As primary dealers’ net Treasury inventories surge to their highest share of the market since 2007 — touching roughly $550 billion, or nearly 2% of the $31 trillion outstanding — the Trump administration’s deregulatory pivot is quietly reshaping who underwrites America’s debt. The shift promises better liquidity and deeper market-making capacity. It also reintroduces concentration risks that should not be papered over with optimism.

In the lexicon of financial markets, there are few numbers with as much quiet authority as the weekly primary dealer position data published by the Federal Reserve Bank of New York. Every Thursday afternoon, at approximately 4:15 p.m., the New York Fed releases figures that reveal how much of the world’s most important fixed-income market the largest banks are actively holding on their books. For much of the post-2008 era, those numbers told a story of retreat — of banks pulling back from Treasury market-making as a thicket of capital rules made the balance-sheet cost of holding government debt increasingly punitive relative to the returns on offer.

That story appears to be changing. According to Financial Times calculations based on New York Fed data, primary dealers’ net Treasury inventories have climbed to approximately $550 billion — their highest level, as a proportion of total Treasuries outstanding, since 2007. That figure, representing nearly 2% of a market that has ballooned to roughly $31 trillion, is not merely a statistical curiosity. It is a structural signal: Wall Street banks are returning to their traditional role as the central nervous system of American government finance, propelled in large part by the most consequential regulatory reform to hit the banking sector since the Dodd-Frank era.

A Market That Outgrew Its Intermediaries

To understand why this moment matters, it is necessary to appreciate just how dramatically the Treasury market’s growth has outpaced the capacity of its traditional intermediaries. As the Bank Policy Institute has documented, since 2007 the stock of outstanding Treasury securities has grown nearly fourfold relative to primary dealer balance sheets. The U.S. government now borrows far more than the financial system was designed — post-crisis — to efficiently intermediate.

The arithmetic of this mismatch is stark. From $2.1 trillion outstanding in 1990, the Treasury market expanded to $5.8 trillion in 2008 and approximately $21 trillion by 2020. Today it approaches $31 trillion. Meanwhile, dealer intermediation capacity — measured not by raw holdings but by their ability to warehouse risk relative to market size — stagnated, constrained by post-crisis rules that treated U.S. government debt with much the same regulatory suspicion as any other leverage-intensive exposure.

This seemingly contradictory situation — where dealers’ market-making capacity decreased while banks’ Treasury holdings increased — can be explained by the dual impact of post-crisis regulations. While capital requirements constrained dealers’ ability to actively intermediate in the Treasury market, liquidity regulations simultaneously incentivized banks to hold more high-quality liquid assets, including Treasuries. As a result, although large banks held more Treasuries, their capacity to provide liquidity and depth to the market did not keep pace with the growth in outstanding Treasury securities. Bank Policy Institute

The consequence was a market that appeared deep — daily turnover reaches some $750 billion according to SIFMA — but proved intermittently fragile, as the March 2020 “dash for cash” catastrophically illustrated. That episode, in which the supposedly most liquid market in the world briefly seized up, forcing the Federal Reserve into an emergency $1.6 trillion intervention, was the clearest possible demonstration that the structural plumbing of the Treasury market had become inadequate.

The eSLR Pivot: Deregulation With a Purpose

The proximate cause of the current inventory surge is identifiable: the enhanced Supplementary Leverage Ratio reform, finalized by the Federal Reserve, the OCC, and the FDIC in late November 2025. The final rule includes an effective date of April 1, 2026, with the optional early adoption of the final rule’s modified eSLR standards beginning January 1, 2026. Federal Register

The eSLR, established in 2014, was conceived as an additional capital buffer for America’s globally systemically important banks — the eight institutions whose failure would, in the regulators’ estimation, send shockwaves through the entire financial system. The logic was sound in the immediate post-GFC environment. But the rule’s blunt architecture — it treated all assets equally, regardless of their riskiness — produced a perverse disincentive. A leverage ratio constraint that is more stringent than any applicable risk-based standards may discourage a bank from engaging in low-risk activities, such as Treasury market intermediation. OCC

