The freight market outlook | Freight Capacity Exit Index (FCEI)

Reporting period:

April 2026. Carrier exit is flat.

32,505 carriers had a cancellation filed, a rate of 76.20 per 1,000 insured carriers, against 75.35 a year earlier.

April 2026 is the newest month whose filings have actually arrived. The months after it are computed and held back rather than shown as a collapse, and the section below names them.

No rate forecast, and no dollar figure anywhere on this page. We hold no rate data at all.

Carrier exit is flat year over year

Every figure in this section counts carriers whose insurer filed a cancellation with FMCSA during the month, once each however many filings they had, against the insured carrier base domiciled in the same state. Both sides are counts of carriers.

Carriers with a cancellation filed, per 1,000 insured carriers. April 2026
StatePer 1,000CarriersInsured baseA year earlierChange
NJ103.971,38813,35095.80+8.17
CA101.415,48854,11593.54+7.87
FL101.212,18821,61899.49+1.72
TX95.014,00442,14593.98+1.03
NC92.781,35514,605108.23-15.45
GA87.091,39616,02989.76-2.66
SC84.235556,58983.90+0.33
DE83.771291,54070.55+13.22
MD83.176768,12889.50-6.33
AZ81.155867,22183.29-2.13
NV79.252803,53377.82+1.43
IN75.9193612,33172.02+3.89
VA75.296748,95270.58+4.71
PA74.951,48319,78771.56+3.39
OH72.651,33118,32175.05-2.40
MS67.372804,15666.83+0.55
TN67.175848,69567.85-0.69
MA66.324526,81564.36+1.96
CT65.052443,75164.61+0.44
IL64.741,61024,86964.10+0.64
CO64.583996,17864.36+0.22
AL63.883545,54265.76-1.88
UT63.392674,21259.37+4.02
NM62.781822,89958.64+4.14
LA62.241943,11765.58-3.34
OK59.343025,08961.02-1.68
OR58.833075,21852.69+6.14
RI58.43701,19856.65+1.78
MI57.565108,86152.02+5.53
MN56.555008,84148.26+8.30
KY54.632755,03460.26-5.63
WA54.5055510,18459.64-5.14
AR53.991883,48257.28-3.29
MO52.994568,60652.44+0.55
WY52.34941,79650.61+1.73
WV51.70851,64450.62+1.09
ID51.601793,46954.68-3.08
NH48.02751,56248.40-0.38
MT45.80851,85639.68+6.12
ND45.781102,40350.43-4.66
NY45.684449,71951.87-6.19
KS45.362275,00448.06-2.69
NE43.672024,62638.95+4.72
WI42.743267,62845.20-2.47
IA41.952906,91341.64+0.31
VT38.513077932.26+6.25
ME37.99531,39537.66+0.34
SD37.68852,25636.35+1.33
AK36.761540832.34+4.43

Source: FMCSA insurance filings. Rates are comparable across periods and geographies and are not national totals.

What the calendar says next

Model output, not measurement. These figures come from the seasonality model the FreightScout product uses to price a lane when a broker has no history of their own on it. The months that move and the direction they move in rest on published sources. How far they move is our own conservative estimate and has not been calibrated against delivered freight.

Each figure is a multiplier against a region's own average month, for September. A region reading 1.10 is expected to run about ten percent above what it averages across the year. The twelve months always average to 1.00, so this says nothing about whether one region is cheaper than another. It says when each one tightens.

