Sunk Cost Fallacy in Stock Ownership

Intermediate | Published: 2025-12-28

Why It Matters

Sunk cost fallacy—the tendency to continue holding an investment because of previously invested resources (money, time, effort), even when continuing is no longer optimal—shows up in portfolios as holding losers to 'get back to even', averaging down on deteriorating positions (escalation of commitment), and refusing to sell 'at a loss' (treating past cost as economically relevant). In real market data, stocks with large unrealized losses (held by sunk-cost-driven investors) continue underperforming by 5-10% over the next year as investors resist selling despite deteriorating fundamentals (Frazzini, 2006).

The practical antidote isn't forcing yourself to "admit mistakes." It's mechanical forward-looking decision rules that evaluate positions based on future prospects only—treating every day as a fresh capital allocation decision, independent of past purchase price.

Definition and Core Concept

Sunk cost fallacy is the tendency to continue an investment or behavior because of previously invested resources (money, time, effort), even when continuing is no longer optimal (Arkes & Blumer, 1985). In stock ownership, past costs are economically irrelevant to future decisions—but psychologically, the price you paid (cost basis) dominates whether you hold or sell.

Two predictable distortions follow:

• Holding losers: "I paid $350, can't sell at $90—that would lock in the loss" (cost basis anchor preventing rational exit)
• Averaging down: "I'll buy more at $50 to lower my cost basis to $60" (escalation of commitment, throwing good money after bad)

The result: you hold deteriorating positions too long (hoping to "get back to even") and add capital to failing investments (to recover sunk costs)—both driven by backward-looking thinking instead of forward-looking analysis.

The Cost Basis Anchor (Why Past Price Dominates Future Decisions)

Sunk cost fallacy isn't System 1 emotion—it's mental framing that makes past costs (which are economically irrelevant) feel psychologically relevant. Rules based on fresh eyes analysis activate System 2 evaluation of future prospects, not past costs.

The mechanism (Heath, 1995): mental accounting amplifies sunk cost fallacy—investors frame stock as "investment to recover" rather than "capital allocation decision." This causes escalation of commitment: adding to losers to "average down" and recover sunk costs, even when thesis has deteriorated.

Staw (1976) shows escalation of commitment—when initial decision fails, decision-makers often increase investment (throwing good money after bad) to justify initial decision, driven by self-justification and avoiding admitting mistakes. In investing: you bought at $70, stock drops to $50, you buy more to "double down" and "prove you were right"—even though thesis may have deteriorated.

Related Concepts (Use These to Think Clearly)

• Sunk cost fallacy: the cognitive error—letting past unrecoverable costs influence future decisions
• Cost basis anchor: the mechanism—psychological attachment to purchase price
• Escalation of commitment: the behavioral manifestation—increasing investment in failing course of action to justify initial decision

A useful causal chain: Cost basis anchor (driver) → Sunk cost fallacy (thought pattern) → Escalation of commitment (behavior) → Holding losers too long + Averaging down (portfolio impact)

Frazzini (2006) shows investors hold losing stocks too long (sunk cost fallacy + disposition effect), causing underreaction to negative news. Stocks with large unrealized losses continue underperforming by 5-10% over the next year as investors resist selling—market is slow to incorporate bad news because sunk-cost sellers are absent.

How Sunk Cost Fallacy Shows Up in Portfolios

Example 1: Holding Loser to 'Get Back to Even' (Meta 2022—when cost basis dominated thinking)

Scenario: You bought Meta (Facebook) at $350/share in September 2021 (pre-metaverse pivot announcement). Thesis: "Dominant social platform with advertising moat—Instagram, WhatsApp, Facebook all printing money."

