The Indian Railways changes its playbook for its coaches | The Daily Brief #451

ELI5 / TLDR

Three stories in one episode. The headline: India’s iconic blue passenger coach has been quietly retired and replaced — first by safer German-design LHB coaches, now increasingly by self-propelled Vande Bharat trainsets, with a Rs 65,496 cr FY27 rolling-stock budget pulling private players like Tata, BHEL and Medha Servo into a once-PSU-only club. Story two: cheap drones have broken the math of modern warfare — a Rs 29 lakh Iranian Shahed forces a Rs 34 cr Patriot interceptor in response, and India’s Rs 2 lakh cr defense package runs straight into a Chinese component dependency it can’t currently dodge. Tidbits: Russian oil holding at 2 mbpd into India, Tata Elxsi snaps a 5-quarter decline, JPM bullish on China property recovery.

The Full Story

The Indian coach, three eras

For about six decades, the Indian passenger train was a single idea: one locomotive at the front, a chain of mild-steel boxes behind it. The original ICF coach was Swiss-derived 1950s technology — heavy, prone to rust, capped at 110 km/h, and joined together by a screw coupling for pulling and side buffers for shock. Two systems with slack between them. Fine on a normal day. Dangerous on a bad one. The bad days came in 2015–16, when a string of derailments saw coaches uncouple, jackknife, and climb on top of each other.

The replacement had been waiting in the wings since 1998, when Rail Coach Factory Kapurthala took a tech transfer from Germany’s Linke-Hofmann-Busch. LHB coaches are stainless steel, lighter, last 35 years against ICF’s 25, brake on discs instead of treads, and — the load-bearing bit — use a single centre coupler that handles pull and push together. In a crash they stay in line. The 2014 Dibrugarh Rajdhani derailed at high speed and lost no lives; a CAG audit specifically flagged that no LHB coach flipped over.

The cost: Rs 75 lakh to Rs 1 cr more per coach. The pace: about 23,000 conventional coaches replaced since 2015 by 2024. ICF flagged off its last conventional coach in January 2018.

LHB raised the speed ceiling to 160 km/h but kept the locomotive-pulls-everything architecture. Vande Bharat changes that. Half the coaches in a 16-coach Vande Bharat are motor coaches with traction motors under the floor — distributed power instead of one engine doing all the lifting. Two driver cabs, no terminal turnaround, regenerative braking that feeds 30% of electrical energy back to the grid, and a 0–100 km/h time of 52 seconds in a 2022 trial. The government claims up to 45% journey-time savings, mostly from acceleration.

The catch is the track. Design speed 180 km/h, max operating speed 160, but only 174 km of Indian track — the Tughlakabad–Agra stretch — actually lets a Vande Bharat run at 160. As of 2025, just 21.8% of the network (about 23,100 km) supports 130 km/h and above. The trains can do more than the rails will allow.

The procurement shift

This is the part of the story that matters for capital allocation. For decades, three PSU factories did everything: ICF Chennai, RCF Kapurthala, MCF Raebareli. They still do — combined output crossed 7,000 coaches in the last fiscal year, with RCF at a record 2,383 and MCF at a record 2,025. ICF, where Vande Bharat was born, is among the largest rail coach makers in the world.

What’s new is who else is now in the room. A consortium of BHEL and Titagarh Rail Systems took an Rs 23,000+ cr contract for 80 Vande Bharat sleeper trains, with 35 years of maintenance bundled in. Titagarh — a wagon maker in its previous life — has stood up a Vande Bharat line at Uttarpara, West Bengal. Hyderabad-based Medha Servo Drives, which built the propulsion system on the original Train 18 prototype, has an Rs 2,211 cr contract for electricals on 44 Vande Bharat sets with a 75% local-content floor. The FY26-27 rolling-stock capital outlay is Rs 65,496 cr — a number aimed squarely at steel mills, wheel plants, brake suppliers, and power-electronics firms.

Exports exist but are modest: ICF has shipped 875+ coaches to Thailand, Bangladesh, Mozambique, Sri Lanka. The promised Vande Bharat exports to Europe, South America and East Asia haven’t materialised in the public record yet.

Drones broke the war ledger

The second story is a single observation with large consequences. Iran’s defense budget is $10 bn. America’s is $850 bn. And yet, in the current Gulf war, Iran’s economics are winning. Each Shahed-136 drone costs about $35,000 (Rs 29 lakh) — moped engine, composite shell, basic GPS, flies at highway speed. Each Patriot interceptor that knocks one out costs $4 million (Rs 34 cr). That’s roughly 114-to-one in Iran’s favour. Lockheed Martin made about 600 Patriot interceptors in all of 2025; in the first week of this conflict, the UAE alone faced 689 drones. Months of stockpile, gone in days.

Ukraine got there first and went cheaper. The dominant Ukrainian frontline weapon is now a $400 FPV racing quadcopter with explosives strapped on, flown by a soldier in goggles. Ukraine made 800,000 drones in 2023, 2 million in 2024, and is targeting 5 million this year. Russia is targeting 1,000 Geran-2 attack drones per day in 2026.

