Thursday, August 17, 2006

Improving groups, first in a series of off the cuff rants

Ah, group rewards -- the source of mucho controversy on Prosper forums. Some randomly selected thoughts on the matter:

A)I agree that sourcing borrowers not on prosper, deserves some sort of compensation. My next idea to push will be asking prosper to separate referral fees, which go to the GL, or whoever successfully refers a borrower to prosper who then gets funded -- from group rewards as they exist today.

in other words, pretend that Hi_Max sources a borrower from off prosper. the borrower hates his group, joins fanafi, gets funded. Hi_Max will at that point in time, get a "referral bonus," just as fanafi will get the current type of group leader rewards.

This furthermore allows for increased specialization, and thus, efficiency -- after all, one can refer lots of people to prosper, and get referral fees -- why bother requiring such people to be group leaders? Thus, affiliate agents recruit for referral fees, and group leaders can concentrate on leading groups -- or if they choose, they can still go for referrals as well.

After all, if one wants referral fees, one has to roughly proxy them by creating a standard non-100% shared group, and then invite friends/craigslisters/people at work to your group, and get them funded -- the referral and GL functions are inseparable.

B)basically, as a simple economics problem, group fee effects on borrowers and lenders have to do w/elasticity of borrower demand vs lender demand for loans -- to the degree that it is a borrowers market, borrowers can take the otherwise GL fee'd slive and set a lower rate, and still get lenders to bid on it. (At present, I'd argue that it is a borrower's market if holding borrower quality constant)

in other words, econ 101 says in an efficient market, imposing a 50 cent tax on widgets on sellers vs on buyers, has the same net economic effect on all parties in the end. The person who bears the brunt of it is the person w/least flexibility re: selling or buying the product, not the person who is technically targeted.

a simple to understand example at the extremes (for simple but middle of the road egs, google demand elasticity, tax, etc...

as always, making unrealistically idealized econ assumptions.

say diabetics all have decent amounts of money in our economy, and REALLY value insulin. if the government places a tax on sellers of insulin, rather than buyers, the sellers need only raise the price of their supplied insulin by almost the amount of the tax to offset their losses -- diabetics will still pay nonetheless.

Why lenders boycott certain groups:

C)Some lenders feel that for the same reasons that they bid on social loans for below-market rates, they should boycott UNJUSTIFIABLY high fee groups -- that is, not because the fees exist, but because those groups are demonstrably not value-additive -- or even value-subtractive, or have in the past been proven to lie to get borrowers to join their group. (these lenders are happy to bid on certain fee-charging groups, in other words.)

D)Other big lenders feel that by reducing the future prominence of certain groups by not bidding on their members (groups that don't add value/charge really high fees), offsets the occasional loans that are still worth it today after return adjustment for fees today. This is not as silly as it sounds -- if I'm unable to place as much $'s today in Prosper as I want because borrowers are getting bid down too far (shortage of good borrowers, surplus of lenders), but anticipate that a year from now, I could/would put 10x what I invest today into prosper, shifting the balance of power away from cruddy groups makes sense -- and some groups seem to have responded to this type of pressure (or possibly shifted their rewards to confuse borrowers, or what have you).

Read more!