Explore the essence of Claim Data

How does ReservePrism analyze claim data

The objective of ReservePrism is to generate synthetic claim data to represent your real claim data. Once you specify the claim file, ReservePrism will first analyze your data and parameterize its Simulation Engine from the fitted optimum statistical properties. ReservePrism can directly estimate many important Claim Level characteristics such as Accident Dates range, Frequency, Severity, Report Lag, Payment Lag, Adjustments, P(0), and Policy Level information such as Limit and Deductible (or SIR), etc. Some other properties such as trend, valuation frequency, etc, will be manually entered based upon your underwriting data and reinsurance treaty terms.


The Case Reserving process, including analyzing incurred loss interpolations, adequacy factors, thresholds, estimated P(0), etc, can be configured to simulate the human behavior as close as possible in the valuation process.

Example 1, Report Lag Fit

Claim file has dates information

ReservePrism Visualizes the Fitting Process

Report Lag Data

From the empirical curves of the report lag (ReportDates-AccidentDates), you pick the desired distribution. ReservePrism then applies maximum likelihood curve fitting method. The result is instantly presented visually along with the KS and ChiSqr test.

Report Lag Fit

Example 2, Severity Fit, Survival Model

Limit, Deductible, and Payment available

ReservePrism Fits the Loss Size Distribution

Loss Data

Payment is censored and truncated from the real loss by Limit and Deductible. Thus a Survival Model (by minimizing the negative loglikelihood) is implemented into ReservePrism to back-fit the Size of Loss distribution from the payment, limit, and deductible. SIR (Self Insured Retention) is also applied as an option besides the deductible.

Report Lag Fit