Editor’s Letter

WHAT’S IN THIS ISSUE AND WHY?

The Journal of Retirement aims to publish high-quality research across those disciplines with the concepts and methods to illuminate retirement problems and solutions. The hope is that we can highlight work grounded in legal studies, organizational behavior, and individual behavior, to name just three, in addition to finance and economics. In this issue, contributors cover retirement modeling, pension plan fund selection, effects of centralizing public pension investments, institutional factors in cross-national pension comparisons and principal-agent conflicts among advisors.

New thinking on modeling contributions and allocations. Over the years, as the wide range of articles in this journal on the subject can attest, reducing uncertainty regarding decisions about retirement savings, spending, and asset allocation is both compelling and difficult. Stretching back to Samuelson and Merton, the developing field of retirement modeling has advanced by incorporating more state variables, better understanding of individual utility, advances in methodology, and greater solution speed. A positive case in point are the first two articles in this issue. In “Retirement Planning: From Z to A,” by Javier Estrada, the solution doesn’t depend on a formal description of utility or consumption smoothing and instead helpfully gives us a closed-form solution for contributions needed to fund retirement withdrawals and a bequest depending on projections of retirement spending and asset returns. The article gives us a sense of the uncertainty involved in these decisions through a set of sensitivity analyses. The companion piece, by John M. Mulvey and Han Hao, entitled “Setting Realistic Goals for Personal Retirement Planning via Micro–Macro Analyses,” operates at two levels. At the macro level, Mulvey and Hao model the self-sustainability of the Social Security Program and look at its sensitivity to changes in interest rates, tax rates, and retirement age. At the micro level, They examine both Social Security and a defined-contribution pension from the perspective of an individual. In the latter case, They use machine learning to identify “normal” and “crash” equity regimes to better understand tail risk and then compare five different asset allocation and saving strategies in simulations, including a target-date glide path and four dynamic strategies. The most complex is a “Dynamic Zone Save” strategy that partially immunizes the portfolio when wealth is above an interim wealth boundary, takes more risk when below a lower interim wealth boundary, and requires additional savings after a crash. This article challenges us to consider formal approaches to dynamic asset allocation and retirement planning.

Active versus passive fund selection. Within the whole retirement planning problem, the article by Gerald W. Buetow, Jr., and Bernd Hanke, “How Long Is Enough?,” contributes to the question of active versus passive fund selection by telling us that plan sponsors can stick too long with underperforming active funds. They propose that using passive funds to replace underperforming active funds can improve subsequent portfolio returns and show this by taking all active US equity funds in the Morningstar universe in each of the nine Russell style boxes and comparing them with their Russell index. When an active fund underperforms over any three or five-year period, it is replaced with the respective Russell index. Then, the subsequent returns of the Russell index are compared with the subsequent returns of the original active fund. They show that (1) replacing the active fund is a better strategy than sticking with it and (2) the shorter three-year replacement window leads to better results than the longer five-year window. These findings may generate a good discussion regarding their applicability to portfolio construction practices and policy, including whether or not the appropriate universe should be actual funds used by plan sponsors after their vetting process rather than all active funds and the comparison should be to actual passive funds, including their expenses, rather than indexes.

Centralizing public plan asset management. During a time when many public pensions are mildly to severely underfunded, sponsors, consultants, legislatures, and the public are all seeking practical ways to improve their status. The article entitled, “State vs. Local Management of Pension Assets: Effects of the Massachusetts Chapter 68 Public Pension Reform,” by Bruce E. Stangle, D. Lee Heavner, Yao Lu, Alex Iselin, and Priyanka Singh, looks at a natural experiment when a 2007 Massachusetts law required underfunded local public pensions to cede investment management control to the state pension. The authors, in a carefully controlled study, were able to compare returns before and after assets were transferred from local to state management as well as the returns that transferred assets would have achieved had they not been transferred. They found a large and significant gain for transferred assets overall due to state management but were not able to distinguish between the effects of changes in asset allocation, manager selection, or reduced expenses. Nevertheless, this article points to a successful tool for achieving incremental improvements in public pension funding status that has not been widely considered.

Cross-national system comparisons. Although we seldom control for all variables, cross-national pension comparisons can help us test assumptions, interventions, and outcomes. In “The Top Three Pension Systems: Denmark, Finland, and the Netherlands” Svend E. Hougaard Jensen, Jukka Lassila, Nikku Määttänen, Tarmo Valkonen, and Ed Westerhout take an institutional-focused look at what are arguably some of most well-designed and operated systems in the world. While they differ in some important constitutional and design elements (e.g., degree of centralization, location on the defined-contribution versus defined-benefit spectrum), they all feature strong involvement and commitment by employee and employer partners, pension management at the sectoral rather than firm level, some flexibility in terms of individual choice about taking benefits, and national political support for ongoing reforms. The authors speculate about whether these features explain the relatively high funded ratios enjoyed by these plans. Nevertheless, each of the three systems face challenges, including legitimacy and the decline of unions, that could affect the support for retirement systems in the future.

Aligning advisors and advisees. While principal-agent conflicts involving commissions and fees for recommending certain financial products have been well documented, John A. Turner and Shelley Giordano, in “AUM-Based Compensation and Financial Advice,” focus on conflicts that can arise in AUM-based (assets-under-management) advisor compensation arrangements. These include taking Social Security benefits prematurely and not fully considering annuities and reverse mortgages, all because they can involve cashing out assets that pay the advisor AUM-based fees. The other is rolling over a 401(k) to a higher-fee IRA. The authors consider each of these in turn and note qualitatively their potential impact on net wealth. They also consider the need for public policies to regulate these conflicts.

TOPICS: Retirement, social security, big data/machine learning

Brett Hammond

Editor