Conference
2021 Academic Research Colloquium
The Center for Financial Planning will host the fifth annual Academic Research Colloquium for Financial Planning and Related Disciplines virtually on November 11-12 and 15-16, 2021.
November 11 - 16, 2021
2021 ACADEMIC RESEARCH COLLOQUIUM FOR FINANCIAL PLANNING
AND RELATED DISCIPLINES
ACCEPTED PAPERS
Consumer Budget Management in the Age of Information Access
Anastasiya Ghosh
University of Arizona
Tucson, AZ, USA
Budgeting is often suggested as a solution for consumer financial struggles. We study a context (budgeting apps) where monitoring budget progress against the budgeted amount can backfire. Using data from several lab experiments, a field study, and a budget tracking application, we demonstrate that setting a budget and having access to spending information can increase consumer spending within the budgeted amount. We argue this happens because consumers use budget as a reference point for spending and access to spending information enhances consumers' certainty in budget standing relative to the reference point. We further show that this effect is dynamic, spending information affecting spending at the end (but not the beginning) of the budget period. We demonstrate two interventions (providing less precise spending information and rolling over available money in the budget to the next period) that attenuate the increase in spending. This research provides insights for consumers on managing their finances better and for financial institutions on how to develop better budgeting tools for consumers.
Predictable Price Pressure
Samuel Hartzmark
Booth School of Business
University of Chicago
Chicago, IL, USA
We present evidence that stock returns, at the market and individual stock level, can be predicted by the timing of uninformed investor cashflows that are known in advance. Standard asset pricing models and the efficient market hypothesis suggest such flows should not significantly influence prices. Aggregate dividend payments (whose timing and amount are announced weeks prior to payment) predict higher daily value-weighted market returns, by 13 b.p. for the top five days per year, and 5 b.p. for the top fifty days. This effect holds internationally, is weaker when asset managers are less likely to reinvest dividends, and is stronger when liquidity is low (i.e. VIX is high). Industries with greater past exposure to dividend price pressure underperform those with less exposure, consistent with a long-term partial reversal. Predictable selling pressure leads to lower returns after blackout periods for firms with higher stock compensation. Back of the envelope calculations suggest price multipliers of each dollar invested in the aggregate market ranging from 1.5 to 2.3. These results suggest that predictable price pressure is a widespread result of money flows, rather than an anomaly.
What is risk? How investors perceive risk in return distributions
Stefan Zeisberger
Radboud University / University of Zurich
Zürich, Switzerland
Most textbook finance literature assumes risk to be defined as the standard deviation of returns (volatility). Academics, financial advisors and regulators use this definition alike and build models and policy around it. However, whether this is consistent with how investors actually perceive risk, has hardly been tested. In a comprehensive series of studies, we present investors with return distributions that have different risk characteristics. Investors have to state their perceived risk for these and make investment decisions. Our method allows us to detect which risk measures they implicitly use. The results hint at the probability of losing being the main driver of risk perception and investment propensity. Volatility plays a less important role. Our findings are robust with regard to color-coding of return distributions, different investor types, personal characteristics including investment experience, the use of monetary incentives, and the presentation format of the return distribution.
Measuring Financial Advice: Aligning client elicited and revealed risk
John Thompson
Wilfrid Laurier University
Waterloo, ON, CA
Financial advisors use questionnaires and discussions with clients to determine a suitable portfolio of assets that will allow clients to reach their investment objectives. Financial institutions assign risk ratings to each security they offer, and those ratings are used to guide clients and advisors to choose an investment portfolio risk that suits their stated risk tolerance. This paper compares client Know Your Client (KYC) profile risk allocations to their investment portfolio risk selections using a value-at-risk discrepancy methodology. Value-at-risk is used to measure elicited and revealed risk to show whether clients are over-risked or under-risked, changes in KYC risk lead to changes in portfolio configuration, and cash flow affects a client's portfolio risk. We demonstrate the effectiveness of value-at-risk at measuring clients' elicited and revealed risk on a dataset provided by a private Canadian financial dealership of over 50,000 accounts for over 27,000 clients and 300 advisors. By measuring both elicited and revealed risk using the same measure, we can determine how well a client's portfolio aligns with their stated goals. We believe that using value-at-risk to measure client risk provides valuable insight to advisors to ensure that their practice is KYC compliant, to better tailor their client portfolios to stated goals, communicate advice to clients to either align their portfolios to stated goals or refresh their goals, and to monitor changes to the clients' risk positions across their practice.
