The Sunk Cost Fallacy

You're 40 minutes into a genuinely bad movie. You stay — "already paid for the ticket." Notice the logic's shape: the ticket money is gone whether you stay or leave; staying only adds two wasted hours to a loss that's already complete. Decisions should weigh futures — but the brain insists that past spending must be "honored" with continued commitment. That's the sunk cost fallacy, and its trading translation is the single most account-destroying behavior in retail markets: averaging down to redeem a loss.

Anatomy of the spiral, ingredient by ingredient (you've met every part): a position falls → selling means the loss becomes real and "wasted" (loss aversion, Chapter 2, plus the verdict-avoidance of Chapter 3) → buying more "reduces my average" (anchoring, Chapter 6 — the entry price steering again) → and the invested money, effort, and public conviction demand honoring (sunk cost, the binding agent). The result feels like strategy — "I'm improving my position" — and is arithmetically something else: increasing your size in the trade that is currently proving you wrong. Every average-down inverts sound position management: winners get no adds; losers get all of them; maximum size is reached at maximum wrongness. (Recall Druckenmiller's flat prohibition in your Legendary Traders school — never average a loser — and Kedia's decade of losses before method: much of that decade, by his own telling, was this spiral.)

The crucial distinction that keeps this honest — because adding lower isn't always wrong: a planned scale-in (thesis intact, adds pre-scripted at pre-set levels, total size pre-capped, written before entry) is a strategy. Averaging down (unplanned, triggered by pain, justified after the fall, size cap = courage remaining) is the fallacy wearing strategy's clothes. The test is one word: when was the add decided? Before the pain, or by it? Your QbarTrade plan field is literally the referee here — if the add isn't in the original plan, it isn't a scale-in; it's a sunk-cost purchase.

Beyond averaging, sunk cost runs quieter rackets: holding a dead stock for years "because I've already waited so long" (time is a sunk cost too); refusing to abandon a strategy you spent months building even as its live results diverge from backtest (research effort as sunk cost — Asness's discipline in your Legendary Traders school is the honest version: re-examine the evidence, not the investment of pride); staying in a losing approach because you've publicly defended it (reputation as sunk cost, compounding Chapter 4's lawyer).

Defense — make the past structurally voteless. (1) The zero-based question (this school's master key, third appearance): "with fresh money, no history, would I open this exact position at this price, this size, today?" Sunk cost cannot survive that sentence — it has no answer to it. (2) Adds require pre-registration: any scale-in must exist, with levels and caps, in the entry plan — else forbidden. (3) Kill criteria at birth: every position and every strategy gets its falsifier written on day one (Chapters 4 and 10), so abandonment is a pre-agreed event, not a live defeat. (4) Rename the loss: booked losses aren't waste — they're tuition already paid; the only genuine waste is paying tuition and refusing the lesson (Kedia again: the decade only became valuable when extracted into a framework).

Offense — trapped averagers are structure. Every average-down cohort deepens the trapped-inventory zones your Market Structure school mapped: each "improved average" becomes a new breakeven anchor, a new layer of future supply waiting overhead. Failing rallies into those layered zones — sold into by relieved averagers finally "getting out flat" — are among the most repeatable fade setups in markets. Their honored sunk costs, your resistance map.