Why we treat it as an error?
I think it’s pretty normal to have a start point and do the adjustment later.
Maybe because you have a start and end level in mind, irrespective of any other information tht may come along to influence your set views.
you start out with an anchor - and do not make any future adjustments to that anchor even if underlying fundamentals have changed. that could be one scenario - which causes an error.
the 2nd situation of error - start with a completely wrong anchor.
_
_
I like this explanation. Spot on!
Yeah, but if your starting point is really stupid, then no adjustments will help.
EG - If I ask you, “Is Mt. Everest taller or shorter than 50k feet?”, you might say, “Shorter.” (And you would be right.) Then I ask, “How tall is Mt. Everest?” You would immediately think, “Hmmm…well, it’s shorter than 50k feet, but probably not much shorter. Let’s go with 10% shorter and say 45k feet.” And you would be wrong.
Mt. Everest is actually 29k feet. But because I “anchored” you with a completely useless number, you picked something close to that, and you were completely wrong.
This actually isn’t completely hypothetical. Motley Fool has done this exact exercise on people, and most people pick a number extremely close to the “anchor”, regardless of where the anchor is. If they say 150k feet, people will guess between 130 and 170. If they say 15k feet, people say between 13 and 17.
Thanks Greenman. Great example.
I don’t think that the material states that it’s wrong to have an anchor and adjust it to take into account the available information. The reason why the anchoring and adjustment bias is an issue is because the analyst does not adjust the anchor to the extent needed. But i guess that if you have performed some quite complicated calculation and just one parameter has changed, it is probably alright to adjust your previous result to take into account this change, without redoing your whole computation if you can argue that it is still relevant (depending on the exact context obviously)