Phase 1: The Decline (September 2021-November 2022)

• Sept 2021: Entry price $350
• Oct 2021: Meta announces metaverse pivot, rebrands to "Meta," commits $10B+/year to Reality Labs
• Feb 2022: First-ever user decline reported, stock drops to $220 (-37%)
• Oct 2022: Q3 earnings miss, metaverse losses accelerating, TikTok competition intensifying → $100 (-71%)
• Nov 2022: Bottom at $90 (-74% from your entry)

How sunk cost fallacy manifests:

Cost Basis Anchor:

• Mental framing: "I paid $350, can't sell at $90—that would lock in -74% loss"
• Backward-looking: "I need to hold until it gets back to $350 to break even" (sunk cost driving decision)
• Psychological pain: Selling at $90 means "admitting I lost $260/share" (feels like failure)

Escalation of Commitment (Averaging Down):

• Nov 2022 thought process: "Meta is down -74%—that's oversold. I'll buy more at $90 to lower my cost basis to $170. Then I only need it to recover to $170 (not $350) to break even!"
• Action: Buy $10,000 more at $90 (already own $10,000 bought at $350)
• New average cost: ((350 + 90) ÷ 2 = 220) per share (on 2x position size)
• Mental framing: "I'm being smart—doubling down when everyone else is panicking"

Forward-Looking Analysis (Ignored):

Correct Questions (Not Asked):

• "If I had $9,000 cash today (current position value at $90), would I buy Meta?" (not "should I hold because I paid $350?")
• "Is Meta at $90 the best use of my capital vs. alternatives (S&P 500, diversified tech, bonds)?"

Thesis Evaluation (November 2022):

• Positive: Still dominant in social (Instagram 2B users, WhatsApp 2B users, Facebook 3B users), massive ad platform scale
• Negative:
  • Metaverse burn: $10B+/year with no clear ROI or timeline (Reality Labs losing money)
  • TikTok competition: Stealing engagement from Instagram Reels, especially Gen Z
  • Apple iOS changes: ATT (App Tracking Transparency) hurting ad targeting, revenue impact ~$10B/year
  • Regulatory risk: Antitrust scrutiny, potential forced divestitures
• Expected return: Highly uncertain—metaverse success is speculative, core ad business facing headwinds

Alternative Use of Capital:

• S&P 500 index: Offering ~10% long-term expected return with diversification (not concentrated bet on Meta turnaround)

Optimal Decision (Sunk Cost Ignored):

Fresh Eyes Test:

• "If Meta were IPO'd today at $90, and I had $9,000 cash, would I buy it?"
• Honest answer (if unbiased by sunk cost): "Uncertain—metaverse is unproven, core business has headwinds, TikTok is real threat. I'd rather own diversified S&P 500."
• Action: SELL Meta at $90, regardless of $350 cost basis. Past cost is irrelevant; only future prospects matter.

What Actually Happened (Meta Nov 2022-Dec 2023):

• Nov 2022: $90 (bottom)
• Dec 2023: $400 (+344% in 13 months)
• Why: Zuckerberg pivoted—cut metaverse burn, focused on AI (Llama models), launched Threads (Twitter alternative), cost cuts improved margins

Does This Validate Holding?

• NO—you got lucky. In November 2022, you couldn't predict AI pivot or Threads success. Your decision to hold was driven by sunk cost ("need to get back to $350"), not forward-looking thesis ("Meta will pivot to AI and margins will expand").
• Process matters, not outcome: Sunk cost fallacy kept you in deteriorating thesis—happened to work out ex post, but process was flawed.

Alternative Outcome (Equally Plausible):

• If Meta had continued metaverse focus, no AI pivot, TikTok kept stealing share → stock $90 → $50 (-44% additional decline)
• S&P 500 (Nov 2022-Dec 2023): +26% actual return
• Opportunity cost: Holding Meta to $50 vs S&P to +26% = 70 percentage point gap
• On $10,000 position: Meta = $5,600, S&P = $12,600 → $7,000 opportunity cost

The practical point: Sunk cost fallacy made your decision "I paid $350, can't sell at $90." The economically correct decision was "Is $90 Meta the best use of $9,000 capital TODAY?" The $350 you paid is GONE—it's economically irrelevant to whether you should hold. Only future prospects matter.

Note: This represents composite pattern observed across millions of retail investors holding Meta through 2022 decline. Reddit/Stocktwits archives show thousands saying "holding until I get back to even."

Example 2: 'Averaging Down' to Recover Sunk Costs (Enron 2001—escalation of commitment)

Scenario: You bought Enron at $70/share in early 2001. Thesis: "Innovative energy trading company, visionary management, expanding into broadband and water—next-generation utility." By August 2001, stock is $50 (down -29%).