The reform recalibrates this. The current fixed two percent eSLR buffer standard for GSIBs is recalibrated to equal 50 percent of a GSIB’s Method 1 surcharge calculated under the GSIB surcharge framework. In plain terms: the largest U.S. banks — JPMorgan Chase, Goldman Sachs, Bank of America, Morgan Stanley, and their peers — now face meaningfully lower capital requirements for engaging in Treasury market-making. FDIC staff estimated that the final rule would lead to an aggregate reduction in Tier 1 capital requirements of $13 billion, or less than 2%, for GSIBs, and a $219 billion reduction, or 28%, in Tier 1 capital requirements for major bank subsidiaries. KPMGABA Banking Journal

That $219 billion reduction at the bank subsidiary level is the operational number that matters most for Treasury market-making. It directly expands the balance sheet capacity available to the dealer desks that sit inside those subsidiaries. A key benefit of the final rule is that it would remove unintended disincentives for banking organizations to engage in low-risk activities, such as U.S. Treasury market intermediation, and reduce unintended incentives, like engaging in higher-risk activities. Davis Wright Tremaine

The Trump administration — and, to their credit, regulators appointed with explicit mandates to revisit post-crisis rules — deserve recognition for acting on what had become, in regulatory circles, an open secret: the eSLR was quietly undermining the functioning of the world’s most systemically critical fixed-income market. The agencies state the changes are intended to serve as a backstop to risk-based capital requirements and to encourage these organizations to engage in low-risk, balance-sheet intensive activities, including during periods of economic or financial market stress. KPMG

What $550 Billion in Net Inventories Actually Means

The approximately $550 billion in net primary dealer Treasury holdings — up from well below $400 billion in much of 2025 — represents genuine re-privatization of a function that had been, by default, increasingly outsourced either to the Federal Reserve (through QE) or to non-bank intermediaries whose capacity to absorb shocks is structurally different from that of regulated banks.

Net inventory, as opposed to gross positions, strips out hedged or offsetting positions and measures the actual directional risk that dealers are absorbing from the market. A higher net inventory means dealers are more willing to be price-makers rather than merely conduits — they are warehousing duration and credit risk on behalf of clients, an activity that requires balance sheet and, critically, regulatory appetite.

Since the beginning of the Federal Reserve’s balance sheet normalization in June 2022, dealers’ intermediation activities in the Treasury and MBS markets have increased. Dealers’ SLR constraints have become less binding as Tier 1 capital generally grew more quickly than total leverage exposure. The eSLR reform accelerates and institutionalizes this trend. Federal Reserve

This matters enormously given what lies ahead on the issuance calendar. The United States faces a staggering wall of debt refinancing over the next several years — trillions in Treasuries maturing and requiring rollover, on top of ongoing deficit financing that shows no credible signs of abating. A Treasury market in which primary dealers have greater balance sheet capacity to absorb new supply is unambiguously better equipped to handle this reality without repeated bouts of yield dislocation.

The Shadow in the Room: Hedge Fund Leverage and Basis Trade Risk

Improved dealer capacity is genuinely good news. It is not, however, a complete story — and intellectually honest analysis requires acknowledging what surrounds this structural improvement.

The decade since post-GFC regulation constrained bank balance sheets has not been a period of reduced risk in the Treasury market; it has been a period of risk migration. The activity that dealers could not profitably conduct moved, as it tends to do in finance, to entities subject to less regulatory friction. In the Treasury market, that migration produced the spectacular — and partly terrifying — growth of the hedge fund basis trade.

