Phantom fleets: 2,610 active US carriers running 3× more trucks than they declared

Investigation · Carrierintel data desk · Six weeks of reporting

The man with two trucks and five hundred and fifty-one VINs

The address on the FMCSA file for GOLLA GROUP LLC is a residential lot in Davenport, Washington, a wheat-country town of 1,700 people on the eastern edge of the state. The carrier's USDOT (number 4,514,820) was issued in 2024. The MCS-150 biennial filing lists 2 power units and 2 drivers. The one named officer on the registration is Willie Charles II Barnes, listed simply as "Manager."

In the twelve months ending this week, those two trucks were stopped at roadside inspections 714 times in 46 different US states. The inspectors recorded 551 distinct 17-character tractor VINs in the power-unit slot, with trailers excluded. 385 of those 714 stops ended with the truck placed out of service: a 54% OOS rate, more than double the national average.

A truck has one VIN. It has one VIN for its entire physical life. So GOLLA GROUP either owns 549 trucks it failed to tell FMCSA about, or, far more likely, the USDOT number is being rented, sold, or otherwise lent to a rotating cast of independent operators who use it to move freight under the cover of someone else's operating authority. In the industry this is called a chameleon carrier. The Carrierintel data desk has spent the last six weeks measuring how widespread it really is. The answer is: much more than anyone has been willing to say out loud.

What the numbers actually say

We joined every US roadside inspection from the last 12 months to each carrier's MCS-150 and to its full operating-authority history. Then we dropped, irreversibly, every carrier whose latest authority event is revoked, out-of-service, or inactive. What's left in this dataset are carriers with valid US operating authority today. The threshold for inclusion is intentionally conservative:

With those filters, the residual national set is 2,610 currently-active US carriers, spread across 64 states and territories. Together they declare 4,330 power units and were inspected with 29,078 unique tractor VINs, representing a real-world fleet roughly 6.7× the size of the paper one. 412 of these carriers run 10× or more. 66 run 25× or more. Seven exceed 100×, led by GOLLA GROUP at 275×.

17,164 of the 54,244 inspections in this set were out-of-service. That's a 31.6% OOS rate, about 50% above the national vehicle baseline, and a direct measurement of what happens when a dozen unrelated operators share one maintenance program (or no maintenance program at all) under a single authority number.

Where the shells live

The popular assumption is that chameleon carriers cluster in New York and New Jersey. The popular assumption is wrong. The largest absolute concentrations of 3×+ phantom fleets sit in big trucking states with high rates of new-entrant LLC formation.

Active phantom-fleet carriers (3×+ tractor-VIN-to-declared ratio) by physical-address state, top 10. 12-month window ending June 2026. Active operating authority only; anything currently revoked, out-of-service, or inactive is excluded. Refreshed nightly from FMCSA inspection-unit and authority-event data.

A few notes on the geography. California and Texas combined account for 826 carriers, nearly a third of the national 3×+ set, which is roughly proportional to their overall share of US trucking. Illinois punches above its weight: many of the worst single-truck shells in the country are small-town Illinois LLCs with Eastern-European officer names.

The names behind the worst twelve

Here are the twelve most extreme phantom fleets active in the United States right now: every one of them with valid operating authority today, every one of them a carrier a broker or insurer could legally book a load with tomorrow morning. We've included the primary officer on each FMCSA registration. These are the human beings whose signatures are on the paperwork.

Source: FMCSA roadside inspection-unit data joined to the carrier census and full authority-event history. 12-month window. VIN counts are unique power units only. "States" is the number of distinct US states where the carrier's DOT appeared at a roadside inspection. "Officer" is the primary named officer on the carrier's FMCSA registration; an empty value means none is publicly recorded.

The patterns inside this table are not subtle:

How the scheme actually works

The mechanics are, depressingly, not complicated. A new LLC is formed and applies for a USDOT and MC number. New-entrant review takes 18 months, and during that window the authority is fully operational. Once the DOT is live, the LLC's owner (or, more often, a broker specializing in this) sells access to the authority to independent operators who, for any number of reasons, cannot or will not carry their own.

Each operator pays a flat weekly fee (typically $300 to $700 in the underground market we've documented) and a percentage of revenue. In exchange they get to put the chameleon's USDOT number on the door of their tractor, pull loads under that carrier's MC authority, and, critically, be covered by that carrier's liability policy. The insurance is the single most valuable thing the shell sells.

When the CSA score gets ugly, which it does fast at these volumes, the shell is allowed to die quietly and a new LLC is formed at the same address by a different relative. The new DOT shows up clean and the cycle starts over. That's why the median age of a top-50 phantom-fleet DOT in this dataset is under 24 months. The shells rotate; the underlying network of operators does not.

Why this matters to everyone else

Phantom fleets aren't a curiosity. They quietly distort almost every downstream system that consumes FMCSA data:

What the data can't tell us (yet)

We can prove the gap exists at scale, with names. The harder reporting question is why any one carrier shows the gap. Three benign explanations exist and we want to name them:

Even after allowing for all three, the residual set of clearly chameleon-style phantom fleets in the United States is in the low thousands, and the worst 500 carriers in this dataset have no legitimate explanation that survives ten minutes of scrutiny.

How to check your own exposure in five minutes

If you're a broker, shipper, insurer, or factoring company, three checks will catch most phantom fleets before they touch your business:

How we reported this

Every figure in this article comes from public FMCSA data: the carrier census (MCS-150), the operating-authority history, and the roadside inspection-unit records that capture a VIN at every stop. We pull those datasets nightly, join them in our own pipeline, and filter to carriers with currently-valid US operating authority. No private data, no interviews used to source numbers; every claim in this piece is reproducible from raw FMCSA files by anyone who's willing to build the joins.

Open the live phantom fleets report to drill into any carrier in the dataset, or look up a specific USDOT in our carrier database . The "Fleet vs Reality" card on every active carrier profile runs the same check, on demand, for any DOT in the country.

This was the kind of analysis that should have existed five years ago. The data was always public. Nobody had bothered to join the right tables, nationally, and then look at who actually signed the paperwork.