The HPCDM Procedure

The HPCDM procedure is a high-performance procedure that estimates a compound distribution model (CDM) by modeling the severity (magnitude) of loss and frequency (count) of loss while aggregating them into one estimate.

The HPCDM procedure takes advantage of a computing environment that enables it to distribute the optimization task among one or more nodes. In addition, each node can use one or more threads to carry out the optimization on its subset of the data. When several nodes are used, with each node using several threads to carry out its part of the work, the result is a highly parallel computation that provides a dramatic gain in performance.

One example of aggregate loss modeling is an insurance company that wants to assess the expected and worst-case losses for a particular business line (such as automobile insurance) over an entire year, given the models for the number of losses in a year and the severity of each loss. Another example is a bank that wants to assess the value-at-risk (VaR)-a measure of the worst-case loss-for a portfolio of assets, given the frequency and severity models for each asset type.

PROC HPCDM estimates the distribution of an aggregate loss by combining the probability distribution model of the frequency of loss events (which you estimate by using the COUNTREG procedure in SAS/ETS®) and the probability distribution model of the severity of each loss (which you estimate by using the SEVERITY procedure in SAS/ETS). The probability distribution model of the aggregate loss is called the compound distribution model (CDM).

At its core, PROC HPCDM uses the Monte Carlo simulation method to generate a random sample of the aggregate loss. The random sample is used to compute empirical estimates of various summary statistics and percentiles of the compound distribution.

PROC HPCDM also supports externally simulated counts for cases in which you cannot use PROC COUNTREG to estimate the frequency model. You can also use PROC HPCDM to estimate the distribution of an aggregate adjusted loss by writing programming statements that compute the adjustments to the gross incurred loss. An example of an adjusted loss is the amount paid to policyholders, which is computed by adjusting the gross incurred loss with policy provisions such as deductibles and limits.


For further details, see the SAS/ETS® User's Guide: High Performance Procedures