The Government-Spending Supercycle: Space, Defense-AI and Nuclear-for-AI

01Executive summary

By mid-2026 three seemingly separate sectors share the same fuel: multi-year public and institutional spending flows. In space, the U.S. Space Force is funding “anti-jam” satellite constellations for resilient communications, and Viasat (VSAT) is one of the selected suppliers under the Protected Tactical SATCOM-Global (PTS-G) program, which carries a total (IDIQ) ceiling of 4 billion dollars. In defense, the FY2026 Pentagon budget, cited at roughly 900.6 billion dollars, is accelerating orders of AI-enabled unmanned systems, with AeroVironment (AVAV) among the direct beneficiaries after its transformative acquisition of BlueHalo. In AI, electricity has become the real bottleneck: data centers demand 24/7, carbon-free baseload power, and this has reignited advanced nuclear, with Oklo (OKLO) as a bellwether after its agreement with NVIDIA and Los Alamos and a public endorsement from Nvidia CEO Jensen Huang.

This report is not a call on a single stock, but a read on an investment theme: public and institutional spending as a cross-cutting engine across three different sectors. For each name we analyze the business model, the financial position (cash, debt, backlog, guidance), the concrete catalysts already announced, the aggregate analyst view, trader sentiment and — above all — the specific risks. The goal is to provide a complete, verifiable map, not a trading signal. Every figure reported is drawn from official press releases, SEC filings or first-tier sources, and where data diverge across sources we flag it explicitly.

02The theme: one engine, three sectors

The common denominator is demand funded by public budgets and institutional mega-CapEx. It is a different kind of demand from purely commercial demand: it tends to have multi-year horizons, to be less sensitive to the short-term economic cycle, and to offer greater visibility on future revenue — provided programs get funded, not reallocated, and are properly executed. Let us look at the three pillars of the theme.

Pillar 1 — The defense budget and the military-space race

The FY2026 Pentagon budget is cited at roughly 900.6 billion dollars, with an explicit priority on signal processing and AI-based capabilities. In parallel, next-generation missile defense — often referenced under the “Golden Dome” label — and the need for jam-resistant military communications are driving huge investment in space. The U.S. Space Force is shifting from a few large, expensive satellites to “proliferated” constellations of smaller, agile, maneuverable satellites, using commercial designs to accelerate timelines and cut costs. The PTS-G program, with its 4-billion-dollar IDIQ ceiling and five selected suppliers, is a textbook example of this transition.

Pillar 2 — Drone densification and AI on the battlefield

Lessons from recent conflicts have accelerated a structural trend: pushing persistent airborne reconnaissance and unmanned systems directly under the control of combat units at the battalion level, increasing the “density” of drones in formations. Programs such as Drone Dominance / Gauntlet have selected multiple suppliers with orders for tens of thousands of low-unit-cost units (on the order of 5,000 dollars per drone), while venture-capital investment in defense tech has grown markedly (over 14 billion dollars raised in 2026, up from about 9.6 billion in 2025). The distinctive element of 2026 is not just the number of drones, but the onboard artificial intelligence: automatic target recognition, operation in environments where the GPS signal is jammed or denied, and greater decision-making autonomy.

Pillar 3 — AI is hungry for power, and nuclear returns to center stage

The third pillar shifts the theme from public spending to institutional mega-CapEx. The four large AI hyperscalers, per industry estimates, plan over 710 billion dollars of capital expenditure in 2026 — a volume of compute the traditional electric grid struggles to power. Nuclear, in particular small modular reactors (SMRs) and advanced microreactors, is cited as one of the few sources able to deliver continuous, directly carbon-free, 24/7 power. It is in this context that Oklo has captured the market’s attention, helped by agreements with NVIDIA and the national laboratories and by Jensen Huang’s public comments that energy is becoming “the bottleneck” for AI.

The numbers behind the spending: where the money goes

A few figures help to size the theme. On the defense side, the FY2026 Pentagon budget of around 900.6 billion dollars is accompanied by a marked increase in private investment in defense tech: over 14.6 billion dollars raised in 2026, up from about 9.6 billion in 2025, with large rounds such as the roughly 5 billion raised by Anduril in May. It is the sign of an ecosystem in which public and private capital push in the same direction.

