Last price
$43.32
Railroad operator; thesis orthogonal.
Live quote sourced from Yahoo Finance. Prices cited in narrative below reflect the original memo date and may be stale.
CSX is a Class I railroad operating freight rail across the eastern United States. The business is infrastructure-dependent (track, locomotives) and regulated (Class I rates). AI may optimize routing, fuel consumption, or maintenance scheduling, but these are marginal. Thesis does not apply: CSX is a utility-like transportation operator, not a software or services company.
CSX is a physical-network utility: a regulated railroad where AI improves internal operations (dispatch, maintenance scheduling, fuel consumption) but does not sell outcomes to customers and does not displace labor in a services-budget sense. The thesis barely applies — transportation economics are commodity-driven and asset-intensive, not labor-capture.
Predictive maintenance can reduce unscheduled downtime and locomotive fuel consumption. Routing optimization (weather, congestion, fuel-efficient paths) is already in deployment. These are real cost saves — and CSX has the data (decades of operational logs, telematics from locomotives) to feed such systems. Cumulative margin lift from full deployment could be 100–200 basis points.
CSX does not sell labor-replacement services; it moves freight. No customer budget is being displaced — CSX retains its own freight-dispatch workers even as AI optimizes their routes. Regulated rates limit pricing-power capture. The services-as-software thesis targets outcome-priced markets ($15T services budgets); CSX operates in freight pricing (commodity-tied, volume-driven). No autopilot customer base.
| Function | AI role | Impact | Thesis fit |
|---|---|---|---|
| Locomotive dispatch | Route optimization, fuel prediction | Margin lift, 50–100 bps | Internal only |
| Maintenance scheduling | Predictive failure detection | Uptime gain, capital save | Internal only |
| Freight pricing | Demand forecasting | Modest pricing power | Not outcome-based |
| Labor displacement | Minimal—crews still needed | Not material | No services-budget capture |
Locomotive fuel efficiency alone is a 5–10% variable-cost item; optimized routes could recover 50–100 bps of system margin.
CSX has 150+ years of routing, weather, and failure history — competitors cannot replicate this without acquisition.
Coal, chemicals, and automotive shipments are cyclical; a recovery from 2025 lows is a secular hedge that has nothing to do with AI.
Class I railroad framework is entrenched; CSX has pricing floors and demand stability (bulk commodities move by rail).
Services-as-software targets labor-displacement budgets. CSX cannot sell outcomes to customers because it is a freight operator, not a service provider.
Coal shipping is structurally declining (energy transition). Automotive and chemicals are cyclical. CSX cannot control demand; margin defense is the play.
Class I rates are negotiated with shippers and regulators. AI-driven efficiency is captured by customers (shipper rate cuts) or reinvested, not passed to shareholders as multiple expansion.
Even as AI optimizes dispatch, crews remain employed. No services-budget-displacement story exists.
CSX is a thesis orthogonal. AI improves internal operations (margin defense), but CSX does not sell outcomes, does not displace customer labor, and does not capture services budgets. Own CSX for regulated-utility dividend and freight-volume cyclical upside; do not own it for Sequoia-thesis reasons. The efficiency gains from predictive maintenance and routing are real but will be shared with customers (lower rates) or absorbed as reinvestment, not flowed to equity.
Thesis does not apply; transportation infrastructure economics dominate.