Underwriting Technology: Art vs. Science

Much of the pricing for insurance products in the property/casualty field depends upon market conditions and the appetite for risk carriers demonstrate. As the U.S. economy has tumbled out of control, many expected the market to move from its current softness to a hard market in which more attention to underwriting would have to be paid. But such expectations have yet to materialize in actuality.

“I would have guessed we would have been in a hard market by now, certainly on the commercial lines side, but by all indications that's not happening, so there continues to be market softness across the board,” says Ellen Carney, a senior analyst with Forrester. “I don't see that changing soon.”

And despite the continuing soft market, carriers have become much more interested in better underwriting tools, Carney observes, adding she has fielded questions from a number of carriers about rating engines. “There is a lot of curiosity about what else is out there,” she says. “[Carriers] are not saying specifically they are concerned about their pricing, but that's clearly what's going on.”

Stephen Forte, research director for Gartner, agrees with Carney the market remains soft, although he believes it has bottomed out. “There are a few aggressive insurers that are trying to raise some of their pricing, but because of the economic malaise we still are in, insurers have to be very careful in terms of price elasticity,” he says.

Forte predicts consumers will shop around and jump to other carriers if premiums start creeping up. “In certain lines of business–more so in excess and surplus–you have seen some increased pricing, but I think it is going to be [too] challenging a remainder of 2009 to start calling this a hard market,” he says. “I don't think we're really going to see that until 2010.”

The industry eventually will return to a hard market, but Forte contends the switch depends on the global economy. “If [the economy] improves to the point where policyholders are apt to pay a little bit more and stay with the company they've been with, [the market will harden],” he says.

Penn National CIO Bill Jenkins also concedes the property/casualty industry remains in a soft market, and the forecasts made by his company are conditions will not change for 18 to 24 months. “Like most in the industry we have fallen short of our revenue goals for the last several years because the market has been soft and as a result of the poor economy,” he says.

THE ECONOMY

Jenkins remarks the economy has wreaked havoc on insurers' surplus. “Return on equity has been down significantly for most carriers,” he says.

With such challenges, expense management has become a major focus for Penn National and most other insurers. “There is a lot of pressure being put on the IT area to get better control [of expenses],” says Jenkins. “We're even passing on projects that have a benefit to the company. It's a matter of what we can allocate and where we can get the best bang for the buck.”

Given the economy, pricing and customer segmentation are important issues for insurers to address. Carriers are asking themselves if it makes sense to hang on to certain risks given how little money they are making–or even losing–with a certain customer base, according to Carney. “I think what we are going to start seeing is pricing that makes it such that policyholders are incented to go find coverage elsewhere if it is too much of a risk for the carrier,” she says.

With losses suffered from investment income, there are few other places insurers can resort to, which is why many analysts have been expecting a hard market. Consumers also are doing more shopping–thanks to the ease of online shopping–to optimize their coverage and remove certain features to drop the price.

“The economy is putting a drag on insurers' ability to price policies to reflect what else is going on in their business right now,” says Carney. “They are looking to get a lot smarter about their underwriting decisions and the customer segment they're going after. We are seeing more science now than art in the underwriting space.”

The market and the economy have forced insurers to return to basics in terms of trying to make money by selling policies, notes Forte. “Their model is built on making money on their investment income, not the premium they take in from policies,” he says. “Their investment income has deteriorated, and so they have to be a little bit smarter when it comes to underwriting and the types of risk they take on and make sure they are pricing risks to exposure.”

With that, some insurers are making investments in underwriting workstations, particularly commercial and specialty lines. “The challenge on the underwriting side is it has been very manual and there aren't a lot of off-the-shelf applications you can use,” he says. “You can cobble some different components together–geo-spatial technologies, risk-assessment technology, rules engines for the front end–but for the most part, it's not like you can go out in the market and buy a policy admin system or a claims system where there are a plethora of vendors and various degrees of functionality. Insurers in general have had to build their own [underwriting] systems or cobble together different components.”

