Although many property/casualty personal lines offerings now are thought of as commoditized products, carriers with older or inferior rating systems have learned the hard way such thinking can be costly. Insurers that don't embrace newer rating systems often are doomed to choke on their competitors' dust.

Retention of old customers and attracting new business are affected dramatically by rating systems–for better or worse, asserts Larry Phillippi, manager of planning and project management at Penn National Insurance. Carriers with older rating systems frequently may acquire new business but at an incorrectly low price. At the same time, competitors with newer rating systems are able to price policies more accurately. “If [the competition] prices new business higher and the new business comes to you because you are lower priced, you probably are pricing it wrong,” he says. “Basically, with older systems you are looking at adverse selection on retention and new business.”

Heavily regulated insurance lines don't give insurers a pricing advantage, according to George Grieve, CEO of CastleBay Consulting, but an upgraded rating system makes implementation of those rate changes much easier. “If you have a difficult-to-use legacy environment, it is going to take you longer [to make the changes], and it is going to cost you more as an IT department to be able to implement those changes,” he says. “The issue is maintenance–the actual cost of making changes. Newer rating engines not only are more accessible because of SOA architecture but they often have nice development interfaces and testing tools.”

The whole issue of pricing for competitive advantage is a difficult one for some P&C lines because of regulation, continues Grieve. Heavily regulated lines such as personal lines automobile and workers' compensation have their prices set by state regulators. There is more price competition in some of the other personal lines and especially in commercial lines, and where there is competition, Grieve maintains, a nimble and accessible rating engine can help insurers.

Making changes to correct rating mistakes also necessitates having a solution that provides flexibility to the end-user. While insurers need to avoid mistakes such as coding errors, getting an algorithm wrong, or rounding incorrectly, Arrowhead Insurance's Stephen Boyd, vice president of information systems, contends “mistakes are going to happen. You want to minimize them, but when they occur, you want to be able to make those changes and get them in production quickly. You don't want a long cycle of redevelopment and extensive testing. You want to be able to push this up to production as soon as possible.”

The flexibility of tools available to insurers in today's market is one reason to upgrade, according to Dave Duden, director in the actuarial insurance solutions group for Deloitte's consulting practice. Such flexibility includes the ability to bring additional data into the rating and pricing process. He also points out there are speed advantages to be gained. “Most of the older rating systems I'm familiar with are doing some sort of batch processing to do the number crunching,” says Duden. Citing improved architecture and databases, he adds, “a lot of the modern systems allow you to do that work in real time and online.”

Some third-party data providers started simply as data collectors, Duden notes, and now offer services insurers can tap into electronically to supplement the carrier's data in real time. Some of these data providers do more than supplement a carrier's data, providing most of the data carriers use for the rating process.

Saul Swartout, senior consultant in the insurance practice at R.E. Nolan, also points to the availability of data as a reason carriers revisit their rating systems and processes. Data providers allow insurers to go to an outside source for information such as where customers live, their credit score, where they lived previously, and how many children they have, among other data points. “All of those things are readily available,” says Swartout. “It's created an avenue for companies to look at different factors than they had before.”

Companies formerly relied strictly on internal data, but Swartout claims due to its limited volume, the credibility of such data was suspect. “Now, you have so much third-party information, you can get that data and start mining it for different rating factors more readily,” he says. “The availability of this data and its relatively minor cost make this an important factor in rating today.”

Rating is one of the most self-contained applications in the insurance process, Grieve explains, which is why other issues such as legacy replacement, rules vs. code, and SOA architecture come into play. “If you get a better and more accessible rating engine, you can support what the business people are trying to do by creating more rules-based algorithms, which means you can change rates quickly,” he says. These factors make rating one of the easiest applications to replace in his view.

Another benefit to upgrading rating systems comes in environments where insurers need to take rates up or down in a timely manner, which Grieve feels should fall under an IT department's maintenance budget. “So many IT shops are spending 50 percent to 60 percent of their revenue on maintenance,” says Grieve. “Part of that maintenance work is ongoing rate changes where carriers write in a lot of different states. There is constant activity in terms of rate changes as well as responding to regulatory changes.”

Insurers with easy-to-use rating engines get pricing done quickly, more accurately, and are more responsive to the market, which can cut the time IT staff spends maintaining the rating system. “A lot of things show up as maintenance work that is just responding to the business needs,” says Grieve.

Some insurers jumped into newer rating technology too early, Duden indicates. Today, the focus is on Web-based systems, bypassing what he calls the client/server generation of products. “Companies are moving right into the Web-based systems,” he says. “The latest trend being the smart client–a browser-based front end where all the processing occurs on the back end.”

The new generation of rating tools for property/casualty carriers is mature, functional, and provides significant benefit over the legacy tools, Grieve agrees.

Swartout claims nearly every company he knows is looking to integrate its systems so those systems can communicate with each other, handle single entry of data, and be a single repository for data. “Double entry is not only very expensive, it leads to higher rates and makes insurers less competitive,” he says. “It also leads to a lot of errors.”

If insurers can go to one place to find information rather than 10, underwriting cycle times start to decrease, which reduces expenses, Swartout adds. “Compounding all that is the realization in the industry there are a lot of similar risks that perhaps don't need to have hands-on underwriting–auto, homeowners, some small business are viewed almost as commodity underwriting these days,” he says.

