Auto Insurance Loss and Policy Analysis

Turn a fleet auto submission - declarations, vehicle and driver schedules, and multi-carrier loss runs - into one cited view with frequency, severity, and loss ratio per power unit, and hand the underwriter a renewal recommendation in minutes.

Insurance

What it extracts

15 extraction fields

Account and Fleet Summary
A one-look roll-up of the whole auto submission - named insured, coverages, carriers, policy periods, valuation dates, power-unit count, claim counts by status, paid/reserves/incurred, largest claim, and loss ratio if premium is provided - so a reviewer can orient before reading detail.
Policy and Coverage Schedule
The policy declarations for every term - effective/expiration dates, carrier, covered-auto symbols, liability limit, collision and comprehensive deductibles, UM/UIM limit, power units, and premium by coverage - so a reviewer sees how coverage, exposure, and price changed year over year.
Vehicle Schedule
The covered-auto inventory (ACORD 127/129 style) - VIN, year/make/model, vehicle class, weight, garaging state, radius, stated value, and coverages applied - which drives physical-damage rating and exposure.
Driver Schedule and MVR Review
The driver roster with motor-vehicle-record detail - license state/status, experience, violations, at-fault accidents, major convictions, and an eligibility flag - because driver quality is a primary predictor of auto loss.
Fleet Safety and Telematics Profile
The fleet's safety and telematics context - DOT/MC number, FMCSA CSA/SMS indicators, reportable crashes, telematics/ELD provider and behavior rates, safety-program elements, and prior-coverage continuity - that explains loss trends and supports renewal credits or debits.
Normalized Loss Register
Every auto claim from every carrier and format normalized into one row-per-claim register - coverage type and Code, cause, status, paid/reserves/recovery, and footed total loss, expense, and incurred - so mixed-format loss runs finally line up.
Loss Summary by Policy Year
Loss experience per policy year aligned to exposure and premium - counts, paid/reserves/incurred, liability-vs-physical-damage split, frequency and incurred per power unit, and loss ratio where premium is known - noting the most recent year is immature.
Loss Summary by Coverage Type
Losses grouped by coverage - Auto Liability versus Physical Damage (collision, comprehensive, glass) - with counts, incurred, average severity, and share of total incurred, so an underwriter sees which coverage drives the account.
Large and Shock Losses
The individual auto claims an underwriter must read - those above the large-loss threshold and isolated shock losses (usually serious bodily-injury or multi-vehicle liability) - with the open reserves that could still develop.
Frequency and Severity Trend
The per-year frequency and average-severity trend normalized per power unit, with a pure-premium (incurred per power unit) view, so different-sized terms compare fairly and the underwriter can read whether the risk is improving or deteriorating.
Cause of Loss Analysis
Auto claims grouped by cause - rear-end, intersection, backing, single-vehicle, comprehensive (glass/theft/weather), third-party bodily injury and property damage - revealing operational or training exposures a cause concentration points to.
Open Claims and Reserve Adequacy
Open exposure and reserve adequacy by year - open-claim ratio, reserve per open claim, and open liability-claim count - flagging books that could develop adversely after binding because auto bodily-injury claims settle slowly.
Data Quality and Validation Flags
The integrity checks an analyst runs before trusting the file - per-claim loss/expense/incurred footing (Total Loss = O/S + Paid - Recovery; Total Incurred = Loss + Expense), coverage/code consistency, policy-period continuity, mixed valuation dates, power-unit reconciliation to the vehicle schedule, limits adequacy, and loss ratio - each flagged Pass, Fail, or Review.
Fleet Renewal Recommendation
A single underwriting decision for the account - preferred, standard, renew with rate increase, refer to senior underwriting, or decline/non-renew - driven by loss ratio and per-power-unit trend, large/shock losses, reserve adequacy, driver quality, and safety profile, with a cited rationale.
Underwriting Summary
The underwriter-facing synthesis at the top of the file - overall experience, exposure-adjusted frequency vs severity, coverage mix and large losses, driver and safety read, data caveats, and the follow-up items to request - tied to the renewal recommendation.

Where it fits

Commercial auto / fleet underwriting intake and renewal

Upstream

Submission intake - policy declarations, vehicle and driver schedules, and loss runs arriving with a new-business or renewal submission, often from multiple prior carriers in mixed formats (PDF, Excel, CSV)

This step

Auto policy and loss-run analysis

Downstream

  • Pricing and rating
  • Quote and bind
  • Renewal / referral / declination decision
  • Loss control referral

What it needs

Documents

  • Auto / fleet policy declarations
  • Vehicle (auto) schedule (ACORD 127 / 129)
  • Driver schedule and motor vehicle records (MVRs)
  • Multi-carrier loss runs (current and prior, 3-5 years)
  • Telematics / FMCSA safety reports (where available)

Systems

  • Email
  • Underwriting workbench or policy admin system
  • Document repository

Prerequisites

  • Loss runs for the account, ideally covering the last three to five policy periods
  • Declarations pages showing limits, deductibles, and power-unit counts
  • The vehicle and driver schedules
  • Earned premium by term, if a loss ratio is required (loss runs rarely include it)
  • The carrier's large-loss / shock-loss thresholds and driver-eligibility criteria, if they differ from the defaults

What it produces

A normalized, cited auto loss register plus policy/coverage, vehicle, and driver schedules; by-year and by-coverage summaries; large/shock-loss and reserve-adequacy flags; data-quality checks; a fleet renewal recommendation; and an underwriting summary

Delivered to

  • Underwriting file
  • Pricing / rating worksheet
  • Renewal, referral, or declination memo

Review model

An underwriter reviews the flagged large and shock losses, the reserve-adequacy and data-quality flags, the driver MVR exceptions, and the cited loss register before the loss history and renewal recommendation are trusted for pricing.

Who uses it

Commercial Auto UnderwriterFleet UnderwriterUnderwriting AssistantManaging General Agent (MGA) UnderwriterLoss Control / Fleet Safety Analyst

Volume fit

Works best for

fleet and commercial-auto underwriting teams processing multi-carrier submissions where declarations, schedules, and loss runs each arrive in a different format and turnaround is measured in days

Too small for

a single private-passenger auto or one-vehicle risk with a single clean loss run and no normalization pain

Grounded in

  • ISO Business Auto Coverage Form (CA 00 01) - covered-auto symbols on the declarations page designate which vehicles each coverage (liability, physical damage, UM/UIM) applies toverified as of 2026-07-22
  • ACORD 127 Business Auto Section (and ACORD 129 for additional vehicles) - standard vehicle and driver schedule fields (VIN, year/make/model, garaging, driver license and MVR data)verified as of 2026-07-22
  • Loss identity: Total Loss = Outstanding (O/S) Loss + Paid Loss - Recovery; Total Incurred = Total Loss + Total Expense; equivalently Incurred = Paid + Reserves (Outstanding)verified as of 2026-07-22
  • Loss ratio = Incurred Losses / Earned Premium; exposure measured per power unit (covered-vehicle car-year)verified as of 2026-07-22
  • FMCSA Compliance, Safety, Accountability (CSA) / Safety Measurement System (SMS) BASICs and financial-responsibility minimums (49 CFR Part 387) for regulated motor carriersverified as of 2026-07-22

Changelog

  • July 2026

based on a production deployment at a fleet & auto insurance provider

Related agents

Browse all agents →

See Auto Insurance Loss and Policy Analysis on your documents

We'll run it against a file of yours and walk through every cited field.