Underwriting
Underwriting
Quick Definition
Underwriting is the process by which an insurance company assesses the risk posed by a prospective policyholder and decides whether to offer coverage, on what terms, and at what price (premium). Underwriters analyze factors like health history, age, occupation, driving record, credit score, and property characteristics to classify applicants into risk categories that determine their premium rate. In 2026, AI and machine learning are reshaping underwriting across both insurance and mortgage lending, with Fannie Mae and Freddie Mac issuing formal governance frameworks.
What It Means
Insurance is fundamentally a risk-pooling business. For the pool to remain financially stable, every policyholder must pay premiums that reflect their actual risk level. Underwriting is how insurers identify and price individual risk, preventing adverse selection where only the sickest or highest-risk people buy insurance, which would make the pool unsustainable.
Without underwriting, the healthiest, lowest-risk people would avoid paying high pooled premiums and self-insure instead, leaving only the highest-risk people in the pool. Premiums would spiral upward until the insurance program collapses.
The same principle applies to mortgage lending. When a lender evaluates a borrower's income, credit score, debt obligations, and property value before approving a mortgage, that process is also underwriting. The lender is assessing the risk that the borrower will default and pricing the loan accordingly through interest rates and PMI requirements.
Types of Insurance Underwriting
| Insurance Type | Key Underwriting Factors |
|---|---|
| Life insurance | Age, health, medical history, family history, tobacco, BMI, occupation, hobbies |
| Health insurance (individual) | Age, location, tobacco use (ACA limits other factors for marketplace plans) |
| Auto insurance | Age, driving record, vehicle type, annual miles, credit score, location |
| Homeowners | Property location, construction type, age, claims history, credit score, proximity to fire station |
| Commercial property | Location, construction, fire protection, business type, claims history |
| Disability insurance | Occupation class, income, health, elimination period selected |
The Underwriting Process: Life Insurance Example
| Step | Description |
|---|---|
| 1. Application | Detailed health questionnaire, financial information, lifestyle questions |
| 2. Medical records request | Attending Physician Statement (APS) for health conditions |
| 3. Medical exam | Paramedical exam: blood pressure, height/weight, blood and urine samples |
| 4. Lab analysis | Tests for cholesterol, glucose, nicotine, prescription drug markers, HIV |
| 5. MIB check | Medical Information Bureau database, shared medical history among insurers |
| 6. Financial underwriting | Verify income supports the coverage amount requested (insurable interest) |
| 7. Risk classification | Assign to rate class; determine premium |
| 8. Decision | Approve as applied, approve with modified terms, approve with rating (surcharge), or decline |
Risk Classification: Life Insurance Rate Classes
| Rate Class | Profile | Premium vs. Standard |
|---|---|---|
| Preferred Plus / Super Preferred | Excellent health, ideal BMI, no family history, non-smoker | 30-40% below standard |
| Preferred | Very good health, minor conditions well-controlled | 15-25% below standard |
| Standard Plus | Good health, some minor issues | 5-10% below standard |
| Standard | Average health, height/weight within range | Baseline |
| Substandard (Table Rated) | Controlled chronic conditions, prior cancer, family history | 25-200% above standard |
| Decline | Uninsurable due to terminal illness, recent cancer, severe conditions | No coverage |
Table ratings: Substandard applicants receive a "table rating." Each table adds 25% to the standard premium. Table 2 equals +50%, Table 4 equals +100%, Table 8 equals +200% the standard premium. An actuary designs these tables based on mortality statistics.
Guaranteed Issue vs. Underwritten Policies
Some policies skip full underwriting:
| Policy Type | Underwriting | Trade-Off |
|---|---|---|
| Fully underwritten | Complete medical review | Best prices; lowest premiums |
| Simplified issue | Health questions only; no exam | Higher premiums; faster approval |
| Guaranteed issue | No health questions; no exam | Highest premiums; graded death benefit; limited coverage amounts |
| Group insurance | Employer-sponsored; minimal individual underwriting | Lower premiums through group rates; may not be portable |
Graded death benefit on guaranteed issue policies: If you die within the first 2 to 3 years, beneficiaries receive only a return of premiums plus interest, not the full face amount. Full death benefit begins after the waiting period. For more on policy types, see our guides on term life insurance and whole life insurance.
Auto Insurance Underwriting: Credit Scores
Auto and homeowners underwriters in most states use credit-based insurance scores:
| Insurance Score | Premium Impact |
|---|---|
| Excellent (760+) | 20-30% discount vs. median |
| Good (700-759) | Near-median rates |
| Fair (640-699) | 10-30% surcharge |
| Poor (below 640) | 30-70%+ surcharge in some states |
Extensive actuarial data shows strong correlation between credit-based insurance score and claims frequency, independent of income. California, Massachusetts, and Michigan prohibit credit-based pricing in auto insurance.
AI and Machine Learning in Underwriting (2026)
Artificial intelligence and machine learning are no longer experimental tools in underwriting. They are embedded across the loan and insurance lifecycle, from borrower intake and underwriting to pricing, quality control, fraud detection, and loss mitigation.
