State of Small-Cap Defense Investing 2026 Pt. III: The Small-Cap Arsenal

Part One made the case that the alpha in small-cap defense lives in the gap between accelerated Pentagon demand and prime contractors that move on legacy timelines.

Part Two walked the contested-domain layer of that thesis: maritime autonomy and counter-UAS, i.e., the names sitting closest to the physical fight in the sectors facing the largest technical and strategic disruption in today’s warfare climate(s).

Today, we’ll move to the layer behind every platform in that fight, the consumables and the eyes:

Munitions get fired and have to be replaced.

Sensors get integrated and have to be sourced.

Both halves of this edition cover the companies that supply what the primes cannot (or will not) manufacture themselves, and the small-caps coming up from behind to resupply and rearm, to ultimately remake and rejuvenate the American defense industrial base.

Part One's most damaging single statistic was that the US fired more than a thousand Tomahawks against an annual procurement as low as 90 missiles, while also burning through roughly half its Patriot stock with replenishment running toward the end of the decade.

A golden oldie: 2015 Tomahawk test footage targeting moving sea targets

That is a multi-year production mandate, and the cleanest way to play a multi-year production mandate is to own the input nobody can substitute.

 The mandate has moved from policy framing to procurement action. On January 6, 2026, Lockheed Martin signed a seven-year framework agreement with the Department of War to lift annual PAC-3 MSE production from approximately 600 interceptors to 2,000.

 On April 10, Lockheed received the first contract under that framework, a $4.7 billion undefinitized contract action to begin drawing the volume forward. The framework is part of the Department of War's broader Acquisition Transformation Strategy (a parallel deal quadrupled THAAD production from 96 to 400 interceptors annually, a massive increase for a munition of that size).

In other words, the demand signal is no longer about whether the ramp happens, but about who supplies the inputs to it.

The input thesis points straight at Park Aerospace, which sits right in the small-cap defense stock sweet spot at $712M market cap. Park is the sole-source qualified supplier of advanced composite ablative materials for the solid rocket motors on the PAC-3 system, and the exclusive North American distributor of ArianeGroup's C2B fabric used in the same ablative applications.

Park Aerospace makes the composite materials that hold these rocket motors together long enough to intercept their target.

The demand math runs straight through to Park's order book. Lockheed delivered 620 PAC-3 MSEs in 2025, a 20% increase over the prior year, and the framework calls for that figure to more than triple by 2030.

Management has guided that the company-level number for Park runs higher still than the headline production rate, because per-interceptor ablative content scales with both the rate ramp and Park's qualified position.

Fiscal 2026 revenue came in around $73 million on the company's year-end results, and management is targeting roughly $200 million by FY2031, supported by a planned $50 million plant expansion to double manufacturing capacity. The balance sheet carries no long-term debt and roughly $63.6 million in cash.

A great deal of the ramp is already priced. PKE trades at roughly 58 to 60 times trailing earnings after a 60% one-year run, and the thesis depends on Lockheed continuing to draw at framework volumes, the ablative content intensity holding as engineering iterates, and Park executing the capacity expansion on schedule. Any single slip on those compresses the multiple fast.

 On the other hand, at just 0.43x beta, Park represents a way to hedge against wider portfolio volatility if you’re building a basket of disruptive, higher-risk defense tech small-cap stocks.

The quirkiest name in this section is National Presto, a holding company that sells pressure cookers under the PRESTO brand and, through its AMTEC subsidiary, is a sole-source supplier of 40mm ammunition to the U.S. Army.

The same company making your mom’s pressure cooker also produces the 40mm munitions fired from this MK-19 automatic grenade launcher.

The defense order book has been stacking: a $139.3 million delivery order earlier this year alongside past delivery and options awards bring the total value of AMTEC's current five-year 40mm contract to $965 million. AMTEC has held the prime 40mm contract since 2005 with no signs of new entrants on the horizon to shake up that status quo.

NPK is right in the middle of small-cap range, trading at ~$915M market cap, and consistently releases a special annual dividend, though total annuel distributions vary. Current yield is 0.77%, representing a 22.8% payout ratio.

Like PKE, NPK is low-beta, meaning it screens as low-beta relative to the basket. Likewise, NPK’s diversified operations include kitchen appliances (microwaves, pots and pans, and more), which help diversify away defense sector risk in the event that Uncle Sam decides to sunset or replace the existing 40mm grenade program, such as how the Pentagon recently built out a program to augment (but not replace!) the classic M67 hand grenade.

A pure picks-and-shovels play sits in CPI Aerostructures, a micro-cap at roughly $66M market cap that builds a range of structural assemblies for military aircraft (both fixed-wing and helos) and missile platforms as a subcontractor to the primes.

 CVU is agnostic to which program wins, which creates a durable position and moat producing components that are simultaneously too low-margin for the primes and too tricky for new entrants.

 Q1 2026 marked an inflection: revenue of $17.4 million, up 13% year-over-year, a return to GAAP profitability at $1.2 million in net income, gross margin of 25.8%, and total backlog of $495 million per the Q1 10-Q.

 Recent contract wins span the picks-and-shovels stack:

The trade-off is the trading profile: CVU is small, illiquid, and historically lumpy on program timing.

