Chapter 23 — Key Takeaways
A one-page field card for commercial auto and fleet underwriting. Numbers are illustrative.
The one thing to remember
The problem is severity, not frequency — and you insure the drivers, not the trucks. Commercial auto loses money for the industry because the rare, enormous bodily-injury verdict lives in a tail that no rate or control can price away. The craft is severity discipline: hard driver selection, adequate rate, adequate limits and attachment, and required-and-monitored controls.
Core claims
- Commercial auto is a severity line. Frequency is flat or falling; the line runs unprofitably anyway because losses have migrated into the tail. One bad loss can exceed a small fleet's lifetime premium.
- Symbols decide coverage. On the business auto policy, the coverage symbol (1 = any auto; 7 = described autos; 8 = hired; 9 = non-owned) determines what's covered. Write liability broad (symbol 1) so the hired-and-non-owned exposure is covered and the umbrella attaches cleanly.
- "Twelve units" is not an exposure. Read the fleet's composition: vehicle type and weight, use, radius, and cargo. Weight is destiny in severity.
- Fleet rating is an inflection point. Above a size threshold, the fleet's own (partially credible) loss experience starts driving its price; below it, you rate off the class. Know which side an account falls on.
- MVRs are necessary, not sufficient. Read for pattern, not count. A recent major violation (DUI, suspension, hit-and-run) on a heavy unit is disqualifying — remove the driver, not a debit.
- Underwrite the process, not just the roster. The roster turns over; the insured's driver-qualification system is what governs the loss future.
- Radius proxies severity drivers — time on the road, speed, fatigue. Verify it against the loss run; it's easy to understate and expensive to get wrong.
- Pull the FMCSA data. For DOT-regulated fleets, the public SAFER/SMS profile (crash history, out-of-service rate, safety percentiles) is a regulator's third-party read that tests the application's story.
- Don't miss hired-and-non-owned. Almost every account has it; few disclose it. Respondeat superior lets a plaintiff reach the employer for an employee's at-fault crash in a personal car on company business.
- Telematics is the best lever — as a program, not a box. It manages the frequency you can influence and produces evidence (a dashcam that exonerates a driver can pay for the whole program). Required and acted on, it's load-bearing; bought and ignored, it's a line item.
The rule of thumb
A clean loss history records the frequency you can see, not the severity you can't. A spotless fleet and a fleet that draws a nuclear verdict next year can be the same risk today. Charge for the tail.
Key terms
Commercial auto · fleet rating · radius of operations · hired & non-owned auto (HNOA) · nuclear verdict · driver qualification file
What you could defend to your manager
"I wrote the Harbor Steel auto line at an adequate, severity-aware rate for heavy haulers on an intermediate radius. I wrote liability broad on symbol 1 so the hired-and-non-owned exposure is covered and the \$10M umbrella attaches cleanly. I removed the one driver with a recent major violation as a condition of coverage, and I required telematics with active coaching and forward-facing cameras — and I'm monitoring that it's actually run. I can't price the nuclear-verdict tail away; nobody can. What I can do, and did, is reduce serious-crash frequency, build the evidentiary record, and make sure the limits and attachment are right when the tail event lands. The clean two-claim history is not a reason to shade the rate — it's the absence of a tail event, not proof one can't happen."