research insightful. To keep up to date with our publications, go to the Low-volatility pages.
Clients are invited to request a copy of this limited edition from their Robeco contact or by sending an email to firstname.lastname@example.org
BY CLICKING ON “I AGREE”, I DECLARE I AM A WHOLESALE CLIENT AS DEFINED IN THE CORPORATIONS ACT 2001.
What is a Wholesale Client?
A person or entity is a “wholesale client” if they satisfy the requirements of section 761G of the Corporations Act.
This commonly includes a person or entity: