Planning for retirement? "Easy as A-B-C" may not be what you want
Submitted by Dr. Sabuhi Sardarli
Associate Professor of Finance | Director - Center for Financial Analysis | Phillips 66 Faculty Fellow Department of Finance
The burden of retirement planning has shifted to individuals as employers have transitioned away from defined benefit plans such as pension plans to defined contribution (DC) plans such as 401(k) and 403(b). As of end of year 2018, $7.5 trillion in assets are housed within these DC plans. The ultimate success of the retirement outcome depends on the employer in the plan-design process and on the individuals in the investment-choice process. Today, 56% of American households have an employer-sponsored retirement plan where they must make investment choices within the plan.
A new research publication jointly authored by Dr. Sabuhi Sardarli, Associate Professor of Finance at Kansas State University, and recently featured on the Wall Street Journal and CNBC’s Squawk Box, suggests that individuals’ choices are subject to a behavioral bias called alphabeticity. Alphabeticity occurs when options which come earlier alphabetically are chosen more frequently than others. This bias shows up in 401(k) plans because employers usually list investment options alphabetically within plan information handouts and websites. The outcome is that a fund within the plan usually receives higher allocation if it is named with an early alphabet letter. More specifically, the study finds that each of the top four funds on an alphabetical list receive 10% more money than each would receive if money was equally allocated to all funds within the plan.
Alphabeticity is known to affect many decision processes. In the days of phone books, it was the reason why one would be more likely to find “AAA Plumber” rather than “XYZ Plumber”. The bias causes the decision-maker to stop the search process when the first acceptable choice is presented, even if continuing the search might result in a better outcome. Historically, this bias been found to be present in large choice sets where the cost of a full search may be too high. Dr. Sardarli’s research confirms the existence of this bias within DC retirement plans, where investors pick from a relatively smaller set, typically consisting of 20-22 mutual funds.
The findings suggest that there is an opportunity to improve retirement plan designs as the current alphabetical presentation results in unintended consequences. Retirement structures can be and have been altered in the past to account for individual biases, allowing for optimal retirement outcomes. For example, one of the most important recent changes to DC retirements was made about a decade ago, as both the industry and regulators realized that the participation rate in 401(k) plans was low because employers required employees to opt-in to a plan manually, a strategy which did not elicit high enrollment. The Pension Protection Act of 2006 and efforts by the Department of Labor changed retirement plans to automatic enrollment, which would instead give employees the option to opt-out.
It is time to consider alternative presentations of investment choices within retirement plans. By sorting funds by characteristics, alphabeticity can be exploited to create more desirable outcomes. Plan architecture matters and plan administrators can use it to assist investors in their choice process. Alternative ordering options to consider are fund expense ratios, historical performance, Morningstar ratings, fund turnover ratios, etc. Finally, it is vital for individuals to be aware of this bias and put in the extra effort to conduct a full search before making important choices.
The paper was published here: ”Alphabeticity Bias in 401(k) Investing” (with T. Doellman, J. Itzkowitz, J. Itzkowitz), 2019, Financial Review, Vol 54(4), 643-677.
It was also chosen as best paper by the editorial board in 2019: https://financialreview.poole.ncsu.edu/awards/