What is anchoring and adjustment?
Anchoring and adjustment is a mental shortcut (heuristic) people use when making estimates or decisions. A particular number or piece of information—the anchor—serves as the starting point. People then adjust away from that anchor to reach a final judgment. Those adjustments are frequently insufficient, so the final answer remains closer to the anchor than it should be.
Key concepts and definitions
– Anchor: any initial value or reference point (a price, forecast, previous estimate, or even a random number) that influences subsequent judgments.
– Adjustment: the change made from the anchor toward a final estimate.
– Heuristic: a simple mental rule that reduces complex reasoning to a more manageable process; useful but prone to bias.
– Conservatism bias: a different cognitive bias in which people underweight new evidence relative to prior beliefs. Anchoring differs because it specifically uses an explicit starting figure as the basis for subsequent estimates.
How anchoring works (short explanation)
When presented with an initial figure, people often take that number as informative—even if it’s irrelevant—and then make incremental corrections. Because those corrections are frequently too small, the final answer stays pulled toward the original anchor. Anchors can be self‑generated (your own prior estimate), suggested by others (a seller’s first price), or the output of a model or tool.
Why it matters (practical effects)
– Negotiations: The first monetary figure offered usually shifts the bargaining range.
– Forecasting and valuation: Analysts can become overly influenced by model outputs or prior forecasts.
– Everyday judgments: Random or irrelevant numbers can change unrelated estimates if they are presented before asking the target question
Evidence and experiments (selected findings)
– Classic studies: Tversky and Kahneman (1974) showed anchoring with arbitrary anchors. In one task participants spun a wheel of fortune that landed on a number; when later asked whether the percentage of African countries in the UN was higher or lower than that number, subsequent numeric estimates correlated with the wheel’s number—even though it was irrelevant. This demonstrated that irrelevant anchors shift judgments.
– Robustness: Hundreds of replications across domains (numeric estimation, pricing, forecasting, legal decisions) show anchoring is pervasive but variable in magnitude. The effect is larger when people lack relevant knowledge or when the anchoring number is presented as plausible or authoritative.
– Mechanisms: Two main explanations are (1) insufficient adjustment—people start at the anchor and fail to correct fully—and (2) selective accessibility—considering why the anchor might be true makes anchor‑consistent information more available, pulling estimates toward it.
– Moderators: Expertise, motivation, and time matter. Experts still anchor, but often less. When people are motivated and have time to think, anchoring effects shrink but do not disappear.
A simple worked example (negotiation)
– Scenario: A seller asks $120,000 for a car. Your independent fair‑value estimate (based on comps) is $100,000.
– Typical anchored outcome: You counter with $110,000 and settle at $115,000. Why? Your adjustment from the anchor ($120k) was only −$10k, which is smaller than the $20k difference between anchor and your independent estimate.
– Quantify the bias: Let A = anchor = 120, T = your true estimate = 100, F = final = 115.
– Adjustment magnitude = A − F = 5 (thousand)
– Remaining anchoring gap = F − T =
= Continued calculation and interpretation =
Remaining anchoring gap = F − T = 115 − 100 = 15 (thousand).
Proportion of the original anchor gap that you moved away from the anchor:
(A − F) / (A − T) = 5 / 20 = 0.25 → you adjusted only 25% of the distance from the anchor toward your independent estimate. Equivalently, 75% of the initial anchoring bias remained in the final price.
A few quick ways to report the bias
– Absolute under‑adjustment = F − T = 15 (thousand).
– Adjustment magnitude = A − F = 5 (thousand).
– Relative adjustment (fraction of gap closed) = (A − F) / (A − T) = 0.25 (25%).
– Percent move away from anchor = (A − F) / A = 5 / 120 ≈ 4.17%.
Why people under‑adjust (concise)
– Insufficient cognitive effort: adjusting requires effort; people stop early.
– Anchor salience: the first number offered is focal and easy to use.
– Confirmation and selective attention: people notice evidence that supports the anchor.
– Uncertainty: when unsure, people weight anchors more heavily.
Practical implications for traders and negotiators
– Price discovery: initial quotes (bids, offers, price targets) can pull others’ valuations toward them.
– Analyst influence: an early published target or model can bias later forecasts.
– Order placement: visible large orders or trade prints can serve as anchors for execution prices.
– Negotiation tactics: first offers systematically shape settlement ranges.
Worked example — analyst research (step‑by‑step)
Scenario: An equity analyst sees an IPO price and forms an independent DCF (discounted cash flow) fair value.
– Anchor (A) = IPO price = $60.
– Independent model fair value (T) = $48.
– Analyst issues a published target (F) = $54.
Compute bias:
– Adjustment magnitude = A − F = 60 − 54 = 6.
– Remaining anchoring gap = F − T = 54 − 48 = 6.
– Original gap = A − T = 12.
– Fraction adjusted = 6 / 12 = 0.5 → analyst moved halfway toward their model but left 50% anchored to the IPO.
How to detect anchoring in your process — checklist
– Before you see any market quotes, write down an independent estimate or range (blind estimate).
– After seeing a quote or external number, record your immediate reaction and the final decision. Compare the two.