How sunk cost fallacy manifests:

Phase 1: Initial Investment (Early 2001)

• Entry: $70/share, $10,000 position
• Thesis: Energy deregulation creating trading opportunities, Enron dominates trading platforms, growth into new markets (broadband, water)

Phase 2: Decline and Averaging Down (August 2001)

• Aug 2001: Stock drops to $50 (-29%)
• Your thought process (sunk cost framing):
  • "Great company, temporary dip—I'll buy more to lower my average cost"
  • "If I buy $10,000 more at $50, my average cost becomes $60 (not $70)"
  • "Then I only need stock to recover to $60 to break even—that's only +20% from $50, very achievable"
  • "Doubling down shows conviction—I believed in thesis at $70, so $50 is even better value"

Escalation of Commitment (Action):

• Buy $10,000 more at $50
• Total position: $20,000 ($10,000 at $70, $10,000 at $50)
• Average cost basis: $60
• Mental framing: "I'm smart—averaging down to get back to even faster"

Forward-Looking Analysis (Should Have Done, But Didn't):

Correct Question (Not Asked):

• "Is Enron at $50 a good investment TODAY?" (not "will it recover to my $60 average?")

Thesis Evaluation (Available Information August 2001):

• Red flags (publicly available):
  • Short-seller reports (Jim Chanos, March 2001): "Accounting is fraudulent, cash flow doesn't match earnings"
  • Bethany McLean article (Fortune, March 2001): "How does Enron make money?" (questioning business model opacity)
  • CFO selling stock: Andrew Fastow sold millions in Enron shares (insider selling signal)
  • Complex accounting: Off-balance-sheet SPEs (special purpose entities) hiding debt
  • CEO resignation (Jeffrey Skilling resigned August 2001 citing "personal reasons"—suspicious after 6 months as CEO)
• Positive case:
  • Revenue growth: Expanding rapidly (though later revealed to be accounting manipulation)
  • Energy trading dominance: Market leader in trading platforms
• Risk assessment: High opacity, accounting concerns, insider selling, CEO departure = major red flags

Optimal Decision (Forward-Looking):

Fresh Eyes Test (August 2001):

• "If Enron were IPO'd today at $50, would I invest $10,000?"
• Honest answer (unbiased by sunk cost): "No—too many red flags (short-seller reports, accounting questions, insider selling, CEO resignation). I don't understand the business model well enough, and smart people are raising serious concerns."
• Action: SELL existing position at $50, do not average down

What Actually Happened:

• Aug 2001: $50 (you averaged down here)
• Oct 2001: Accounting scandal revealed, stock $15 (-70% from Aug)
• Nov 2001: Bankruptcy filing, stock $0.26 (-99.5%)
• Total loss: $20,000 invested → $87 recovery value = $19,913 loss (99.6%)

Quantified Cost of Sunk Cost Fallacy:

Scenario 1: Sell at $50 (No Averaging Down)

• Initial investment: $10,000 at $70
• Sell at $50: (10{,}000 × (50 ÷ 70) = 7{,}143) recovery
• Loss: (10{,}000 - 7{,}143 = 2{,}857)

Scenario 2: Average Down at $50 (Sunk Cost Fallacy)

• Initial: $10,000 at $70
• Added: $10,000 at $50
• Total: $20,000
• Bankruptcy: $87 recovery
• Loss: (20{,}000 - 87 = 19{,}913)

Incremental Cost of Averaging Down:

• Extra capital deployed: $10,000 (capital that could have been deployed elsewhere or held in cash)
• Incremental loss: (7{,}143 - 87 = 7{,}056) (recovery difference)
• Total cost of sunk cost fallacy: $17,056 ($10,000 new capital lost + $7,056 worse recovery on original position)

The durable lesson: Averaging down feels like "doubling down on conviction," but when thesis has deteriorated (red flags piling up: accounting concerns, insider selling, CEO resignation), it's actually "doubling down on a mistake." Sunk cost fallacy drove you to add $10,000 to a failing position to recover initial $70 investment—instead of evaluating "Is $50 Enron a good use of $10,000 capital TODAY?"