Mortgage Underwriting AI Governance
In April 2026, Fannie Mae issued Lender Letter LL-2026-04, establishing a formal governance framework for any approved seller or servicer using AI or machine learning in connection with loans sold to or guaranteed by Fannie Mae. The framework applies broadly to internally developed tools and vendor-provided systems, with an effective date of August 6, 2026.
Freddie Mac acted earlier through Guide Bulletin 2025-16, issued December 3, 2025, with an effective date of March 3, 2026. Freddie Mac's requirements are codified in Section 1302.8 of its Single-Family Seller/Servicer Guide, imposing detailed governance, audit, and security requirements for AI and ML systems.
Both entities require AI governance to be embedded in enterprise risk management and overseen by senior leadership. Fannie Mae requires written policies governing AI and ML use that are reviewed at least annually. Freddie Mac goes further, requiring AI policies be approved by senior management including the CIO, CTO, CISO, or Chief Risk Officer.
Freddie Mac LPA Machine Learning Savings
In May 2025, Freddie Mac announced automations to its Loan Product Advisor (LPA) underwriting system utilizing machine learning. Lenders who maximize automation through Freddie Mac are originating loans that are $1,500, or 14%, cheaper and tend to have a 5-day shorter loan production cycle time. A fully digitized mortgage process can help save up to 40% in costs.
Freddie Mac's ACE (Automated Collateral Evaluation) waivers have saved families more than $2 billion in appraisal costs since 2017. LPA Choice feedback messages have enabled lenders to qualify an additional 18,000 borrowers for a mortgage.
Credit Score Modernization
The GSEs are transitioning from Classic FICO to modernized credit scoring models:
| Development | Date | Impact |
|---|---|---|
| VantageScore 4.0 approved for limited lender use | July 2025 | Considers rent payment history, trended data |
| FICO Score 10T historical data published | December 2025 | Enables lenders to analyze new model |
| GSE selling policies updated for VantageScore 4.0 | April 2026 | Lenders can choose model on loan-by-loan basis |
| FICO 10T and additional VantageScore 4.0 data published | July 2026 | Broader adoption expected |
Both VantageScore 4.0 and FICO Score 10T consider trended credit data and additional sources like rent payment history, providing more precise risk assessments than Classic FICO. This matters for credit score evaluation in both mortgage and insurance underwriting.
Reinsurance: Underwriting at Scale
Insurance companies themselves purchase reinsurance, which is insurance for their own risks:
- Primary insurer underwrites and issues policies to consumers
- Primary insurer transfers a portion of large risk concentrations to reinsurers (Munich Re, Swiss Re, General Re)
- Reinsurers assess the insurer's overall book of business and pricing adequacy
- This chain allows primary insurers to underwrite policies they could not support on their own balance sheet
ACA Limitations on Health Insurance Underwriting
The Affordable Care Act (ACA) dramatically restricted health insurance underwriting for marketplace plans:
| ACA Rule | Description |
|---|---|
| Guaranteed issue | Insurers must cover everyone during open enrollment, regardless of health history |
| Community rating | Cannot vary premiums by health status |
| Age rating (3:1) | Oldest enrollees can only be charged 3x the youngest |
| Tobacco surcharge | Up to 50% surcharge for tobacco users (only allowed factor beyond age/location) |
| Essential health benefits | All plans must cover a defined set of benefits |
This ACA framework creates a structured risk pool rather than individual underwriting. Healthy people subsidize sick people within the system, with subsidies to make coverage affordable.
Common Mistakes to Avoid
- Lying on the application to get a better rate: Insurance applications include clauses that allow the insurer to contest the policy during the first two years (the "contestability period"). If the insurer discovers misrepresentation, they can deny claims or rescind the policy entirely. Be honest on applications, even about health issues that seem minor.
- Assuming guaranteed issue is a good deal: Guaranteed issue whole life insurance charges the highest premiums and includes graded death benefits that pay only a return of premium if you die in the first 2 to 3 years. If you can qualify for fully underwritten term life insurance, you will get far more coverage for less money.
- Ignoring credit-based insurance scores: In states where it is legal, your credit score can impact auto and homeowners premiums by 30 to 70% or more. Improving your credit can reduce insurance costs as significantly as it reduces borrowing costs.
- Not shopping multiple insurers: Each insurer underwrites differently. One may decline you while another offers preferred rates. Being declined by one insurer does not mean all insurers will decline. Always compare quotes from multiple companies.
