C.H. Robinson is North America's largest freight broker. Shippers (Walmart, an auto plant, a produce grower) need goods moved. Roughly 450,000 small trucking companies have empty trucks. Robinson sits in between, buys the truck's capacity, sells it to the shipper at a markup, and keeps the spread. This dashboard explains that business from first principles: how the spread is earned, why it whipsaws with the freight cycle, how "Lean AI" changed the cost base, why a $604M Texas jury verdict is now the dominant debate, and what the stock is worth under bull and bear cases.
Assume you know nothing about logistics. Here is the whole industry in five ideas.
About 70% of U.S. freight by weight moves by truck. The supply side is hundreds of thousands of tiny carriers: Robinson's own network has ~450,000 contract carriers, and roughly 72% of its truckload shipments are hauled by carriers with fewer than 100 trucks Bernstein via AlphaSense primer. A shipper with 5,000 loads a week cannot phone 450,000 owner-operators. A one-truck owner-operator cannot find Walmart. Someone has to match them.
Robinson never owns the truck. For each load it quotes the shipper a sell rate and then finds a carrier willing to move it for a buy rate. The difference is called net revenue or adjusted gross profit (AGP).
Why "gross revenue" is misleading: the $16.2B headline includes the $13.5B Robinson simply passes through to carriers. Analysts value the company on the $2.7B of AGP and on operating income as a percentage of AGP (Robinson calls this "adjusted operating margin"). When you see "NAST margin 40.9%", that is operating income ÷ AGP, not ÷ gross revenue.
The DAT dry van spot rate (a benchmark from load-board operator DAT) is the market's real-time price of a truck. In August 2026 the dry van spot linehaul rate was ~$2.25 to $2.32 per mile ex-fuel and the contract rate ~$2.42 DAT, Aug 2026. In mid-2025 spot was ~$1.80.
This is the single most important mechanic in the stock.
The mirror image is true in a downturn: when rates collapse (2023), contract sell rates stay high for a while and the spread fattens even as volume falls. That is why Robinson's net revenue margin moves inversely to spot rates in the short run.
Drag the spot-rate shock. The chart shows how buy cost jumps immediately while the contract sell rate catches up over three quarters, and what happens to gross margin per load. Illustrative mechanics calibrated to 2Q26 (price +25.5%, cost +29%, margin 13.1%).
3PL = third-party logistics provider (an outsourcer that arranges transport). NAST = North American Surface Transportation, Robinson's trucking segment. Global Forwarding = ocean and air freight, plus customs clearance. Cass Freight Index = a monthly gauge of U.S. shipment volumes from freight-bill payer Cass. Full glossary at the bottom.
A 121-year-old produce broker that became a $16B logistics platform. Asset-light means ~85% of operating expenses are variable purchased transportation, which is why it has never posted an operating loss through a freight downturn CFRA via primer.
Manufacturers, retailers, food companies, auto plants, energy firms. The largest customer is only ~2% of revenue 10-K FY2025. Management is prioritising "key verticals": automotive, retail, healthcare, energy and tech, where cross-border, drop-trailer and seasonal-surge complexity favours a scaled broker Investor Day, Deutsche Bank.
Mostly small fleets and owner-operators. Robinson is their sales force: it finds them loads, pays them fast (often with cash advances), and gives them a mobile app (Navisphere Carrier / Driver) to book freight and upload paperwork. For a one-truck operator, Robinson's load board is a primary way to find backhauls Expert call via primer.
Robinson must check each carrier's authority, insurance and FMCSA safety rating. After Montgomery (May 2026), how well it vets carriers is now a litigable question in state courts, which is the crux of the legal section.
An internally built, multi-modal transportation management system launched in 2012 and funded by a 2018 pledge to spend $1B over five years on technology StoneX via primer. It is the system of record for shippers (Navisphere Vision: tracking, cost per mile analytics), carriers (Navisphere Carrier) and drivers (Navisphere Driver: geofenced status updates, proof of delivery).
