When Asymmetry Changes Everything: Rethinking Factors in a Frictional World
Much of modern asset pricing theory is grounded in factor-based explanations of expected returns. Since the landmark Fama and French (1993) paper, which introduced value and size as explanatory risk premia, hundreds of factor strategies have emerged in both academic research and practical investing. These factors are typically constructed using symmetric, long–short portfolios—buying the ‘good’ side (for example, cheap or profitable stocks) and shorting the ‘bad’ side (such as expensive or low-profitability stocks). This construction is elegant, tidy, and statistically convenient. However, it rests on a crucial and often unspoken assumption: that symmetry is both achievable and meaningful in practice.
In reality, most investors operate under constraints. Shorting is expensive, operationally complex, and in many cases outright prohibited. Capital must be allocated, trading incurs costs, and leverage is often limited or restricted. These frictions are largely absent from the academic literature, which tends to abstract away from the practical realities of portfolio implementation. But what happens if we relax the assumption of symmetry and impose more realistic conditions? That is precisely the question addressed by van Vliet, Blitz, Baltussen and Swinkels (2025) in their latest research paper Factoring in the Low-Volatility Factor. The results are not only revealing but transformative—especially in reshaping how we understand the role and robustness of the low-volatility factor.
The Problem with Symmetry in Factor Investing
The dominant approach in academic finance is to evaluate factors as symmetric long–short portfolios. This methodology, used by Fama and French (1993), Carhart (1997), and many subsequent studies, creates a tidy decomposition of returns attributable to a particular characteristic. By going long the ‘top decile’ and short the ‘bottom decile’, researchers construct zero-cost portfolios that supposedly isolate the pure effect of a factor. However, this construction assumes several conditions that do not reflect the realities of portfolio management.
First, shorting is not costless. Securities lending fees, collateral requirements, and risk controls all reduce the viability of the short side. Second, portfolios in the real world are long-only. Whether due to regulation (e.g., UCITS constraints), institutional mandates, or investor preferences, most capital is deployed on the long side. Third, transaction costs and market impact are material. High-turnover strategies that perform well in academic backtests may fail when frictions are considered. Lastly, capital allocation is not free. Unlike academic constructs, portfolios must be funded, and risk needs to be controlled in a real-world setting.
Van Vliet et al. (2025) argue that we must evaluate factors not in isolation or abstraction, but under conditions that mimic how investors actually operate. They introduce a framework for analysing 14 well-known equity factors using long-only portfolios, capital allocation constraints, and real-world trading costs. The goal is to identify which factors retain their power once symmetry and costless implementation are stripped away.
The Low-Volatility Anomaly: CAPM Expectations vs Empirical Reality
One of the most well-established empirical anomalies in finance is the low-volatility effect. According to the Capital Asset Pricing Model (CAPM), the expected return on an asset should increase linearly with its beta—its sensitivity to the overall market. The formal representation is:
E(Rᵢ) = Rf + βᵢ × (E(Rₘ) − Rf)
Where E(Rᵢ) is the expected return on asset i
Rf is the risk-free rate
βᵢ is the asset’s beta
E(Rₘ) is the expected return on the market
In theory, higher-beta stocks should command higher expected returns, compensating investors for bearing greater systematic risk.
Yet, in practice, this relationship is consistently violated. Numerous studies show that stocks with lower beta or volatility tend to offer higher risk-adjusted returns than their high-beta counterparts. This empirical contradiction—known as the low-volatility anomaly—has been observed across decades, countries, and asset classes, including equities, bonds and real estate investment trusts (REITs). One of the most influential contributions in this space is the study by Frazzini and Pedersen (2014), who documented that portfolios composed of low-beta stocks exhibit Sharpe ratios that exceed those of high-beta portfolios. Importantly, this pattern persists across time and geography, suggesting that it is not an artefact of sampling error or data mining.
Betting Against Beta (BaB): A Theoretical Construct with Practical Barriers
In an effort to explain the low-volatility anomaly, Frazzini and Pedersen (2014) introduced a long–short factor strategy called ‘Betting Against Beta’ (BaB). The core idea is simple but powerful: investors constrained from using leverage—institutional investors, mutual funds, and retail clients—attempt to achieve higher returns by overweighting high-beta assets. This behaviour drives up the prices of high-beta stocks, lowering their expected returns. Conversely, low-beta stocks are under-owned and undervalued, creating an opportunity for excess returns. The BaB strategy seeks to exploit this mispricing by going long low-beta assets and short high-beta ones.