WEIRD SCIENCE?

Carney is hearing questions from carriers about analytics, rating engines, and core applications, including underwriting. “The commercial lines side has definitely picked up,” she says. “The pendulum has switched for the carriers. Technology vendors offering commercial lines solutions probably are in a good position to exploit that. I think we are going to see a boon in the application side.”

The need continues for underwriting on the personal side, Carney points out, noting there has been a lot of investment made there for a while. “There's been a flurry of investment going around that, and it will continue, but there's a microscope on the commercial lines to take advantage of what's happening,” she says.

In the last five to seven years, there has been an evolution in the P&C sector regarding pricing, indicates Claudine Modlin, a senior consultant with the EMB consultancy, with carriers moving away from rudimentary pricing methods and developing a more statistical approach. Led by the personal lines insurers, she has seen a number of medium and small carriers begin to use these methods.

“What we're seeing is taking things that historically have been art and bringing more discipline to them so they are more science,” says Modlin.

The debate over the best methods in underwriting stems from a cultural mindset, asserts Forte. “When I was a commercial lines underwriter, we had master underwriters who had been doing commercial underwriting for 30 years,” he says. “They had all that information analysis data trapped in their head. It was all experience, which made them very valuable.”

The challenge carriers face is when they bring in new underwriters and want to pass that knowledge along. When examining a risk, Forte explains, younger underwriters need to know immediately the types of exposures and hazards that are there. The underwriters also have to assess, first, whether the carrier is going to write the risk; second, how they are going to write the risk to protect the company against exposures; and third, how to price the policy correctly.

“In general, underwriters have been a bit resistant to automation and technology,” says Forte. “It's threatening in a way. If you put in these processes that are automated, companies might not need as many underwriters. Obviously, as the aging work force continues to put a strain on the insurance industry and new people come in and are looking for more automation, it definitely will switch to where we get more products, services, and automation around commercial underwriting. The rule of the commercial underwriter is not going away, but it certainly is going to change. They shouldn't be stuck behind a desk. They should go out being more collaborative with their distribution channels and being good account managers to try to build and grow the business.”

Forte comes down on the side of underwriting as more art than real science and believes most insurance software vendors realize carriers consider their underwriting efforts a center of excellence.

“There have been a few [vendors] that have tried to sell systems or build systems, but it hasn't happened because insurers are unique in terms of how they underwrite, their workflows, and their processes,” says Forte. “Vendors have decided there are other areas where there is more demand. With underwriting, there hasn't been the opportunity.”

Underwriting is a blend of art and science, in Jenkins' view, but of late Penn National's underwriters have been edging more toward the science. “The types of appetite we develop, based on the results of our predictive analytics, do become subjective in a lot of ways,” he says. “We have to rely on the data to help us make decisions. Underwriting becomes an art, especially with larger risks. But the meat and potatoes of our business is [small] commercial lines, and we see that leaning toward more science than art.”

USAGE-BASED INSURANCE

Modlin is convinced usage-based underwriting for personal auto will be the next wave in underwriting coverage. “If you think about some of the variables such as age and gender–factors that generally drive the auto premium today–a lot of those are just proxies for how you operate the vehicle,” she says.

With usage-based insurance, the insurer places an on-board diagnostic device in the car and tracks whether the car is being used, where it is being used, and the driver's driving practices.

“It can track a variety of information, and depending on the device you would invest in, it can track either a few or multiple things,” says Modlin. “Carriers now can analyze that information and say that at the end of a six-month policy, this is what your premium will be based on how you drove that car. There's a lot more appeal [to such a product] than people would have thought.”

There are issues of privacy and Big Brother that still have to be addressed, but with Progressive utilizing the program in several states, Modlin is beginning to see more work in this direction here in the United States.