Because the data is proprietary, Robinson trains pricing and routing algorithms on 37 million shipments a year of transaction history, which sub-scale competitors cannot replicate TD Cowen, Jefferies.
| Line | $M | What it means |
|---|---|---|
| Gross revenue | 16,233 | Everything shippers paid |
| − Purchased transportation | (13,504) | Paid out to carriers, ocean lines, airlines |
| = Adjusted gross profit (AGP) | 2,729 | The spread. 16.8% of gross revenue |
| − Personnel expense | ≈(1,300) | 11,855 employees. Biggest controllable cost |
| − SG&A (ex D&A) | ≈(500) | Offices, tech, insurance & claims (~$60M), other |
| − D&A | (103) | Mostly capitalised software |
| = Adjusted operating income | 834 | 30.6% of AGP; 5.1% of gross revenue |
| − Interest, tax (18.8%) | (216) | Net debt ~$1.2B |
| = Net income / EPS | 618 / $5.09 | 121M diluted shares |
Robinson historically ran as a federation of branch offices, each with its own customer relationships and pricing. Bozeman collapsed this into one go-to-market, one pricing engine, and centralised carrier procurement (separate from sales, unlike TQL where the same person does both) UBS, Expert calls.
Over 30 proprietary generative-AI agents now quote, classify, book, track and handle exceptions: 3M+ tasks done without a human. Quote response fell from 17–20 minutes to under 32 seconds; RFQ response coverage rose from 60% to 100% Mizuho, CFRA. Token cost is only ~$1.0–1.5M a year GS HQ visit, Aug 2026.
The "North Star" set at the December 2024 Investor Day: grow 2026 adjusted operating income by $220M vs 2024, raised in October 2025 to $336M, giving the $964M–$1.04B 2026 range. The low end is modelled to deliver ~$6.00 EPS even with a −3% market Press release, Deutsche Bank.
NAST is the trucking business and ~70% of AGP. Global Forwarding is ocean, air and customs, ~25%. Robinson Fresh (produce) and Managed Solutions (4PL/TMS) make up the rest. Europe Surface Transportation was sold to sennder in February 2025.
What: truckload and LTL brokerage across the U.S., Canada and Mexico. ~$11.6B annual gross revenue, more than 50% larger than #2 TQL and ~3x RXO Deutsche Bank. Roughly 14–15% share of a fragmented U.S. truck brokerage market Jefferies.
2Q26: revenue $3.59B (+23.1%), AGP $469M (+8.6%), gross margin 13.1% (−170bp), adjusted operating income $192M, margin 40.9%. Volume +1.5% (TL +0.5%, LTL +2.0%) vs Cass shipments −3.3%: the 13th straight quarter of outgrowing the market.
Inside the quarter: truckload AGP per shipment held flat despite a 34% surge in dry van spot rates (Deutsche Bank called this "highly impressive"); LTL AGP +21.7% on +19.5% AGP per order as shippers pushed freight from TL to LTL to dodge rising truckload costs.
What: Robinson does not own ships or planes. It buys container slots and air cargo capacity in bulk from carriers (Maersk, CMA CGM, airlines) and resells them to importers and exporters, then clears customs. Built via the 2012 Phoenix International deal ($635M) and APC Logistics (Australia, 2016).
2Q26: revenue $897M (+12%), AGP $189M (+0.7%), gross margin 21.1% (−250bp), operating income $63M, margin 33.4% (+470bp y/y). Air AGP +23.4% (AGP per ton +33.5%, tons −7.5%) as armed conflicts tightened air capacity; ocean AGP −2.8% on softer rates UBS, Bernstein.
Why it matters: forwarding is where the Lean AI playbook is only in the "first inning" (vs "third inning" in NAST), so bulls see the most productivity runway here Evercore. It is also the segment exposed to tariffs and transpacific volumes.
Robinson Fresh ($47M AGP in 2Q26) is the 1905 legacy: buying, selling and marketing fresh produce for grocers and foodservice. A new GFSI-certified produce logistics centre opened in South Texas in May 2026 for cross-border volumes.