Key Points to Remember
- Underwriting is the risk assessment process that determines whether and at what price to insure someone
- Rate classes (preferred plus through substandard) reflect actuarial risk; better health means lower premium
- Credit scores significantly affect auto and homeowners premiums in most states
- Guaranteed issue policies skip underwriting but charge far higher premiums and have waiting periods
- The ACA eliminated individual health underwriting for marketplace plans, replacing it with community rating
- AI and machine learning are now embedded in mortgage underwriting, with Fannie Mae and Freddie Mac issuing formal governance frameworks in 2026
- Freddie Mac's LPA machine learning automations save lenders $1,500 per loan and 5 days in cycle time
- VantageScore 4.0 and FICO Score 10T are replacing Classic FICO, incorporating rent and trended credit data
Related Concepts
- Insurance Premium: The price that underwriting determines, based on risk classification
- Actuary: The professional who designs the risk tables and statistical models that underwriters use
- Term Life Insurance: The most common product requiring full underwriting
- Whole Life Insurance: Permanent insurance with underwriting, cash value, and higher premiums
- Coverage: What the underwritten policy protects against
- Credit Score: A key underwriting factor for auto, homeowners, and mortgage lending
- Mortgage: The loan product whose underwriting is being transformed by AI and ML governance
- PMI: Insurance that lenders require when mortgage underwriting identifies higher LTV risk
Frequently Asked Questions
Q: Can an insurer deny me coverage based on underwriting? A: For individually underwritten life, disability, and long-term care insurance, yes. For ACA-compliant health insurance, no (guaranteed issue). For auto and homeowners, insurers can decline in some circumstances such as very poor driving record or extremely high-risk property, but must comply with state regulations on acceptable underwriting criteria. Being declined by one insurer does not mean all insurers will decline. Shop multiple companies.
Q: Does applying for life insurance hurt my credit score? A: No. Life insurance underwriting does not involve a hard credit inquiry. Insurers may check a credit-based insurance score for some policies, but this is a "soft pull" that does not affect your credit score. Health and lab checks go through the Medical Information Bureau (MIB), not credit bureaus.
Q: What is the difference between a life insurance medical exam and a doctor's physical? A: The paramedical exam for life insurance is specifically designed to gather data for underwriting, not to diagnose or treat. It measures height, weight, blood pressure, and collects blood and urine samples. The examiner is a trained technician, not your doctor, and the results go to the insurer, not your medical record. You can request a copy of your results. The exam is free to you and is paid for by the insurer.
Q: How is AI changing mortgage underwriting in 2026? A: Freddie Mac's Loan Product Advisor now uses machine learning to automate portions of underwriting, saving lenders up to $1,500 per loan and reducing cycle times by 5 days. Fannie Mae issued Lender Letter LL-2026-04 in April 2026 requiring formal AI governance frameworks from all sellers and servicers, effective August 6, 2026. Both GSEs are also transitioning to VantageScore 4.0 and FICO Score 10T, which consider rent payment history and trended credit data for more precise risk assessment.
Related Terms
Insurance Premium
An insurance premium is what you pay to keep your policy active. Learn what drives premium costs, 2025 price data, and how to lower your rates without losing coverage.
Actuary
An actuary is a professional who uses mathematics, statistics, and financial theory to assess and quantify risk for insurance companies and pension funds, calculating premiums, reserves, and the financial impact of uncertain future events.
Rider
A rider is an optional add-on to an insurance policy that expands or modifies your coverage. Learn which riders are worth the cost and which are oversold in 2026.
MBS
A mortgage-backed security is a bond-like investment backed by a pool of home loans, paying investors principal and interest as homeowners make mortgage payments, with agency MBS guaranteed by Fannie Mae, Freddie Mac, or Ginnie Mae.
Auto Insurance
Auto insurance covers financial losses from car accidents, theft, and vehicle damage, required by law in nearly every US state, with mandatory liability coverage protecting others and optional collision and comprehensive coverage protecting your own vehicle.
Insurance Claim
An insurance claim is a formal request to your insurance company for payment or coverage of a loss or medical expense covered by your policy, triggering the insurer's obligation to investigate and pay according to the policy terms.
Related Articles
What Is a Debt-to-Income Ratio and Why Do Lenders Care?
Your debt-to-income ratio is one of the most important numbers lenders look at. Here is how it is calculated, what it means for borrowing, and how to improve yours.
What Financial Independence Looks Like When You Have a Chronic Illness
More than 38% of Americans live with a chronic health condition. The financial impact is enormous: medical costs, lost earnings, and higher insurance needs. Here is what FIRE looks like when your health is not guaranteed.
Short Term Rentals vs Long Term Rentals: Which Makes More Financial Sense?
Short-term rentals generate 30 to 80% more gross revenue than long-term rentals in most US markets. But after expenses, the net advantage narrows to 20 to 35%. Here is the 2026 comparison with real market data.
House Hacking: How to Live for Free While Building Equity
House hacking lets you live in one unit of a multifamily property while renters pay your mortgage. With an FHA loan and 3.5% down, you can buy a $350,000 duplex for $12,250. Here is how it works in 2026.
What Medicare Actually Covers (And What It Does Not)
Medicare is not free and it is not comprehensive. Before you retire, you need to know exactly what Parts A, B, C, and D cover, what the costs are in 2026, and where the dangerous coverage gaps are.