As of 2025, Treasury basis trades are estimated to account for $1 to $2 trillion in gross notional exposure, with a significant concentration among large hedge funds. The mechanics are straightforward: hedge funds buy Treasury bonds in the cash market while simultaneously shorting the corresponding futures contract, financing the long position through the repo market and extracting the spread between cash and futures prices — typically a few basis points — amplified through leverage. Data suggests that hedge fund leverage in this market can range from 50-to-1 up to 100-to-1. WikipediaBetter Markets

According to the Fed’s most recent Financial Stability Report, average gross hedge fund leverage has reached historically high levels since the data first became available in 2013 and is highly concentrated. The top 10 hedge funds account for 40 percent of total repo borrowing and have leverage ratios of 18 to 1 as of the third quarter of 2024. Hedge funds now represent approximately 8% of all assets in the U.S. financial sector, but their footprint in the Treasury market — through cash positions, futures, and repo — is disproportionately large. Federal Reserve Bank of Cleveland

The interaction between a more capacitated dealer sector and a heavily leveraged hedge fund sector is not purely benign. Dealers are the prime brokers who finance most of the repo lending that sustains the basis trade. A dealer sector newly emboldened by eSLR reform may, paradoxically, become more willing to extend leverage to basis traders — adding a layer of procyclical amplification to the very market they are meant to stabilize. A rapid unwinding of leveraged positions could create a feedback loop: selling pressure drives price dislocations, which in turn triggers further deleveraging. Hedgeco

The March 2020 episode remains instructive. When volatility spiked and repo conditions tightened, hedge funds were forced to unwind basis positions simultaneously, transforming a liquidity-enhancing strategy into a liquidity-consuming crisis. The Fed’s emergency intervention prevented a complete seizure — but it also reinforced the moral hazard implicit in the market’s current architecture: the Treasury market is too important to fail, and everyone in it knows it.

A Geopolitical Dimension: Who Underwrites the Safe Asset

This debate does not occur in isolation from global capital flows and the geopolitics of the dollar’s reserve currency status. For decades, the implicit assumption was that demand for U.S. Treasuries — from foreign central banks, sovereign wealth funds, and global investors seeking the ultimate safe asset — would reliably absorb U.S. issuance at reasonable yields. That assumption is under pressure.

Foreign holdings of U.S. Treasuries, while still substantial in absolute terms, have been declining as a share of the market. The share held by the Federal Reserve has also contracted sharply as quantitative tightening proceeded. The result is a market increasingly reliant on domestic private investors — which is to say, increasingly reliant on precisely the primary dealers and non-bank intermediaries whose capacity the eSLR reform is designed to expand.

In this context, the re-privatization of Treasury market-making represented by the $550 billion in dealer inventories is not merely a domestic banking story. It reflects a structural rebalancing of who underwrites American sovereign debt — away from foreign central banks and the Federal Reserve, toward Wall Street firms operating under incentive structures that are ultimately profit-driven rather than policy-driven.

This matters particularly for the longer-dated end of the yield curve. Primary dealers, unlike the Federal Reserve or long-term foreign investors, are not natural buy-and-hold owners of thirty-year bonds. They are intermediaries who manage duration risk actively. A market more dependent on dealer intermediation is a market more sensitive to the balance sheet cost of holding duration — which means it is a market more sensitive to the regulatory environment that determines that cost. The current eSLR may limit banks’ ability to buy U.S. Treasuries at moments of market distress, particularly as the amount of U.S. debt continues to balloon. Brookings

Benefits Are Real, But They Are Not Risk-Free

It would be intellectually unfair to portray the eSLR reform as a deregulatory gift to Wall Street dressed in public-interest clothing. The case for reform is, in important respects, genuinely compelling — and has been made not merely by bank lobbyists but by serious scholars of financial market structure, including former Federal Reserve regulators.

As the Brookings Institution’s Daniel Tarullo argued — notably, a former Fed governor not known for regulatory permissiveness — the eSLR as designed created real disincentives for the largest banks to perform their intended function in the Treasury market, particularly during stress episodes when their capacity was most needed. The reform addresses a genuine structural flaw, not merely a banker’s wish.

The Federal Reserve’s own analysis confirmed that dealer intermediation capacity was projected to be tested by the ongoing increase in Treasury supply. Every additional billion dollars of dealer balance sheet capacity directed toward Treasury market-making is, in a meaningful sense, a contribution to the smooth functioning of the mechanism through which the U.S. government finances itself — and, by extension, through which the global dollar system maintains its coherence.