Seasonal multiplier by region and trailer, September. Model output.
RegionDry VanReeferFlatbedStep DeckPower OnlyLTLIntermodalTankerHopperOther
Pacific Southwest1.061.021.031.03no driver1.06no driverno driverno driverno driver
Pacific Northwest1.061.091.081.08no driver1.06no driverno driverno driverno driver
Mountain West1.040.981.061.060.981.040.980.980.98no driver
Southwest Desert1.060.971.031.03no driver1.06no driverno driverno driverno driver
Texas and Gulf1.060.961.031.03no driver1.06no driverno driverno driverno driver
Plains1.040.981.061.060.981.040.980.980.98no driver
Upper Midwest1.040.981.061.060.981.040.980.980.98no driver
Lower Midwest1.06no driver1.031.03no driver1.06no driverno driverno driverno driver
Southeast1.060.961.031.03no driver1.06no driverno driverno driverno driver
Florida1.060.931.031.03no driver1.06no driverno driverno driverno driver
Mid-Atlantic1.06no driver1.031.03no driver1.06no driverno driverno driverno driver
Northeast1.040.981.061.060.981.040.980.980.98no driver
Western Canada1.040.981.061.060.981.040.980.980.98no driver
Central Canada1.040.981.061.060.981.040.980.980.98no driver
Atlantic Canada1.040.981.061.060.981.040.980.980.98no driver

No carrier rate is published

We are not publishing a carrier rate. There is no dollar figure on this page and there is not one behind it either.

What would put one here is a cross brokerage cost series, pooled across enough brokerages that no single book shows through it. That layer is built and it is empty. It stays empty until enough brokerages are contributing and running the same lanes, which is a question of how many customers we have rather than of engineering.

Connor's read: a higher floor, not a spike

This section is operator judgment, not analysis. Everything in it is one person's read on the market, written by hand. It is not model output, it is not derived from any figure elsewhere on this page, and no method produced it. Weigh it as you would weigh any operator's opinion.

Struck 2026-08 by Connor Miller, FreightScout.

Operator judgment

A higher floor, not a spike

55%
The call
Capacity keeps leaving faster than freight does. Coverage gets steadily harder on lanes that were already thin, and contract rates reset up through bid season without a spot blowoff. Brokers who bid last year's numbers get squeezed on award day.
What would have to be true
Tender rejections hold above 10 percent, net authority counts keep falling, and driver supply stays constrained by enforcement rather than by wages.
What would prove it wrong
Rejections fall under 8 percent for two straight months, or authority counts flatten and turn positive.

Operator judgment

Capacity crunch

25%
The call
Driver exits accelerate past what anyone modeled and coverage on secondary lanes gets genuinely hard. Shippers with weak route guides pay whatever the market asks, and spot re-tests its highs.
What would have to be true
Enforcement keeps pulling drivers out with nothing replacing them, rejections push past 15 percent, and carrier exits accelerate month over month rather than holding steady.
What would prove it wrong
Capacity re-enters. Authority grants turn positive, or rejections stall in the low teens instead of climbing.

Operator judgment

Demand air pocket

20%
The call
Industrial and construction demand contracts far enough that soft freight offsets tight capacity. Fewer trucks, but even less to haul, so coverage stays easy and the reset never arrives.
What would have to be true
Manufacturing and construction contract measurably, and volumes fall faster than capacity does.
What would prove it wrong
Volumes hold flat while capacity keeps shrinking, which is the base case.

What he is watching

The indicators this view gets revised against. Named in advance, so a revision is a change of mind with a reason rather than a quiet edit.

  • Net carrier authority
  • Insurance cancellations by state
  • Tender rejection trend
  • Fall contract bid prints

What is not on this page, and why

The months we are holding back

FMCSA filings for a month keep arriving long after that month closes. A recent period is not a quiet market, it is an incomplete one, and publishing it would show a sharp drop that quietly corrects upward later.

Months held back, and how much of each has arrived
PeriodArrived so farShare of a normal monthStatus
May 202618,07055.9%Held back
June 20261,8815.9%Held back
July 2026290.1%Held back

The rate forecast we are not making

The obvious next section is a forward view of national carrier rates with a confidence split. We are not publishing one, and the reason is not caution. We hold no rate data at all, our pricing engine prices lanes from history rather than forecasting them, and carrier exit measures attrition rather than price. A scenario split would also assert a fitted model with a track record, and we have published no forecasts, so we have no forecast error to report.

How this is built

Carrier exit is counted from FMCSA insurance filings. The full methodology, including the sample floors, the arrears gate and what is withheld, is at /methodology.

The seasonal model reads no brokerage data of any kind, which is why it can be shown in public.