Thread: Insurance Rant
View Single Post
VincentVega VincentVega is offline
Registered
 
VincentVega's Avatar
 
Join Date: Aug 2002
Location: MD
Posts: 5,747
Quote:
I'm an actuary. I price life insurance and annuity products. I also provide valuations of contingent events whose outcomes are not known. I'm not an auto insurance actuary, but the general idea is the same.

As more information becomes available, we gain more knowledge of a risk. Imagine you don't know if a die is weighted or not. So you roll the die over and over again and record your observations. You might start with the assumption that out of 6000 rolls, 1000 will be 1, 1000 will be 2, etc....

Say you observed 9 6's out of 10 rolls....you would think that the die is quite loaded!

You can then weight together your prior assumption (that all sides are equal) with the new information using statistical methods (Bayesian or Buhlmann credibilty). Of course the 6000 observations gets more weight and the 10 observations gets less weight. IIRC you need 1083 observations to gain full credibility (assuming certain random loss distribution shapes)

The point is, you showed them your cards
Interesting.

I was very happy with my State Farm agent for ~20 yrs, only ins I ever had. When I moved they told me they wouldnt cover my new place. I ended up saving some $$ with the new company bundled car/house but do miss the great service. I would call the old agents office and they basically asked, "what car did you buy now?", very friendly folks.
Old 03-29-2013, 11:43 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #30 (permalink)