OptiNod Academy
Anchoring and averaging down — Fixating on entry price pushes a valid stop away
Delayed stops and averaging down often come from treating your entry price as the market's reference. Three figures in trade records reveal the cost.
Buying more of a losing position to lower its average cost makes your entry price, rather than the market, the reference point. The market does not know that price, but attachment to it pushes a tested stop farther away.
Anchoring is a judgment bias identified by Amos Tversky and Daniel Kahneman in 1974. People take an initial number as a reference and struggle to move away from it. The sunk-cost fallacy, described by Hal Arkes and Catherine Blumer in 1985, is the tendency to continue an activity because money has already been spent, even when its future expectancy is negative. In trading, the two meet at the same point: your entry price becomes the anchor, and the money put into the position becomes a sunk cost.
Averaging down is often called “buying a bargain” or “managing the average price.” Buy more below your original entry and the average falls, so a rebound reaches break-even sooner. The arithmetic of the average is correct. The problem is that this is rarely a staged buying plan made in advance; it is often a rushed reaction after a loss, intended to preserve the hope of break-even.
The cost is invisible in the lower average price. The distance to break-even shrinks, but position size and total loss risk grow. Suppose a trade planned a 5% stop, skipped it, added an equal-sized buy, and kept falling. Combined losses can become many times the planned 1R. A lower average feels reassuring while much larger risk sits inside the same position.
A tested stop comes from market structure and volatility; your purchase price was never an input. Once profit and loss are judged relative to your entry, however, the stop feels like “the place where the loss becomes real,” so it is avoided. The planned exit fails to occur. Three numbers in the trade record measure how far this habit pushes stops away.
Your entry price is a reference the market does not know
Anchoring fixes judgment to a reference point. People experience gain or loss relative to one, and in trading it is usually the price paid. At a market price of $90,000, a buyer at $95,000 sees a $5,000 loss per unit; a buyer at $100,000 sees a $10,000 loss. The market price is the same, but each entry makes the loss feel different.
The market does not know where you bought. Support, resistance, prior swing highs and lows, and volatility are visible to everyone; your entry price is personal. Price moves through market structure regardless of it. Using your entry as the reference for stops and exits means making decisions around a level the market itself does not care about.
This connects to the value function of prospect theory and loss aversion. It treats gains above and losses below a reference point differently, with more risk seeking in the loss region. If entry price becomes the reference, all prices below it fall in that psychological loss region. Averaging down is a characteristic action there.
Sunk costs make an exit feel like creating a loss
The sunk-cost fallacy brings money already spent into a current decision. A rational decision asks only about expectancy from this point forward. An unrealized loss has already occurred economically; the remaining question is whether the position's future expectancy is positive. Sunk-cost thinking drags the amount already lost into that calculation and makes closing feel like causing the loss.
The mark-to-market loss is in the account whether or not the stop is executed. Stopping does not create a new loss; it prevents the loss from growing beyond a level defined by the rule. Under the sunk-cost fallacy, it feels reversed: clicking the stop seems to create the loss. The trader avoids it and waits for price to return to break-even.
Averaging down emerges when anchoring and sunk costs combine. Attachment to break-even and reluctance to realize a loss make another buy feel more attractive than an exit. A lower average looks closer to recovery, while putting in more money makes abandoning the position emotionally harder.
Averaging down lowers the average while increasing total R at risk
Consider a decline in early 2026. Bitcoin was around $95,000 in mid-January and reached $97,924 on January 14. Suppose a trader bought at $95,000 with a stop near $90,000, about 5% below. Following the plan would exit there for −1R. Instead, at the stop level the trader holds and buys the same quantity again near $88,000 on January 20, whose low was $87,896. The average entry falls to $91,500 and break-even looks closer, but the position doubles and the stop rule is gone.
The decline continued. Bitcoin closed at $78,741 on January 31 and reached $60,000 by February 6. If both purchases were sold near $62,000, the first lost about 34.7% and the second about 29.5%. Define 1R as the first purchase's $5,000 planned stop distance: the combined loss is roughly 12R. A trade that would have ended at −1R near $90,000 became about −12R after one unplanned addition. The lower average did nothing to reduce risk. It fell from $95,000 to $91,500 while total potential loss grew from one R to more than ten. Averaging down reduces the distance to break-even but increases the total R at risk. This is the structure discussed in systematic staged buying.
The usual plan afterward is “sell when I get back to even.” The new $91,500 average becomes the target, and the gap below it seems temporary. Yet recovering from $62,000 to $91,500 requires about a 47.6% rise. The increase needed to recover a decline is larger than the decline itself, as explained in drawdown recovery math. Increasing a losing position also resembles Martingale: grow the bet after losses in hopes that one rebound repairs them all. One sustained decline can threaten the account.
Three figures measure anchoring
Calling anchoring a state of mind gives little diagnostic guidance. Compare trades with unplanned additions against the stop rules using three figures.
First, average realized R for trades with and without averaging down. Tag every unplanned addition and compare the two groups. If averaging down was supposed to avoid losses but those trades show markedly lower average R, the difference measures the habit's cost.
Second, the stop-violation rate: among losing trades, the share exited later than the planned stop. Compare the stop fixed before entry with the actual exit. A high rate means break-even hopes displaced the stop rule.
Third, tail asymmetry: the largest single loss divided by average loss or planned 1R. In an account where stops are followed, this may generally be near one or two. With anchoring, the largest loss can be several times average, and a −12R event makes that tail. One such addition can undo many properly stopped trades.
Base stops on market structure to remove the entry anchor
Anchoring begins when entry price becomes the reference. Shift the stop and exit reference from your purchase to market structure. Define the stop from structure and volatility before entry so break-even has less opportunity to intrude.
- Fix the stop before entry: Set it from structure—such as a previous swing low or invalidation level—and volatility such as ATR, and submit the order before opening. Do not improvise it afterward.
- Do not add at the stop: When the planned stop is reached, close rather than buy more to lower the average. Stop placement comes from the market, not the entry price.
- Keep staged entries within a total R budget: If multiple purchases are planned, define each price and the final stop before entering, with the combined possible loss within a 1R budget. Calculate position size for all tranches together.
- Tag averaging down: For each trade, record the planned stop, actual exit, unplanned additions and realized R.
- Review weekly: Compare mean R of added versus unadded trades, stop-violation rate and largest loss divided by planned 1R.
The objective is to make the stop based on market values rather than personal entry, so the exit assumed in the backtest also happens at the planned level live.
Two pitfalls
Equating planned staged buying with averaging down. Staged buying with a fixed total R budget, tranche prices and a final stop all defined before entry need not involve anchoring. The distinction is whether the additional purchase was part of the plan or a rushed decision after seeing an unrealized loss. Buying to defend break-even after the loss is the harmful form.
Believing a lower average means lower risk. A lower average shortens the distance to break-even but increases total position size and potential loss. It makes recovery look easier without reducing what can be lost at once. Risk is the loss on the entire position if price reaches invalidation, not the average entry figure.
Remove entry price from the stop decision
Anchoring and averaging down move the decision's reference from market structure to your personal entry. That price matters to you, not to the market. A stop based on it can be repeatedly postponed to protect break-even, leaving a few outsized losses in the record. Set stops from structure and volatility and track mean R for added trades, stop violations and tail asymmetry. “I will sell when it gets back to even” then becomes a measurable behavior to correct. The starting point is to remove your entry price from the stop rule.