Underwriting Class Estimator Methodology
Life insurance carriers label their underwriting classes differently. The same applicant can be Super Preferred Non-Nicotine at one carrier, Preferred Best at another, and Preferred Select at a third. This page explains how we normalize 18 real carriers to a shared class taxonomy so placement probabilities are directly comparable.
The Problem: No Industry-Standard Class Taxonomy
Every life insurance carrier publishes a field underwriting guide describing its preferred criteria: build tables, blood pressure thresholds, cholesterol cutoffs, family history lookback windows, tobacco-free requirements, and a long list of other factors. These guides determine which tier an applicant lands in, and tier placement drives premium.
The complication is that carriers do not share a common naming convention for their tiers. Four examples from the carriers modeled here:
- Symetra names its top non-tobacco tier Super Preferred Non-Nicotine.
- Prudential names its top non-tobacco tier Preferred Best.
- Securian names its top non-tobacco tier Preferred Select.
- John Hancock names its top non-tobacco tier Super Preferred Non-Smoker.
Each name refers to the best non-tobacco class the carrier offers. But the underwriting criteria behind those names are not identical. One carrier may require five tobacco-free years for the top tier while another requires three. One may allow BMI up to 30 in the top tier while another caps at 28. An advisor comparing two quotes without knowing these details cannot fairly assess carrier fit.
The Canonical Class Taxonomy
To make carriers comparable, we define a shared taxonomy with ten members. Every carrier's native class name maps into one canonical slot at import time.
| Tier group | Canonical slot | What it represents |
|---|---|---|
| Non-tobacco | Top tier | Best non-tobacco class the carrier offers (Preferred Best, Super Preferred, Preferred Select). |
| Preferred | Second non-tobacco class; typically Preferred or Preferred Non-Smoker. | |
| Standard plus | Third non-tobacco class where offered. Some carriers (Lincoln, Protective, North American, MassMutual) skip this tier. | |
| Standard | Base non-tobacco tier. Every modeled carrier publishes this tier. | |
| Tobacco | Preferred tobacco | Best tobacco class (Preferred Smoker). Not every carrier offers a preferred tobacco tier. |
| Standard tobacco | Base tobacco tier. | |
| Substandard | Mild rating | Table 2 to Table 4 range, roughly 125 percent to 200 percent of standard mortality. |
| Moderate rating | Table 6 to Table 8 range, roughly 250 percent to 300 percent of standard. | |
| Severe rating | Table 10 and above. Not every carrier offers ratings this high. | |
| Decline | Decline | Applicant would be declined by this carrier at current underwriting rules. |
Every carrier module includes an explicit mapping from its native class names to these canonical slots. The mapping is validated at import time; a carrier that publishes fewer non-tobacco tiers maps its top two tiers to Top tier and Preferred, and leaves Standard plus empty for that carrier.
Data Sources
Underwriting rules for every carrier come from publicly available field underwriting guides. No proprietary or under-NDA documents are used. Each carrier module records the source document title, publication date, and guide version stamp. When a carrier publishes a new guide, the module is updated and the version incremented.
Mortality probabilities come from the Society of Actuaries 2015 Valuation Basic Table (VBT) with MP-2021 mortality improvement projections. These are public actuarial standards used across the US life insurance industry. The SOA updates these tables periodically; the Lumis Life engine tracks the update cadence and applies the current standard when the applicant's assessment is run.
How the Estimator Produces Probabilities
For a given applicant, the estimator produces a probability distribution across the ten canonical tiers for each carrier. The output is a full distribution, not a point estimate. The top-two placement probability shown on the carrier pages is the sum of probabilities assigned to the top two non-tobacco tiers.
The distribution is computed in two steps. First, a base mortality model assesses the applicant's overall risk and produces an initial class-weight distribution. Second, each carrier's overlay applies that carrier's specific underwriting rules, shifting probability mass toward or away from tiers based on carrier-specific thresholds (build tables, BP cutoffs, cholesterol floors, family history windows, tobacco requirements). The final distribution reflects both the applicant's underlying risk and the carrier's documented preferences.
