- View SourceHi, all members and visitors!
I skirted making a letter last month, being away on holiday
Here is the 42nd month newsletter, I advanced it by one day,
seeing that the group is really quiet, what makes you so
pensive ? ;-))
1) Monthly stats
The membership was active again in the last 2 months with
ca 100 new members. Slower activity on the messages side,
140 in two months, but some on quite fundamental topics
Now 1250+ members, 5800+ messages.
Our sister list, Finance-academy picked up some speed
reaching the 170+ members area.
2) Further projects
As I already said, and was comforted by the polls results,
we could think about how to improve our "club-like" vocation
for subgroups of members having similar
* topics of interest (specific fields of research, of
* professional issues,
* geographical proximity
* or whatever.
One way could be to start a "webring" of related sites and fora,
some already existing and others to be created along those
This is not too difficult to do technically, but I didn't
advance much on this idea, not seing exactly what could be
useful for members. Any thoughts you could express would be
appreciated by the group, I'm sure.
Same thing for any proposal for new resources to add to the
group's files or the FAQ (see below section 3). A member
took contact and brought an idea that seems feasible and that
he is working on. If it can be launched, a message will
3) Group's resources, and their access
3a) Accessibility of resources
As attachments to messages are not kept anymore in yahoo
archives, the "file" section of the group gets more importance
But for access to polls, links, polls, files, you need a
*yahoo ID / profile*. To get one, click "account info" (on top
of the group' homepage or your "my groups" page)
Of course, you dont need to fill in all information yahoo asks,
but no need to be too shy either.
3b) Your free BF resources: a short list
(regularly updated, and at your fingertips 24/7):
* Your "BF gallery" / FAQ with 300+ BF keywords.
Some definitions are short but I hope sufficient, others are
written as small articles. Completion and update are done
regularly, usually thanks to the group's debates.
Any suggestion to enhance or enter definitions is welcome.
* Your BF Group's archives:
(and then use "search")
* Your BF Bookmarks
* Your behavioral stockpricer
Wishing the best to all, and another month of fruitful info
- View SourceWell, not too much messages in here these days...
I did not participate either, due to some professional constraints: PhD in
I'm working on the detection of cognnitive biases, especially after earnings
annoucements. I know, it doesn't seem original, but what I'm trying to do is
to classify the recent annoucement, according to a list of the most
important (individual) cognitive biases: cognitive dissonance, availability
bias, hindsight bias, anchoring, overconfidence and representativeness
This classification suggests that depending on the past earnings surprises,
the recent surprise may be interpreted differently, thus leading to
different biases and several post-anouncement drifts.
For example, after a series of bad surprises, a recent surprise is
underestimated and under-reaction appears. Each bias has it's own definition
and is currently tested. My aim is to find out if it is possible to identify
special reactions (through CARs and/or BHARs) for each bias using the same
data set and methodology.
I'm currently facing the problem of bias superposition. After a recent null
surprise, average CAR are not significantly different from zero; as
expected. But if I take into account the prior surprise, it appears that an
important prior surprise is followed 1 quarter later by CARs of the opposite
sign; correction of the prior over-reaction I guess. Does anybody know how
to empirically remove the effect of this correction ? For some biases, I
have to condition on prior information and thus might have corrections in
the same direction "biasing my results". (if I don't condition on the past
surprise, I may think that these corrections cancel each other out, leaving
only the effect of recent bias...)
I'd be thankful for any hint or reference.
Have a nice week-end