The risks are real, however. Concentration risk — the clustering of market-making capacity in a small number of very large institutions — does not disappear simply because those institutions now face lower capital charges. The interaction with the basis trade’s leverage ecosystem remains a source of systemic fragility. And the eSLR reform is, as regulators themselves have acknowledged, a first step in a broader sequence of capital recalibrations that could, if not carefully managed, erode the genuine resilience that post-GFC regulation achieved.

What Comes Next: The Test Will Be in the Stress

The surge in primary dealers’ net Treasury inventories to their highest share of the market since 2007 is, on balance, a structurally constructive development for the world’s most important fixed-income market. It represents a meaningful correction to a regulatory framework that had become misaligned with the realities of a $31 trillion Treasury market, and it comes at precisely the moment when the U.S. government’s borrowing needs are most acute.

But the lesson of the past two decades in financial markets is that structural improvements can also create conditions for structural complacency. The real test of this re-privatization will not come in the benign equilibrium of 2026, when balance sheets are expanding and regulatory headroom is fresh. It will come in the next episode of acute market stress — the next March 2020, the next moment when the basis trade unwinds and repo markets freeze and duration holders seek the exits simultaneously.

In those moments, the question will not be whether Wall Street banks increased their Treasury holdings when times were good. It will be whether they maintained their intermediation function when maintaining it was expensive, risky, and deeply uncomfortable. The eSLR reform gives them the capacity to do so. Whether they will choose to is a question that capital regulation, incentive design, and ultimately financial culture will answer together — and not in advance.

SpaceX’s Falcon Heavy returns to flight on April 27, 2026, launching the ViaSat-3 F3 Asia-Pacific satellite from LC-39A. Only its 12th mission in history, this rare flight completes Viasat’s global broadband constellation and reshapes the GEO vs. LEO satellite broadband competition. Here’s what it means for the new space economy.

At 10:21 a.m. Eastern Time on Monday, April 27, 2026, the most powerful operational commercial rocket on Earth — and one of its rarest fliers — ignites its twenty-seven Merlin engines simultaneously at Kennedy Space Center’s storied Launch Complex 39A. The ground shakes the way the ground is supposed to shake near a rocket: not from a single source, but from a column of fire wide enough to seem geological, to seem geological. Falcon Heavy’s triple-core frame, generating more than 5.1 million pounds of thrust, clears the tower in a wall of sound. Then, minutes later, comes the signature spectacle — two side boosters separating and wheeling back toward Cape Canaveral in precise, mirror-image arcs, landing on Landing Zone 2 and Landing Zone 40 with the kind of choreography that still, somehow, feels impossible. The central core flies on, burns everything it has left, and falls into the Atlantic. Its sacrifice is the price of orbiting a six-metric-ton satellite to geostationary transfer orbit.

This is Falcon Heavy’s twelfth flight in its eight-year operational life. Twelve. The number is almost deliberately understated for a vehicle of this capability. And that rarity — the extended eighteen-month hiatus since its previous mission, NASA’s Europa Clipper in October 2024 — is itself a story worth telling, because it reveals as much about where the commercial space economy is heading as the launch it frames.

A Rocket Reserved for Giants

Understanding why Falcon Heavy flies so seldom requires understanding what it is and what it isn’t. Falcon Heavy is not SpaceX’s everyday workhorse; that role belongs to Falcon 9, which has become perhaps the most routinely astonishing piece of engineering in contemporary aviation history, completing an extraordinary 165 launches in 2025 alone. Falcon Heavy is something else: a vehicle summoned for missions too massive, too energetic, or too classified for a standard Falcon 9 to handle. It is the draft horse you bring out when the load demands it and put back in the barn when ordinary work resumes.

At a listed price of approximately $97 million per launch in its reusable configuration — and roughly $150 million in fully expendable form — Falcon Heavy is already a relative bargain compared to the now-retired Delta IV Heavy, which cost ULA customers between $350 and $400 million per flight. But the market for truly heavy payloads simply isn’t large enough to sustain monthly cadence, and SpaceX has never pretended otherwise. The vehicle was designed for a specific tier of mission: very large commercial communications satellites, deep-space science flagships too heavy for a single Falcon 9, and high-orbit national security payloads demanding maximum throw weight. When those missions come, Falcon Heavy flies. When they don’t, it waits.