“This is not just for the big [insurers],” she says. “We are starting to get calls from mono-line, mono-state carriers. For the little guys to gain competitive advantage, they can go usage based and be on par with the big guys. There is an immense amount of data to be collected, sorted, and analyzed. I think the landscape is going to look different in a couple of years.”

Serhat Guven, also a senior consultant with EMB, adds usage-based coverage has drawn positive reviews among regulators. “What regulators like about it is it is explanatory and it is directly tied to the consumer,” he says.

With even older cars having on-board diagnostics ports, the technology is not a problem. “There is a technology cost, but as more players get into the market and the idea is growing, the technology is getting cheaper,” Guven says. “I think this is the truly revolutionary way to price auto insurance because it really captures the idea of driving behavior. A lot of commercial carriers are utilizing this, not just for the insurance rates but because it is a good way to manage your fleet.”

PRICE OPTIMIZATION

When insurers establish their rates, Modlin explains, there always is the process after the fact when companies examine the rates and how those rates will affect policy renewals or the company's competitive position.

“In the past, that information was incorporated into the pricing process almost judgmentally,” she says. “Now, companies are pulling together data on what influences retention, renewals, and new-business conversion and feeding statistical models to that information to understand such factors as price elasticity that influenced the renewal decision or purchase decision.”

Price optimization incorporates that information along with the technical analysis of losses to come up with optimal premiums at the individual policy level that will achieve the volume targets and/or profitability targets carriers want to achieve.

“It's adding more discipline and science,” says Modlin. “We're continually trying to convince [carriers] they have always incorporated competitive positioning and retention information in their rate-setting process. This just does it in a more systematic way.”

Price optimization uses notions of customer value and elasticity models to identify the right rate for the risk, adds Guven. “Price optimization is not about trying to squeeze every dollar out of the insured; it's about putting a science behind the rate selection process,” he says. “Most insurers deviate from the actuaries' indications because of competitive concerns and market concerns. Those concerns are being introduced in more technical and sophisticated ways through elasticity modeling, customer value studies, and things such as that.”

ANALYTICS

Penn National began work on analytics for commercial lines about a year ago, relates Jenkins, and the carrier has completed the workers' compensation line of business. He reports Penn National is utilizing the results from the model to evaluate the book of business it has been writing to see whether certain coverages are profitable. The carrier now is working on the BOP line of business to be followed by commercial auto.

“We have done a lot of work around personal auto and homeowners as they relate to predictive analytics and have seen a lot of benefits because of it,” says Jenkins. “The results from the model help us develop rules for automated underwriting and also allow us to adjust our rating engine and underwriting processes. This allows us to expand our underwriting risk selection or retract it in certain areas.”

Jenkins sees some improvement in the personal lines business and credits this progress to the investments made in upfront automated underwriting using predictive analytics. With that success, Jenkins indicates his IT department is getting a lot of pressure from the underwriting and executive staffs to move analytics into commercial lines.

One hindrance to that proposal is the carrier's legacy systems, which challenge the carrier's goal of being easy to do business with. “We are trying to redesign the underwriting processes and expand the use of the Web and the portal upfront to build in functionality for the distribution channels and policyholders,” says Jenkins. “Where our personal lines have shown positive growth and good benefits from our systems, we are moving into commercial lines as quickly as we can to refresh those systems.”

Underwriters have “seen the value we've achieved in personal lines,” he says, “and the thought process is we can mirror image a lot of the work we've done in personal lines for the commercial lines.” He points out this will open the underwriting appetite as it relates to pricing and evaluating the best risks.

Penn National is using the Insbridge rating tool from Oracle in personal lines with the idea of moving it to the commercial business as the carrier goes forward. “It has given us the ability to get more sophisticated in the number of rates we are able to develop,” says Jenkins.

“One of the things that can hurt CIOs is being a business inhibitor,” he concludes. “When you go back to the business side and say it is going to take a year to make changes, that's not acceptable. Before, legacy rating engines took months to put in new rates. The new systems allow us to make changes in hours or days.”