A top-two placement probability of 70 percent means: if this applicant applied to this carrier, roughly seven times out of ten the case would be offered one of the carrier's top two non-tobacco classes. The remaining thirty percent of the probability mass is distributed across Standard Plus, Standard, and any applicable substandard ratings.
What the Probability Does Not Mean
Probabilities are estimates based on published field guide rules. They do not represent a quote, an offer of insurance, or a guarantee that a carrier will make a specific offer. Several real-world factors can shift the actual offer relative to the estimator output:
- Attending physician statements (APS). Carriers order medical records for applicants with any non-trivial health history. APS findings can confirm, expand, or surface conditions beyond what the applicant disclosed.
- Paramedical exam labs. Blood pressure readings, cholesterol values, A1C, liver enzymes, and other paramedical measurements often differ from the applicant's self-reported values. The actual lab readings drive the actual tier placement.
- Financial underwriting. Very large cases require documentation of income and net worth. This does not typically shift the class, but it can delay or affect issue.
- Current carrier appetite. Carriers adjust their willingness to take on specific risk profiles based on portfolio mix and reinsurance capacity. These adjustments happen continuously and are not reflected in the static field guide.
- Underwriter discretion. Borderline cases often land on an underwriter's desk. Two underwriters reading the same file can reach different decisions. An experienced broker or BGA can occasionally push a borderline case into a better tier.
The estimator is a pre-shop analytic, not a replacement for an informal inquiry through a Brokerage General Agency (BGA) or a formal application. Its value is in helping the advisor or applicant understand which carriers are structurally favorable for the profile before investing the time and effort to submit.
Limitations and Edge Cases
Product-line scope
Several carriers publish distinct preferred ladders for different product lines. Term ladders may differ from universal life, indexed universal life, or variable life ladders. Where the modeled rules apply to one product line only, the affected carrier pages carry a scope caveat. Transamerica and Pacific Life are current examples where the modeled ladder is term-specific.
Impaired-risk ceilings
Carriers differ in how high they will rate a case before declining. Some cap at Table 8 (roughly 300 percent of standard); others go to Table 16 (roughly 500 percent). Severe impaired-risk cases should be shopped to carriers with deeper substandard capacity; the carrier pages note substandard ceilings where public guides disclose them.
Reinsurance and facultative cases
Very large face amounts or complex impaired-risk cases may be submitted facultatively (for individual underwriter review at a reinsurer). Facultative underwriting is outside the field guide framework and is not modeled here.
Non-US applicants
The estimator assumes a US-resident applicant underwritten in a US jurisdiction. Non-resident foreign national business follows entirely different underwriting rules and is not covered by the modeled field guides.
Update Cadence
Each carrier module records a guide_version stamp (typically YYYY.MM format). Updates happen when a carrier publishes a new field guide, which for most carriers occurs every twelve to twenty-four months. Version history is tracked in the Lumis Life changelog. The canonical class taxonomy itself is versioned separately; the current taxonomy version is 2.0.0.
Material changes to a carrier's rules (for example, a new substandard ceiling, a revised BMI table, or a change in the tobacco-free window) trigger a re-run of the underlying estimator logic and a bump to the carrier's guide version. Advisors running the estimator always see the current version stamp on the per-carrier result.
Transparency and Accountability
Every top-two placement probability on the carrier pages and in the live estimator can be traced back to: (1) the applicant's assessment inputs, (2) the current carrier guide version, and (3) the canonical taxonomy version. Rerunning the same applicant with the same inputs and the same versions produces the same output. When any input changes, the estimator captures the new inputs and the new version stamps together.
If an advisor or carrier believes a specific rule has been misinterpreted, the carrier-specific module source is the authoritative record. Corrections are welcomed and trigger a guide-version bump with a changelog entry.
About the Author
Lumis Life was built by Jeff Ting, a Fellow of the Society of Actuaries (FSA) and a CFA Charterholder. The underwriting estimator combines actuarial modeling with structured carrier knowledge to make carrier selection more analytical and less reliant on broker-level tribal knowledge. Read more on the about page.