Publicado por Ryan Park Grant emAssine nosso blog
As in “I love loss development triangles!” (Wait, what were you thinking?) I love them so much I made a couple hundred thousand of them*. And, I am proud that our Comparative Triangles℠ are now available as the inaugural Research topic in our new Guidewire Live℠ University.
Don’t know what in the world I’m talking about? Even within insurance companies, many people associate triangles with actuarial voodoo performed behind the priestly curtain. But triangles are just a useful way of comparing how claims of differing ages have matured. The format is easier to see in pictures than words, but each cell shows the status of the claims created in one calendar quarter at the end of each subsequent quarter. Because older claims have had more time to develop, they have more quarters of data—leading to the triangle shape as older loss quarters are stacked on younger ones. You can evaluate how claims have developed by reading across a row, or evaluate how your company’s performance has improved by reading down a column.
Actuaries and other analysts have long known the benefits of examining their data in this manner. But until now, insurers have not had access to comparative data that is as recent and granular as their own, using identical categories and definitions**. That’s what Guidewire Live offers: outside industry data that is truly comparable to yours.
What exactly makes our Comparative Triangles special?
We build each one entirely from granular transactional data, not surveys or roll-up reports.
We classify each exposure by a standardized line of business and coverage, state or province, and whether the loss was marked as a catastrophe—using definitions that are standard across companies.
We fill our triangles not just with a couple traditional metrics like indemnity paid & incurred but with detailed calculations of closing percentages, expenses, salvage & subrogation recovery, and number of payments or reserve changes.
We normalize for company size by showing your company average per exposure against the “average of averages” for other insurers with exposures in the exact same LOB, coverage, state, and catastrophe category—thus comparing apples to apples regardless of the other insurers’ overall size or mix of business.
We provide graphs and other visualizations of the data, its underlying quality, and whether your company performed better or worse than the group average.
We think that adds up to a lot of value. But we would like to know what you think! Let us know how & why you use triangles and whether ours help you do your job better.
How can you get started?
I’m glad you asked:
• If you are already participating in Guidewire Live, we will send you a spreadsheet file with your own company’s data compared to the benchmarks. Please live-research guidewire.com (email)target="_blank" us to get your own tailored version.
• If you aren’t yet a member of our 'cool club', you can get started by using our sample triangles. This file shows a fictional company’s data for a small number of states and coverages. But even better, see your own data by joining Live as soon as possible. It might take less time than getting a new report from your IT department!
Either way, I invite you to dig in and see what you can find. And I’m sure you will soon love triangles too!
Note: all links in this post will take you to the login page for Guidewire University. Registration is required to view the specific examples cited.
No exaggeration. Over twenty companies send us data, each with about a dozen coverages in as many states, and we show twenty-seven metrics by Cat/NonCat/All, plus all the benchmarks. It adds up fast.
** I am politely ignoring the difficulties that some insurers encounter in extracting their own data in a consistent and useful format from brittle or multifarious legacy systems.
TagsAssine nosso blog
Navigate what's next.