What brings it back today is the final satellite of Viasat’s ambitious ViaSat-3 program: the ViaSat-3 F3 spacecraft, destined for the Asia-Pacific region, built by Boeing, and configured with a Ka-band payload designed to add more than one terabit per second of broadband capacity to Viasat’s global network. At approximately 6.6 metric tons, ViaSat-3 F3 is too heavy for a Falcon 9 to lift to the transfer orbit Viasat needs — particularly one favorable enough for the satellite’s electric propulsion to complete the journey to geostationary orbit on a reasonable timeline. As confirmed by Viasat’s own leadership, Falcon Heavy’s superior performance means the spacecraft can be delivered to an orbit just below geostationary apogee with only about three degrees of inclination — cutting weeks off the months-long electric orbit-raising process compared to what an Atlas V delivery required for ViaSat-3 F2.

The Mission in Detail: Engineering a Global Network

The technical architecture of this mission rewards attention, because it illustrates exactly why some satellite programs still require the big rocket rather than the commercially expedient one.

ViaSat-3 F3 will be deployed to geosynchronous transfer orbit — an elliptical orbit with a perigee in the low tens of thousands of kilometers and an apogee near geostationary altitude — approximately five hours after liftoff from LC-39A. From there, the spacecraft’s all-electric propulsion system takes over, gradually raising and circularizing the orbit over the course of roughly two months until ViaSat-3 F3 arrives at its reserved slot at 158.55 degrees East longitude, directly above the Pacific Ocean at geostationary altitude of 35,786 kilometers. Once in position, Viasat expects rigorous bus and payload testing before a commercial service entry expected by late summer 2026.

The satellite itself is a remarkable piece of engineering: a fully flexible Ka-band broadband spacecraft designed to direct its capacity dynamically, rather than assigning fixed amounts of spectrum and power to fixed geographic beams as earlier generations of GEO satellites did. In the words of Viasat’s vice president of space systems, Dave Abrahamian, the constellation’s hallmarks are “a huge amount of absolute capacity, but also the flexibility to put it wherever you need it, whenever you need it.” Traditional satellites — including Viasat’s own earlier generations — operate more like fixed highway lanes: once built, the bandwidth goes where the beams point, regardless of where demand actually flows on any given day. ViaSat-3 F3 is architected to be more like a managed network, allocating spectrum and power dynamically in response to real-time demand.

This flexibility matters enormously for the commercial aviation market, which constitutes one of Viasat’s primary revenue streams. Airline routes shift seasonally and commercially. Demand spikes during peak travel periods and across high-traffic corridors. A satellite that can concentrate capacity over the North Pacific during the morning push and redistribute it over Southeast Asian leisure routes in the afternoon represents a fundamentally different commercial proposition than one locked into static beam patterns.

For the booster side of the mission, SpaceX will fly side boosters B1072 and B1075 back to Cape Canaveral Space Force Station, landing at LZ-2 and the recently commissioned LZ-40 respectively. B1075 carries a flight heritage that includes SDA orbital transport missions, multiple Starlink deployments, and an international synthetic aperture radar spacecraft. Their recovery is not merely theater — it is the economic logic underlying SpaceX’s cost model, allowing the amortized cost of booster manufacturing to be spread across multiple flights. The central core, carrying nothing but a nearly empty propellant load by the time it has done its work, will be expended — a trade-off SpaceX has consistently made on GTO missions demanding maximum performance from the vehicle’s core stage.

Completing the Constellation: What ViaSat-3 F3 Means for Viasat

The ViaSat-3 program has not had an easy journey. When ViaSat-3 F1 arrived in orbit in May 2023, engineers discovered an antenna deployment anomaly that severely constrained the satellite’s throughput — reducing it to an estimated 5 to 10 percent of its intended capacity. For a company that had bet heavily on this generation of satellites to compete against the rising LEO constellations, the setback was consequential. Customers noticed. Starlink, with its terrestrially-derived latency characteristics and rapidly growing coverage, captured aviation connectivity contracts that Viasat had hoped to retain.

