Financial Analysts Journal, Second Quarter, 2021, Vol. 77 No. 2

The second quarter 2021 issue starts with a Perspective on securing retirement with “Retirement Income Sufficiency through Personalised Glidepaths.” The authors challenge our professional objective to save and invest “to retirement” rather than “through retirement.” The old problem of amassing wealth at retirement date is conventionally solved with the gradual de-risking of portfolios as the clients age. Instead, these authors demonstrate how we could and should personalize glidepaths to securely provide income during retirement based on individual career paths and longevity risks rather than arbitrary demographics. It’s time for retirement planning to “get personal”—see how in this article. For a good partner to this article, consider “Targeting Retirement Security with a Dynamic Asset Allocation Strategy,” published in the third quarter 2020 issue.

In the first quarter 2021 issue, “Reports of Value's Death May Be Greatly Exaggerated” presented a retrospective view on the value style and the proposed solution involving inclusion of intangible assets in traditional value metrics. In this issue, authors from Bridgeway Capital Management add an industry dimension. In their article, “Equity Investing in the Age of Intangibles,” they classify industries using a measure of “intangible intensity” constructed from intangible assets, innovation capital, and organizational capital. They too examine whether the value relevance of book value and earnings has declined over time but they make the distinction for high- and low-intangible-intensity industries and they find significant divergence. The old value ratios are of declining importance for high-intangible-intensity companies such as pharmaceuticals and software. But no such decline was observed for the low-intangible-intensity group which includes for example utilities and food and beverage. Outside of the US, price-to-book and earnings ratios may even be of increasing relevance in these low-intangible-intensity sectors. The bottom line is: Value isn’t dead and the intangible intensity of the sector matters when it comes to re-evaluating value ratios.

The next article focuses on a particular kind of tail risk, namely litigation risk. Authors from the University of St. Gallen show that firms with better ESG performance are (not surprisingly) less likely to be sued at all. But these ESG performers also experienced substantially less loss in market value and better recovery post lawsuit than their peers. In this article, “Risk Mitigation of Corporate Social Performance in US Class Action Lawsuits,” the authors demonstrate the litigation buffer created by good corporate social performance.The implications of this for investors is clear and has particular relevance for those investing in sectors where litigation risk is high as a rule. For those who are more broadly interested in legal and scandal risks, this article also provides a good roadmap of previous studies in this subject area.

In 2017, Antti Petajisto exposed the “Inefficiencies in the Pricing of Exchange-Traded Funds” in this Journal, for which he earned that year’s Graham and Dodd award. Four years later, an article in this issue shows that algorithmic trading has closed the gap. Our authors from universities in New York and Shanghai claim that high-frequency algorithmic trading in the ETF market is not to be feared but encouraged. In their article, “Active Trading in ETFs: The Role of High-Frequency Algorithmic Trading,” they document the big increase in high frequency trading activities in a rapidly growing ETF market over time. Across a large sample of ETF funds, they measure departures of ETF prices from their net asset value and how long this mispricing persists. They find that, if there is high frequency trading activity in an ETF, the ETF’s price is more likely to be accurate. High frequency trading reduces the persistence of mispricing by maintaining price efficiency through intraday arbitrage and reducing the spread which in turn facilitates arbitrage activities by authorized participants.

The next two articles in this issue concern improved performance measurement: the first for bond funds and the second for hedge funds.

Separating alpha and beta with a linear regression against a benchmark in the usual way to determine management success is flawed in a bond fund environment. We know that maturity or duration and consequent interest rate risk are nonlinearly related to expected return. Authors from Germany propose a simple solution: “Maturity-Matched Bond Fund Performance,” where benchmarks are matched to the declared duration of the funds being assessed. The authors demonstrate the extent to which the conventional measurement of alpha is biased and they show that popular fund ratings as well as investor flows follow this biased measurement. Their solution requires funds to report their duration—which not all do—but it is simple in the sense that the non-linearity doesn’t have to be modeled. They take it to the next level by demonstrating another patch to the problem using betas and r² for funds that don’t declare their duration. This is a must-read for bond fund selectors and a good companion read to an article published in the second quarter last year: “Looking under the Hood of Active Credit Managers.”

The problem of separating skill (or alpha) from hedge fund performance is even trickier given how opaque hedge funds are with regard to their strategies and leverage. In their article, “Identifying Hedge Fund Skill by Using Peer Cohorts,” authors from Australia and the UK demonstrate the benefit of arranging hedge funds into cohorts as a first step using cluster analysis. The usual factor models can be used to separate individual hedge fund alphas from their cohort benchmarks. The authors show that this method is a substantial improvement on the established seven-factor model: It is better at isolating the unique component in fund performance not associated with the strategy and related factor exposures, and the resulting alpha is more persistent—that is, the performance is better for longer—out of sample.

Finally, authors from AQR offer a “Grand Unified Theory” of portfolio construction in their article, “Enhanced Portfolio Optimization,” or EPO. The problem with mean–variance optimization in practice is well known and this article presents the solution. It’s an elegant and seemingly simple approach of shrinking correlations. In the first section, followers of portfolio theory will find the context and connections of EPO to other well-known enhancements to the portfolio construction problem as well as detailed proof of the approach. The second section provides the “how-to” for portfolio constructors who want to try this out and the final section is full of the evidence of the success of this solution in various portfolio construction settings.