On drones, programs such as Drone Dominance / Gauntlet have selected about twelve suppliers for prototype orders worth on the order of 150 million dollars, with the goal of delivering tens of thousands of units (about 30,000) at a maximum unit cost around 5,000 dollars — a paradigm shift toward quantity and low cost. On the AI side, the four large hyperscalers plan over 710 billion dollars of CapEx in 2026, and the baseload power demand that follows is the bridge connecting the data-center world to advanced nuclear. These are orders of magnitude that explain why the theme has captured investors’ attention.

03Viasat (VSAT) — Space / Military SATCOM

$VSAT daily chartSource: Finviz — informational only, not a recommendation.

The business model

Viasat is a global satellite communications company headquartered in Carlsbad, California. After completing its acquisition of Inmarsat in May 2023, the company combines connectivity for aviation, maritime, governments and the military, operating across a broad range of frequencies and orbits. The segment most relevant to this report is defense and advanced technologies: here Viasat acts as both an end-to-end satellite system manufacturer and an operator, a relatively rare combination that lets it position itself as a single partner to the U.S. Space Force and to allies seeking “dual-use” space communication systems.

The space catalyst: the PTS-G program and the Swarm 1 Delivery Order

In May 2026, Viasat and Intelsat General Communications received a combined roughly 437.7 million dollars under the U.S. Space Force’s Protected Tactical SATCOM-Global (PTS-G) program, with about 150 million dollars of FY2026 research, development, test and evaluation (RDT&E) funds obligated at award. The PTS-G program carries a total (Indefinite Delivery Indefinite Quantity, IDIQ) ceiling of 4 billion dollars split among the selected suppliers, which include — beyond Viasat and Intelsat — Astranis, Boeing and Northrop.

On June 11, 2026, Viasat announced it was selected as prime contractor for the Swarm 1 Delivery Order: it will produce and deliver a dual-band (X/Ka) maneuverable geosynchronous (GEO) mini-satellite, and also provide ground stations and operations support, plus five years of operations and sustainment services covering tracking, telemetry and command (TT&C), network operations and cybersecurity requirements. The award moves the program into the production phase, building on the 2025 Delivery Order 1, in which Viasat matured the design of a resilient, low size, weight and power GEO satellite. Per the official release, production and delivery of the system will support reaching initial operating capability (IOC) no earlier than 2029. The mini-GEO satellite architecture will leverage technology already developed for the ViaSat-3 fleet, aiming for rapid, cost-effective deployment.

One important distinction is worth noting: the official press release for the Swarm 1 Delivery Order does not disclose a specific dollar amount for that order. The 437.7 million cited above refers to the combined May 2026 award (Viasat plus Intelsat), not to the individual June Swarm 1 order. This is a distinction to hold firmly so as not to inflate the figures.

Beyond PTS-G, recent news flow confirms Viasat’s “government” trajectory: on June 1, 2026, the company was selected by Lockheed Martin to support NOAA’s next-generation aircraft with a hybrid satellite communications platform. It is a smaller contract than PTS-G, but it illustrates the diversification of institutional demand.

The financial position (fiscal year 2026, ended March 31, 2026)

Viasat published its fourth-quarter and full fiscal-year 2026 results on May 28, 2026. Full-year revenue was about 4.6 billion dollars, with quarterly revenue of roughly 1.2 billion, up about 2% year over year. The figure most cited by management was the record backlog: about 4.07 billion dollars as of March 31, 2026, up 15% year over year, supported by record contract awards. Adjusted EBITDA came in at about 1.55 billion dollars, with free cash flow of 177 million excluding the 420-million one-time Ligado payment.

The delicate point in the balance sheet is debt. As of March 31, 2026, the company had about 1.75 billion dollars in cash and equivalents, against net debt of about 4.84 billion dollars (improved from 5.59 billion a year earlier). The ratio of net debt to adjusted EBITDA fell to about 3.1 times. This is a significant level of financial leverage, which makes Viasat sensitive to interest rates and refinancing costs, and which explains why part of the bear case focuses precisely on the capital structure, beyond the quality of the contracts. It should also be noted that, even with revenue growth, fourth-quarter earnings per share missed consensus estimates.

What bulls see

Supporters of the stock highlight the combination of an all-time-high backlog (4.07 billion), the prominent position in the PTS-G program, and the ability to offer complete solutions — from the satellite to the ground station to multi-year operational services. In a scenario where the Space Force accelerates the shift to proliferated constellations, Viasat could convert a growing share of the 4-billion IDIQ ceiling into actual revenue, adding visibility to an already broad revenue profile (4.6 billion per year). The improvement in net debt (from 5.59 to 4.84 billion) and free cash flow generation, however modest, are cited as signs of a deleveraging trajectory.

