Defence
The first analytical act in defence is classification. A titanium ingot maker and a radar-electronics designer are both "defence companies" and should never carry the same multiple. Get the layer wrong and every comparison that follows is wrong.
Value the company by its layer in the chain and its end-user, then underwrite the one variable that actually moves margin: import content falling, and whether the order book is production or development work.
Classify twice: by end-user, then by layer.
Companies divide first by end-user exposure — Army, Navy, or Air Force — and second by operational layer: Materials, Components, Subsystems, or complete Platforms. This is not taxonomy for its own sake. A raw-material supplier must be valued on materials-industry metrics, not benchmarked against a complex electronics or subsystem manufacturer.
| Layer | Army | Navy | Air Force |
|---|---|---|---|
| Platforms | Tanks, armoured vehicles, artillery | Naval vessels and construction support | Fighter aircraft, helicopters, UAV platforms |
| Subsystems | Fire control, missiles, battlefield electronics | Torpedoes, sonar, submarine combat systems | Radars, avionics, missiles, electronic warfare |
| Components | Hydraulics, pneumatics, valves | Precision components, underwater hydraulics, heat exchangers, propulsion | Actuators, landing gear, pumps, control systems |
| Materials | Specialty steel, aluminium, titanium | Steel and titanium alloys for submarines | Titanium, aluminium alloys, composites |
Order type decides margin; import content decides the re-rating.
Production orders beat development orders. Development work carries execution risk and thin economics; series production is where margin sits. The mix between the two inside a headline order book matters more than its size.
Payment terms differ sharply by customer. Foreign contracts typically use letters of credit. Advance payments in domestic tenders are essentially unheard of — which is why working-capital days in this sector run at 120–200, far above normal industrials. Where raw materials are supplied by the customer as part of the contract price, or a price-escalation clause exists, the risk profile improves materially.
Watch PSU insourcing. As PSUs internalise low-value assembly and outsource selectively to MSMEs, private suppliers' margin sustainability comes into question. Ask where in the chain the company's work would sit if the PSU chose to bring it in-house.
How the economics vary by layer
| Metric | Platform & assembly | Systems integration | Design-IP led |
|---|---|---|---|
| Order book / sales | 3.5–5.5x | 3.0–4.2x | 3.0–4.0x |
| R&D / revenue | 3–8% | 4–6% | 8–10% |
| EBITDA margin | 16–24% | 19–21% | 27–30% |
| Working capital days | 180–200 | 150–190 | ~120 |
| Exports (% revenue) | 6–12% | 5–18% | 20–25% |
The pattern is the point: the highest margin, the lowest working-capital intensity and the highest export share all sit with the most design-IP-intensive layer, and the heaviest capital demands sit with platform assembly. Ranges are indicative across the listed peer set and move each quarter.
Import content by segment — the localisation thesis
| Segment | Import content | Key imported subsystems | Direction |
|---|---|---|---|
| Missile systems | 35–45% | RF/IR seekers, propulsion fuel, actuators | Falling toward ~25% by FY30 via seeker tech transfer |
| Radar systems | 20–25% | GaN/GaAs MMICs, HF transistors, EW processors | Below 15% with fab upgrades from 2027 |
| Avionics & aircraft | 40–50% | Engines, FADEC, radar altimeters | Engines the biggest dependency until an indigenous turbofan works |
| C4ISR / digital | 10–20% | Processor ICs, FPGAs | Stable; India gaining export competitiveness in system design |
| Space & ISR payloads | 25–30% | Radiation-hardened chips, optical sensors | Gradual decline as private satellite component fabs scale |
IP, white space, and export access.
Design IP
Owning the design rather than assembling to someone else's is what produces 27–40% EBITDA margins and low working capital. Moving from component supplier to solution provider is the re-rating event.
High-barrier white space
Titanium, superalloys, energetic materials, aero-engine components — segments that are import-dependent, hard to enter, and where indigenisation policy creates a captive demand curve.
Export access
Exports diversify away from a single monopsony buyer and carry better payment terms. Private companies already account for the majority of India's defence exports.
The demand backdrop is explicit. The Technology Perspective and Capability Roadmap sets out where procurement is heading: electronic warfare and software-defined radios for the Air Force; ammunition and anti-tank guided missiles for the Army; weapons systems and next-generation destroyers and frigates for the Navy; and remotely piloted aircraft plus EW and communications across all three. Externally, the EU's ReArm Europe programme — over €800bn through 2030 on ammunition, missiles and EW — pulls Indian component, processed-material and radar-electronics suppliers into a second demand pool.