Another factor affecting the rating issue, continues Swartout, is a growing realization rate structures are starting to soften. Premium is made up of two components, he explains: the expense factor and the cost of risk. “As they reduce rates, insurers are looking at both those components to see whether there is a savings,” he says. With integrated rating and data systems, insurers can save on expenses, he suggests. Carriers also get a better view of the quality of the risk, which Swartout equates ultimately to the cost of carrying that risk. “If you can cleave out a better portfolio than your competitors, your rate structure can be more competitive,” he says. “Rate structures decrease because of competition and capacity. Companies are looking at how they can stay in the same lines but have better rates and be able to do the rating faster.”

Separate systems, such as claims and underwriting, have had plenty of information, remarks Swartout, but these systems were siloed and integrated through huge legacy systems. “It was a very inefficient way to manipulate an underwriting process, especially if you make a change to your rating structure,” he says. “[Insurers] had to find a way to integrate those disjointed systems.”

Modern systems depend on integrated underwriting and information collection, states Swartout. “You put in a person's name, and the system can go out to third-party vendors, acquire all the history for that person, and integrate the data into the rating system,” he says. “Ten years ago, to do that same thing, you had to go to your own system to see whether you had that person [in the system], then go to a different place for a VIN and a different place for claims history. The cost for accumulating information for rating when you have disjointed systems is very high.”

Carriers need to emphasize the ability of systems to interface when they look at new systems, Phillippi advises. Companies adept at rating collect and keep more data than older systems are able to handle. If carriers keep their current administration system and integrate it with a new rating engine, they still will have hard data issues they are going to have to get around, he adds. “Newer systems should be able to integrate much more effectively,” says Phillippi.

Arrowhead has a system called Arrowhead Exchange, which is a portal for agents and brokers developed in-house. On any given day, Boyd explains, Arrowhead has between 10,000 and 12,000 agents on the system, using it for quoting, binding of new business, endorsements, and making payments.

One of the first issues insurers have to deal with is getting their rates up properly, comments Boyd. “In some cases, we have agents and brokers who are not using a comparative rating platform so they need us to provide those rates upfront,” he says. “In other cases, maybe the comparative rater was in error and we need to verify the rate or correct it. You can't compete, especially in personal lines, without being able to offer a bindable rate right there and then.” Arrowhead's application process for personal auto takes five to 10 minutes, reports Boyd, with 99 percent of personal auto business done through Arrowhead Exchange. The system is rated through the DRC product Decision Maker.

Boyd has a staff of business analysts who work well with the system, he affirms. “We like it because it's easy to maintain and set up programs,” he says. “Once you have a base structure in place for a given line of business, rolling out new programs for new states is a pretty straightforward task.”

Arrowhead started with personal auto on the rating system and then added homeowners. “Over the course of the last three years it's expanded beyond personal lines. We do workers' compensation through the product and a number of professional liability, commercial auto, and general liability coverages,” says Boyd.

Standardizing on one rating platform has made it easier for Boyd's IT staff. “It's just one application I have to train my staff on to set up and manage,” he notes. “Its flexibility allows us to use it in multiple lines of business. It has helped us get to a point where we can have one point-of-sale platform.”

Standardizing point of sale with one rating engine was critical, according to Boyd. “On the policy side, partly because of our approach to go after niche markets, there isn't one policy solution that can do everything we need,” he says. Thus far, Arrowhead has plugged in the DRC product as a policy admin rating solution as well as a point-of-sale piece for three divisions.

When companies try to enter a new state, a new business line, or add new coverages, that often has been the straw to break the IT department's back. Carriers can't modify their old rating system to handle these changes, so instead they build a new rating system, put the new product on the new system, and leave the older business running on the old rating engine, observes Duden. “It's an effective strategy if they build the new system to handle multiple lines,” he says. “But I've seen people take the band-aid approach where [the new system] handles only one line. If they develop a system to handle multiple lines, it is a good strategy.”

Duden claims he sits on the fence on the build-vs.-buy issue. “I do see a trend toward building now. It has to do with control and competitive advantage,” he says, adding insurers feel the only way they can gain control is to build the product in-house.

“With some of the tools out there it's practical to think of building,” continues Duden, “whereas five or 10 years ago people would look at you funny if you said you wanted to build a rating system.” He credits Web services and SOA for this change in attitude. “Most insurance companies have a pretty fully developed SOA strategy in place,” he says.

The movement to build rating systems is reaching down into the middle market. “Five years ago I might have said the middle market needs to buy,” indicates Duden. “I do think even some of the smaller companies are looking to build or buy a toolkit [to assist in the build process]. Insurance companies, historically and going forward, want that in-house control of IT. With options on development now, I think you can do some of these things in-house, maybe with some outside help.”

Penn National's Phillippi believes carriers need to take a systematic approach to figuring out what needs to be done. The company's business goals and objectives must be set out–both the current state and where the business side wants to go. The current rating system needs to be evaluated to assess what the system can and cannot do. A market scan then can be conducted to determine what products are available, Phillippi advises. Requests for information from vendors will help carriers ascertain what the best implementation options are. “Based on what you view as the pressure points in the current systems, how do you move away from there in terms of replacing parts?” he asks. “The goal is to try to match those options with what you think the best implementation path is for your company,” says Phillippi. “What you are trying to do at a granular level is develop the benefits of the different options and what you think costs and time frames are going to be. You look at all the scenarios and the benefits of each. When you are done, you hope to feed the costs and benefits into each of the options to see what gives you the biggest bang for the buck.”

Rating engines have been the most commonly replaced core asset in P&C systems over the last five years, in Grieve's view, because doing so has become a relatively easy task. “It's a lot easier to do than replace a claims system or a policy administration system,” he says. “It's still not trivial, but quite a lot of companies have done it. The return on your investment is a lot faster, and you can see the benefits are relatively measurable in a short time frame.”

Special Reports /

At What Price?

Related Stories

Recommended For You