When Paper Losses Become Real Problems: A Robo‑Advisor's Struggle with Client Aversion
A real-world example of Loss aversion in action
Context
Leafline Capital is a mid-size robo-advisor managing $120M in client assets. The firm rolled out an automated tax‑loss harvesting (TLH) feature designed to improve clients' after‑tax returns by systematically realizing small losses and replacing positions with equivalent exposure.
Situation
The TLH feature was enabled as an opt‑in recommendation via in‑app messaging and email, with data showing a projected 0.8–2.2% increase in after‑tax returns depending on client tax brackets. Despite clear projected gains, a substantial share of clients clicked 'decline' or ignored educational material explaining how harvesting losses now leads to larger after‑tax wealth.
The bias in action
Clients focused on the immediate feeling of realizing a loss — seeing 'loss realized' in account activity — and interpreted that as personal failure. Many equated harvesting with 'locking in losses' rather than understanding the tax benefit and future reentry positions. Advisors reported that when presented with account statements showing realized losses, clients expressed regret and asked to disable TLH even when models projected higher long‑term returns. The company saw decision patterns consistent with loss aversion: the psychological pain of recording a loss outweighed the rational expectation of tax‑efficient gains.
Outcome
Within 12 months of the rollout, 18% of eligible clients actively declined TLH and another 12% ignored the recommendation (effectively declining). The cohort that refused TLH underperformed similar clients who accepted it, and refusal was associated with higher churn and lower referrals. Management paused further behavioral features while they reworked the messaging and defaults.


