Twice Shy: The Hidden Cost of Never Going Back to a Stock That Burned You

by AdvisorAnalyst.com Editorial Team

Every investor has one. A name they owned, a thesis they believed in, a loss they absorbed, and a vow never spoken aloud but quietly kept: never again. The stock did not go to zero. The company did not commit fraud. The fundamentals may have long since recovered. But the door stays shut.

Clare Flynn Levy, CEO and founder of Essentia Analytics and a former fund manager who ran over $1 billion in pension assets, names this phenomenon directly and traces its consequences with the kind of analytical precision that comes from sitting on both sides of the equation. Writing in "Once Bitten: Why Investors Won't Buy Back a Stock That Burned Them,"¹ Levy argues that repurchase bias is not a personality flaw but a documented, quantifiable drag on investment performance, and one that active managers can no longer afford to carry.

The Bias Has a Name, and the Data Are Unambiguous

Levy points to research by Strahilevitz, Odean and Barber (2011), who studied more than 660,000 individual investors and found they were one-half to two-thirds more likely to repurchase a stock they had previously sold for a gain than one sold for a loss. The driver is not deteriorating fundamentals. The driver is regret. Levy puts it plainly: investors are "treating the stock like an ex you've blocked."

The pattern holds among professionals. Du, Niessen-Ruenzi and Odean (2024) found that a stock sold by a mutual fund manager at a loss was roughly 20% less likely to be repurchased than one sold at a gain. The most revealing detail in that study: when fund managers changed jobs, they still avoided the names that had burned them at the old shop, even when the investor base had completely turned over and no institutional memory of the original loss existed. The wound traveled with the manager, not with the portfolio. Levy connects this to the broader concept of the snakebite effect, a term from behavioural economics for the post-loss risk aversion that persists "long after the odds stop justifying the caution."

The Competitive Dimension

Levy situates repurchase bias within a structural challenge for active management. The IPO market has been relatively stagnant in recent years, already narrowing the investible universe. Repurchase bias shrinks it further, silently. Every name placed on a permanent blacklist is one fewer opportunity the portfolio can ever access.

The competitive threat sharpens this. Levy observes that the growing share of non-human market participants do not suffer from repurchase bias unless deliberately designed to. The machine, as she puts it, "feels no shame or embarrassment." It would maintain a watchlist of every exited position, re-run the thesis on schedule with no memory of the pain, and ask a single question: is this one of the best opportunities available right now? It would not care if the answer was yes on a name that lost the portfolio money in the past.

This is the asymmetry active managers need to reckon with. The human carries scar tissue the algorithm does not.

The Discipline in Practice

The article is not a diagnosis without a prescription. Levy outlines a process that mirrors the objectivity of the machine while remaining executable by a human. When a position is exited, it goes on a watchlist immediately. It gets reviewed at defined intervals, at minimum annually. At each review point, the manager asks a single question, framed deliberately: "not 'how do I feel about this name?' but 'based on the facts, if I had no history with it at all, how much would I want to own it today?'"

The framing matters. Levy argues that a re-entry produced by a documented, repeatable process is not churn. "It's discipline made visible, and it's a far easier story to tell a client than a gut call." Investors, she notes, do not punish deliberate process-led decisions. They punish decisions that feel arbitrary.

The piece connects directly to Levy's prior essay on cutting losers, positioning both articles as two halves of the same discipline: a process for knowing when you are wrong, and a process for staying open to something you were once wrong about. The distinction between the two is important. Cutting a loser is an act of intellectual honesty. Getting back in is an act of intellectual courage.

Levy closes with a line that reorients the entire emotional architecture of the problem: "The ex you blocked actually wronged you. The stock never did, your own regret did."

Five Key Takeaways for Advisors and Investors

1. Repurchase bias is empirically documented and measurable, not anecdotal. Academic research confirms professionals and retail investors alike are significantly less likely to buy back a name sold at a loss, regardless of its current fundamentals.
2. The bias is personal, not institutional. Fund managers carry repurchase aversion across jobs, even when the original investors are long gone. Awareness of this portability is the first step toward correcting it.
3. A shrinking investible universe makes the bias more costly. With IPO activity constrained, voluntarily excluding previously held names compounds an already-tighter opportunity set.
4. Non-human competitors do not share this bias. AI-driven participants will reconsider every prior position on its current merits, with no emotional residue. Active managers who cannot do the same are ceding a structural edge.
5. Process is the antidote. A formal watchlist with scheduled, criteria-based review transforms an emotionally charged re-entry into a defensible, client-explainable decision rooted in discipline rather than impulse.

Footnote:

1 Flynn Levy, Clare. "Once Bitten: Why Investors Won't Buy Back a Stock That Burned Them." Essentia Analytics, 24 June 2026,.