Assymetric information and Reputation in Banking

[fgascon@hp845.econo.uniovi.es (F]

Dear Network companions,

Im currently working on a model dealing with assymetric information
and reputation in banking. Does anybody know of any model concerning
assymetric information or better than that assymetric information in
banking?


Thank you in advance



Fernando Gascon
E-mail: fgascon@hp845.econo.uniovi.es

[Segel Jonathan D <4jds@qlink.que]

I am not sure if this is what you had in mind by assymetric information,
but some of the models I have been working on recently have dealt with
assymetric information in the context of modelling market share change.

I have posted to the group in case anyone else is interested... If not
surf on!


Information and reputation were handled in my model as follows:

1) Customers of one supplier may or may not be aware of the existance of
other suppliers. The proportion of customers who are aware of the other
supplier is represented as a factor which is influenced by the amount of
advertising expenditure by the other supplier. In general awareness
only increases and converges to some limiting value. I represented
the evoked set (ie. the suppliers who are considered) for each suppliers
customers.

2) The reputation of each supplier is a composite of their reputation on
a number of attributes. Each attribute is influenced by three factors:
Customers experiences, marketing activity, and image halo effects from
other attributes.

The first two of these factors are roughly comparable
to the internal and external factors in the Bass diffusion model of
innovation. That is, customer awareness is changed both by marketing
activities, and by word of mouth.

Note that it is important to distinguish who is doing the perceiving and
who is doing the experiencing. For my purposes, looking at market share
change, what I am interested in is the repution of an alternative
supplier B from the perspective of a customer who currently uses supplier
A. I am therefore interested in what A thinks of B based on marketing
activity of B and halo effect from what else I know about B, and also
what I hear from people who are currently customers of B about their
experiences. Therefore the experience effect on reputation is a
function of not only the effectiveness of word of mouth as a
communication channel, but also how many customers B has.

3) In system dynamic models I have usually used aggregate
representations. That is, you create a store which represents aggregate
customer perceptions say. However, where there is a broad range of
individual levels within that aggregate, you have a problem. Say the
average perception is 5 out of 10 but it ranges from 2 to 9. Representing
just the average as an aggregate measure is often not adequate, because
it implies that all members have the average level.

There are a number of ways of dealing with this. I used the following:

-To account for heterogenity in customer preferences and behaviour, I
segmented customers (and the model) into separate partitions. There is
therefore a separate model effectively for each segment. This is
particularily important where you need to differentiate between the
different needs of say business vs individual customers.

-To account for heterogenity in customer perceptions, I assumed that
customers perceptions were normally distributed about a mean. This
allowed me do do calculations such as find out how many customers think
that B is better than A which is then an input to the calculation of
market share change.


Hope this gets you started... My system is crashing in one minute, must go...

Jonathan Segel.
4jds@qlink.queensu.ca
-