Behind defence's 320 percent run, a ₹6 lakh crore order book quietly cleared the analysts

The number is 320 percent. Fourteen listed defence-PSUs and their ancillary ring, cohort-weighted, over 18 months ending early 2026. Screens lit. Analysts revised. Every observer with a geopolitical map reached for the obvious story.

The obvious story is wrong.

The Cohort, and Why the War Frame Misfires

The fourteen names span the full supply chain of Indian state-owned defence manufacturing. The core: HAL, BEL, BDL, BEML, Mazagon Dock Shipbuilders (MDL), Garden Reach Shipbuilders (GRSE), and Cochin Shipyard (CSL). The ancillary ring extends the frame: Astra Microwave, Apollo Micro, Solar Industries, Premier Explosives, Data Patterns, MTAR Technologies, and Paras Defence. Cohort-weighted, 320 percent total return over 18 months. That is not rounding error.

War trades are structurally different. They spike on a border escalation, add 8 to 12 percent to the cohort in a session, and mean-revert when the diplomatic channel reopens. Enough repetitions and the experienced book treats those spikes as a fade, not a buy. The current move did not behave that way.

It built over six consecutive quarters. Drawdowns within the rally were shallow. Re-ratings held. These companies were sitting on disclosed order books their valuations had not yet priced in. The market was simply wrong for several quarters and corrected slowly. Geopolitical commentary supplied the narrative. The order books supplied the math.

In this cohort, the math was doing the work.

Order Books as the Actual Signal

HAL's order book closed FY25 at ₹1.25 lakh crore. BEL closed at ₹74,000 crore. MDL at ₹38,000 crore. BDL at ₹19,000 crore. These are disclosed backlog figures from company filings, not estimates, not analyst projections. They represent committed programmes with phased deliverables and payment milestones tied to capital appropriations.

Defence order books carry a different risk profile from commercial backlogs. Once the Indian Armed Forces commit to a programme, execution almost always completes on the multi-year schedule. Payment cycles are tied to parliamentary capital allocations, not a corporate customer's cash flow. Default rates are structurally low. That is a feature of the asset class, not an assumption built into the model.

HAL's ₹1.25 lakh crore is the largest single disclosed backlog among Indian defence manufacturers. The work spans Tejas production, ALH overhaul cycles, Sukhoi maintenance and repair operations, and LCH Prachand deliveries. The AMCA programme sits above all of it as an optionality layer; it is not embedded in the current backlog number. That distinction matters when assessing risk.

BEL's ₹74,000 crore covers Akash missile system electronics, battlefield surveillance radars, electronic warfare suites, sonar systems, and secure communications architecture. The Akash production commitment alone is multi-year. BEL's scope has widened as indigenisation requirements push systems integrators toward domestic electronics suppliers, a structural tailwind that is policy-driven and not discretionary.

MDL's ₹38,000 crore spans the Scorpene production run, P17A stealth frigates, and the P75I submarine programme. BDL's ₹19,000 crore covers the Astra beyond-visual-range missile, Akash missile stockpile production, and the ATGM family.

Aggregate the four disclosed books and the combined backlog is ₹2.56 lakh crore, this desk's arithmetic from the four individual company-reported FY25 disclosures. Against a FY24-FY29 procurement plan sized at approximately ₹6 lakh crore, that aggregate represents forward revenue already committed and absorbed into production schedules. The rest of the plan has not yet been awarded.

That is the actual signal. The rest is noise.

The Capex Spine: What ₹6 Lakh Crore Commits

The Ministry of Defence's FY24-FY29 procurement plan runs to approximately ₹6 lakh crore. This figure is the structural anchor beneath the entire cycle thesis, not a ceiling.

Defence capital budgets differ from discretionary state spending categories. They are tied to parliamentary appropriations and service acquisition plans cleared at Cabinet level. Rolling back a committed procurement plan requires a deliberate policy reversal at the highest level, not a marginal fiscal squeeze that a finance secretary can execute unilaterally. That political cost is a feature, not a detail. It distinguishes this backlog from, say, a state road-building allocation that can evaporate in an interim budget.

The market historically discounted these commitments with a cyclical-government-spending haircut. The re-rating is that haircut coming down, quarter by quarter, as companies execute against existing backlogs and demonstrate the pipeline is being replenished rather than depleted. Once the execution track record repeated across three or four quarters, the discount rate fell. The price followed.

The Lag Mechanic: Two to Three Quarters Behind the Data

Order-book disclosures land in quarterly filings and investor-day presentations. The market's default posture in this sector has been scepticism: apply a heavy discount for execution risk, programme slippage, and potential budget revision. That discount erodes only when a company converts backlog to revenue at scale and confirms the pipeline is not a one-cycle event.

