How predictive analytics is stabilizing renewals

Every broker I talk to has a version of the same story. A client group renews its stop-loss coverage, the carrier prices it based on last year's claims and then somewhere around month eight or nine, a handful of catastrophic cases blow up the numbers. The carrier raises rates at renewal, sometimes dramatically. The broker is left explaining why the plan that looked stable 12 months ago is suddenly unaffordable. And the employer is left wondering if self-funding was the right decision at all.

This story plays out thousands of times a year across the country. And the frustrating part is that in many cases, the warning signs were there. We just didn't have the right tools to see them.

I founded my company believing that advanced AI and clinical intelligence could give carriers, employers and their advisers a much earlier window into group health risk. We recently completed a retrospective study with a national stop-loss carrier that put that belief to the test and the results were eye-opening, even for us.

Here's what we found. The carrier had underwritten 19 employer groups covering nearly 17,000 members. Six of those groups ultimately lost money for the carrier — meaning the claims they paid out exceeded the premiums they collected. When we ran those same groups through our platform before any of that happened, we flagged four of the six money-losing groups as our highest-risk classification. Four groups that represented just 21% of the total cohort drove 52% of the underwriting losses. The single worst group generated a $2.3 million loss and had a risk score 1.5 times higher than any other group in the study. 

If the carrier had acted on those signals by pricing them to reflect the actual risk, their loss ratio would have dropped from 83.6% to 55%. And if the benefits consultant and employer knew what risks were likely to emerge next year in their membership, they would have the tools to manage that population proactively rather than reactively through preventive care programs, an aligned network and the most appropriate pharmacy benefit manager partners. 

I'm not telling this story to sell you on technology. I'm telling you this because it changes how I think about what brokers and benefit advisers actually need to do their jobs well.

When a carrier prices a group today, they're largely working from trailing claims data, basic demographics and actuarial adjustments that have been refined over decades. That's not a criticism; it's simply the toolkit that has been available. But trailing claims data tells you what happened, not what's likely to happen next. A group can have a quiet year and still be sitting on a member population with chronic conditions, high-cost specialty drugs in the pipeline or disease patterns that create enormous volatility risk going forward. Conversely, a group with a terrible claims year may actually be a stable, predictable risk — one bad event skewing the picture.

The ability to tell those two stories apart before underwriting is where predictive analytics changes the game. And I think it matters enormously to brokers because your clients rely on you to see around corners that they can't.

Think about what better risk intelligence means in practice. It means fewer renewal shock conversations. It means being able to advise a client on whether self-funding is genuinely appropriate for their population, not just based on last year's claims, but based on a forward-looking picture of their group's risk profile. It means having more confidence when you recommend a stop-loss carrier that their pricing reflects reality and that your client isn't quietly sitting in a portfolio that's headed for a correction.

It also means something that often gets overlooked in these conversations: carriers that can accurately identify the truly volatile groups don't have to conservatively load everyone's premiums to protect themselves. When volatility is visible, it gets priced appropriately and the groups that aren't volatile can actually see better pricing as a result. Better risk selection isn't just good for carriers. It's good for the clients who deserve competitive rates and don't get them because someone else's tail risk is quietly embedded in their renewal.

The health benefits industry is in the middle of a real reckoning around data and transparency. Employers are asking harder questions. Regulators are paying closer attention. And brokers are being asked to add more value in an environment that's getting more complex every year.

The good news is that the tools to do that are finally catching up to the problem. We don't have to fly blind into renewal season anymore. The warning signs were always there. Now we can actually read them.