The Negativity Bias and Perceived Expected Return Distributions: Evidence from a Pandemic
Harry Turtle
Colorado State University
Fort Collins, CO, USA
We hypothesize that the negativity bias—the thesis that negative aspects have a more powerful impact than positive aspects on attention, learning, decision-making, and perceived risks—can help explain why most individuals hold strongly bearish views of short- and long-term expected equity return distributions, why individuals exhibit heterogeneous beliefs, and stock market under-participation. Using variation in the perceived risk of mortality from the 2009 H1N1 swine flu pandemic as a measure of the negativity bias, we find strong support for our hypothesis.
The Effects of Stock Ownership on Individual Spending and Loyalty
Michaela Pagel
Columbia Business School
Columbia University
New York, NY, USA
We show that when individuals own stock from a certain company, they increase their spending in that company’s stores. We use data from a FinTech app that opens brokerage accounts for users and rewards them with stock when they shop at pre-selected stores. For identification, we use the staggered distribution of brokerage accounts over time and quasirandomly distributed stock grants. We also show that loyalty is the dominant psychological mechanism behind our findings, that weekly spending in specific stores is strongly correlated with retail stock holdings of that company, and that stock rewards increase overall investment activity.
Are Millennials Wary of the Stock Market? A Cohort Analysis of Stock Holdings
Zhujun Cheng
The Ohio State University
Columbus, OH, USA
Participation in the stock market dramatically dropped during the Great Recession. There have been concerns that certain demographic groups, especially Millennials, have not since been back to the stock market. Using the Survey of Consumer Finances (SCF), we create seven cohorts based on birth years and distinguish age and cohort effects on stock holdings outside and in retirement accounts. Our results provide evidence for strong cohort effects for stocks outside of retirement accounts. Controlling for age, not only Millennials but all cohorts had lower stock holdings outside retirement accounts in 2016 compared to 2007. We did not detect any cohort effects for stocks in retirement accounts.
Who Pays the Price for Bad Advice?: The Role of Financial Vulnerability, Learning and Confirmation Bias
Julie Agnew
The College of William and Mary
Williamsburg, VA, USA
Abstract: What kinds of people will pay bad financial advisers? We show that experimental participants (n=2003) with a proclivity toward confirmation bias are more susceptible to bad advisers. We give participants a sequence of signals of adviser quality that can be clear or ambiguous, depending on each participant’s ability to discern bad advice. Rational participants set aside ambiguous signals and do not use them to update beliefs about advisers. Biased participants treat ambiguous signals as favoring their priors, and update accordingly. Younger, more trusting, more impulsive, less financially literate and less numerate participants are most vulnerable to paying a poor-quality adviser.
Roth Conversion Strategies for Different Generations
Jenny Gu
University of Dallas
Dallas, TX, USA
This paper studies Roth conversion strategies for different generations under the impact of recent tax reforms. Conventional wisdom to choose between Traditional IRA and Roth IRA is that Roth IRA is preferred when higher tax bracket is expected in retirement. However, the challenge is to predict the tax rates in decades, and market risk and inflation risk add more uncertainties. Each generation has disparate priorities on retirement planning and different early withdrawal potentials. When future tax policy is uncertain, it is prudent to hedge the risk by allocating assets into accounts with different tax treatment. Previous studies have mixed arguments on the relative advantages of Roth conversions, and financial planning scholars have suggested it would be advantageous to have both types of IRA accounts. In this paper, we specifically examine whether the “back-door” Roth IRA conversion is advantageous for each generation and how to reach the optimal tax diversification at different ages. Tax diversification, one of the major retirement planning strategies, refers to the strategic allocation of investment assets among multiple investment accounts with varying taxation. (Silver, 2013, Wendling and Tacchino, 2018). A taxdiversified retirement portfolio reduces a retiree’s tax burden during retirement years. Finally, recent tax reforms have profound impacts on retirement planning strategies. The 2017 Tax Cuts and Jobs Act (TCJA) and 2019 Secure Act added new dimensions to consider when tax-diversifying the retirement portfolio. We incorporate all these factors into our study and explore why a proactive Roth conversion strategy is important at current time.