Defence price action historically lags order-book disclosure by two to three quarters. The current move is that lag closing. Not a fresh impulse on new news. The data was already on the table. The repricing took time.

This matters for how you size and hold a position in the cohort. If the thesis is structural, the entry window is not defined by a single event. The move that looks parabolic from the outside is, in cycle terms, a re-rating that was slower than it should have been. Momentum traders and cycle investors have arrived at the same position through different analytical frames, with different intended holding periods. The cohort accommodates both, for now.

The Policy Floor: Indigenisation and the iDEX Architecture

The 75 percent indigenous-content target in defence procurement by FY27 is the policy floor beneath the cycle thesis. It is a procurement condition, not a preference. The requirement means the programme mix inside the ₹6 lakh crore plan must route spending through domestic manufacturers as a contractual matter.

The listed PSUs sit at the top of that domestic manufacturing stack. The ancillary ring sits below them in the supply chain. The relationship is structural: the prime assembler sources from domestic suppliers. When volume through the prime rises, volume through the ancillaries rises alongside it.

iDEX, launched in 2018, added a private-defence innovation channel that feeds procurement demand into the same supplier tail the listed names occupy. The scheme has seeded several of the ancillary names now trading in the cohort. Critically, the listed PSUs are not threatened by this. They are the primes. The private names are tier-two and tier-three, fitting into the supply structure below the prime rather than competing with it directly.

Reversing the policy floor would require simultaneously unwinding the procurement plan, the indigenisation mandate, and the iDEX architecture. That is a high political cost for any government and an unlikely rollback scenario.

Export Optionality: The Layer Analysts Under-Model

Export programmes represent the upside layer in the order-book story that most analyst models still treat as a future footnote rather than a present input.

BrahMos is contracted to the Philippines. Pinaka multi-barrel rocket systems are contracted to Armenia. ATAGS artillery guns are going to Bhutan. Tejas is in active discussions with multiple Latin American buyers. These are named, in-progress programmes, not aspirational pipelines on a government presentation slide.

None of them individually reshapes the near-term revenue story for HAL or BDL in a material way. The aggregate trajectory matters more than any single deal. India has moved from aspirational to executed on defence exports. The Philippines BrahMos delivery is a reference customer. Reference customers open subsequent conversations that a sales pipeline alone cannot.

If Tejas sales to Latin America close over the FY26-FY28 window, HAL's ₹1.25 lakh crore order book receives a top-up that draws on a separate government budget, not the domestic procurement plan. Additive, not merely supportive. Export margins carry different economics from domestic supply, and the diplomatic risk of delivery disruption is real. But the optionality is named, it is in progress, and it is not in consensus models yet. That gap is where the mispricing sits.

Where the Cohort Thesis Fragments

The 320 percent aggregate masks sharp dispersion within the cohort. The cycle thesis is strong at the portfolio level. It is conditional at the single-name level, and that conditionality is where positioning risk concentrates.

BEL and HAL convert order backlogs to revenue with demonstrable efficiency. BEL's electronics scope is largely modular; sensor systems, radar arrays, and communication gear can be assembled in parallel with fewer single-programme dependencies. HAL's production runs are long-cycle, but the order mix across multiple platform categories provides some smoothing against any one programme slipping a quarter or two.

BEML is a different case. Defence mobility is project-specific. A mine-protected vehicle order converts over 12 to 18 months. Gaps between large contract awards are visible in the revenue line. BEML's non-defence segments (metro coaches, mining equipment) add further variability that is harder to model than a named programme backlog. There is no disclosed order book for BEML in the available data; that absence itself signals something about forward visibility relative to HAL and BEL.

The ancillary tail is lumpier still. Solar Industries, Premier Explosives, and MTAR Technologies carry concentrated programme exposure that can dominate quarterly results. A single large order inflates the reported growth rate; a delivery delay compresses it just as fast. These are not stories to hold through a programme disruption on margin.

The risk panel for the cohort runs to three scenarios. First: a defence capital budget cut would appear in the union budget; capital allocations resist marginal fiscal squeezes, but the mechanism exists and any reduction flows directly into order-book replenishment rates. Second: programme cancellation or delay shows up in order-book run-down without replacement inflows; the AMCA is the named risk here, and a delay resets the multiple priced for AMCA conversion without touching HAL's existing ₹1.25 lakh crore backlog, though the market may not make that distinction cleanly. Third: as private-defence JVs mature under the indigenisation framework, margin pressure on PSU primes becomes a possibility if the JV structure shifts pricing power toward the private partner over time.

None of these scenarios breaks the thesis at the cycle level. Each one resets the multiple on the specific name before it resets the structural story.

Cycle conviction at the portfolio level; single-name discipline underneath. That is the position the order books are actually offering.