Note: Enron collapse wiped out $74 billion in market value. Millions of investors exhibited sunk cost fallacy—holding and averaging down to "get back to even" instead of exiting when red flags emerged.

Quantified Decision Rules (Defaults, not prescriptions)

These are starting points to counter measurable sunk cost fallacy in portfolio decisions. Adjust for your investment process, but maintain the discipline of forward-looking analysis.

Fresh Eyes Buy Test (Sunk Cost Neutralizer)

Before holding any position, ask: "If I had cash equal to position value, would I BUY this stock TODAY?"

Rationale: Every day you hold a stock is a fresh decision to allocate capital to that stock. If you wouldn't buy it today with fresh cash, you shouldn't hold it with "stale" cash. Cost basis is irrelevant.

Professional-grade upgrade:

• Monthly review (set calendar reminder): For each position, write:
  • "Would I buy TODAY: Yes / No"
  • If "No" for any position → SELL immediately (regardless of cost basis, gain/loss status)
• Hide cost basis during review (cover cost basis column in portfolio tracker) to prevent sunk cost influence
• Document reasoning: If "Yes," write 1-sentence forward-looking thesis. If "No," sell and document why (prevents post-hoc rationalization).

Interpretation:

• Healthy: Can confidently answer "Yes, I'd buy at current price" for all holdings (forward-looking, no sunk cost influence—holding because of future prospects, not past costs)
• Warning: Answer is "Maybe" or "I'd rather own something else but don't want to realize loss" (sunk cost creeping in—past loss influencing decision)
• Critical: Answer is "No, but I'm waiting to get back to even" or "No, but I've already lost so much" (pure sunk cost fallacy—decision driven by past cost, not future prospects)

Example: You hold Tesla at current price $200, down from cost basis $300. Fresh eyes test: "If I had $20,000 cash, would I buy Tesla at $200?" If answer is "No, I'd rather buy S&P 500," then SELL Tesla—regardless of $300 cost basis.

Cost Basis Blindness Exercise (Mechanical Practice)

Hide cost basis when evaluating positions—analyze stock based ONLY on current price and future prospects.

Rationale: Cost basis is psychologically powerful but economically irrelevant to forward decision. Hiding it during analysis forces forward-looking thinking.

Professional-grade upgrade:

• Practice (quarterly):
  • Cover cost basis column in portfolio tracker (use Post-It note or Excel hide column)
  • Analyze each position without seeing purchase price
  • Make hold/sell decision for each
  • Then reveal cost basis—did it change your decision? If yes, sunk cost is influencing you
• Portfolio aggregator modification: Use tools (Personal Capital, etc.) that let you hide cost basis view. Set as default view.
• Accountability: Share analysis with investment partner before revealing cost basis (prevents unconscious peeking)

Interpretation:

• Healthy: Position decisions based on forward-looking analysis (valuation, thesis strength, opportunity cost vs. alternatives)—cost basis revealed but doesn't change decision
• Warning: Mention cost basis during analysis ("I'm down 30%") but try to separate it from decision (partial sunk cost influence—aware of it, fighting it)
• Critical: Decision explicitly driven by cost basis ("Can't sell until I get back to even," "Need to average down to lower break-even point," "I've already lost 50%, what's another 10%?")—pure sunk cost fallacy

Practical note: If revealing cost basis changes your decision, that's diagnostic—sunk cost is influencing you. The goal is cost basis irrelevance to hold/sell choice.

Averaging Down Pre-Mortem (Escalation Circuit Breaker)

Before averaging down (buying more of losing position), answer: "If this stock IPO'd TODAY at current price, would I buy it?" If No → don't average down.

Rationale: Averaging down is rational when thesis is intact AND current price offers compelling value (forward-looking). It's sunk cost fallacy when driven by "lowering break-even point" or "recovering initial investment."