The setback also complicated Viasat’s financial position at a moment when the company was simultaneously integrating its transformative 2023 acquisition of Inmarsat — a deal that expanded the company’s maritime and government connectivity business dramatically but also loaded the balance sheet. ViaSat-3 F2, the second spacecraft in the constellation targeting the Americas and EMEA regions, flew on a ULA Atlas V and has been progressing through in-orbit testing, with its reflector deployment now completing after challenges posed by the spring eclipse season. As Viasat’s latest confirmation notes, F2’s final deployments are expected to complete over the coming weeks — meaning the company is, finally, beginning to see its multi-year, multi-billion-dollar satellite program deliver on its intended architecture.

ViaSat-3 F3 completing the constellation closes a strategic gap that has left Viasat without full global high-throughput coverage since the program began. The Asia-Pacific region — home to some of the world’s busiest aviation corridors, fastest-growing maritime trade routes, and largest underserved broadband markets — has been waiting for this capacity. As Abrahamian told Spaceflight Now, “We have a number of airline customers in the APAC region that are really anxious to get this capacity online so they can start serving their customers better.” When F3 enters service, the ViaSat-3 constellation will represent a genuinely global, high-capacity, dynamically flexible broadband network — something no single competitor can claim across every orbit regime.

The Broadband Wars: GEO Renaissance or Rearguard Action?

Here is where the analysis must become honest about the headwinds rather than merely celebrating the engineering achievement.

Viasat’s strategic context is brutal. Starlink has grown to more than two million subscribers, and its low-Earth orbit architecture delivers latency characteristics — typically below 40 milliseconds — that geostationary satellites, orbiting at altitudes 60 times higher, cannot physically replicate. The laws of physics impose a minimum round-trip delay of roughly 550 milliseconds on GEO communications; for most broadband applications this is acceptable, but for latency-sensitive traffic including video conferencing, interactive gaming, and real-time financial transactions, it represents a structural disadvantage no amount of throughput can fully compensate.

Amazon’s Project Kuiper presents a different competitive threat: well-capitalized, backed by Amazon Web Services infrastructure, and designed from the outset for the enterprise and consumer markets where Viasat has historically been strongest. Kuiper has struggled with deployment pace — the program had launched only 78 satellites by mid-2025, far behind the FCC’s schedule — but Amazon’s financial resources and strategic motivation to protect its cloud business by owning connectivity infrastructure represent a long-term competitive pressure that will not diminish.

And yet. It would be a mistake to write GEO satellites out of the connectivity story, for several reasons that the ViaSat-3 program crystallizes.

First, coverage economics. A single geostationary satellite at 35,786 kilometers altitude covers roughly one-third of the Earth’s surface. A LEO constellation providing equivalent global coverage requires hundreds to thousands of individual spacecraft, each with a design life measured in years rather than decades. The capital efficiency of GEO for serving large geographic areas — particularly over oceans and sparsely populated territories where ground infrastructure is limited — remains compelling. ViaSat-3 F3’s coverage of the Asia-Pacific region, from a single orbital position, encompasses an area that would require a significant fraction of a LEO constellation to replicate.

Second, the defense and government market. Viasat has historically derived substantial and growing revenue from U.S. and allied government customers who value the satellite’s dedicated capacity, security architecture, and the ability to integrate with existing military communication networks. ViaSat-3 F3 explicitly introduces “new forms of resilience for US and international government customers,” per Viasat’s official launch confirmation. The national security satellite broadband market values characteristics — including resistance to jamming, controlled access, and sovereign oversight — that a commercially operated LEO megaconstellation does not automatically provide.

Third, the multi-orbit future. The most sophisticated satellite operators today are not choosing between GEO and LEO. They are building hybrid architectures that leverage the throughput and geographic efficiency of GEO alongside the latency characteristics of LEO, using intelligent ground terminals and network management to route traffic dynamically. Viasat’s own NexusWave service integrates its GEO capacity with OneWeb’s LEO network for maritime customers. The ViaSat-3 constellation, as it reaches full operational capability, becomes a cornerstone of this hybrid strategy rather than a standalone product competing head-to-head against Starlink on latency.