Eponymous Hedge Funds
TriTrinh
NEOMA Business School
Paris, France
Using a relatively common phenomenon of eponymy in the hedge fund industry where funds are named after their founder-managers, we examine if eponymy is associated with skilled managers signaling their ability. Our results suggest that eponymous fund managers are neither necessarily skilled nor outperform their non-eponymous peers. In contrast, eponymous funds take higher risk which lead them to have lower Sharpe ratios and information ratios, hence worse risk-adjusted performance. Moreover, we do not find any evidence of an increase in reputational costs and benefits associated with eponymy. Fund investors neither reward nor punish eponymous managers for good and bad performance, respectively, relative to their non-eponymous peers. Overall, these results fail to support a signaling-based explanation of eponymy and highlight the need for exploring other rationales behind the eponymy decision of hedge fund managers.
Racial Animosity and Black Financial Advisor Underrepresentation
Derek Tharp
University of Southern Maine
Lewiston, ME, USA
This study investigates whether racial animosity across metropolitan markets is associated with Black financial advisor underrepresentation. Using a dataset of all U.S. securities-licensed individuals (N = 642,543), we first estimate the racial and ethnic composition of the industry using an algorithm that accounts for name, gender, and location. Second, we use a dataset enhanced by a commercial vendor to restrict the analysis to only those identified as working as financial advisors (n = 237,435). Using racially charged Google search queries as a proxy for racial animosity, we find that greater racial animosity is associated with greater Black advisor underrepresentation. We estimate lower underrepresentation of 0.9 percentage points when comparing markets with the highest and lowest levels of animosity. For the average market with an estimated 11.4% Black advisor representation, an increase of 0.9 percentage points would represent a 7.9% increase in Black advisor representation.
Goal Setting and Saving in the FinTech Era
Antonio Gargano
University of Houston
Houston, TX, USA
We study the effectiveness of soft, self-designed commitment devices, i.e. saving goals, in increasing individuals' savings using data from a FinTech App. We establish that setting goals increases individuals' saving rate and show that the effect is causal using a difference-in-differences identification strategy that exploits the random assignment of users into a group of beta-testers that can set goals and a group of users that cannot. We also show that the increased savings within the App do not come at the expense of reduced savings outside the App. We explore the economic channels of our results by matching App user survey responses to their behavior and highlight the importance of a monitoring channel, consistent with models where agents experience disutility from falling short of their goal.
Everyone else is making a mistake: Effects of peer error on saving decisions
Elizabeth Perry
Harvard Business School
Harvard University
Cambridge, MA, USA
The power of social influence to affect our choices is well-documented in the behavioral science literature. However, social influence has been less reliable in financial contexts, where multiple studies have found that it either makes no difference or even backfires. This paper describes three interventions to increase retirement saving among nearly 10,000 federal employees. We explore whether letting people know about the mistakes of others can be effective in financial contexts. All three interventions led to significantly more people adjusting their savings rates compared to the baseline, raising new questions about the nuances of social influence.
Race, Gender, and the Likelihood to Take A Financial Planner’s Advice
Miranda Reiter
Texas Tech University
Lubbock, TX, USA
This study examined the likelihood of consumers to take a financial planner’s advice based on race and gender using an experimental design with a sample of Black and White respondents. Findings suggest that financial planners’ racial or gender backgrounds do not affect whether a client will follow a planner’s advice. However, women respondents were more likely than men respondents to report that they would follow the advice given by a financial planner.
Does Saving Cause Borrowing?
Michaela Pagel
Columbia Business School
Columbia University
New York, NY, USA
We study whether savings nudges have the unintended consequence of additional borrowing in high-interest credit. We use data from a pre-registered experiment that encouraged 3.1 million bank customers to save via SMS messages and train a machine learning algorithm to predict individual-level treatment effects. We then focus on individuals who are predicted to save most in response to the intervention and hold credit card debt. We find that these individuals save 5.7% more (61.84 USD per month) but do not change their borrowing: for every additional dollar saved, we can rule out increases of more than two cents in interest expenses.