Professional-grade upgrade:

• Before adding to any losing position, write:
  • "Thesis status: Intact (fundamentals solid, temporary macro headwind) / Deteriorated (competitive threats emerging) / Broken (core assumption invalidated)"
  • "Would buy as new position TODAY: Yes (compelling 15%+ expected return at current price) / No (better opportunities elsewhere)"
  • "Why current price is attractive (forward-looking): Trading at 12x P/E vs historical 18x, margins improving, new product cycle starting"
• Only proceed if: Thesis = Intact AND Would buy = Yes AND Can articulate forward-looking thesis (not "to lower cost basis")
• Red flag phrases (if you hear yourself say these, STOP):
  • "I'll average down to get back to even faster"
  • "I'll lower my break-even to $X"
  • "I've already lost so much, might as well double down"
  • "I need to recover my initial investment"

Interpretation:

• Healthy: Averaging down only when thesis is intact AND current price offers compelling value (forward-looking analysis—adding capital because opportunity is attractive, not to recover sunk costs)
• Warning: Averaging down to "lower break-even point" or "recover faster"—mild sunk cost influence, but thesis is still evaluated and articulated
• Critical: Averaging down explicitly to "recover initial investment" or "get back to even faster" or "I've already lost X%, what's a bit more?"—pure sunk cost fallacy, escalation of commitment, throwing good money after bad

Example: Stock down -40% from your cost basis. Pre-mortem: "Thesis status: Deteriorated (competition intensifying, margins compressing). Would buy as new position: No (I'd rather own index). Action: Do NOT average down—selling more likely correct."

Counter-example: Stock down -30% from cost basis. Pre-mortem: "Thesis status: Intact (temporary macro headwind, fundamentals solid). Would buy as new position: Yes (compelling value at 12x P/E vs historical 18x). Action: Averaging down is rational—buying more because it's attractive, not to recover sunk cost."

Mitigation Checklist (tiered)

Essential (high ROI on forward-looking decisions)

• Monthly Fresh Eyes Test: For each position, ask "Would I buy TODAY?" — if No, sell regardless of cost basis
• Hide cost basis during reviews: Cover cost basis column when analyzing positions (prevents sunk cost anchor)
• Pre-commitment rule: Write "I will NOT hold positions to 'get back to even'—cost basis is irrelevant to future prospects"
• Averaging down pre-mortem: Before adding to loser, confirm thesis intact AND would buy as new position

High-impact (structural forward-looking discipline)

• Quarterly portfolio audit: Review all positions with cost basis hidden—make hold/sell decisions, then reveal cost basis to check for sunk cost influence
• Exit criteria pre-commitment: Before buying any stock, write "I will sell if revenue growth turns negative for two consecutive quarters, or if primary competitor gains 5%+ market share" (not "I will sell when I get back to even")
• Opportunity cost comparison: For each loser, calculate "If I sold and bought S&P 500, what's expected return difference over next year?"
• Accountability partner: Share hold/sell decisions with someone before acting (reduces self-justification bias)

Optional (advanced sunk cost resistance)

• Tax-loss harvesting discipline: Automatically sell losers in taxable accounts for tax benefit (forces exit, prevents sunk cost paralysis)
• Portfolio rebalancing automation: Auto-rebalance quarterly based on target allocation (removes discretionary sunk-cost-driven decisions)
• Investment policy statement: Pre-commit to decision rules in writing when emotionally neutral (refer to IPS during sunk cost temptation)

Detection Signals (how you know sunk cost fallacy is affecting you)

• You use phrases like "I'll hold until I get back to even" or "I need to recover my investment" (pure sunk cost language)
• You can't articulate forward-looking thesis for positions you hold, only "I paid $X and it's now $Y" (backward-looking framing)
• You're averaging down on positions where thesis has deteriorated (escalation of commitment to justify initial decision)
• You refuse to sell losing positions but would sell winning positions easily (disposition effect + sunk cost fallacy)
• You feel more pain from realizing $5,000 loss (selling loser) than from holding position down $10,000 (unrealized loss feels less real)
• You check cost basis before making hold/sell decisions (letting past cost influence future choice)

Measurement Framework (make it measurable)

Fresh Eyes Buy Test Pass Rate

Method: Monthly review—count positions where you answer "Yes, I'd buy TODAY" vs. "No, but holding to recover."

