"Conflicts of Interest Between Employers and Employees in Mutual Fund Families: Evidence from 401(k) Plans", Job Market Paper
In recent years, 41 of the 150 largest U.S. mutual fund families have faced lawsuits from their own employees, who allege that their employers populate company 401(k) plans with expensive and underperforming proprietary funds. I show that fund family employees reduce their allocations to proprietary funds, and this reallocation intensifies following lawsuit filings. In addition, I find that fund families respond strategically: while they remove the specific funds targeted in litigation, they do not add external funds as replacements. Instead, they substitute proprietary funds with similar expense ratios and performance characteristics, preserving their ability to generate revenue from employee assets. Lastly, I document that the style-adjusted expense ratios of proprietary funds in fund families’ own 401(k) plans are higher, and style-adjusted returns are lower, than those of proprietary funds offered in their external clients’ plans. These conflicts of interest are more pronounced among publicly traded fund families.
Monash University; Silicon Prairie Conference (2025)
"Conflicts of Interest among Affiliated Financial Advisors in 401(k) Plans: Implications for Plan Participants", (with Gjergji Cici and William Bazley)
Institutional features of 401(k) plans can give rise to conflicts of interest between plan participants and financial advisors that advise them. We study one such conflict that arises when advisors are affiliated with the plan’s recordkeeper. Using a large dataset of 401(k) plans, we find that affiliated advisors reduce investment performance by steering participant flows to proprietary funds. We observe no similar effects for unaffiliated advisors. Additionally, affiliated advisors provide no significant benefits in terms of participation rates, administrative fees, or diversification. Given the increasing prevalence of advisors within 401(k) plans, our findings have relevant implications for households, plan sponsors, and policymakers.
Revise and Resubmit at The Journal of Finance
SFA (2025, scheduled); FMA (2025, scheduled); FBA (2025)
“Learning by Losing in Prediction Markets”
I study whether prediction market traders learn from their losing trades using complete transaction records of 1.5 million Polymarket wallets across ten topical domains from January 2023 through February 2026. I find that cumulative losing trades predict significantly higher subsequent returns, consistent with traders correcting past errors, while cumulative winning trades predict lower returns, consistent with overconfidence. Decomposing returns by bet side reveals that losses on YES (or NO) bets specifically improve future returns on that same side, suggesting bet-type-specific learning. Exploiting variation across domains such as Sports, Politics, Crypto, and Finance, I show that losing trades improve returns within that domain but do not transfer to other domains, and traders active in many domains simultaneously exhibit smaller marginal benefits from additional losses, consistent with attention dilution. Investigating mechanisms, I find that traders reduce trade volatility after accumulating losses, while winning trades lead to increased trading activity. I further document a systematic wealth transfer from new entrants to seasoned traders that widens monotonically over time.