Managed Solutions / TMC ($32M AGP) sells 4PL managed transportation and Navisphere TMS technology. Low capital, sticky, and a funnel that feeds contract and spot freight into NAST.
Europe Surface Transportation (road brokerage in Europe, once headquartered in Poland via the 2012 Apreo deal) was sold to Berlin-based digital forwarder sennder in February 2025, removing a low-margin, sub-scale business and part of the reason 2025 gross revenue fell 8.4%.
| Target | Segment | Closed | Value ($M) | Rationale |
|---|---|---|---|---|
| American Backhaulers | NAST | 1999 | 136 | Early truck brokerage scale |
| Phoenix International | Global Forwarding | Nov 2012 | 635 | Doubled ocean forwarding; foundation of GF |
| Freightquote.com | NAST (LTL) | Jan 2015 | 365 | Digital LTL matching, SMB shippers |
| APC Logistics | Global Forwarding | Sep 2016 | 225 | Oceania region |
| Milgram & Co. | Global Forwarding | Aug 2017 | 50 | Canadian customs brokerage |
| Prime Distribution Services | NAST | Mar 2020 | 223 | Retail consolidation, warehousing |
| DeSpir Logistics | NAST | Jun 2026 | 75 | Secure, high-value cargo (pharma, data-centre gear) |
| Europe Surface Transportation (sold) | — | Feb 2025 | n/d | Divested to sennder; exit sub-scale European road |
Robinson's physical footprint is small for a $16B company because it owns almost nothing. Its people sit in leased offices; the trucks, ships and planes belong to others. The map shows employees by region (10-K FY2025), key hubs, and the trade corridors that drive Global Forwarding.
Everything in Robinson's P&L collapses to three variables. Volume (shipments) is mostly the freight cycle plus share gains. Spread per shipment (AGP per load) is pricing discipline versus the spot/contract squeeze. Cost-to-serve is people and technology, which is where Lean AI changed the equation. FY2025 actuals from the Jefferies and UBS models.
The most important thing that happened to Robinson between 2022 and 2026 was not the freight cycle. It was that the company cut headcount by a third while growing volume faster than the market, by automating the quote-to-cash workflow with in-house AI agents. Whether this is a permanent, structural change or a cyclical cost cut that gets rehired in an upturn is the central fundamental debate.
A three-layer stack: applications on top, a custom "harness" in the middle that injects logistics context and enforces constraints, and interchangeable LLMs and clouds underneath. Big models are only invoked when necessary, which keeps token counts low and grounds outputs in real shipment data Jefferies. Robinson calls the agent ecosystem the "Agentic Supply Chain" and the "Always-on Logistics Planner" Evercore.
2024 program: $45.7M of charges to flatten management and consolidate real estate. 2025 program: three years, $50–75M total, $30.4M taken in 2025, aimed at AI-driven process optimisation and office footprint 10-K.
Watch three numbers each quarter:
Management's own scorecard: 2026 adjusted operating income of $964M–$1.04B versus $553M in 2023 with a smaller company, on a market that shrank.
Trucking is a textbook commodity cycle. Rates rise, carriers buy trucks and hire drivers, capacity floods in, rates collapse, carriers go bankrupt, capacity leaves, rates rise again. The 2022–2025 downturn was the longest on record (Bernstein: trucking "ran below cost for three years"). In 2026 the supply side finally broke, which is why spot rates are up 34–36% year over year while shipment volumes are still falling.
