Your Renewal Is Priced on Evidence: Trucking Liability Insurance Decisions Before 2027
Liability premiums keep outrunning crash rates. How carriers and brokers should set retention, excess limits, and safety evidence before 2027 renewals.
The renewal you lose in October
If your auto liability renews January 1, the submission is roughly 45 to 60 days from going out. That is where most of the money gets won or lost. A bad loss run usually isn't what costs a fleet. More often it's a fleet that bought cameras, ran coaching, tightened hiring, and still walks into renewal with nothing an underwriter can actually price. The safety program exists. The evidence lives in five systems, two spreadsheets, and a safety director's memory. The underwriter sees a loss run, a CSA snapshot, and a schedule of units, and prices the fleet like everyone else in its size band.
In 2026, being priced like everyone else is expensive. The decisions that matter this quarter are how much risk to keep, where your tower stops, and what proof you can put in front of a market. Brokers have their own version of the problem after this spring's Supreme Court ruling, and hybrid shops get both.
The market you are renewing into
The numbers have stopped being a cycle and started looking structural.
Premiums are still the fastest-rising cost line. The industry's annual operational-costs benchmark (July 2026 update) puts average truck insurance premiums at 10.6 cents per mile for 2025, up 3.9% from 10.2 cents in 2024. First-quarter 2026 data showed premiums up another 6.4% over 2025, the largest increase of any line item in that quarter.
Premiums are rising while crash rates fall. A May 2026 industry research report on trucking insurance costs found liability premiums rose 18.6% from 2021 to 2024, while heavy-duty truck-involved crash rates fell 2.6%. Among surveyed fleets, per-mile liability losses rose 33.1% over the same period.
Excess is where the pain concentrates. The same report found per-mile premiums for the $5 million to $10 million layer rose 34% (to 1.58 cents) and the $10 million to $15 million layer rose 45% (to 1.05 cents) from 2021 to 2024.
Insurers are still losing money on the line. Commercial auto combined ratios have been above 100 every year since 2014 except 2021, per insurance rating agency data cited in that report: 109.2 in 2023 and 107.2 in 2024. Some improvement showed up in 2025, but liability remains the drag.
Size still buys price. In 2024, fleets of 5 to 25 trucks paid 20.3 cents per mile in premiums, fleets of 26 to 100 paid 16.1 cents, and fleets over 1,000 trucks paid 6.6 cents.
Then there's the severity backdrop. A litigation-tracking report counted 135 corporate verdicts above $10 million in 2024, totaling $31.3 billion, with a median of $51 million. The federal minimum for general freight has been $750,000 since 1985. Your primary policy covers a small slice of what a bad venue can do.
So frequency isn't what's pushing price. Severity and litigation are. A fleet that can only show fewer crashes is answering a question the underwriter isn't asking. The question is whether you can show you control the events that turn into eight-figure claims, and whether you could defend them in a deposition.
Decision 1: How much risk to keep
The most useful finding in the May 2026 research is easy to miss. Fleets that kept more risk in their primary layer had lower combined liability losses and premium costs, regardless of size. Fleets that reduced total purchased coverage saw an average 2.4% drop in combined losses and premiums the following year, after inflation.
That is not a blanket case for raising your deductible. Retention works when three things are true:
You can fund it. A higher self-insured retention moves premium dollars into reserves. If a $250,000 retention would wipe out a quarter of your cash, you've traded premium for solvency risk.
You can handle claims. Retained losses mean you, your TPA, or your defense counsel control early response. The fleets that benefit get to the scene, preserve video, and set reserves within days, not weeks.
Your loss data is granular enough to price your own risk. If you can't tie losses to driver, lane, customer, equipment, and time of day, you're guessing at retention just like the market is guessing at you.
Asset carriers with stable mileage and a working claims function are the natural candidates. Small fleets with lumpy mileage usually aren't, unless they're joining a captive or group program, and that requires the same data discipline.
Decision 2: Where the tower stops
Excess limits are now the hardest call on the renewal, because those layers repriced fastest and they're what responds to a verdict. Too little excess bets the company on venue. Too much spends margin protecting against exposure your network doesn't actually carry.
Answering that well takes something most shops don't track: miles and stops by state and, ideally, by county. Venue exposure isn't evenly spread. Texas HB 19 (2023) let commercial vehicle defendants bifurcate trials so the liability and damages phases are tried separately. Florida HB 837 (2023) cut the negligence statute of limitations from four years to two and moved the state to a modified comparative negligence bar. Several other states, Georgia among them in 2025, have since passed their own tort packages. Underwriters are starting to price those differences, but unevenly, and only if you show them where your trucks actually run.
A regional carrier that runs 70% of its miles in reformed states has a different tower argument than one cutting through the nation's most aggressive plaintiff venues every day. If your lane data can't produce that split in an afternoon, you can't make that argument.
Decision 3: What evidence you can actually prove
Underwriters increasingly treat camera coverage, telematics, and CSA trends as pricing inputs, not nice-to-haves. The May 2026 research found that six specific safety technologies were statistically associated with lower per-mile liability losses. But installed technology is table stakes. What earns credit is proof that the technology changes behavior.
