Anchoring to a Previous Stock-Price High: The Refusal to Sell Below It

Tversky and Kahneman’s 1974 paper on anchoring showed that arbitrary numerical anchors influence numerical judgments even when the anchor is known to be irrelevant. A roulette wheel stopped on 65 raised subsequent estimates of the percentage of African nations in the UN. A roulette wheel stopped on 10 lowered them. The mechanism operates below conscious reflection.

Applied to equity comp, anchoring creates a specific pathology. An employee watches their stock hit $220 in January 2022. The stock falls to $110 by October 2022. Two years later, the stock trades at $170. The employee cannot bring themselves to sell because $170 is “still below the high.” The implicit frame: the stock is underpriced until it returns to $220.

This is incoherent. The $220 print was a specific moment in a specific market regime. It may not represent any sustainable valuation of the business. It has no forward-looking information content. But it operates as a psychological anchor on every subsequent sell decision.

This article describes the anchoring trap, why it is particularly strong for vesting stock, and what actually breaks it.

Why the Previous High Sticks

Four factors make the previous high a strong anchor:

1. Salience. The high was probably seen by the employee at the time, remembered, possibly celebrated. It has emotional weight.

2. Aspirational target. The high represents the “peak” of what the employee’s equity was worth. Selling below it feels like accepting a reduced outcome.

3. Narrative alignment. The high often coincides with a positive narrative about the company (strong earnings, successful launch, market enthusiasm). The narrative fades; the number persists in memory.

4. Bidding ask-price mechanism. When asked “what would you sell at?” the employee’s brain retrieves the anchor first and works downward. The initial retrieval skews the final number.

The result: a sell threshold that is fundamentally above the current market price and has no connection to the current value of the business.

The Efficient-Market Argument Against Anchoring

If the stock is now trading at $170 with full market information, the $170 price is a rough summary of what participants with access to current information believe the stock is worth. The previous $220 price was a summary of what participants believed in a different market regime, with different interest rates, different competitive context, and different growth assumptions.

Neither price is “correct” in an absolute sense. Both are equilibrium points given their moments. But the $170 price has the advantage of incorporating all information that has accumulated since $220. The $220 is dated.

A rational decision-maker uses the current price as the reference for “what can I get” and the forward distribution of prices as the reference for “what might I get by holding.” The previous high has no role in either.

Anchoring overrides this by inserting $220 into the mental model as the “true” or “deserved” price. The employee waits for the market to return to the anchor, rather than recognizing that the anchor has no claim to being the right price.

Case Study: The 2022 Tech Drawdown

A concrete example. Many tech stocks peaked in November 2021 and drew down 50-80% by October 2022.

• Meta: $378 (Sep 2021) to $88 (Nov 2022), a 77% drawdown. Back to $375 by Feb 2024.
• Shopify: $169 (Nov 2021) to $24 (Oct 2022), an 86% drawdown. At $100 by early 2026.
• Peloton: $162 (Jan 2021) to $6.40 (Dec 2022), a 96% drawdown. Still at $6 in early 2026.
• Zoom: $588 (Oct 2020) to $60 (Oct 2022), a 90% drawdown. At $85 in early 2026.
• DocuSign: $310 (Sep 2021) to $40 (Oct 2022), an 87% drawdown. At $90 in early 2026.

An employee at any of these companies with a 2021-era anchor was waiting for the stock to “come back” through 2022, 2023, 2024, and 2025. Meta employees got the recovery. Peloton employees did not. Zoom employees got partial recovery.

The anchor didn’t know which company would recover. The employee didn’t know either. Holding based on the anchor was a bet, not a forecast. The bet paid off for Meta employees and failed for Peloton employees. A diversified alternative was indifferent to which was which.

A Peloton employee with $1M of concentrated stock at the $162 peak, anchoring on that price, held through the 96% drawdown waiting for recovery. The stock has not recovered. The employee ended with $60K of stock. A diversified alternative over the same four years would have grown to roughly $1.3M-$1.4M under ordinary returns. The anchoring cost: roughly $1.25M.

The Vest-Date Anchor Problem

For RSU employees, the vest-date FMV is the tax basis and a natural candidate anchor. “I was taxed at $180 per share, so selling below $180 feels like a loss.”

The vest-date anchor is almost always wrong as a sell threshold. It is a tax-reporting number. It is not a forecast of future price. It is not a statement of fair value.

A rational frame: vest-date FMV is the basis for computing capital gain. Selling at $160 after vesting at $180 produces a $20-per-share capital loss, which is usable against other gains under §1211 or carried forward under §1212. The vest-date FMV is a reference point for tax computation, not a hurdle to clear before selling.

Many employees combine the vest-date anchor with the recent-high anchor and get a compound distortion. “I was granted at $15 in 2019. The stock peaked at $220 in 2022. Current vest-date FMV is $180. It’s now at $140.” The employee has three reference points (grant, peak, vest) and refuses to sell below any of them. The sell threshold becomes $220, which the stock may never see again.

What Actually Works Against Anchoring

Research on debiasing anchoring suggests three interventions that work:

1. Explicit consider-the-opposite. Before making the decision, write out the case for the opposite conclusion. If you believe the stock is worth $220, write out why it might be worth $100. The exercise often shifts the anchor.

2. Reference to current-price-based analogies. Substitute a similar stock trading at a similar multiple. If your stock is at $170 and a peer trades at $185 with similar growth and margins, the peer is the anchor, not the previous high.

3. Remove the anchor from the visible field. Stop looking at the stock’s chart. Do not pull up historical prices. Make decisions based on current price and diversification frame only.

Structural interventions that bypass anchoring:

1. 10b5-1 plan. Pre-committed sell schedule. The anchor doesn’t get to vote because the plan sells whether or not the current price is “acceptable.”

2. Exchange fund contribution. Contribute concentrated stock to an exchange fund. The fund receives stock at current price, converts your exposure to diversified. The anchor attaches to the new fund position, which is diversified and stable.

3. Staged selling triggered by percentage of position, not price. “Sell 10% of the position every quarter regardless of price.” Removes the anchor entirely.

The “I’ll Sell When It Gets Back To…” Language

Listen for the specific language. “I’ll sell when it gets back to $X” is anchoring announced out loud.

The question back: “What if it never gets to $X? What if it takes 7 years? What would you wish you had done today looking back from 7 years from now?”

The question rarely gets answered in the moment. It plants the frame for later.

Frequently Asked

Is anchoring the same as loss aversion? Related but distinct. Loss aversion is the 2:1 pain asymmetry around the reference point. Anchoring is what sets the reference point. Together they create the hold-to-breakeven trap.

Does anchoring affect gain positions too? Yes, inversely. Employees with a stock trading above a previous high often sell too quickly, anchoring to the peak as “overvalued” when fundamentals may support the new price. The bias cuts both ways.

Can a financial advisor help me remove the anchor? Sometimes. A good advisor reframes the decision in current-price terms. An advisor who validates the anchor (“yes, let’s wait for $220”) is not helping.

Does dollar-cost-averaging out work? Yes. DCA out of a concentrated position ignores the anchor by treating the decision as a schedule rather than a price call. 10b5-1 plans are the formal version.

What about selling some and holding some to “split the difference”? Better than holding everything. The middle option captures some diversification benefit while preserving some upside. Explicitly size the “held” portion based on your forward belief about the stock, not based on the previous high.