Theoretically, BaB offers a clean and elegant solution to the anomaly. However, in practice, it is difficult to implement. Shorting high-beta stocks introduces costs and risk. Leveraging low-beta positions requires borrowing, which comes with margin requirements and operational hurdles. Furthermore, beta and volatility are related but not equivalent. Beta reflects relative movement with the market, whilst volatility measures absolute risk. It’s possible for stocks with similar volatilities to have very different betas, depending on their correlation with the market — and vice versa. As such, BaB remains mostly a theoretical construct rather than a widely adopted strategy.
Instead, what has gained traction in institutional portfolios is low-volatility investing via long-only, rules-based strategies. These are practical, implementable, and scalable. Index providers such as MSCI and S&P have developed low-volatility indices, and major ETFs—such as the iShares Minimum Volatility and Invesco S&P 500 Low Volatility funds—track these indices. These strategies avoid shorting and leverage entirely, making them far more accessible to everyday investors.
The Case for and Against Low-Risk Investing
Robert Novy-Marx has offered a rigorous critique of betting-against-beta (BaB) strategies, arguing that their apparent success is largely an artefact of flawed market assumptions. In his 2014 paper ‘Understanding Defensive Equity’, he contends that BaB does not work because of beta itself, but because it unintentionally avoids overpriced, low-alpha stocks that dominate the market portfolio. By shorting the market, BaB effectively shorts these high-beta, low-alpha glamour (growth) stocks, meaning that the strategy's excess returns stem not from beta exposure but from a hidden alpha effect. He further argues that low-volatility long-only strategies—often used as long-only proxies for BaB—are messy composites with undesirable exposures, such as negative tilts to the profitability and value factors, and high sensitivity to interest rate risk. These portfolios, frequently dominated by large, expensive utilities or consumer staples, appear stable but are fundamentally weak, and their concentration and turnover can make them costly and fragile in practice. Novy-Marx instead proposes profitability as a cleaner, more intuitive alternative: his 2013 research on gross profitability shows that highly profitable firms consistently outperform, and tend to have low beta anyway—so BaB, in his view, is an inefficient proxy for the profitability premium.
However, van Vliet challenges the practicality of Novy-Marx’s conclusions, especially within long-only portfolio constraints. He stresses that low-volatility strategies, though sometimes associated with less profitable stocks, still deliver superior risk-adjusted returns when implemented thoughtfully, especially in combination with value or quality tilts. Van Vliet also critiques Novy-Marx's desire for factor purity, noting that low-risk strategies capture a blend of return drivers—including value, momentum, and interest rate exposure—that evolve over time and diversify portfolio risk. Importantly, he cautions against abandoning low-volatility investing in pursuit of a single explanatory factor, arguing instead for a pragmatic, multi-factor approach that reflects the real-world constraints and needs of institutional investors. Van Vliet contends that volatility and profitability measure fundamentally different dimensions of risk, and therefore serve different purposes in portfolio construction. His core point is that volatility is about price risk and profitability is about business risk. Thus, within long-only portfolios, low-volatility is not a flawed substitute for profitability, but a robust diversifying factor premium in its own right.
What Survives in a Frictional, Long-Only World?
Van Vliet’s results are telling. When factors are tested using realistic, long-only implementations, many of the popular anomalies lose their potency. Momentum, for instance, relies heavily on the short side and suffers from high turnover and transaction costs. Value fares somewhat better, but its performance diminishes once frictions are imposed. Profitability remains moderately effective, but not dominant.
Only low-volatility consistently performs well across all constraint types. It delivers strong Sharpe ratios in long-only, capital-allocated, net-of-cost implementations. It is not only statistically robust but also economically implementable. This aligns with earlier findings by Blitz and van Vliet (2007), who noted that low-volatility strategies outperformed not in spite of their defensive characteristics, but because of structural and behavioural inefficiencies in the market.
