Behind the 320 percent defence-PSU rally, a ₹6 lakh crore order book quietly compounded

Fourteen stocks. Eighteen months. Roughly 320 percent, cohort-weighted.

The instinct in most newsrooms is to write the war story. The instinct is wrong.

The defence-PSU rally that carried HAL, BEL, BDL, BEML, MDL, GRSE, CSL, and a cluster of private ancillaries (Astra Microwave, Apollo Micro, Solar Industries, Premier Explosives, Data Patterns, MTAR Technologies, Paras Defence) to those returns is a capex-cycle trade. It has a procurement spine, named order books, a measurable revenue-conversion lag, and a policy moat that the war-narrative framing consistently underweights. Every element of that construction is visible in disclosed financials. None of it requires a geopolitical bet.

The procurement spine: ₹6 lakh crore as the demand floor

The Ministry of Defence's FY24-FY29 planning document puts the procurement envelope at approximately ₹6 lakh crore. That number is not a wish list. It is the demand floor the tape is discounting forward.

To understand why, consider what an allocated procurement plan does to a PSU order book. It converts latent ministry intent into binding payment schedules, milestone-linked disbursements, and multi-year supply commitments. The listed cohort sits at the fulcrum of that conversion. Unlike private defence integrators operating under newer frameworks, the incumbents carry decades of qualification history on the platforms the plan is funding. They do not have to win the contract; in most cases, they are the contract.

The ₹6 lakh crore figure is also not uniformly distributed across the five-year window. Front-loading in FY25 and FY26, with ramp acceleration into FY28-FY29, creates a revenue-print schedule that sophisticated buy-side analysts can model with reasonable confidence once the order-book disclosures arrive. The tape is pricing that schedule. It is not pricing a news cycle.

Named order books: the evidence, not the anecdote

End-FY25 disclosures are on record. HAL reported an order book of ₹1.25 lakh crore. BEL closed FY25 with ₹74,000 crore. MDL carried ₹38,000 crore. BDL stood at ₹19,000 crore.

Translate those into revenue cover and the picture sharpens. HAL's book at ₹1.25 lakh crore represents several years of forward revenue at recent run-rates. BEL's ₹74,000 crore positions the electronics major to execute well into the back half of the decade without winning a single additional contract from today. MDL's ₹38,000 crore in shipbuilding commitments is structurally different from a manufacturing order book because shipbuilding cash flows are milestone-spread over five to ten years per vessel, but the size alone confirms that the yard is fully committed through the mid-2030s on existing programmes. BDL's ₹19,000 crore is the leanest of the four named names, yet still represents a pipeline running significantly beyond a single fiscal year.

The point is not that these numbers are large in isolation. The point is that they are already contracted. They are not contingent on future procurement decisions, future budget lines, or future geopolitical conditions. Revenue off these books will convert regardless of what happens on the border, in the newsroom, or in Parliament.

The lag thesis: what the tape is actually paying for

There is a structural pattern in defence-PSU price behaviour that has repeated across previous procurement cycles. Order-book disclosures and price movements do not coincide. Prices have historically moved two to three quarters after the order-book print, as the market processes what large order additions mean for the revenue schedule two to four years out.

The current 18-month rally is that lag closing.

When HAL's order book expanded materially, the sell-side took roughly two quarters to recalibrate revenue models. When BEL's electronic warfare and radar orders accumulated post-FY22, the price action followed with a similar delay. The mechanism is not mysterious. Institutional capital moves on revised earnings models, not on press releases. Revenue models take time to reflect the full scope of multi-year order additions, particularly when delivery schedules are milestone-driven and not cleanly divisible by fiscal year.

The retail and prop traders who caught this cycle early did so by reading the order-book disclosures rather than the news tickers. The 320 percent cohort return is the compensation for that analytical front-running. The cycle is not over simply because the lag has closed; it continues as long as the procurement plan keeps adding to the books at a pace that the current multiples have not fully absorbed.

Indigenisation: the policy moat that locks demand in

The indigenisation mandate is structural, not aspirational. The 75 percent indigenous content target by FY27 is a procurement filter, not a preference. It means a significant fraction of the ₹6 lakh crore plan is legally required to source from domestic vendors.

For the listed cohort, this operates as a demand fence. Foreign OEMs cannot simply offer a better platform at a lower price and capture the order; they must either meet the indigenous content threshold through local manufacturing or joint ventures, or accept exclusion from the procurement round. The incumbents in the listed cohort have the supply chain depth, the vendor qualifications, and the existing platform integration to meet that threshold routinely.

The iDEX channel, launched in 2018 and scaled significantly since, extends this dynamic to smaller private-sector innovators. iDEX routes early-stage defence technology demand into the domestic ecosystem through challenge grants and procurement commitments, which means private ancillaries in the cohort (Data Patterns, MTAR, Paras Defence, Astra Microwave) are beneficiaries of the same policy architecture, not just the large PSUs. The indigenisation mandate does not fracture demand away from the cohort; it keeps it inside the domestic ecosystem that the cohort anchors.