Every platform in this series, the drone, the interceptor, the satellite, the counter-UAS system, runs on the same invisible layer: components that manage radio-frequency and visual signals (electro-optical, infrared) where the same picks-and-shovels logic applies.

Whichever airframe or missile or drone or whatever next-gen hardware platform wins its slot in the procurement budget, the contents of the RF and video feed boxes inside it are sourced from a relatively short list of small-cap specialists.

M-tron Industries sits at the radio-frequency end of the picture. The company makes microwave and millimeter-wave filters, precision oscillators, and integrated microwave assemblies for radar, electronic warfare, missiles, and satellite communications, with aerospace and defense now around 70% of revenue.

The numbers are clean, reinforcing the bull case and M-tron’s positioning in a changing procurement climate. Q1 2026 revenue rose 15.4% year-over-year, gross margin came in around 44.9%, the balance sheet carries roughly $52 million in cash and no debt, and the backlog represents roughly 17 months of sales.

Management has guided to doubling electronic-warfare and radar revenue in 2026. The stock has returned more than 500% since its 2022 spinoff and still trades cheaper on a forward multiple, around 40x, than its closest peer Frequency Electronics (FEIM) at roughly 60x, despite stronger growth and margins. MPTI’s momentum is continuing, with an approximately $425M market cap following a recent rights offering.

MPTI is still a roughly $425M market cap name with a 52-week trading range factor of about 2.5x, which is the volatility profile of a small-cap (still, its 0.71x beta is fairly smooth considering its size and exposure). Regardless, position sizing matters here with the same demand tailwinds that have driven the stock working in reverse if procurement timing slips.

LightPath Technologies sits at the infrared end. The company makes thermal optics and imaging systems built around its proprietary BlackDiamond chalcogenide glass, licensed exclusively from the US Naval Research Laboratory. That license is a real moat for a small-cap, particularly in the context of China's tightening export controls on germanium (the legacy material for IR optics) as BlackDiamond is germanium-free.

 The catalyst is concrete and the window is short. LPTH supplies the thermal imaging seeker for Lockheed Martin's QuadStar interceptor, Lockheed's bid for the U.S. Army's Next Generation Short Range Interceptor (NGSRI) program, the planned replacement for the Stinger across man-portable and vehicle-mounted short-range air defense.

Per LPTH's July 2024 qualification milestone release, an initial 10,000-system production order at $5,000 to $10,000 per system represents $50 million to $100 million in initial revenue for LightPath if Lockheed is selected over Raytheon, with the company framing the life-of-program opportunity at $500 million to $1 billion as the Stinger replacement rolls out across short-range air defense (SHORAD) formations.

 The fundamentals have inflected to match. Q3 FY2026 revenue hit $19.151 million, up nearly 100% year-over-year. Backlog hit $111M at investor day, with a hefty percentage tied to defense, surveillance, and public-safety end markets. The company has $55M in cash and expects to be operating cash positive in the near future. Market cap sits around $940M.

LPTH remains unprofitable on a GAAP basis and trades on a rich sales multiple (~14x trailing and “just” 9.5x forward P/S). Ultimately, this is a margin-inflection bet that pays off if the NGSRI selection lands with Lockheed, though management’s increasingly diversified customer base, as seen above, helps soften the blow if Lockheed drops the ball.

Espey Manufacturing is a century-old maker of ruggedized, MIL-spec power supplies, transformers, magnetics, power distribution equipment, and high-power radar systems: the unglamorous backbone that radar, electronic warfare, and high-energy systems all require. As power budgets on platforms climb (directed-energy weapons, high-power radar, larger sensor stacks), Espey is a quiet way to own the power layer.

Q3 FY2026 results were strong on the margin side: net sales of $11.4 million, EPS of $0.99 against $0.63 a year earlier, and nine-month net income up to $7.8 million from $5.2 million. The balance sheet runs net cash with no drawn debt, and ESP trades around 15 times earnings, cheap by the standards of this sector.

The order side is mixed: backlog sits at $137.1 million, essentially flat year-over-year, and nine-month new orders came in at $30 million against $75.1 million in the prior-year period. The conversion of backlog into earnings is working, but management’s conversion of opportunity into new orders has cooled.

Four near-term items will tell you whether the picks-and-shovels thesis in this edition is playing out:

PAC-3 contract drawdowns: The April 10 $4.7 billion undefinitized contract action was the first and the cadence/size of subsequent contracts against the seven-year framework is the demand signal Park Aerospace's ramp depends on.

NGSRI selection: The Army's choice between Lockheed's QuadStar and Raytheon's competing interceptor is the binary catalyst for LightPath. Watch for downselect signals over the next two quarters.

AMTEC option exercises: National Presto's current five-year 40mm contract has further option years, with the pace of option exercise is the cleanest read on whether 40mm procurement is being treated as accelerated or steady-state.

 The Acquisition Transformation Strategy's expansion: PAC-3 and THAAD were the first programs structured under the framework. If other munitions or sensor programs get the same treatment, the picks-and-shovels names supplying into them re-rate together.

In Part IV, later this week, we’ll get a little more future-forward and look at next-gen propulsion, small-cap space stocks, edge manufacturing, and more. Stay tuned.

Hit reply with thoughts, corrections, or names worth a closer look. If you’re seeing this outside of your inbox, shoot me an email here. Keep digging, and see you next time.