1 / 38

Motor Postcode Zoning Working Party

Motor Postcode Zoning Working Party. Duncan Anderson Roger Massey Saideh Charlton Chris Spiller Dave Coughlan (Chair) Kim Pape Mark Harrison James Tanser Stephen Jones Richhard Verrall. Aims. Understand market Data availability Possible approach. Importance of zoning.

zahir-ellis
Download Presentation

Motor Postcode Zoning Working Party

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Motor Postcode Zoning Working Party Duncan Anderson Roger Massey Saideh Charlton Chris Spiller Dave Coughlan (Chair) Kim Pape Mark Harrison James Tanser Stephen Jones Richhard Verrall

  2. Aims • Understand market • Data availability • Possible approach

  3. Importance of zoning • Review of postcodes • 3rd biggest factor • Needed to limit scope • Theft only 10 - 15%

  4. Market Analysis

  5. Investigating market rates • EL Systems Præmium • "Average" risk consideredfor every UK post code district • 18 comprehensive quotesobtained from 15 insurers • Some possible interactions also examined • Limited • districts only • no direct writers • off screen discounts

  6. 1 10 2 11 3 12 4 14 5 15 6 16 7 16 8 17 9 17 10 18 11 19 12 19 13 19 14 20 15 20 16 21 17 22 18 125 Post code categorisation Number of categories Insurer

  7. 1 10 2.12 2 11 1.97 3 12 2.29 4 14 3.30 5 15 2.20 6 16 2.18 7 16 2.19 8 17 2.67 9 17 2.12 10 18 2.16 11 19 2.70 12 19 1.93 13 19 2.31 14 20 3.17 15 20 2.17 16 21 3.13 17 22 2.76 18 125 2.26 Post code categorisation Highest premium /lowest premium Number of categories Insurer

  8. Interactions  

  9. Interactions  

  10. 1.06 1.05 1.04 1.03 1.02 1.01 1.00 High High Vehicle Group Area Low Low Example interaction

  11. Loadings by postcode • For each company, "average loading" for a district defined as premium for district divided by unweighted average of premiums across all districts • Districts ordered by average"average loading" • Average loading plotted against district

  12. Loadings by postcode

  13. Loadings by postcode

  14. Comparison of categorisations

  15. Comparison of categorisations

  16. Comparison of categorisations 2/8

  17. Comparison of categorisations 3/8

  18. Comparison of categorisations 4/8

  19. Comparison of categorisations 5/8

  20. Comparison of categorisations 6/8

  21. Comparison of categorisations 7/8

  22. Comparison of categorisations 8/8

  23. Data, Method and Results

  24. Information available • Complex • Appropriate • Up to date • Higher level view

  25. Data • Internal data • External • Combined/Grouped

  26. Interesting views • Drive around and look • Social deprivation index • Subjective views

  27. Methods • Size of exercise • Consider residual risk • Three main types of approach • - Smoothing • - Credibility • - Market related

  28. Smoothing “Based on underlying assumption that areas “close” each other have similar risk” • Adjacency?

  29. Credibility “The risk associated with a postcode is a mixture of its own experience and some other estimate of expected” Appropriate?

  30. Market Related • “everybody doing it, so I should too” • Realistic?

  31. Approach • Smoothing - “weighted distance” • - “Spatial models” • Credibility - “Credibility Method” • Market - “Average Market” • - “External provider” • Other - “Modern Heuristic”

  32. Approach • Split data 70/30 • Standardise data • Use residual risk to zone •  squares as test statistic

  33. Simple results • All proved better than null model • Little to choose between results • Combination of “credibility and weighted distance” is the best

  34. Comparison of methods

  35. Implementation • Solid risk cost • Improved understanding

  36. Implementation - issues • Competitive position • Existing customers • Systems

  37. Conclusion • Market is not uniform in approach • Suggest approach as mixture ofmethods

  38. Motor Postcode Zoning Working Party Duncan Anderson Roger Massey Saideh Charlton Chris Spiller Dave Coughlan (Chair) Kim Pape Mark Harrison James Tanser Stephen Jones Richhard Verrall

More Related