Consider:
– Institutional constraints on leverage
– Investor preference for lottery-like stocks
– Benchmark-relative incentives
– Behavioural biases favouring glamour and growth
In this light, low-volatility’s premium starts to look rational (Baker, Bradley, and Wurgler 2011). These frictions and preferences conspire to drive capital towards high-volatility, high-beta stocks—many of which are overpriced relative to fundamentals. This creates a durable mispricing that long-only low-volatility strategies can exploit without the need for shorts, leverage, or complexity.
Empirical Evidence from Van Vliet et al. (2025)
One of the most compelling illustrations of this idea comes from page 48 of Factoring in the Low-Volatility Factor. The chart displays Sharpe ratios for several equity factors implemented in long-only portfolios, adjusted for transaction costs and capital allocation limitations and opportunity costs. This setup reflects the most realistic investment conditions faced by institutional managers and individual investors alike.
The chart shows clearly that low-volatility has one of the highest risk-adjusted returns in this frictional environment. Many traditional factors—especially those with high turnover or that rely heavily on the short leg—see their Sharpe ratios shrink significantly or even turn negative. In contrast, low-volatility remains strong. This is not a statistical illusion or a backtest quirk. It is robust, replicable, and consistent with what we see in real portfolios and live ETF performance.
This evidence reinforces the view that any serious attempt to explain the cross-section of returns—especially one grounded in long-only implementation—must give central weight to the low-volatility factor.
Conclusion: Elevating Low-Volatility from Anomaly to Anchor
The ongoing debate between theorists like Novy-Marx and pragmatists like van Vliet underscores an important truth in empirical asset pricing: even the most elegant models struggle to fully explain how returns emerge in the real world. Whilst Novy-Marx raises valid concerns about the compositional flaws and alpha attribution of low-volatility strategies, van Vliet reminds us that practical investing must grapple with constraints, frictions, and investor behaviour. Both views offer insights—but they also reflect different priorities: theoretical purity versus empirical implementability.
What emerges from van Vliet et al. (2025) is not just a defence of the low-volatility anomaly, but a broader reframing of how we evaluate factors in a world of limitations. Their research reveals that when you abandon the symmetry assumption and impose realistic constraints—no shorting, limited leverage, transaction costs—many celebrated factor strategies falter. Low-volatility, in contrast, continues to shine not because it is perfect, but because it is resilient, scalable, and aligned with real-world incentives and frictions.
Of course, it is hard to know who is definitively right. Asset pricing is an evolving field, full of competing frameworks and persuasive narratives. Each new paper brings fresh evidence, but we must be cautious not to overreact to the latest findings simply because they are new. Recency bias—our tendency to overweigh the most recent information—can distort judgment just as much as any model misspecification. What matters most is consistency, robustness, and applicability over time.
In that light, the low-volatility factor may not be the most exciting anomaly, nor the most theoretically pure. But it endures. And in a noisy, frictional world, that may be the strongest signal of all.
References
Baker, Malcolm, Brendan Bradley, and Jeffrey Wurgler. 2011. ‘Benchmarks as limits to arbitrage: Understanding the low-volatility anomaly’. Financial Analysts Journal 67 (1): 40–54. https://doi.org/10.2469/faj.v67.n1.4
Blitz, David, and Pim van Vliet. 2007. ‘The volatility effect: Lower risk without lower return’. Journal of Portfolio Management 34 (1): 102–113. https://doi.org/10.3905/jpm.2007.684759
Carhart, Mark M. 1997. ‘On Persistence in Mutual Fund Performance’. Journal of Finance 52 (1): 57–82. https://doi.org/10.1111/j.1540-6261.1997.tb03808.x
Fama, Eugene F., and Kenneth R. French. 1993. ‘Common risk factors in the returns on stocks and bonds’. Journal of Financial Economics 33 (1): 3–56. https://doi.org/10.1016/0304-405X(93)90023-5
Frazzini, Andrea, and Lasse H. Pedersen. 2014. ‘Betting against beta’. Journal of Financial Economics 111 (1): 1–25. https://doi.org/10.1016/j.jfineco.2013.10.005
Novy-Marx, Robert. 2014. ‘Understanding Defensive Equity’. NBER Working Paper No. 20591. Cambridge, MA: National Bureau of Economic Research. https://doi.org/10.3386/w20591
Pim van Vliet, David Blitz, Guido Baltussen, and Laurens Swinkels. 2025. ‘Factoring in the low-volatility factor’. SSRN Working Paper. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5295002