| Indicator | Latest | Read-through |
|---|---|---|
| DAT dry van spot (linehaul) | $2.25–2.32/mi, +26% vs 9-yr avg | Cooled from the June peak but near the top of the range DAT Aug 2026 |
| Spot vs contract, y/y | +36% vs +21% | Spread narrowing from June; still wide = contract repricing ahead MS Sep 2026 |
| Cass shipments | +2.1% y/y (Aug); −4.8% (Jul) | First positive print in over a year; needs a second month to confirm the demand leg Cass via FRED |
| Cass expenditures | +18.7% y/y (Aug) | Rates doing the work: implied cost per shipment +16% y/y |
| MS Truckload Freight Index | Highest for Sep in ~30 yrs | Off early-June peak to "sustainable but elevated" MS |
| Tender rejections | "Down" (shippers) | Route guides holding; contract rates "flattening" MS shipper survey |
| Class 8 orders | 16.8–18.2k (Aug), +31–42% y/y | 2026 build slots sold out; 2027 pre-buy ahead of EPA engine rules. New trucks = future capacity ACT, FTR |
| Capacity exits | ELP/CDL enforcement, DOT crackdown | English-proficiency and licence enforcement removing drivers; small carriers' insurance costs rising post-Montgomery MS, JMP, UBS |
| 2026 shipper rate expectation | +7.1% ex-fuel | Highest since July 2018 MS survey |
| Management's FY26 spot assumption | +34% y/y | Raised from +17%; expects another 4Q holiday spike MS 2Q26 |
Spot buy rates jump faster than contract sell rates. Robinson's NAST gross margin fell to 13.1%. Gross revenue booms (+23%) but AGP grows slowly (+8.6%, all from LTL). Landstar (100% spot) and asset-based carriers with owned trucks capture the upside immediately. Estimates for 3Q26 get trimmed (UBS 3Q26E $1.69 → $1.64).
The 2026–27 bid season resets 70% of the book at higher sell rates (Stephens: +10–15% ex-fuel; UBS models +9–10% TL base pricing in 2027). If spot plateaus, the spread widens and AGP per load rises 2–3 quarters after the spot surge. UBS: NAST gross margin back to ~14% in 2027; Jefferies models 2027 AGP +9% and EPS +26% to $7.70.
So far this is a supply-driven upcycle with Cass volumes still negative. GS: "the next lever of spot-rate growth must come from volume demand." If demand recovers, Robinson's fixed-cost base (personnel ÷ AGP at 44%) converts volume into 100%+ incremental margins. If demand stays weak and 2027 Class 8 deliveries add capacity, the cycle rolls over: UBS warns a second year of falling spot rates is "a meaningful headwind".
Historically, freight brokerage had low barriers to entry: a phone, a load board and a $75,000 surety bond. Robinson's argument is that scale, data, technology and now regulation have turned a commodity service into a network business. The bears say a spread business with 15,000 competitors can never be a fortress.
$11.6B of NAST gross revenue is >50% larger than #2 TQL ($7.4B in 2025) and ~3x RXO. With 37M shipments a year and 450,000 carriers, Robinson sees more price points on more lanes than anyone. That transaction density lets it (a) find a truck faster in a tight market, (b) price contracts more accurately, and (c) offer carriers denser backhaul opportunities, which attracts more carriers. Deutsche Bank calls this a "self-reinforcing flywheel". Jefferies: the industry is consolidating to a top 5–10.
Navisphere is fully owned, so Robinson keeps every byte of shipment, pricing and carrier-behaviour data and trains its own models on it. The AI layer is architected to be cloud- and LLM-agnostic, with a custom harness that injects logistics context and keeps token spend at ~$1–1.5M a year Jefferies, GS HQ visit. Enterprise shippers integrate Navisphere into their own TMS, which creates switching costs Expert call.
The May 2026 Supreme Court ruling makes brokers suable in state court for negligent carrier selection. Insurance underwriters will now price carrier-vetting rigour into premiums. Management told Goldman three things follow: small brokers cannot get insurance; those that can cannot afford it; shippers move to large, well-insured brokers GS HQ visit, Aug 2026. Jefferies expects the long tail of 15,000+ brokers to shrink. Bernstein agrees on consolidation but says the winners are asset-based carriers, not brokers.