What bears see

Bears stress that a record backlog does not translate immediately into cash: PTS-G revenue will materialize over several years and initial operating capability is expected no earlier than 2029. Meanwhile, revenue growth remains low single-digit (about +2% in the quarter) and earnings per share missed expectations. Above all, net debt near 4.84 billion and leverage at 3.1 times EBITDA make the stock vulnerable to high rates and to any program delays. It is a profile in which execution — on debt and on contracts — matters at least as much as the quality of the strategic positioning.

04AeroVironment (AVAV) — Defense / AI-enabled drones

$AVAV daily chartSource: Finviz — informational only, not a recommendation.

The business model and the transformation via M&A

AeroVironment is a leading maker of small and medium unmanned aircraft systems (UAS), with a long history of supplying tactical drones to the U.S. military and allies. Over recent cycles the company has undergone a transformation via acquisitions: the integration of BlueHalo (and of businesses such as Empirical) substantially broadened its scope, adding capabilities in space, cyber, directed energy and other advanced defense technologies. The result is a larger, more diversified company than the pure drone maker of a few years ago, but also one that is more complex to integrate and to value.

The defense catalyst: the U.S. Army P550 contract

On June 3, 2026, a roughly 117.3-million-dollar U.S. Army contract was publicized for 82 P550 long-range reconnaissance drones. The order was placed on March 20, 2026 via the Army’s UAS Marketplace Basic Ordering Agreement, administered by Army Contracting Command at Redstone Arsenal, Alabama, with completion estimated around July 23, 2026. The P550 combines up to about five hours of endurance with vertical takeoff and landing (VTOL), modular payloads and onboard artificial intelligence for target recognition and operation in GPS-jammed or GPS-denied environments.

The strategic logic is consistent with “Pillar 2” of the theme: putting persistent airborne reconnaissance directly under battalion control, reducing reliance on higher-echelon aviation assets and increasing drone density across combat formations. The P550 is therefore a concrete example of how the defense budget translates into specific orders for a listed supplier.

The financial position and guidance

AeroVironment reported fiscal year-end results around June 30, 2026, with a record fourth quarter: revenue of about 641.6 million dollars and adjusted earnings per share of about 1.84 dollars, both above expectations, supported by contributions from BlueHalo and Empirical. Full-year revenue came in near 1.98 billion dollars, with three-year growth close to 49%. The backlog stood at about 1.2 billion dollars with a book-to-bill of about 1.4 — a sign of visibility on future revenue. Guidance for fiscal year 2027 points to revenue between 2.125 and 2.225 billion dollars.

Despite the record numbers, the market reaction was mixed, and this is where the picture becomes more nuanced. Several research houses — including BTIG, KeyBanc and Clear Street — cut their price targets toward the low-to-mid 200-dollar range, from prior levels as high as about 330 dollars, citing in particular the loss of the SCAR program, a slower pace of contract awards stretching into fiscal years 2027-2028, and weaker margins in the space, cyber and directed-energy segments. In other words, operational execution is solid, but the market is questioning the cadence of new orders and the integration of the acquisitions.

Analysts and sentiment

Per a survey of about 17 analysts, the average rating on AVAV is “Buy”, with an average 12-month price target cited around 276 dollars. However, this average hides significant dispersion: some houses revised their targets down toward 200-230 dollars after the earnings report, while others keep higher targets. On the trader-sentiment side (comments on Reddit, Stocktwits, X), the stock is perceived as one of the purest “stories” on the drone and defense-AI theme, but also one of the most volatile — over the past year the price has gone through very wide swings. We note that these comments come from non-professional traders and are not equivalent to institutional analyst opinions.

Bull and bear case

05Oklo (OKLO) — Advanced nuclear for AI

$OKLO daily chartSource: Finviz — informational only, not a recommendation.

The business model: the Aurora reactor

Oklo develops sodium-cooled fast nuclear microreactors called Aurora, rated between 15 and 75 MW and also fueled by recycled nuclear fuel. The design draws on the heritage of Experimental Breeder Reactor II, an experimental reactor active in Idaho from 1964 to 1994. Oklo’s model is vertically integrated and different from that of a mere technology supplier: the company aims to build, own and operate the reactors, selling power directly to customers through multi-year supply agreements. This “build-own-operate” approach can generate recurring revenue over time, but it requires substantial capital and a long development phase before producing cash.