Revisiting Covered Calls and Protective Puts: A Tale of Two Strategies
Bryan Foltice
Butler University
Indianapolis, IN, USA
This paper examines the historical risk-adjusted returns of two hedging strategies designed to minimize downside market risk: Protective-puts and covered-calls, using US market data from 1993 to 2020. Here, we find that covered-call strategies significantly outperform the buy-and-hold strategy on a raw and risk-adjusted basis over the entire sample and these excess returns appear to remain persistent over time. We also find the opposite results hold for the protective put strategy: This strategy not only significantly underperforms the buy-and-hold strategy from a raw and risk-adjusted return standpoint, it actually significantly increases the probability of incurring losses each month. Finally, we evaluate the overall utility of various covered call strategies for loss averse investors, using the standard prospect theory utility function. Here, we find that out-of-the-money covered-call options yield the highest utilities for investors with less than average loss aversion, while in-the-money covered call options become more favorable as loss aversion increases.
What Do the Portfolios of Individual Investors Reveal About the Cross-Section of Equity Returns?
Sebastien Betermier
McGill University
Montreal, QC, CA
We construct a parsimonious set of equity factors by sorting stocks according to the socio-demographic characteristics of the individual investors who own them. The analysis uses administrative data on the stockholdings of Norwegian investors in 1997-2018. Consistent with financial theory, a mature-minus-young factor, a high wealth-minus-low wealth factor, and the market factor price stock returns. Our three factors span size, value, investment, profitability, and momentum, and perform well in out-of-sample bootstrap tests. The tilts of investor portfolios toward the new factors are driven by wealth, indebtedness, macroeconomic exposure, age, gender, education, and investment experience. Our results are consistent with hedging and sentiment jointly driving portfolio decisions and equity premia.
How do I ask for Help? Psychological Barriers to Improving Wellness through Financial Professional Use
Lynnette Purda
Queen's University
Kingston, ON, CA
Working with a financial professional has been documented to provide psychological and financial wellness benefits yet many individuals are reluctant to engage professional help. We explore potential barriers to working with a financial advisor, focusing on previously mixed findings on the role of self-efficacy in this context. We unpack self-efficacy into two unique components: a belief in one’s ability to independently manager their own financial affairs (financial self-efficacy or FSE for short) and a separate level of confidence in the ability to successfully engage with a financial professional who can advise or undertake these tasks (Financial Advice Seeking Self-Efficacy or FASSE). We are motivated by evidence that consumers struggle to comprehend various aspects of the financial services industry and hypothesize that this struggle acts as a barrier to advice-seeking behavior. We find that the two aspects of self-efficacy play different roles in encouraging the use of a financial advisor and that advisor adoption is highest among those individuals who hold both a favorable attitude towards financial professionals and those with higher levels of FASSE. Moreover, our experiments show that FASSE can be improved with relatively small changes to the presentation of financial professional’s qualifications and the services they offer. Given suggestions that financial advice can help overcome low levels of financial literacy, our findings have significant implications for financial professionals and financial inclusion which plays prominently in the United Nations sustainability goals.
Exploring the moderating role of self-esteem: How does the interplay between internal and external support factors impact young adults’ financial resiliency?
Josh Harris
Kansas State University
Manhattan, KS, USA
A growing number of households face financial hardships ranging from the loss of utility coverage, inability to pay medical bills, and eviction. Various studies have pursued the efficacy of increasing financial knowledge, financial capability, or self-efficacy to improve financial outcomes. The current study explores self-esteem as an internal factor supporting financial resiliency in young adults as measured across multiple time points from the public-use data set of the National Study of Adolescent to Adult Health (Add Health). Add Health is a longitudinal survey conducted across five waves, with the first wave taking place between September 1994 and December 1995 when the population was in grades 7 through 12. The current study utilizes data from waves I, III, and IV. Wave III data was collected from August 2001 to April 2002, when the respondents were 18 to 26 years old. The latest wave used in this study, IV, was conducted in 2008 and 2009 when the original population was 24 to 32 years old. The population of interest is young adults who experienced a financial hardship (e.g., utility service being shut off due to inability to pay) during the Wave III time period. Beyond the moderating effect of self-esteem on young adults’ financial resiliency, the study explores the interaction between self-esteem and external sources of support - parents, peer, and school – finding positive relationships between these perceived sources of external support and positive financial outcomes. Early results find no conclusive support for improving self-esteem as a means to improving financial outcomes, but positive support is found for self-esteem for those young adults who experienced some level of public assistance.