What to answer before underwriting.
- →Layer and end-user. Which layer does this company genuinely occupy, and against which peer set should it be valued? Are Army, Navy and Air Force exposures disclosed separately?
- →Production vs development. What share of the order book is series production versus developmental work, and how does margin differ between them?
- →Import content by product line. What is it today for missiles, radar, avionics — and what is the credible path down? This is the margin and multiple story.
- →New vs legacy programmes. What share of order inflow comes from new programmes versus legacy ones? New programmes signal relevance to the next procurement cycle.
- →Contract clauses. Is there a price-escalation clause? Are raw materials supplied by the customer within the contract price? Is it a yearly contract with customer-borne inflation?
- →Advance and payment terms. What percentage of order value is received as advance, and does the export book use letters of credit?
- →Export vs domestic. Split between exports and domestic PSU/MoD/DRDO work, and how concurrent are repeat orders?
- →Technology transfer. Any ToT arrangements with domestic or international partners, and what do they give away versus acquire?
- →PSU insourcing risk. How sustainable are margins as PSUs internalise low-value assembly and outsource to MSMEs?
- →Tender quality. For high-value tenders, what is the genuine win probability, and how much of the pipeline is above a 50% threshold?
What to monitor, quarter by quarter.
| KPI | Calculation / source | Benchmark or read-through |
|---|---|---|
| Order book / sales | Closing order book ÷ TTM revenue | 3.0–5.5x across listed peers; the revenue visibility measure |
| Order inflow | New orders booked in the period | Lumpy — use trailing four quarters, never a single quarter |
| Production vs development mix | Order book split | Production-weighted books convert at better margin and lower risk |
| Import content % | Imported input value ÷ total inputs | The direction matters more than the level — falling = margin tailwind |
| EBITDA margin | P&L | 16–18% at platform/subsystem assembly; 27–40% for design-IP businesses |
| R&D ÷ revenue | P&L disclosure | 3–10% band; higher spend precedes IP-led margin expansion |
| Working capital days | Inventory + debtors − creditors | 120–200 is normal here; the trend against peers is the signal |
| Cash conversion cycle | Days | Best-in-class around 90 days; the sector's key differentiator |
| Advance as % of order value | Contract terms | Higher advances fund execution and cut working-capital drag |
| Export share of revenue | Segment disclosure | 5–25% across peers; diversifies away from single-buyer risk |
| Execution / conversion rate | Revenue ÷ opening order book | Chronic slippage against guidance is the sector's recurring disappointment |
| Capex vs order visibility | Capex plan ÷ order book | Capacity built ahead of confirmed orders is the classic value trap |
How the thesis breaks.
- !Execution slippage. Order books convert slower than guided — engine delivery delays being the canonical example. A large book is not revenue.
- !Valuation ahead of earnings. Steep multiples relative to expected growth leave no room for a single missed quarter.
- !Margin dilution from bought-out content. Rising bought-out components and competitive intensity trend margins down even as revenue grows.
- !Single-buyer concentration. The MoD is a monopsony. Budget reprioritisation or programme cancellation has no commercial offset.
- !Working-capital strain. No advances domestically plus 150–200 day cycles means growth consumes cash.
- !FCF constrained by growth capex. Some of the fastest-growing names will not generate free cash for many years — size the position accordingly.
- !Misclassification. Valuing a materials supplier on electronics multiples is the most common and most expensive error in the sector.
The figures, and where they stand.
| Metric | Value | Note | Basis |
|---|---|---|---|
| Defence export target | ₹500 bn (~$5.8bn) | Government target by FY29 | FY29E |
| Private share of defence exports | ~65% | Of total defence exports | Research note |
| Foreign procurement share | 12% | India still the 2nd-largest arms importer globally | 9M FY25 |
| Import content — missiles | 35–45% | Targeted to ~25% by FY30 | FY25 |
| Import content — avionics | 40–50% | Engines the persistent dependency | FY25 |
| Import content — radar | 20–25% | Below 15% expected from 2027 | FY25 |
| EBITDA margin range | 16–40% | Lowest at platform assembly, highest at design IP | Peer set |
| Working capital days | 120–200 | Structurally high; no domestic advances | Peer set |
| Best-in-class cash conversion | ~90 days | Among domestic defence names | Research note |
| Order book / sales range | 3.0–5.5x | Across the listed peer set | Peer set |
| EU ReArm Europe programme | >€800 bn | Through 2030; ammunition, missiles, EW | To 2030 |
| War Wastage Reserve | 40 days | Stocking requirement driving ammunition demand | Policy |