Your loss run is the only document in an equipment submission that was written about you by somebody else. It is also the one most owners have never read line by line. Reading it properly takes an afternoon, and it changes what you are able to argue at renewal.
What the document is, and who wrote it
A loss run is a report generated by one insurer listing the claim files it opened under your policies. That definition carries two limits worth stating out loud, because both of them mislead owners every year.
The first is that it is per insurer. It knows nothing about the years another company wrote you, so a fleet that has changed markets holds a history split across documents that never reference each other. A missing year is not a clean year; it is a year on somebody else’s report.
The second is that every substantive field on it is a judgment made by a claims professional who was not standing there. The amounts are estimates until the file closes. The cause description is a category somebody chose from a list. The dates are the dates the office recorded, not necessarily the dates things happened. None of that is dishonest, and all of it is arguable.
Read the reserves before you read the payments
Owners look first at what was paid, which is the least useful column on an open file. What an underwriter reads is incurred — paid plus the reserve still standing.
A reserve is the insurer’s current estimate of where a file will end up. On a recently reported claim it is often set generously, because the cost is unknown and under-reserving is the more expensive mistake for the insurer to make. That estimate then sits on your record, in full, being read as though it were a settled fact.
Which makes the open files the most productive place to spend your afternoon. A file that has been open for a long time with no activity, or one carrying a reserve that no longer resembles the damage, is a file worth asking the adjuster about. A reserve that comes down before your submission goes out lowers what every market sees, and nothing else on the document responds to a phone call that way.
Two dates, and the distance between them
Almost every loss run carries a date of loss and a date reported. Owners read the first and underwriters read the gap.
A short gap says the yard knows what its machines are doing and that information reaches the office. A long one says something was discovered late — which on an equipment file usually means the machine sat somewhere unattended, or that the operator did not mention it, or that a supervisor decided to see whether it mattered. All three are read as control problems rather than as luck.
The gap is also the one column you can improve without changing anything about how you work, and it improves prospectively: reporting on the day, every time, produces a record that reads better in three years regardless of what the claims themselves cost.
The cause code is a label somebody chose
Every file carries a cause or peril category, and equipment claims are miscoded more than most. Theft, vandalism, collision, overturn, water, fire, mechanical breakdown and “other” are not always distinguishable from a first phone call, and the code recorded at the outset frequently survives to the report unchanged.
It matters because coding drives how a market reads the pattern. Three collision files on a mixed fleet look like a hauling problem and invite questions about the transit and trailer transport side of the program. Three theft files look like a storage and access problem. Three overturn files look like an operator problem. The same three incidents, coded differently, produce three different underwriting conversations.
So read the description against your own memory of the event. Where the label genuinely does not fit what happened, say so in writing when you submit — you are unlikely to have the code changed, and you are very likely to have the explanation read.
Count the files before you total them
Frequency and severity are separate readings of the same list, and they are weighted unequally. One large file is generally read as an event; several small ones are read as a pattern, because severity contains a great deal of luck and frequency contains almost none.
This has a practical consequence most owners find counter-intuitive. A modest claim you could have absorbed yourself may cost more across the following renewals than it returned, precisely because it lands on the frequency count identically to a serious one. That arithmetic is the honest argument for a deductible set where you actually intend to use it — and it is the same arithmetic behind the damage waiver comparison, where the question is which record a small rental loss lands on.
Count the files per year, then per machine, then per operator if the data allows. A history that looks scattered by year is sometimes concentrated in one machine or one crew, which is a far better story than a random one because it names something you can fix.
Closed with no payment is not the same as nothing
Files closed without payment sit on the report looking like clutter, and they are doing real work in both directions.
Read favorably, they show a fleet that reports events promptly and lets the insurer decide, which is exactly the behavior an underwriter wants and rarely sees. Read unfavorably, and unexplained, a row of them says incidents keep happening on your sites. The difference between those two readings is entirely a matter of whether anyone told the underwriter which it was.
Real-World Scenario: An owner requests loss runs for a renewal and finds four files across the period, three of them small and one still open with a substantial reserve against a track machine damaged in a yard incident. He reads only the paid column, sees very little paid, and assumes the record is quiet. The market reads incurred, sees the open reserve as the dominant number on the document, and prices accordingly. The repair had actually been completed months earlier for far less than the reserve standing against it, and one call to the adjuster before submission would have closed the file — but nobody made the call, because nobody read past the paid column.
Recoveries and subrogation change the story late
Where another party caused the loss, the insurer may pursue them and recover part or all of what it paid. Recoveries show up on the report as offsets, and they arrive long after the file made its first impression.
Two things follow. Check whether a subrogated recovery has been credited against a file you remember as somebody else’s fault; an uncredited recovery overstates your history. And notice that a waiver of subrogation you signed in a contract can prevent that recovery entirely, which is one of the places contract language and loss history meet — worth reading alongside how not-owned exposure works in rented, leased or borrowed equipment coverage.
The errors that are actually fixable
Separate factual errors from judgments, because only the first category moves.
Factual and worth challenging: a claim listed against a machine you never owned, a duplicated file, a loss belonging to a different insured with a similar name, a loss recorded in the wrong policy period, or a machine described by the wrong serial. Bring documentation and ask for a corrected report.
Judgment and not worth challenging as an error: how the cause was coded, how the reserve was set, whether the file should have been opened. Those are answered with explanation rather than with correction, and explanation is what the submission narrative is for.
Reading the whole set at once
Lay every year side by side, from every insurer, and read down rather than across. What you are looking for is direction: whether frequency is rising or falling, whether the reporting gap is shortening, whether the same machine class keeps appearing, and whether the files that hurt are on the equipment floater side, the general liability side, or in workers compensation.
Those three read very differently. Equipment files raise questions about storage, hauling and operators. Liability files raise questions about the work and the sites. Crew injury files raise questions about supervision, and they are the ones most likely to push the discussion toward an umbrella layer or toward how the not-owned side of the program is built on rented and leased equipment.
Then write the note. Our track loader cost guide explains why a single machine class can dominate a history, and the Texas cost guide works the renewal question from the state side. If you want your own loss runs read before a market sees them, send them through the quote form, or read how we work first.