Foregone Consumption and Return-Chasing Investments
Ben Charoenwong
National University of Singapore
Singapore
Existing behavioral economics and finance research has documented that individuals make mistakes in various contexts such as under diversification in investment allocations, time inconsistent savings behavior, and consumption errors; and, that biases seem exacerbated among those with less cognitive abilities (D’Acunto et al. 2019). However, most of the existing research focuses on a small number of decisions, such as how investors trade within their brokerage account or whether mortgage refinancing responds to changes in mortgage rates. A question remains how different individual decisions interact at the household portfolio level. As stated in Campbell (2006), “the study of household finance is challenging because household behavior is difficult to measure, and households face constraints not captured by textbook models.” Further, the process by which individuals form expectations and how the expectations map to real economic outcomes like consumption, savings, and risk-taking are relevant to policy makers considering different fiscal and monetary policy tools. Such analysis has typically not possible due to the lack of comprehensive data.
AlphaPortfolio: Direct Construction Through Reinforcement Learning and Interpretable AI
Will Cong
SC Johnson College of Business
Cornell University
Ithaca, NY, USA
To best facilitate financing planning practice, we directly optimize the objectives of portfolio management via reinforcement learning---an alternative to conventional supervised-learning-based paradigms that entail first-step estimations of return distributions, pricing kernels, or risk premia. Building upon breakthroughs in AI, we develop multi-sequence neural network models tailored to distinguishing features of economic and financial data, while allowing training without labels and potential market interactions. The resulting AlphaPortfolio yields stellar out-of-sample performances (e.g., Sharpe ratio above two and over 13% risk-adjusted alpha with monthly re-balancing) that are robust under various economic restrictions and market conditions (e.g., exclusion of small stocks and short-selling). Moreover, we project AlphaPortfolio onto simpler modeling spaces (e.g., using polynomial-feature-sensitivity) to uncover key drivers of investment performance, including their rotation and nonlinearity. More generally, we highlight the utility of deep reinforcement learning in finance and invent ``economic distillation'' tools for interpreting AI and big data models.
Off Target: On the Underperformance of Target-Date Funds
David Brown
University of Arizona
Tucson, AZ, USA
Target-date funds (TDFs) are popular vehicles that provide investors with an evolving asset allocation to meet their needs at some future date (e.g., retirement). While TDFs provide investors with extensive diversification and active rebalancing, TDFs are also a type of fund-of-funds. As such, investors pay multiple layers of fees as most TDFs charge fund-of-funds’ fees and also hold funds that collect additional fees. We show that TDFs are easy to emulate with a portfolio of cost-efficient exchange-traded funds (ETFs) and we coin these portfolios Replicating Funds (RFs). RFs substantially outperform TDFs, exhibit low tracking error, do not suffer from cash drag, and require infrequent rebalancing. Our analysis shows that TDF sponsors collectively charged nearly $2.5 billion in excess fees in 2017 alone. We provide a normative rule-of-thumb for investors to construct their own RFs using low-cost ETFs.
Understanding and neutralizing the expense prediction bias: The role of accessibility, typicality, and skewness
Chuck Howard
Texas A&M University
Consumers display an expense prediction bias in which they underpredict their future spending.
We propose that this bias occurs in large part because:
- 1) consumers base their predictions on typical expenses which come to mind most easily during prediction,
- 2) typical expenses approximate the mode of a consumer’s expense distribution rather than the mean, and
- 3) expenses display strong positive skew with mode < mean.
Accordingly, we also propose that prompting consumers to consider reasons why their expenses might be different than usual increases predictions – and therefore prediction accuracy – by bringing atypical expenses in the right tail of the distribution to mind. 9 studies (N = 5,644) provide support for this account of the bias and the “atypical intervention” we develop to neutralize it.