On the regulatory front, Oklo aims to begin commercial power generation by the end of 2027, having secured a site use permit from the Department of Energy (DoE) at the Idaho National Laboratory and continuing its engagement with the Nuclear Regulatory Commission (NRC). The licensing path for a nuclear reactor, even a small and advanced one, is by definition long and subject to uncertainty: it is one of the stock’s main risk factors.

The AI catalyst: the NVIDIA and Los Alamos agreement, and the Huang effect

On April 23, 2026, Oklo announced a strategic agreement with NVIDIA and Los Alamos National Laboratory (LANL) to advance critical nuclear infrastructure, AI-enabled research and nuclear fuel development — including plutonium-bearing fuels — in support of so-called “AI factories”, with reference also to the federal Genesis Mission. The agreement combines Oklo’s sodium fast-reactor platform, NVIDIA’s AI and accelerated-computing infrastructure and LANL’s expertise in materials science and nuclear fuels. On the news, Oklo shares rose on the order of 19% in a single session.

Subsequently, public comments by Nvidia CEO Jensen Huang — that energy is becoming “the bottleneck” for AI, with data centers turning into “gigawatt factories” the current grid cannot support — sparked a rally across the entire nuclear space, with Oklo up about 24% in a single wave. It is important to keep the two catalysts distinct: the April 23 agreement (a concrete corporate event) and Huang’s words (a narrative sector endorsement). The second amplified the first, but they are two different events.

One aspect often cited by commentators is Oklo’s ties to prominent figures in the technology and AI ecosystem, which has helped the stock’s visibility but also introduces a “narrative” component into the price. It is an element to weigh carefully: the story is compelling, but the valuation of a pre-revenue company remains anchored above all to expectations about future events, not to results already achieved.

Customer pipeline and financial position

Oklo’s customer pipeline is cited at around 14 gigawatts of potential electricity demand, with agreements involving names such as Meta (up to about 1.2 GW of future capacity in Ohio), Switch and Equinix. At its first-quarter 2026 results, the customer backlog was cited at around 1.5 GW following new data-center power agreements. It should be stressed that these are largely pipeline and framework agreements, not revenue already booked: Oklo remains a pre-revenue company on the commercial-generation front.

On liquidity, the position is solid for a company at this stage: cash is cited at around 2.54 billion dollars after an “at-the-market” (ATM) offering of about 1.18 billion completed in May 2026, which provides — per industry estimates — a runway on the order of seven to eight years at the current cash burn rate. This war chest reduces near-term funding risk, but it also implies that much of the valuation reflects future expectations, not results already achieved. It is the classic profile of a high-binary stock: the potential is large, but the road to positive free cash flow is long and capital-intensive.

Analysts and sentiment

Analyst consensus is broadly constructive but dispersed. Price targets across a sample of about 15 analysts sit in a wide range: some houses keep “Buy” ratings with high targets (for example Canaccord around 125 dollars), others are more cautious (Goldman Sachs with a “Hold” rating and a lower target, JPMorgan with “Neutral”), while Bank of America initiated coverage with a “Buy”, calling Oklo an “early leader”. Some consensus estimates point to median targets in the 89-99-dollar area, but the spread reflects precisely the uncertainty on execution timing. On the retail-sentiment side, Oklo is one of the most-discussed names in the AI-power theme; part of the narrative has also concerned its ties to prominent figures in the AI ecosystem. As always, non-professional traders’ comments should be taken as such, not as institutional opinions.

06Snapshot comparison of the three names

The table below summarizes the three profiles. It highlights how different they are, even under the same theme: from a mature, indebted company (VSAT), to a fast-growing but program-dependent one (AVAV), to a long-dated pre-revenue bet (OKLO). The figures should be read together with the verification notes in the previous sections.

07What to watch in the coming months

Beyond individual announcements, the theme will hinge on each company’s ability to convert positioning and pipeline into concrete results. Here are the most useful descriptive indicators to monitor for each of the three names, without this constituting any trading guidance.

Viasat (VSAT)

• New PTS-G delivery orders: additional awards within the 4-billion IDIQ ceiling would give more substance to the backlog.
• Debt trajectory: progress on deleveraging (net debt and debt/EBITDA) and on free-cash-flow generation.
• Progress toward IOC: production and delivery milestones on the road to initial operating capability, expected no earlier than 2029.