Formula: (Positions passing fresh eyes test) / (Total positions)

Interpretation:

• Healthy: 100% pass rate (all holdings pass "would buy TODAY" test—no sunk cost influence, purely forward-looking)
• Warning: 80-95% pass rate (1-2 positions held due to sunk cost—mild influence, addressable)
• Critical: <80% pass rate or any position explicitly held to "get back to even" (sunk cost fallacy dominating decisions)

Practical note: If any position fails fresh eyes test ("No, wouldn't buy today"), sell immediately. Every day holding it is a decision to buy at current price—act accordingly.

Cost Basis Influence Score

Method: Review last 10 sell decisions. Count how many were influenced by cost basis (sold because "got back to even" vs. "thesis deteriorated").

Formula: (Sells driven by thesis/opportunity cost) / (Total sells)

Interpretation:

• Healthy: >=90% (9-10 of 10 sells driven by forward-looking analysis—cost basis irrelevant)
• Warning: 70-89% (2-3 of 10 sells driven by "got back to even" or "hit stop loss tied to cost basis"—mild sunk cost influence)
• Critical: <70% or frequent phrases like "sold when I got back to even" (sunk cost fallacy routinely influencing exits)

Example: Review your last 10 sells. If 3 were "sold because it finally got back to my cost basis," sunk cost score = 70% (warning—letting past cost drive 30% of sell decisions).

Averaging Down Justification Quality

Method: Review averaging down decisions over past year. For each, check: Was thesis intact? Would you buy as new position?

Formula: (Rational averaging down) / (Total averaging down instances)

Interpretation:

• Healthy: 100% (all averaging down based on intact thesis + forward value, not sunk cost recovery)
• Warning: 50-80% (some averaging down driven by "lower break-even" logic—mild escalation tendency)
• Critical: <50% or any instance of "averaging down to get back to even faster" (sunk cost fallacy driving capital allocation)

When Holding Losers Might Be Acceptable (the nuance)

Sunk cost fallacy explains most poor hold decisions, but not all holding of losing positions is irrational. Holding losers can be acceptable when:

Legitimate reasons:

• Thesis intact + Temporary mispricing: Stock down due to macro/sentiment, but fundamentals unchanged AND you'd buy more with fresh cash (not just holding because you paid more)—forward-looking conviction, not sunk cost
• Tax-loss harvesting timing: Holding loser in taxable account until late December to harvest loss (tax planning), but still would sell even if capital gain (tax consideration, not sunk cost)
• Concentrated position with capital gains tax: Selling would trigger large tax bill, economically rational to hold despite mild thesis deterioration (tax cost vs. opportunity cost trade-off—quantified, not emotional)

The test: Can you articulate why holding is optimal based ONLY on future prospects and opportunity cost—without referencing past purchase price?

If your answer is "I paid $350 and need to get back to even," that's sunk cost. If your answer is "Thesis is intact, current price offers 15%+ expected return, better than alternatives, tax cost of selling is $10K which exceeds $5K opportunity cost of holding," that's forward-looking (defensible).

Case Studies (Sunk Cost Fallacy at Scale)

AT&T + Time Warner Acquisition (2018-2022—Corporate Sunk Cost Fallacy)

Context: AT&T acquired Time Warner for $85 billion (2018), betting on vertical integration (telecom + content). Strategy: Own pipes (AT&T network) AND content (HBO, CNN, Warner Bros.) to compete with Netflix/Disney+.

Thesis Deterioration (2018-2021):

• Streaming competition intensified: Netflix, Disney+, Amazon Prime all spending billions on content
• Debt burden: $85B acquisition loaded AT&T with debt, limiting flexibility
• Synergies failed: Telecom + content integration didn't create value
• HBO Max losses: Streaming required massive cash burn to compete

Sunk Cost Fallacy Manifestation:

• 2019-2021: AT&T held Time Warner assets despite deteriorating thesis
• Management rhetoric: "We're committed to the strategy" (self-justification—can't admit $85B mistake)
• Escalation: Invested billions more into HBO Max streaming to justify acquisition (throwing good money after bad)
• Backward-looking framing: "We paid $85B, can't admit mistake and sell at a loss"

Forward-Looking Decision (Ignored 2019-2021):