Four archetypes compete with Robinson: other large brokers, digital-native brokers, asset-based truckers with brokerage arms, and giant integrators. Each was hit differently by the 2022–2025 freight recession and by the new liability regime.
| Competitor | Archetype | Scale | Model vs Robinson | Recent development |
|---|---|---|---|---|
| TQL (private) | Large broker | $7.4B rev 2025 | Sales-driven; same rep sells to shipper and buys from carrier. Robinson separates the two. | Grew from $3.4B (2019) to $7.4B (2025) UBS. #2 in U.S. truck brokerage. |
| RXO (RXO) | Large public broker | ~$1.6B/qtr | 73% contract (similar to CHRW). Tech-enabled consolidator. | Bought Coyote from UPS for $1.025B; net leverage 3.7x vs CHRW 1.3–1.6x; trades ~33–43x 2027 EPS, a premium to CHRW Deutsche Bank, StoneX, Benchmark. |
| Landstar (LSTR) | Agent-based, 100% spot | — | Every load reprices in real time, so it captured 2Q26's spot surge immediately while CHRW's contract book squeezed Jefferies. | GS also cut its Landstar multiple post-Montgomery. |
| Uber Freight, ex-Convoy, Transfix | Digital-native | — | VC-funded, app-first, priced to win share. | Convoy collapsed in 2023; Wolfe says failures of digital-first entrants structurally limit new threats. Ex-Uber Freight CCO still argues legacy brokers struggle "from man to machine". |
| J.B. Hunt, Knight-Swift, Schneider, Werner | Asset-based with brokerage | — | Own trucks and drivers; self-insure working layers; brokerage arm fills gaps. | Bernstein's thesis: post-Montgomery risk repricing favours these over asset-light brokers. Insurance ~5%+ of revenue vs CHRW <1%. |
| Echo, Arrive, Worldwide Express, Coyote | Mid-tier brokers | $1–5B | Sub-scale versions of CHRW; 40% of the market is brokers under $1B revenue. | The consolidation pool. Deutsche Bank sees CHRW with $850M–$2B of M&A capacity. |
| Amazon | Vertically integrated | — | Owns warehouses, trucks, planes, cloud; now selling B2B logistics. | Flagged by an energy-sector expert as a long-run threat to 3PLs. |
| Expeditors, Kuehne+Nagel, DSV, Flexport | Forwarders (vs Global Forwarding) | — | Same buy-bulk, resell model for ocean and air. | CHRW's GF is a mid-sized player; Jefferies lists a GF sale or spin as a potential catalyst. |
Until May 2026, a federal law (the FAAAA) shielded brokers from state lawsuits claiming they negligently hired an unsafe carrier. That shield is gone. Ten weeks later a Texas jury handed Robinson the largest broker verdict in history. This section explains what happened, what it could cost, and why analysts disagree by $80 a share about what it means.
| House | Assumption | EPS impact |
|---|---|---|
| UBS (Jun 26) | Insurance $60M → $120M | Offset by >3pp NAST GP; net positive |
| Wolfe (Jul 26) | Each +100% in liability cost = $50M pre-tax | −$0.35 per doubling |
| Bernstein (Aug 26) | $120M 2027 op income headwind | −$0.78 |
| TD Cowen (Jul 26) | Premium inflation from 2027 renewals | 2027E cut to $7.25 |
| Stephens (Jul 26) | Claims/insurance escalation | 2027E trimmed to $7.20 |
About a quarter of Robinson's AGP comes from moving containers and air cargo across borders, mostly Asia to North America. Forwarding earns a spread on freight rates, so it is a levered bet on transpacific volumes and on ocean and air rate volatility.
Robinson signs volume commitments with ocean carriers and airlines at wholesale rates, then sells slots to thousands of importers at retail, plus clears customs (a fee business). Gross margin runs 21–25% of gross revenue, higher than trucking, because service content (documentation, consolidation, customs) is larger.
Volatility is the friend and enemy. Ocean AGP swung from +64% to −53% year over year within the 2021–2023 cycle (Jefferies chart). In 2Q26 air AGP rose 23% on +33% AGP per ton as armed conflicts tightened air capacity, while ocean AGP fell 2.8% on softer rates and customs AGP fell 9.4%.