AeroVironment (AVAV)

• New-order cadence in FY27: this is the point raised by the analysts who cut targets; an acceleration in awards would ease the concern.
• Guidance execution: ability to hit the FY27 revenue range of 2.125-2.225 billion and to defend margins in the space, cyber and directed-energy segments.
• BlueHalo integration: actual synergies and no negative surprises from integrating the acquisitions.

Oklo (OKLO)

• NRC licensing progress: each regulatory step reduces regulatory risk and is among the most relevant catalysts.
• Pipeline conversion: turning the roughly 14 GW of pipeline and the backlog into firm agreements and, in time, into revenue.
• Milestones toward 2027: tangible progress on the first plant and on the path to commercial generation, plus any new data-center agreements.

08Cross-cutting risks

The theme is backed by real spending flows, but its very “public/institutional” nature introduces specific risks worth spelling out.

• Political and budget risk: defense and space programs depend on appropriations that can be revised, delayed or reallocated with budget cycles and shifting priorities. An IDIQ ceiling is not the same as guaranteed spending.
• Contract execution: an award is not the same as booked revenue; there are milestones, ceilings that may not be fully drawn, and possible slippage. The VSAT case (IOC no earlier than 2029) and the AVAV case (slower order cadence flagged by analysts) illustrate this well.
• Regulatory and technology risk (especially OKLO): advanced nuclear requires NRC approvals and long timelines; delays, scenario changes or technical issues can weigh disproportionately on a company still without commercial revenue.
• Capital structure (especially VSAT): high net debt and leverage around 3 times EBITDA make the stock sensitive to rates and refinancing costs.
• Dilution (especially OKLO): pre-revenue companies fund development through equity issuance (such as the 1.18-billion ATM in May 2026), which strengthens cash but dilutes existing shareholders.
• Valuation and volatility: names tied to “hot” themes (AI-power, drones, space) can move sharply on news and sentiment, with swings well above those of the broader market.
• Competition: in all three sectors competition is intense (other PTS-G suppliers, other drone makers, other SMR operators), and market share is never won once and for all.

09Scenarios (descriptive, not recommendations)

The scenarios below are reasoning tools, not forecasts nor trading signals. They map how the theme’s conditions could evolve, without assigning probabilities or suggesting actions.

10Bottom line

Space, Defense-AI and Nuclear-for-AI are not three separate stories: they are three faces of the same engine — long-term public and institutional spending. VSAT taps military SATCOM contracts with a broad revenue base but an indebted balance sheet; AVAV rides the densification of AI-enabled drones with robust growth but dependence on programs and order cadence; OKLO is a bet on nuclear as baseload for AI, with solid cash but still no commercial revenue. Each has clear catalysts — PTS-G, the P550, the NVIDIA/LANL agreement — but also specific risks tied to execution, budgets, regulatory timelines and valuations.

11Methodology, sources and data limitations

This report follows a simple principle: every relevant figure must rest on a primary or very high-quality source. Information on contracts, programs, financials and catalysts was gathered from official company press releases, filings with the SEC (such as 8-K forms), investor-relations pages, government and defense agency communications, and first-tier financial news sources. Where multiple sources agreed, the figure was treated as solid; where they diverged, we flagged it explicitly rather than arbitrarily picking one.

A few limitations must be stated clearly. First, prices and market capitalizations change quickly and, for VSAT, AVAV and especially OKLO, the sources consulted reported meaningfully different values depending on the time of day and the calculation method: this is why we labeled them as approximate and to be verified live before any use. Second, a crucial element for this type of stock: a contract award is not the same as revenue already booked. Defense and space contracts unfold over several years, are subject to milestones and to ceilings (IDIQ) that are not necessarily fully drawn; in Viasat’s case, for example, the PTS-G system’s initial operating capability is expected no earlier than 2029.

Third, Oklo is a pre-revenue company on the commercial-generation front: its customer pipeline, however large (on the order of 14 GW), consists largely of framework agreements and intentions, not booked revenue, and the path to commercial generation (an end-2027 target) remains subject to regulatory approvals and execution risks. Finally, the sentiment sections synthesize the opinions of non-professional traders gathered on platforms such as Reddit, Stocktwits and X: they are not equivalent to institutional analyst opinions and should not be interpreted as such. In short, the document’s goal is to provide a complete, transparent map that the reader can verify independently against the cited sources, without ever turning into a trading signal.