• Correct question: "Is owning Time Warner the best use of AT&T's capital TODAY?" (not "we paid $85B, have to make it work")
• Thesis evaluation:
  • Negative: Content business facing fierce competition, high cash burn, uncertain profitability, debt burden limiting telecom investments
  • Alternative: Sell Time Warner, focus on core 5G telecom (growth opportunity), reduce debt
• Expected return: Time Warner drag on AT&T stock, better to exit and focus core business

Outcome (2022):

• AT&T finally spun off WarnerMedia (merged with Discovery), valuing WarnerMedia at ~$43 billion
• Implied loss: Acquired for $85B (2018), spun off at $43B (2022) = -50% value destruction
• AT&T took $67 billion impairment charge (admitting $67B value destroyed)
• Sunk cost cost: Delaying spin-off by 3-4 years (2018-2022) destroyed additional shareholder value:
  • Debt interest: ~$3B annually on acquisition debt
  • Opportunity cost: Capital tied up in failing strategy instead of 5G investment

Quantified Impact:

• AT&T shareholders: Stock -50% (2018-2022) while S&P 500 +60% over same period
• Relative underperformance: 110 percentage points
• Sunk cost fallacy (holding Time Warner 2018-2022 to avoid "locking in loss") cost shareholders ~$100 billion in market cap destruction vs. exiting sooner

The lesson: Sunk cost fallacy at corporate scale—$85B acquisition failing, but admitting mistake (selling at loss) felt harder than continuing doomed strategy. Delaying sell decision to avoid "realizing loss" destroyed more value than immediately exiting would have. Same pattern as retail investors holding losers: backward-looking ("we paid $85B") dominated forward-looking ("is this our best use of capital today?").

Source: AT&T investor presentations, WarnerMedia spin-off filings, financial press coverage 2018-2022.

Retail Investor Sunk Cost (Frazzini 2006—Market-Wide Evidence)

Context: Academic study analyzed investor trading patterns across thousands of stocks and millions of trades. Tested whether investors hold losing stocks (large unrealized losses) longer than winning stocks, and whether this creates market inefficiency.

Manifestation:

• Stocks with large unrealized losses (investors down -30% to -60%) were held at higher rates than stocks with gains
• Investors resistant to selling losers (sunk cost fallacy + disposition effect)—"waiting to get back to even"
• When negative news hit losing stocks, investors held instead of selling—underreaction to bad news (refused to "lock in loss")

Market Impact:

• Stocks held by sunk-cost-driven investors (large unrealized losses) were slow to incorporate negative information
• Continued underperformance: Stocks with large unrealized losses underperformed by 5-10% over the next year as investors resisted selling despite deteriorating fundamentals
• Efficient market delayed: Sunk-cost sellers absent from market, preventing price discovery

Quantified Impact:

• Investors holding losers (driven by sunk cost fallacy) underperformed market by 5-10% annually by refusing to sell deteriorating positions
• On $100,000 portfolio, 5-10% drag = $5,000-$10,000 annual opportunity cost from sunk cost behavior
• Over 10 years: ~$80,000-$150,000 foregone wealth from systematic sunk cost fallacy

The lesson: Sunk cost fallacy is measurable across millions of investors and thousands of stocks. Holding losers too long (to avoid "realizing loss") creates systematic underperformance—stocks you hold despite deteriorating thesis underperform stocks you'd buy with fresh eyes. The aggregate impact is massive: sunk cost behavior creates market inefficiency (delayed price discovery) and individual underperformance (opportunity cost of capital tied up in losers).

Source: Frazzini, A. (2006). The Disposition Effect and Underreaction to News. The Journal of Finance, 61(4), pp. 2017-2046.

Common Rationalizations and Reality Checks

"I can't sell now—that would lock in the loss"

Reality: The loss is already real in your portfolio value (net worth is down). Selling doesn't "make it real"—it just converts unrealized loss (invisible) into realized loss (visible). The economic reality is identical; only the psychological framing changes.

Counter: The stock market doesn't know or care what you paid. The loss exists whether you sell or hold. The question is: "Is holding the best use of my capital going forward?" If No → sell. The past is gone; only the future matters.

