Anchoring is when you rely on the first piece of information you receive in a situation, even if the information is irrelevant. Instead of interpreting new information objectively, you use a reference point as an anchor, blurring you from sufficiently updating the situation.
Anchoring was first introduced by Daniel Kahneman and Amos Tversky who made a bunch of experiments. In one of them, psychologists asked volunteers to guess the % of African nations in the United Nations. Before guessing, they had to spin a roulette wheel which, they were told beforehand, had nothing to do with the question but was predetermined to land on either the number 10 or 65. The volunteers whose wheel stopped on 10 guessed an average of 25%, while the ones who landed on 65 guessed an average of 45%.
In another study, Kahneman and Tversky asked another group of volunteers to quickly compute the product of the numbers one through eight, either as 1 x 2 x 3 x 4 x 5 x 6 x 7 x 8 or in reverse as 8 x 7 x 6 x 5 x 4 x 3 x 2 x 1. The volunteers who got the sequence starting with 1 gave a median estimate of 512. The volunteers who got the reverse gave a median estimate of 2,250. The correct answer was 40,320.
In both experiments, the volunteers anchored their minds to the first information they received. In the UN question, their response was affected by an irrelevant number on a roulette wheel. In the product question, they anchored their minds to the smaller numbers appearing first in the sequence.
Anchoring is rooted in human cognition and its causes are still debated. One certain thing is that it’s related to confirmation bias. Confirmation bias is a form of anchoring. Once we’ve established a point of reference about a situation, we tend to focus on the stuff that’s consistent with our established belief.
We do this because of two human traits: 1) we love making hypotheses to make sense of the world and 2) we use association to understand what we see. In hypothesis testing, you start by evaluating that your hypothesis is correct. Assuming it’s not, you move on to another hypothesis, but not before accessing all the relevant attributes of the anchor itself. So when you evaluate a new hypothesis, you automatically look for ways in which it’s similar to your original hypothesis, the anchor.
Anchoring pops up in day-to-day life in subtle ways. Marketers use it to trick you into spending more money, negotiators use it to get in a stronger position, and stock investors fall prey to it to their detriment. Here are a bunch of such examples:
- When a marketer sets a $9.99 price tag instead of $10, they anchor your mind the $9 rather than the $10.
- When a marketer places a “limit 10 per customer” sign above a shiny product, they want you to anchor to the number 10 so you buy more than you otherwise would have if the sign said “no limit”.
- When a salesman first shows you their most expensive product and then a less expensive one, they want you to think of the less expensive one as the inexpensive one.
- When you meet someone new, you anchor your opinion of them to your first impression.
- When the first price is offered in a negotiation, the anchor is usually set, and the consecutive counterbids go from there.
- When a restaurant prices some of the items on its menu at expensive levels, it makes its mid-priced meals look inexpensive.
- When faced with a larger meal portion, you tend to eat more.
- When a stock splits 5:1, it looks cheaper at, for instance, $20 than $100.
- When a stock reaches a 52-week low, it looks cheaper than if it hit an ATH at the same price.
As for the latter two, in financial markets, anchoring can be pervasive, powerful, and dangerous. Holding on to a bad stock because you anchor to the price you originally bought it at rather than what the current fundamentals tell you is an example. But there are also more subtle ways anchoring can creep into your process: if your valuation is dependent on past sales volume or a certain past commodity price, even if the business model has evolved or the world has changed, you may use the wrong anchor.
In another case, an economist, whose job involves predicting the stock market over the next 12-month period, often anchors to the given absolute level of the index. For instance, if we’re in a 7-year bull market and the index is at 4,500, the economist is biased toward predicting values close to this value instead of considering standard deviations which may be fairly large. And they anchor not only on the price but also on the sentiment. It’s difficult to imagine a change in sentiment.
It reminds me of this quote by Nassim Taleb in Fooled by Randomness:
Anchoring to a number is the reason people do not react to their total wealth, but rather to differences of wealth from whatever number they are currently anchored to. This is in major conflict with economic theory, as according to economists, someone with $1 million in the bank would be more satisfied than if he had $500 thousand but this is not necessarily the case.
Anchoring is usually put in a negative light, as a bias, but there are many areas of life where the mental process of anchoring can be useful. For instance, if you want to estimate the orbit of Mars, you could start with the orbit of the Earth (365 days), then know that Mars is about half the size of Earth, and adjust the days upward from there (the correct answer is 687 days). Or, now knowing about it, you can use it to your advantage in a negotiation by setting the higher price first to anchor the mind of the counterpart. But it’s not only numerical: some people can consciously change their emotional state in an instant by creating a positive anchor. An athlete may use a visual image of shooting the perfect basket to get back in “the zone”, and a professional speaker may use a fixed routine to get in the right mindset before entering a stage.
You don’t want to get rid of your mind’s tendency/ability to anchor but you want to recognize when anchoring is going to be dangerous and when it’s going to be beneficial to your goals. To use anchoring to your benefit, practice self-awareness.
Anchoring is part of my collection of mental models.