The US has noticed late. The GAO found in June 2025 that the Pentagon takes 12 years on average to field a new weapon. In November 2025 Defense Secretary Hegseth declared the existing acquisition system “dead.” In December 2025 the US fielded LUCAS — the Low-Cost Uncrewed Combat Attack System — reverse-engineered from the Shahed in 18 months.

The bind, even for the US, is the supply chain. Motors, batteries, sensors, ESCs — overwhelmingly Chinese. DJI alone has 70% of the global commercial drone market.

For India, this is sharper. In Operation Sindoor, Indian Israeli-made drones cost close to $1 mn each, against Pakistan’s Turkish Bayraktars and Chinese CH-4s. India has the Nagastra-1 at $500, and IdeaForge holds about 50% of the domestic UAV market. A Rs 10,000 cr drone PLI is being floated for the upcoming budget. But the camera-sensor payload — 60–70% of a drone’s bill of materials — is still imported, mostly from China. India’s Rs 2 lakh cr Defence Acquisition Council package in March is the largest ever, but the supply-chain dependency it has to outrun has a 20-year head start.

Tidbits

Russian oil sales to India staying above 2 mbpd through April-May after the US sanctions waiver renewal — March hit a record 2.25 mbpd, half of India’s total imports, with Indian refiners paying a $7-9/bbl premium over Brent. Tata Elxsi posted Q4 PAT of Rs 220 cr, up 27.8% YoY, ending a 5-quarter decline streak; Rs 75 dividend declared. JP Morgan turning constructive on Chinese mainland equities, citing Hong Kong real estate spillover, wealth effect, and housing affordability at 2016 levels.

Key Takeaways

Coach generations: ICF (mild steel, 110 km/h, retired Jan 2018) → LHB (stainless, 160 km/h, single centre coupler, Rs 75 lakh-1 cr premium per coach, ~23,000 replaced since 2015) → Vande Bharat (distributed traction, 0-100 in 52s, 30% regen, 79 services live as of April 2026).
Sleeper Vande Bharat: Launched January 2026 on Guwahati–Howrah, 823 berths across 16 coaches. 24-coach version in design. 260 sleeper sets planned.
Track is the bottleneck: Only 21.8% of Indian rail (~23,100 km) supports 130+ km/h. The 160 km/h max is achieved on a 174 km stretch (Tughlakabad–Agra).
FY27 rolling-stock capex: Rs 65,496 cr. PSU output last fiscal year: ICF + RCF (2,383) + MCF (2,025) = 7,000+ coaches.
New vendor stack: BHEL + Titagarh consortium — Rs 23,000+ cr for 80 Vande Bharat sleepers including 35-year maintenance. Medha Servo Drives — Rs 2,211 cr for electricals on 44 trainsets, 75% local content. Titagarh’s new Uttarpara line.
Exports: ICF has shipped 875+ coaches to Thailand, Bangladesh, Mozambique, Sri Lanka. No Vande Bharat overseas sale on record yet.
Drone economics: Shahed-136 at $35,000 vs Patriot interceptor at $4 mn = 114x cost asymmetry. Lockheed made ~600 Patriot interceptors in all of 2025; UAE faced 689 drones in week one of the Gulf conflict.
Indian drone landscape: Nagastra-1 at $500. IdeaForge ~50% domestic UAV share. Proposed Rs 10,000 cr drone PLI. Rs 2 lakh cr Defence Acquisition Council package approved March 2026.
Choke point: 60-70% of drone BOM is camera/sensor payload, mostly China-sourced. DJI commands 70% of the global commercial drone market.

Claude’s Take

The coach story is the better story, even if the drone piece grabs the headline. It is a clean case study of how India does industrial upgrading: import the technology, run it under licence in a PSU, broaden the vendor pool over a decade, then start vertically integrating the high-value bits. ICF coach to LHB to Vande Bharat is essentially the same playbook ISRO ran on satellites — slower than the Chinese version, but the manufacturing base is real, the IP increasingly is, and the cost curve eventually bends. Worth tracking Titagarh and Medha as the second-order beneficiaries of a Rs 65,000 cr/year buyer who is also the regulator.

The track-versus-train mismatch is the honest punchline. You can keep building faster trains; if the rails, signalling, and gradients aren’t upgraded, the new fleet caps out at the old fleet’s speed. This is the kind of thing that doesn’t get fixed in budget cycles — it gets fixed over a decade or doesn’t get fixed at all.

The drone segment is well-reported but slightly over-tidy. The 114x cost ratio is real but rhetorically loaded — Patriots aren’t fired only at $35K Shaheds, and the Israeli/American C-UAS stack now includes much cheaper interceptors (lasers, gun systems, smaller missiles) that the segment doesn’t engage with. The deeper point holds though: production rate has overtaken sophistication as the binding constraint, and India is on the wrong side of that curve as long as its drone supply chain runs through Shenzhen.

Score: 7/10. Solid macro framing, good numbers, two distinct stories told well in 20 minutes. Loses a point for the drone segment cherry-picking its cost ratio and another for compressing the procurement-reform angle on coaches into a paragraph when it’s the most investable thread in the episode.