A submission-ready evidence package usually includes:
Camera and telematics coverage by unit, with exceptions explained (units out of service, new adds pending install).
Event-to-coaching closure rates. How many flagged events were reviewed, coached, and closed, and how fast. An open queue of 400 unreviewed events is worse than no cameras, because it shows you knew.
BASIC trends with root cause. Not just your percentile, but what you changed after a run of HOS or maintenance violations and what happened after.
Driver hiring and retention data. MVR thresholds, exceptions granted, who approved them, and turnover in the first 90 days.
Loss runs with your own annotations. Reserves, closure status, and what changed operationally after each serious claim.
The failure mode is producing this by hand every year. The fleets that renew well can pull it on demand because the evidence is captured as a byproduct of daily work, not reconstructed in November.
Brokers: Montgomery changed the selection file
On May 14, 2026, a unanimous Supreme Court held in Montgomery v. Caribe Transport II that the FAAAA's safety exception preserves state negligent-selection claims against brokers. The preemption defense many brokers relied on is gone for claims tied to motor vehicle safety.
For brokerage leaders, that turns the carrier selection decision into a discoverable record. Renewal for contingent auto and broker liability coverage will increasingly ask what your selection standard is, how consistently you apply it, and what happens when someone overrides it at 4:45 p.m. on a Friday to cover a load.
The questions to answer before your submission goes out:
What safety criteria gate a carrier from being tendered, and are they enforced by the system or by memory?
When a rep overrides a gate, is the reason recorded, and who has permission to approve it?
Can you reconstruct, for any load, what you knew about that carrier at the time of tender?
Do your contract minimums for carrier insurance reflect the verdict environment, or a number set years ago?
A selection standard you can't prove you applied is worse than a looser one you followed every time. Plaintiffs' counsel will compare your written policy to your actual tenders.
Asset, broker, and hybrid: different exposures, same problem
Asset carriers own the primary exposure and the tower decision. Their lever is retention plus evidence: keep what you can fund and handle, and prove the controls that limit severity.
Brokers own selection risk and contingent exposure. Their lever is a defensible, consistently applied qualification process and contract terms that push appropriate limits to carriers.
Hybrid shops carry both, and are the most likely to have mismatched standards. If the brokerage desk tenders to carriers that would never pass the asset side's own driver standards, that inconsistency becomes an exhibit. Hybrids should decide on purpose whether the two standards should match, and document why if they don't.
Where your operating system has to carry the weight
No two fleets have the same risk profile. Lane mix, freight type, customer facilities, driver pool, equipment age, and claims history combine differently in every shop, and underwriters are finally pricing that variation. The operations that benefit are the ones whose workflow reflects their actual risk, not a generic template.
In practice that means configuring the system you dispatch and tender from to capture evidence as work happens:
Carrier and driver qualification gates that match your written standard, with override reasons and approval permissions recorded on the load.
Lane and stop tagging by state or venue, so miles by jurisdiction is a report, not a project.
Safety event statuses that move from flagged to reviewed to coached to closed, with timestamps and owners.
Claims linked to loads, drivers, and customers, so loss analysis can drive retention and pricing decisions.
Renewal-ready reporting your broker or risk manager can pull in hours, not weeks.
None of that is exotic. It's just your own rules, encoded where the work happens instead of in a binder.
A 60-day renewal plan
Weeks 1 and 2: Pull miles by state and county, current loss runs, and open reserves. Decide whether retention is on the table this cycle.
Weeks 3 and 4: Close or explain the coaching backlog. Document BASIC root causes. Brokers should audit the last 90 days of tenders against their written selection standard.
Weeks 5 and 6: Build the evidence package and model two or three tower structures with your agent or broker, including a higher retention option.
Weeks 7 and 8: Go to market with a narrative, not just a loss run. Show the underwriter why your risk is not the average of your size band.
FAQ
Why are trucking insurance premiums rising when crash rates are falling?
Because claim severity, not crash frequency, is driving cost. Industry research found liability premiums rose 18.6% from 2021 to 2024 while truck-involved crash rates fell 2.6%, as per-mile liability losses rose 33.1% on litigation and larger verdicts.
Should a trucking company raise its deductible or self-insured retention?
Only if it can fund the retained losses, handle claims quickly, and analyze its own loss data. May 2026 research found fleets with more retained primary risk had lower combined costs, but retention without cash and claims capability just moves risk onto the balance sheet.
How much excess liability coverage does a fleet need in 2026?
It depends on where your trucks run and what they haul. Excess layers repriced fastest from 2021 to 2024 (up 34% to 45% per mile), so tie limit decisions to miles by venue and freight severity rather than to a peer benchmark.
What did Montgomery v. Caribe Transport II decide for freight brokers?
The Supreme Court ruled 9-0 on May 14, 2026 that state negligent-selection claims against brokers are not preempted by federal law when they concern motor vehicle safety. Brokers should expect their carrier selection process and override history to be scrutinized in litigation and at renewal.
What do underwriters want to see in a trucking insurance submission?
They want proof that safety controls change outcomes. That means camera and telematics coverage, event-to-coaching closure rates, BASIC trends with root causes, driver hiring exceptions, and annotated loss runs, all producible on demand.