Gross revenue tells you about the freight cycle. AGP tells you about the spread. Operating income tells you about the cost base. The story of 2010–2026 is a business whose spread margin drifted down for a decade, whose profit was hostage to the cycle, and which since 2023 has broken the link between AGP and headcount. Historical data from the Bernstein, Goldman Sachs and Morgan Stanley models; forecasts as labelled.
| Item | $M |
|---|---|
| Cash | 155 |
| Receivables + contract assets | 3,286 |
| Goodwill & intangibles | 1,485 |
| Total assets | 5,844 |
| Accounts payable & accrued transport | 1,875 |
| Total debt | 1,685 |
| Equity | 1,628 |
| Net debt | 1,530 |
| Net debt / TTM EBITDA | 1.6x |
| Total liquidity | ~900 |
| Item | Guidance | Street |
|---|---|---|
| Adj. operating income | $964M–1,040M | $950–980M |
| Personnel expense | $1.25–1.35B (high end) | ~$1.35B |
| SG&A | $540–580M | ~$560M |
| D&A | $95–105M (low end) | ~$98M |
| Capex | $65–75M | ~$70M |
| Tax rate | — | ~19–21% |
| Adj. EPS | "$6+" at low end | $6.10–6.35 (MS $5.73) |
| Industry volume assumption | −3% (low end) | Cass 1H26 −4.7% |
A simplified 2027 earnings model built from the driver tree. Start from a Jefferies-style base (2027E AGP $3.14B, EPS $7.70) and move the levers the analysts argue about. The output shows operating income, EPS, and the implied share price at your chosen multiple.
At $152 Robinson trades at roughly 24–25x 2026E and 20x 2027E consensus EPS, a premium to its 10-year average of ~20x but well below its 2025–26 peak of 27–33x. Price targets run from $91 (Morgan Stanley, a DCF on "normalized" EPS near $5) to $237 (Evercore). The spread is a disagreement about 2027 EPS ($6.07 to $8.20) multiplied by a disagreement about the multiple (12x to 37.5x).
Each cell is an implied share price. Rows span the bear-to-bull 2027 EPS range in the notes; columns span the multiples analysts actually apply. Current price $152 highlighted.
The table above uses price divided by each calendar year's actual EPS, which exaggerates swings because a cyclical's EPS is itself extreme at tops and bottoms. This version uses the blended 12-month forward consensus multiple, the measure the market trades on. Bands from Bloomberg BF P/E screen, 24 Sep 2026: 5-year average 20.8x, +1 SD 24.9x, −1 SD 16.8x, current 20.9x on about $7.15 of forward EPS.
| Forward EPS | −1 SD 16.8x | Mean 20.8x | +1 SD 24.9x | Peak ~30x |
|---|---|---|---|---|
| $6.50 (bear drift) | $109 | $135 | $162 | $195 |
| $7.15 (today) | $120 | $149 | $178 | $215 |
| $7.70 (Jefferies 2027E) | $129 | $160 | $192 | $231 |
| $8.20 (UBS 2027E) | $138 | $171 | $204 | $246 |
| Broker | Rating | PT ($) | Method | 2026E EPS | 2027E EPS | One-line view |
|---|---|---|---|---|---|---|
| Evercore ISI | Outperform | 237 | — | — | — | Decoupling headcount from volume; 13 straight quarters of share gains; "third inning" of AI |
| JMP Securities | Mkt Outperform | 235 | 28–29x 2027E | 6.35 | 8.15 | Lower cost-to-serve drives structural margin gains; 35% premium to S&P P/E |
| UBS | Buy | 230 | 28x 2027E | 6.22 | 8.20 | Street-high 2027; spot surge shows up in AGP 2–3 qtrs later; Montgomery a net tailwind |
| Susquehanna | Positive | 226 | ~30x 2027E | — | 7.45 | Positive revision cycle from lean/AI transformation |
| Jefferies | Buy | 220 | 28.5x 2027E | 6.10 | 7.70 | Scale, tech, regulatory moat; bull $300 / bear $130; would buy verdict weakness |