"I'll just wait until it gets back to even, then I'll sell"

Reality: "Getting back to even" is a psychological milestone with zero economic meaning. The stock doesn't owe you recovery to your cost basis. If thesis has deteriorated, waiting for recovery to "$X" is waiting for randomness, not investing.

Counter: The stock's future path is determined by fundamentals (earnings, growth, competition)—not your cost basis. If thesis is broken, "getting back to even" might take years, decades, or never. Opportunity cost: capital tied up in loser instead of deployed in better opportunities.

"I've already lost 50%—what's another 10%?"

Reality: This is escalation of commitment—past losses making you more willing to accept future losses (reversed risk aversion). If you wouldn't buy the stock today, you shouldn't hold it—regardless of past losses.

Counter: Every day you hold is a fresh decision. If stock is down -50% and thesis has deteriorated, the next -10% is just as costly as the first -50%—it's your capital being destroyed. Sunk costs are irrelevant; protect remaining capital by exiting.

"Averaging down shows conviction—I'm being contrarian"

Reality: Averaging down is rational when thesis is intact and price is attractive (forward-looking). It's sunk cost fallacy when driven by "lowering break-even" or "recovering initial investment." Distinguish conviction from stubbornness.

Counter: Before averaging down, answer: "If this were IPO'd today at current price, would I buy it?" If No → don't average down. Conviction requires forward-looking thesis, not just "I was wrong at $70, so I'll double down at $50."

Next Step (educational exercise)

Run Fresh Eyes Portfolio Review right now (takes 10 minutes):

1. List all current positions (stocks, ETFs, funds)
2. Hide cost basis (cover column, don't look at purchase price)
3. For each position, answer:
   • "If I had cash equal to position value, would I BUY this stock TODAY?"
   • Write: Yes / No
4. Reveal cost basis—did it change your answer? (diagnostic: if seeing cost basis changes decision, sunk cost is influencing you)
5. For each "No": Write 1-sentence reason why you wouldn't buy today (thesis deteriorated, better opportunities, overvalued, etc.)
6. Action: SELL all positions where answer is "No"—regardless of cost basis, gain/loss status

Interpretation:

• All "Yes": Healthy—portfolio driven by forward-looking conviction, not sunk costs
• Any "No, but I'm holding to get back to even": Sunk cost fallacy—sell immediately
• Cost basis changed your answer: Sunk cost influencing decisions—practice cost basis blindness monthly until you can ignore it

Action item: If you have any position failing fresh eyes test, sell it this week. Set monthly calendar reminder to repeat fresh eyes review. Sunk cost fallacy costs 5-10% annually—eliminating it is highest-ROI portfolio improvement.

Related Articles

• Disposition Effect and Taxable Accounts
• Anchoring on Purchase Price Mistakes
• Loss Aversion and How to Counter It
• Mental Accounting in Household Portfolios

---

References

Arkes, H. R., & Blumer, C. (1985). The Psychology of Sunk Cost. Organizational Behavior and Human Decision Processes, 35(1), 124-140. (Sunk cost fallacy is the tendency to continue an investment because of previously invested resources, even when continuing is no longer optimal—past costs are irrelevant to future decisions but psychologically influence choices)

Frazzini, A. (2006). The Disposition Effect and Underreaction to News. The Journal of Finance, 61(4), 2017-2046. (Investors hold losing stocks too long, causing underreaction to negative news—stocks with large unrealized losses continue underperforming by 5-10% over the next year as investors resist selling)

Heath, C. (1995). Escalation and De-escalation of Commitment in Response to Sunk Costs: The Role of Budgeting in Mental Accounting. Organizational Behavior and Human Decision Processes, 62(1), 38-54. (Mental accounting amplifies sunk cost fallacy—investors frame stock as 'investment to recover' rather than 'capital allocation decision,' causing escalation of commitment)

Staw, B. M. (1976). Knee-Deep in the Big Muddy: A Study of Escalating Commitment to a Chosen Course of Action. Organizational Behavior and Human Performance, 16(1), 27-44. (When initial decision fails, decision-makers often increase investment to justify initial decision, driven by self-justification and avoiding admitting mistakes)