Author Topic: Pricing Premiums  (Read 1315 times)

0 Members and 1 Guest are viewing this topic.

Offline bytemaster

Actually, we can adjust payouts after the fact.   

At any given point in time we have the following numbers:

Total Paid In == Shares Purchased
Total Awards

We can thus calculate the maximum award per share as:

Max Award per Share = Total Paid In / Total Awards

For each share you track how much has been paid to it. 

The Max Award per Share changes over time... once a particular share hits this limit they stop getting payouts until the Award per Share goes up (ie: if risks go down).

So now you have a situation where you do not need a prediction market to estimate the risk.  We know the risk and the maximum payout.   Events may conspire that some people may receive payouts while the risk is low... and then the risk goes up.   This does not effect the solvency as risks change over time.  It just means that the value of future insurance shares goes down.

So what happens when you have a large claim and the reserve is empty?   It means you get paid out over time up to the Max Award per Share.    On average though the time you have to wait shouldn't be very long and it all depends upon clustering of claims.   If there have been no claims for a while then there is some reserve built up and you can get paid immediately.   If there is a cluster of claims that consumes the reserves you may have to wait until there is a lull in the claims.  On average thought he claims + income will be equal.... so you can estimate what the worst case delay on payment could be based upon the variance of the events and the size of claims.   

With a large enough insurance pool such as auto insurance, you could expect to be paid out almost immediately barring some weather event that results in a short-term spike in payout demand.  These short-term spikes will boost the long-term average and thus the average new claim rate during normal times will be less than average income. 








For the latest updates checkout my blog: http://bytemaster.bitshares.org
Anything said on these forums does not constitute an intent to create a legal obligation or contract between myself and anyone else.   These are merely my opinions and I reserve the right to change them at any time.

Offline CWEvans

  • Full Member
  • ***
  • Posts: 183
    • View Profile
The premium (P) that one pays should be equal to the probability (p) of the event times the cost (c) of the event, plus a share of fixed administrative overhead (π):

P = p(event)*c(event) + π

A DAC eliminates the need to hire administrative staff, such that π→0.

However, since probability of a future event cannot be known, only estimated, the optimal premium cannot be known ex ante. Given that hindsight is 20/20, we can calculate what the premium should have been, ex post, but that is not helpful.

A prediction market seems to be the best way to set the premium, and perhaps to determine the payout. Whether every Tom, Dick, and Harry should be able to participate in the prediction market, or only certified claims adjusters is the question.

Perhaps we need a separate reputation market, such that the opinions or votes of each market participant is weighted by his, her, its, or their reputation scores?