Human Capital Risk and Portfolio Choices: Evidence from University Admission Discontinuities
Philippe D'Astous
HEC Montreal
Montreal, QC, CA
Theory suggests that increasing idiosyncratic, uninsurable labor income risk causes individuals to reduce the risk in their financial assets. Ceteris paribus, this would lead to a negative association between the levels of labor income and financial risks. This relationship is difficult to measure empirically because labor income and financial risks are jointly determined, and both driven by unobservable preferences. Risk tolerant individuals tend to choose a riskier career and hold riskier portfolios, leading to an upward-biased estimate of the effect of earnings risk on risky assets holdings. We overcome this identification problem by exploiting a discontinuity built into the Danish national university admissions system which provides quasi-random assignment of similar applicants to programs with different earnings risk profiles. We exploit the fact that university programs have a causal impact on students’ subsequent earnings processes, and use earnings volatility as one feature of the earnings process that proxies for income risk. We show that such increase in income risk reduces risky asset holding and stock market participation. Entering a program whose enrollees subsequently experience volatile earnings causes students to have more volatile earnings and, ceteris paribus, to hold fewer risky assets and be less likely to participate in the stock market.
A competency modeling approach to assessing risk tolerance
Sarah Fallaw
Datapoints
John Grable
University of Georgia
Athens, GA, USA
A client’s psychological risk tolerance, or the client’s enduring personality characteristics related to the ability to withstand the ongoing losses and gains, is a complex and multidimensional set of constructs. In a series of studies, we demonstrate the reliability and validity of the Investor Profile, a measure of psychological risk tolerance that includes a combination of various psychological predictors, including self-esteem, self-efficacy, knowledge, emotional stability, risk personality, and risk preference in the prediction of clients’ behavior related to market downturns. This competency modeling approach to measuring psychological risk tolerance provides the client with a better understanding of how the client might react to market downturns in the future while also providing the financial professional with specific areas that can be the focus of ongoing coaching and guidance to improve an investor’s decision-making related to selling equities during a downturn in the market.
Saving More With Less: Optimal Plan Design with Automatic Enrollment and Employer Match Contributions
Zhikun Liu
Empower Retirement
Greenwood Villae, CO, USA
Encouraging DC participants to save more for retirement is a common goal among financial advisors and retirement plan sponsors. In this study, we explore how various factors affected participant saving decisions. Using an administrative data set of approximately 157,000 participants who recently enrolled in an employer-matched 401(k) plan, we find that increasing the default deferral rate (as part of an automatic enrollment scheme) can significantly improve retirement savings. In other words, for plan sponsors interested in increasing participant deferral rates without changing the existing employer contribution structure, simply increase the default savings rate seems like a relatively powerful and straightforward way to accomplish this objective. This study disentangles the impacts of two most commonly discussed factors that can potentially increase participant’s retirement savings rate: Automatic deferral rate and employer contributions. Other factors that are significantly associated with participant’s saving decisions include age, salary, equity percentage, maximum employer match rate, etc.
Insurance Alchemy: The Fixed Rate Annuity Anomaly
David Blanchett
QMA
Newark, NJ, USA
Fixed rate annuities, also called multi-year guaranteed annuities, or MYGAs, offer expected returns protected by state guaranty associations similar to other safe investments such as CDs, money market accounts, or Treasury bonds. Yields on MYGAs have risen considerably, in relative terms, in recent years compared to historical levels. MYGAs offered by highly rated insurers had a 100 bps yield premium above Treasury bonds as of December 2020, with premiums exceeding 200 bps from lower rated insurers. Expected MYGA returns were approximately 100 bps higher than yields on similarly rated corporate bonds and have significantly lower default risk. We explore this apparent anomaly and discuss opportunities for sophisticated investors.
Retirement Income Beliefs and Financial Advice Seeking Behaviors
Wade Pfau
The American College for Financial Services
King of Prussia, PA, USA
This investigation identifies and validates a series of salient behavioral finance and psychological constructs that influence retirement income planning. We show how these scales are related to each other as well as retirement income concerns and investment behaviors. We also describe how four investment personas can be linked with the Advisor Usefulness and Retirement Income Self-Efficacy scales to successfully identify preferred financial implementation methods. This can assist individuals in more readily recognizing their relative strengths and weaknesses when implementing a retirement income strategy, and financial professionals can present advice in a manner that addresses a client’s concerns and preferred implementation.