| Deutsche Bank | Buy | 220 | — | — | — | Execution stands out; consolidator with $850M–2B firepower; legal overhang raises the bar |
| Stephens | Overweight | 215 | — | — | 7.20 | "Cycle vs Montgomery"; contract +10–15% offsets claims inflation |
| StoneX | Buy | 205 | 27x 2027E | 6.35 | 7.60 | Mid-cycle margins and share gains despite legal risk |
| Wolfe Research | Outperform | 196 | 23x 2028E | — | (2028E 8.50) | Multiple cut from 25x for insurance; each doubling of liability cost = −$0.35 EPS |
| Goldman Sachs | Neutral | 180 | 23.5x Q5–Q8 | 6.20 | 7.30 | Impressed by tech moat at HQ visit; multiple trimmed post-Montgomery; needs volume recovery |
| CFRA | Hold | 165 | 22x 2027E | — | 7.50 | Legal-risk discount to 25.1x 3-yr avg; DCF shows 25% downside if productivity plateaus |
| Bernstein | Market-Perform | 151–153 | 21.5x 2027E / 23.5x NTM+1 | ~5.65–6.0 | 6.68–7.05 | Risk repricing favours asset-based; $120M / $0.78 insurance hit to 2027 |
| TD Cowen | Hold | 152 | 21x 2027E | 6.05 | 7.25 | "Brokers enter a new operating paradigm"; premium inflation from 2027 |
| Morgan Stanley | Underweight | 91 | DCF 7% WACC, 2% g ≈ 17x normalised | 5.73 | 6.07 | Normalised EPS "closer to $5 than $10"; bull $225 / bear $65; UW since 2013 |
The Street is unusually polarised: price targets from $91 (Morgan Stanley) to $237 (Evercore) on a $152 stock. Both camps agree on the facts (2Q26 was a beat, margins hit targets, a $604M verdict is on appeal). They disagree on whether Lean AI is structural and whether Montgomery is a moat or a wound.
Who: Evercore ($237), JMP ($235), UBS ($230), Susquehanna ($226), Jefferies ($220), Deutsche Bank ($220), Stephens ($215), StoneX ($205), Wolfe ($196).
Who: Morgan Stanley ($91, Underweight), Bernstein ($151–153, Market-Perform), TD Cowen ($152, Hold), CFRA ($165, Hold), Goldman Sachs ($180, Neutral).
| Question | Bull answer | Bear answer | What would settle it |
|---|---|---|---|
| Is the 60% productivity gain permanent? | Yes: AI agents, 10x quote capacity, no rehiring needed | Partly: incentive comp and rehiring return in an upturn | Personnel ÷ AGP falling toward 44% in 2027 while volumes grow |
| Net effect of Montgomery on EPS? | Positive: +$60M insurance offset by >3pp NAST GP from share shift (UBS) | Negative: −$120M / −$0.78 (Bernstein), spread never recovers | 2027 insurance renewals; NAST volume vs Cass in 4Q26–1H27 |
| 2027 EPS? | $7.70–8.20 (Jefferies, UBS) | $6.68–7.25 (Bernstein, TD Cowen) | Contract repricing in the 2026–27 bid season; NAST gross margin back above 14% |
| Right multiple? | 28–30x on structural growth, positive revisions | 21–23.5x, broker premium over truckers should narrow | Final judgment and appeal path in Lipe; other broker verdicts |
| Who wins share post-Montgomery? | Large brokers (Jefferies, DB, Wolfe channel checks) | Asset-based, self-insured carriers (Bernstein) | CHRW TL volume growth vs KNX/SNDR/JBHT brokerage and asset volumes |
Ranked by the product of likelihood and earnings impact as discussed across the source notes. Colour: red = high, orange = medium, green = lower.
Figures labelled "E" are broker estimates as of the note date. Where sources conflict (e.g., Bernstein's TTM-basis EPS vs company adjusted EPS) the company-reported adjusted figure is used and the difference footnoted. This dashboard is a research aid, not investment advice.