Export optionality: under-modelled, not in the base case

The named export programmes represent a layer of optionality that equity models have been slow to price.

BrahMos to the Philippines is the most advanced transaction in the public record. Pinaka to Armenia and ATAGS to Bhutan are cited programmes. Tejas is in active discussions with Latin American buyers. None of these are certainties. Export defence contracts carry long negotiation timelines, sovereign financing complexity, and maintenance ecosystem requirements that can stall or extend delivery schedules by years.

The right framing is not that exports will double the addressable market. The right framing is that each named transaction, if it closes, adds order-book volume that was not part of the base-case domestic procurement model. BrahMos production has domestic capacity implications that feed back into BEL's radar and seeker integration work. Pinaka volume at scale affects Solar Industries and Premier Explosives in the cohort. The linkages are not speculative; they follow from platform architecture. The export programmes are under-modelled because analysts have been conservative in assigning probability weights. That conservatism is not irrational; it is appropriate given the track record of defence export timelines. But it does mean the cohort's upside scenario, if two or three named programmes close, is not in consensus numbers.

Capex-to-revenue conversion: where the cohort disperses

The 320 percent aggregate return conceals a wide performance gap inside the cohort. Not every name converts order-book additions to revenue at the same pace or with the same margin profile.

BEL and HAL have demonstrated relatively efficient conversion. BEL's electronics and radar programmes carry high component standardisation, which compresses delivery timelines. HAL's MRO segment (Sukhoi servicing, ALH maintenance) converts to cash faster than a new-build programme because the platform base is already in service and the maintenance schedule is contractually locked. Both names have shown the supply-chain depth to execute large orders without significant programme slippage.

BEML is the named contrast. Defence mobility, mining equipment, and metro coach segments each carry different execution rhythms. Large orders can sit in the book for extended periods before revenue recognition events. The lumpiness is not a structural failure; it is a product of programme complexity and customer-side (typically the Army or state transport authorities) delivery acceptance schedules. But it means BEML's order book translates to revenue on a timeline that is harder to model quarter by quarter, and the stock's multiple has reflected that uncertainty relative to BEL and HAL through parts of the cycle.

The cohort return average, in short, is not a guide to individual name selection. The dispersion between efficient converters and lumpy converters is the real analytical work.

The risk register: three cycle-killers, each real

None of these risks is a disclaimer to satisfy compliance. Each is a mechanism that can materially alter the thesis.

Budget reallocation is the fastest-acting risk. The ₹6 lakh crore procurement plan is a planning document, not a constitutional commitment. A Union Budget that shifts allocation toward revenue expenditure (salaries, pensions, operational costs) at the expense of capital procurement reprices the demand floor instantly. Order books already contracted are protected; forward-order expectations are not. A single budget round can create a two to three quarter gap in new order additions that the market will price as a growth pause before the next plan cycle restores momentum.

Programme delay operates more slowly but can be more damaging to specific names. AMCA is the cited example. Advanced fighter programmes routinely exceed their original timelines because technology development risk accumulates at the airframe-engine integration stage. A slippage in AMCA does not cancel the programme or the eventual order. It shifts revenue from the current planning window into the next one, compressing multiples on HAL and its supply chain partners in the specific quarters where the programme was expected to contribute. Duration-sensitive investors price this as a discount; long-cycle investors treat it as irrelevant. The disagreement creates volatility around programme milestone announcements.

Private-public JV margin pressure is the structural risk that the current cycle has largely deferred. As private defence integrators scale under iDEX and direct industrial licensing, they build the qualification history and supply-chain depth to compete for new platform programmes rather than only sub-system supply. Where a PSU previously held sole-source status on a programme, a credible private competitor changes the negotiation dynamic with the MoD and can compress the margin the PSU earns. This has not materially affected the listed cohort through the current cycle, because the incumbents' platform qualifications have not yet been replicated at scale. The risk grows over time as private investment in defence manufacturing compounds.

The frame that survives the headlines

The 14-name cohort's 18-month return is a data point that will attract the geopolitics narrative indefinitely. Every border incident, every escalatory news cycle, will be offered as a retroactive explanation. The explanation is flattering to the storyteller and wrong as analysis.

The actual trade is this: a ₹6 lakh crore procurement plan created a demand floor. Named companies with documented order books converted that floor into revenue visibility stretching three to seven years out. A historical two to three quarter lag between order-book disclosure and price discovery closed, as it always does. An indigenisation mandate locked that demand inside the domestic cohort and out of reach for foreign OEMs. Export optionality added probability-weighted upside that consensus has not fully priced.

Track the order-book deltas at the next disclosure cycle. Watch the execution ratios, not the execution speed. The budget line in the February statement is the single most important data point. The rest is tape.