Why is bayesian rigidity called that? I assume it has something to do with the mathematician but I am not making the connection. Also, what the hell are “organizational barriers” which cause structural chronic market inefficiencies. Schweser doesn’t really go into it as far as I can tell. I will be exhibiting some serious “bayesian rigidity” if I can pass this test and be finished.
A Bayesian approach means that you update your expectations as new information arrives. “Bayesian rigidity” refers to analysts (or other market participants) sticking to their prior expectations more than they should in the face of new information. Think of it as under-reaction to new information.
I think it’s a really lame term…
JoeyDVivre Wrote: ------------------------------------------------------- > I think it’s a really lame term… I’m sure we could come up with a better (snarkier) definition of Bayesian Rigidity. But I’m not feeling the inspiration tonight.
JoeyDVivre Wrote: ------------------------------------------------------- > I think it’s a really lame term… Joey: I have never heard of this before and only see it in one CFAI self-pub. Has anyone else used this term?
My buddy who works for a market neutral equities hedge fund used that term once.
I like the term. It is concise. And it simultaneously describes the problem while offering the “solution” or terms of the problem --namely that Bayesian principles can be used to avoid this rigidity / irrationality.
I prefer “Bayesian D-Bagginess”
But surely the “Bayesian” thing to do IS to update your expectations, hence being too rigid in the face of new info should be called “Unbayesian” or “Non-Bayesian” rigidity?
I have heard the term ‘sticky priors’ all the time. Is this the same?
So does anyone know what an "organizational barrier is (in the context of structural chronic inefficiencies)? Also, thanks to Busprof.
The inability to short, or capital control in a country would be considered organizational barriers that might result in chronic mispricing of assets.
DanLieb Wrote: ------------------------------------------------------- > I have heard the term ‘sticky priors’ all the > time. Is this the same? Af far as I know, yes.
Bayes’ wife used the term “Bayesian rigidity” after she finally got him to muster the courage to ask his doc for some Viagra.
Absolutely not true. Bayes was a priest and not married. In fact, he probably died as a virgin. At least he got to play around with numbers…?
busprof Wrote: ------------------------------------------------------- > A Bayesian approach means that you update your > expectations as new information arrives. > “Bayesian rigidity” refers to analysts (or other > market participants) sticking to their prior > expectations more than they should in the face of > new information. Think of it as under-reaction to > new information. Sounds the same as the heuristic driven biase due to Anchoring and Adjustment. Recall that anchoring is the natural tendency to rely too heavily on a single (first or early) trait or data and extrapolate the future based on the anchoring image of that data and not making appropriate adjustments fast enough to new information rec’d. Can someone confirm my understanding?
I’m so sad you bring this thread back to the CFA material. But yes. Your understanding is correct. It’s the same thing with a fancy word.
Baysian rigidity can result in sticky-posts, but not sticky-priors… because the rigidity must preceed the stickyness.
Additionally, it does appear that Bayes had no wife. He had a theorem but no wife… which would you prefer if they were mutually exclusive? http://www.york.ac.uk/depts/maths/histstat/bayesbiog.pdf Figure 2 on page five shows him alone on the family tree. Just a lonely math nerd/religious scholar from a wealthy cutlery family.
busprof Wrote: ------------------------------------------------------- > A Bayesian approach means that you update your > expectations as new information arrives. > “Bayesian rigidity” refers to analysts (or other > market participants) sticking to their prior > expectations more than they should in the face of > new information. Think of it as under-reaction to > new information. I think you are right. But I am still confused on the difference between Bayesian Rigidity and anchoring trap. Sounds the same thing.