Anchoring
Anchoring to a reference price (e.g. your cost basis).
Anchoring (or the anchoring effect) is a cognitive bias in which people give excessive weight to the first available reference value — the anchor — in subsequent decisions. In investing, the most common anchor is the purchase price: once you have paid $300 for a stock, every subsequent assessment tends to be measured against that number, even though it is irrelevant to the stock's future value.
The bias is particularly strong with round numbers and visual references: the all-time high, the IPO price, an analyst's price target, the value at which you intended to sell. The anchor does not need to be rationally relevant to have an effect: experimental studies show that even random numbers inserted before an estimate influence the estimate itself.
Worked example
Studies of retail investor behavior (Shefrin & Statman, 1985; Odean, 1998) show that losing positions are held in portfolios on average 2–3 times longer than winning ones. The anchor is the cost basis: the investor waits 'to break even' before selling, even when the stock has no fundamental reason to recover.
The price you paid is irrelevant to the market. If you bought Tesla at $400 and the price is now $200, the market does not know you are down 50%. The rational question is not 'when will it get back to $400?' but 'at $200, is Tesla worth more or less than it costs to buy it today?'. Separating cost basis from valuation is the first step to avoiding anchored decisions.
When it's used
Recognizing anchoring is useful in two specific contexts. First, in managing existing positions: identifying whether you are holding a stock because it still has value or because you are unwilling to crystallize a loss relative to your cost basis. Second, in interpreting analyst price targets: if an analyst's target has remained unchanged while the price has fallen 30%, the analyst may be anchored to their previous target rather than having updated it on the basis of new data.
Limits
Anchoring is not always irrational: in some cases the historical price is a reasonable proxy for value (e.g. a company trading below book value for temporary reasons). The problem arises when the anchor is used as a shortcut rather than an updated valuation. Eliminating it completely is impossible — reducing it requires process discipline: reassessing every position starting from current fundamentals, not from P&L.
Frequently asked
Does anchoring affect analyst price targets too?
Yes. Academic studies show that price targets are updated incompletely relative to new information — they tend to move less than they should after important events. Target revisions are often an adjustment around the previous target, not a neutral reset.
How do you avoid anchoring bias?
The most effective technique is the 'blank slate evaluation': before reviewing a position, close the P&L display and ask yourself 'if I were building this portfolio from scratch today, would I include this stock at the current price?'. If the answer is no, the anchor is driving your decision.
Are anchoring and loss aversion the same bias?
Related but distinct. Anchoring is an estimation bias: you give too much weight to a reference point. Loss aversion is a valuation bias: losses weigh more than gains. In practice they reinforce each other: you are anchored to the cost basis (anchoring) and avoiding selling to prevent crystallizing a loss (loss aversion).
Related terms
Educational definition. Not financial advice.