It’s the ultimate irony. Insurance carriers are in the business of risk management, assessing risk while predicting revenue and expenses and then pricing for that. However, the mechanisms many carriers use to generate the rating systems their pricing depends on expose them to risk.
The mechanisms I’m talking about are those that use one system to build pricing models, such as Microsoft Excel, and a completely separate technology, such as Java, to translate that model for use by the enterprise. If that describes how you build your rating system, you lack the agility you need to be competitive. And it could be costing you considerable profits.
Here’s a quick look at just some of the more serious risks you face.
A single point of failure
Actuaries tend to use spreadsheet tools like Excel to build pricing models. An actuary will assemble the model as a series of complex machinations in an Excel workbook. The work entails incorporating multitudes of reference tables on which assumptions and algorithms are based. The complex effort can take months to build and may be refined over many years.
Each actuary has his or her own system for creating a pricing model. They live with their Excel models every day and can navigate them with ease. The problem is, no one else can.
As most people know, when someone builds something intricate in Excel, the only person who can successfully manipulate it is the one who built it. The consequence is that when your pricing model resides in Excel, or a similar tool, your actuary becomes the single point of failure in your entire rating system. It’s very difficult for anyone else in your organization to maintain your pricing model, change it, understand it, test it, debug it, or verify its accuracy. That’s very risky.
Long and costly implementation and testing
A pricing model tool like Excel poses yet another risk in that the process of deconstructing the calculations in Excel and replicating them into a rating engine is often cumbersome and complex. It requires extensive collaboration between the actuary who built the assumptions in Excel and the IT developer tasked with converting them into a “black box” rating engine for the enterprise. In essence, the developer has to recreate the entire model for a different platform.
Multiple levels of parallel testing between the two platforms, along with quality assurance and regression testing, add even more hours and expense. And if changes need to be made to the pricing model along the way, extensive back and forth again ensues.
The entire process can take anywhere from six to nine months to complete and costs between $500,000 and $700,000 to build and integrate an IT-supportable rating engine for a single product, such as Short Term Disability or Private Passenger Auto. As a consequence, most carriers with legacy rating systems update their pricing only quarterly at best, and sometimes only annually.
Lack of agility
In today’s insurance landscape, market conditions and risk profiles are becoming increasingly dynamic. For example, you may need to make geographical adjustments, accommodate pricing changes dictated from a state regulator, or update plan variables from a filing perspective. Integrating such changes takes time. And it takes far longer if the process for refactoring your rating engine lacks agility.
When you can't respond nimbly, either you're overpriced and miss your sales opportunity or you're underpriced and are missing out on profits. Under both conditions, you're creating a financial strain on the business.
The bottom line is, the dysfunctional process for building your rating system leaves you at a disadvantage in a highly competitive market. And that may be the biggest risk of all.
A better way to create rating engines
In my mind, an ideal solution to building a rating engine is one that enables you to build the pricing model just once. You shouldn’t have to replicate the logic in a different technology. I’m talking about a single platform for pricing and rating that everyone can use. Here at EIS Group we have created this type of platform. Stay tuned, because in Part 2 of this post, we’ll look at what an agile rating system can do for your business.
Geoff Walton is a program director at EIS Group. Read more blogs by Geoff here.