| 摘要 |
A method is disclosed for determining the prior net benefit of obtaining data relating to an individual risk in an insurance portfolio, via a survey or similar procedure. A risk model is developed at the individual risk level for mathematically estimating the probability of expected loss given a set of information about the risk. The risk model is incorporated into a profitability model. A probability distribution relating to the type of survey information to be obtained is developed, which is used for determining the gross value of obtaining the information. The method produces as an output a quantitative estimation (e.g., dollar value) of the net benefit of obtaining survey data for the risk, calculated as the gross value of the survey less the survey's cost, where the benefit of the survey relates to a quantitative increase in predictive accuracy resulting from incorporating the survey data into the predictive model.
|
