Close Close
Samantha Chow. (Photo: EIS Group)

Life Health > Life Insurance > Permanent Life Insurance

How an Insurer Predicts Your Client's Future Now

Your article was successfully shared with the contacts you provided.

Samantha Chow wants to help life insurers size up their clients better, and faster, when they’re applying for coverage. As the life, annuities and health markets lead at EIS Group, a San Francisco-based software company, she connects the people in charge of underwriting clients with the technology specialists in charge of improving the underwriting process.

Chow, who joined EIS in 2020, has a bachelor’s degree from Columbus State University and a master’s degree in business from the University of Phoenix. She started out as a field automation specialist with Aflac, in 1997. She later held other research and analytical posts at Aflac, The Klages Group, Lebhar-Friedman, New York Life and Aite Group.

We asked Chow how she sees the state of life insurance underwriting now, in the wake of all of the changes brought on by advances in underwriting data sources and technology.

THINKADVISOR: Can an insurer really assess a life insurance applicant’s risk level with “just seven quick questions”?

SAMANTHA CHOW: For a lot of people, the data streams that go into this system come from reflexive questioning. A handful of health questions get broken down into many little health questions. You may start out with seven questions, but those can lead to more, in-depth follow-up questions, and a total of upwards of 50, or even 100, separate questions.

That’s where the value of in-depth reflexive questioning comes in. A reflexive-questioning-based system starts by asking, “Have you ever been diagnosed with cancer?” If you answer “no,” you go to the next question.

If you answer “yes,” the system asks, “How long ago?” and “What kind of cancer?” and you may see more dropdowns.

One question in a dropdown might be, “Did you have to have chemotherapy?” If so, the system asks, “How many times?” and “How often do you get scans?”

That’s where a lot of the data is coming from.

In addition, insurers are now using more third-party data than ever and third-party predictive models, which produce health scores to support automated processing.

What types of outside data streams are flowing in?

Some of the newer carriers are doing true accelerated underwriting, with more third-party data streams — vendors like MIB, Motor Vehicle Records and LexisNexis Risk Solutions — used for identity verification, or prefilling some application sections.

There are all sorts of reinsurers that offer data sources or algorithms. Those algorithms also take into consideration the prescription databases, MIB data and Motor Vehicle Records data, and your Fair Credit Reporting Act, or FCRA data, including data from LexisNexis and other vendors.

Companies like SelectX can help with the reflexive questioning, as well as with algorithms to support the reflexive questioning. You can also have a third-party deal with data elements from reinsurers, like RGAX, Hannover Re or Swiss Re’s Magnum. Milliman has all kinds of data; they can get down to the ZIP Code level and do predictive analytics around that.

So the streams of data coming in are all going to help the insurers evaluate the risk or associate the applicants to a risk classification, if the streams are used correctly.

What kinds of additional data streams are you adding or thinking about adding, and why?

Think about accelerated underwriting versus automated underwriting versus simplified issue versus fully underwritten medical underwriting. Simplified issue is typically a couple of yes/no questions; automated acceptance/automated decline is a bit deeper. It might have more reflexive questioning, but it still results in a yes or no.

Then, outside of automated, you’ve got pure medical underwriting, which may have some reflexive questioning. Pure medical underwriting programs collect blood, urine, body mass index readings, and those kinds of things. You’re going to have a paramedical exam.

And then you have accelerated underwriting, which is where a lot of the data streams come into play.

It’s a lot of “if this, then that” statements. Every time you say yes to something, you might trigger the need for additional third-party data.

If you go through these health questions, and answer no to all of them, why would the insurer request any additional data about you? But if you start answering yes to some of these questions, then the insurer may want to go beyond MIB and motor vehicle records to gather the prescription data. And if they have found something a little bit suspicious, they may want to go to third-party data from LexisNexis, until they’ve got enough information to underwrite you and assign you to a proper risk class for pricing.

Or maybe they go further and get some medical underwriting. Maybe you need a paramedical exam, or urine, to confirm that you’re a nonsmoker, but that is determined by the data collected and only used if there is uncertainty.

That’s the real path; that’s how to use these data streams, and how they can help improve risk evaluation and reduce the cost of underwriting, all while improving the customer experience.

How do technology choices affect how all of this works?

EIS is built with an extremely open application programming interface (API) architecture. That’s the key to integrating with external third-party data sources. It’s also the key to integrating with external reinsurers and their underwriting workbenches.

And you need to be in the cloud; you need strong open-API architecture. You need really, really minute microservices-based technology, as well, to integrate with some of these outside systems, to create the seamless, in-the-cloud, real-time assessment of risk, so that you can make decisions immediately and give a good customer experience.

It’s not necessarily up to us to add these data streams; it’s up to the carrier to understand which data streams or data elements they want to add and the value the streams and elements bring to them so they can be competitive in the market without assuming more risk than they have in the past.

That’s where a company like EIS comes in. Having the right core technology lets carriers be flexible in the type of underwriting they offer, based on the product and market they’re trying to serve.

That’s where personalization comes from. And that’s key to the success of any insurer, life insurer or otherwise, and very key to the success of underwriting here.

Obviously, life insurers underwrite life insurance policies. Do they actually underwrite annuity applications, too? If so, do they use any extra data streams?

There are insurers that will underwrite annuities, but they’re few and far between. The reason they do it is to benefit the annuitant or the contract owner.

Here’s an example: I’ve got cancer; I’m only expected to live for the next 10 years, and I want to put $100,000 in a single-premium immediate annuity.

If the SPIA issuer knew that when they underwrote me, they would give me a shorter expected lifespan, and I would get more money back on a monthly basis than if I had a standard level of risk.

Here’s another example: I’m 65. I’m not a smoker, and I’m super healthy.

Because annuity pricing is based on standard mortality, which will include male/female predictions, it doesn’t make sense for me, the consumer, to want to go through any type of underwriting.

Underwriting in this situation is only valuable for those who are ill or expected to live a shorter life than the standard mortality table predicts.

So, for annuities, they’re going to use these standard data streams for prefill purposes, identity verification, and that kind of thing. That’s not for underwriting, but rather for filling out the contract, doing their due diligence, and making sure they’ve got the right person for the right reasons.

Pictured: Samantha Chow. (Photo: EIS Group)


© 2024 ALM Global, LLC, All Rights Reserved. Request academic re-use from All other uses, submit a request to [email protected]. For more information visit Asset & Logo Licensing.