Marginal Impact Analysis

A Marginal Impact Analysis enables you to view the effect that a specific loss results set (the target or new portfolio/contract) has on another loss results set (the reference portfolio). For example, an underwriter may want to assess the potential consequences of adding a new contract or portfolio to an existing portfolio.

By default, the Marginal Impact Analysis excludes expiring contracts from the reference portfolio. You can choose to include these contracts when you are configuring the analysis. You can also specify a range to calculate a window-based average for return period (RP) losses.

Marginal Impact combines two loss results sets, providing users with six outputs. Touchstone stores summary EP data (aggregate and occurrence), event totals, and distribution based on the geographic resolution used to save the loss results.

To generate the exceedance probability (EP) curve for each policy, the engine sorts the losses in each row from largest to smallest. The largest loss can be assigned an EP of 0.01% (1/10,000), meaning that it is equaled or surpassed only once in the 10,000 simulation years. The second largest loss is equaled or surpassed twice, corresponding to an EP of 2 in 10,000 years, or 0.02%. The 20th, 40th, and 100th largest losses have EPs of 0.2%, 0.4%, and 1%, which correspond to the 500-, 250-, and 100-year return periods, respectively. By assigning an EP to each loss, the engine can derive a full EP curve.

You can use the Reference Portfolio results to determine capacity, and the Marginal Impact results to determine pricing.

In addition, the Scenarios tab displays the same six types of information for a set of Lloyd's® Realistic Disaster Scenarios (RDS), Extreme Disaster Scenarios (EDS), and the Top 10 tab displays the top 10 Marginal Impact loss points.

You cannot run Marginal Impact analyses on old event sets. Click here for information about working with modeled losses from previous product versions.

You cannot run Marginal Impact, Loss Group, or CAT XOL analyses on the results of Non Catastrophe Peril Analyses.

Marginal Impact results are not viewable by peril or model, even if the results sets in the analysis included peril or model information.

You can view the regional marginal impact provided that you have saved the loss results by an appropriate geographical resolution. For example, to view results for Florida Wind, you must have saved the results by state.

Since Marginal Impact Analyses can consume a lot of disk space, you may want to delete them on a regular basis.

In Portfolio Mode, inputs to the analysis can be Detailed Loss Analysis or Loss Group Analysis results. You determine which of these inputs is the target (or new) portfolio and which is the reference portfolio. The results sets may be on different databases and may be in different projects, but must be on the same server. In addition, all the loss results must be for the same number of simulated years. Further, when you combine two or more loss results for the same model, the model versions must be compatible. The system makes sure that the losses can be consolidated.

In Underwriting Contract Mode, you choose a target contract and a reference portfolio, and Touchstone compares the most recent Detailed Loss Analysis results for this target (or new) contract with the reference portfolio.

For underwriters to run Marginal Impact Analyses in Underwriting Contract Mode, a user with appropriate permissions must first run a Detailed Loss Analysis in Portfolio Mode on the master portfolio exposure view. The results of this analysis serve as the reference portfolio for the Marginal Impact Analyses that underwriters run in Underwriting Contract Mode. Users with appropriate permissions can run multiple Detailed Loss Analyses and/or Loss Group Analyses in Portfolio Mode on the master portfolio exposure view, thereby providing multiple reference portfolio choices for underwriters running Marginal Impact Analyses in Underwriting Contract Mode.

Marginal Impact Analysis

Related topics