1. CDS Convective Offerings Helping Insurers understand their Risk Presented by: Dan Munson Founder, VP of Sales & Marketing CDS Business Mapping, LLC. ©Copyright 2011, CDS Business Mapping, LLC. All Rights Reserved.

2. Product Need Hail Storms and Tornadoes are becoming very expensive to insurance companies Hail Storms and Tornadoes are becoming very expensive to insurance companies Carriers need a way to manage their exposure to hail storms and tornadoes Carriers need a way to manage their exposure to hail storms and tornadoes Identifying areas prone to these events is not enough; these are broad areas, but individual storms actually impact very small areas Identifying areas prone to these events is not enough; these are broad areas, but individual storms actually impact very small areas Insurers need to understand where they are most vulnerable Insurers need to understand where they are most vulnerable

3. 5 Worst Tornadoes of All Time As Ranked by the Weather Channel 1. 1925 Tri-State Tornado 696 people dead; 219 mile path 2. 1974 Super Outbreak 315 dead; 23 F4’s; 6 F5’s 3. 2011 April Outbreak 346 Dead; 4 F5’s 4. 2011 Joplin, MO 5. 1999 Moore, OK

4. News on Losses 2011 Industry Losses ~$7B (source: AIR) Industry Losses ~$7B (source: AIR) Allstate $1.4B Allstate $1.4B State Farm $1B+ State Farm $1B+ Travelers $1B Travelers $1B American Family $1B American Family $1B Alfa $425-475M Alfa $425-475M CinFin $290M CinFin $290M State Auto $75-85M State Auto $75-85M Source: News Reports

5. Historical Data CDS Models based upon data from 1990-Present CDS Models based upon data from 1990-Present Early years were underreported Early years were underreported Visual sightings missed events in rural areasVisual sightings missed events in rural areas In early 90’s, Doppler radar became widespreadIn early 90’s, Doppler radar became widespread Leads to much higher averagesLeads to much higher averages

6. The Bias is Larger with Hail

7. Model Results The model transforms this… Into this…

8. CDS Tornado Model

9. CDS Hail Model Only events with winds >57 MPH were used

10. CDS Straight Line Winds Model Only events with winds >57 MPH were used TX OK KS NE SD MN IA MO AR LA MS AL FL GA SC NCTN KY ILINOH WI MI VA PA NY NJ DE

11. Policy Concentrations Matter Too We can definitely identify areas susceptible to convective events We can definitely identify areas susceptible to convective events But that doesn’t take into account the distribution of policies But that doesn’t take into account the distribution of policies Our new analysis does - It combines the tornado/hail scores with concentration analysis Our new analysis does - It combines the tornado/hail scores with concentration analysis

12. CDS Tornado Model Oklahoma Let’s look specifically at Oklahoma

13. OK Policies Overlaid We then overlay the policies

14. Policies Scored We then ran the policies through the Tornado Model to score each policy. You can see the shading of the points matches the Tornado Model

15. Oklahoma City Policies Scored You can see that the scores for each policy match the tornado scores

16. Close up of Oklahoma City Obviously, there are lots of policies here, with high scores. But where is the largest concentration of policies with high scores?

17. Aggregations 5 mile buffers were created around each policy. The sum total number of “tornado points” around each policy was calculated. So a score of 5 gives 5 points, etc. So its not just counting the policies, its using the scores as well

18. Policies shaded by Total Points You can now see the areas with the highest concentrations and the highest tornado scores

19. Identifying the Highest Risk Areas The red policies are the 10 locations with the highest concentrations within a 5 mile radius Highest Risk – 328 Policies within 5 miles; $41m TIV

20. Closer Zoom Use this information to consider moratoriums and to understand how much capital is at risk

21. Moratorium Example 3 mile buffer around the 10 policies with the highest number of total points. Company could stop writing in this area.

22. Analysis Options This can be done using any or all of the following This can be done using any or all of the following TornadoesTornadoes Hail stormsHail storms Straight line windsStraight line winds Hail & tornado scores can be customized (size of hail, state specific, etc.) Hail & tornado scores can be customized (size of hail, state specific, etc.) Any radius or multiple radii can be used Any radius or multiple radii can be used

23. RMS Convective Modeling RMS Modeling RMS Modeling Location Location YearBuilt:01/01/1971 YearBuilt:01/01/1971 OccupancyType:37 OccupancyType:37 Stories:4 Stories:4 BuildingClassScheme:RMS BuildingClassScheme:RMS OccupancyTypeScheme:ATC OccupancyTypeScheme:ATC BuildingClass:4B2 BuildingClass:4B2 PercentComplete:100 PercentComplete:100 Coverages Coverages Building: 5,000,000.00 Building: 5,000,000.00 Contents: 500,000.00 Contents: 500,000.00 BI: 500,000.00 BI: 500,000.00 Convective Storm Analysis Convective Storm Analysis ReturnPeriod:250 ReturnPeriod:250 PML PML ReturnPeriod:250 ReturnPeriod:250 ReturnPeriodDescription:250(0.40%) ReturnPeriodDescription:250(0.40%) EpType:OEP EpType:OEP LossName:Ground Up Loss LossName:Ground Up Loss Loss:$257,516.00 Loss:$257,516.00 AAL: $5,399.8 AAL: $5,399.8

24. Derivative Products Lists by Zip/County Lists by Zip/County Average and high scoresAverage and high scores Size of hail/tornado scale Size of hail/tornado scale Relative State Scores Relative State Scores

25. What else we’re working on Policy Aggregate Module – Track Aggregates in real time Policy Aggregate Module – Track Aggregates in real time Hail Damage Areas – Providing maps immediately after hailstorms for claims to confirm damage Hail Damage Areas – Providing maps immediately after hailstorms for claims to confirm damage Real Time Weather Real Time Weather Hail, hurricane, tornado, etc.Hail, hurricane, tornado, etc. Warning areas Warning areas Tracks Tracks

26. Conclusions The CDS Convective Analysis offers a new way to manage hail/tornado risk The CDS Convective Analysis offers a new way to manage hail/tornado risk CDS can identify the specific areas most at risk CDS can identify the specific areas most at risk CDS can quantify the amount of TIV within these areas CDS can quantify the amount of TIV within these areas This gives insurers actionable information to manage capacity and exposure This gives insurers actionable information to manage capacity and exposure

27. Contact Information If you have any questions, please contact: Amanda Brax Sales Associate 847-246-6666abrax@cdsys.com