The Economics of Reusability and the Launch Market’s Quiet Monopoly

Step back from the satellite payload for a moment and consider the launch vehicle. Falcon Heavy’s twelfth flight in eight years is, by any conventional measure, an extremely low flight rate for a rocket of this capability. Yet SpaceX has maintained a 100 percent mission success rate across all twelve flights, and the booster recovery on dual RTLS missions has become so routine that it barely registers as remarkable. This combination — extreme reliability at very low cadence — reflects a deliberate commercial strategy that deserves scrutiny.

There is, in practical terms, no alternative to Falcon Heavy in the current market for very large GEO satellites requiring maximum performance to orbit. ULA’s Delta IV Heavy was retired in 2024. Ariane 6, which was originally scheduled to launch ViaSat-3 F3 before development delays and the post-Ukraine reshuffling of launch manifest assignments moved the spacecraft to Falcon Heavy, offers an alternative for European and international customers — but it has struggled to achieve reliable launch cadence and its payload capacity to GTO falls below Falcon Heavy’s peak performance in expendable or partial-recovery configurations. Blue Origin’s New Glenn is operational but has experienced anomalies in early missions, limiting customer confidence. ULA’s Vulcan Centaur serves the national security market but does not offer the throw weight that Falcon Heavy provides.

This effectively means SpaceX holds a de facto monopoly on western heavy-lift launch services for the largest GEO satellites. That is not a comfortable position for an industry that values competitive tension to discipline pricing and incentivize innovation. Viasat, to its credit, originally sought Ariane 6 specifically to maintain European launch options and reduce dependence on SpaceX. The inability of European industry to deliver that alternative on schedule — a consequence of years of chronic underinvestment in European launch infrastructure and the disruption caused by Russia’s elimination from commercial launch markets after 2022 — left Viasat with no practical choice but to return to SpaceX.

The concentration of launch capability matters for industrial policy reasons as much as commercial ones. NASA’s decision to launch Europa Clipper on Falcon Heavy, saving an estimated $2 billion compared to the Space Launch System, was fiscally prudent but also highlighted how completely the U.S. government’s civil launch needs have become dependent on a single private company. When that company is also developing Starlink — a direct commercial competitor to satellite operators like Viasat — the dependency creates tensions that regulators and policymakers are only beginning to grapple with seriously.

Critical Perspectives: Concentration, Fragility, and the Starship Shadow

Any honest assessment of today’s launch must acknowledge the risks embedded in the picture it presents.

Market concentration is the most obvious concern. SpaceX’s dominance of the launch market — executing approximately half of all orbital launches worldwide in recent years, including virtually all U.S. commercial and government heavy lift — is without precedent in the space age. The company’s technical excellence is not in question. But technical excellence is not a sufficient safeguard against the risks that concentration creates: single points of failure in supply chain, the potential for pricing power to increase as competition diminishes, and the strategic complications that arise when a launch provider’s commercial interests are entangled with those of its customers. The European Space Agency and its member states have been reckoning with these consequences since Ariane 6 fell behind schedule; the U.S. government has been slower to act.

The ViaSat-3 F1 lesson is also worth carrying forward. A single antenna deployment anomaly on a satellite that cost hundreds of millions of dollars and several years to build reduced its throughput to a fraction of its designed capacity. For programs predicated on multi-terabit capacity, this kind of single-point failure can be financially devastating. The space insurance market absorbs some of this risk, but it cannot absorb the strategic cost of arriving at the GEO broadband market years late and at a fraction of expected capacity. The resilience of the ViaSat-3 program — its ability to absorb the F1 setback and continue toward F3 launch — reflects the financial depth that came with the Inmarsat acquisition. Smaller satellite operators would not survive an equivalent anomaly.

The Starship era represents a more fundamental disruption lurking behind today’s Falcon Heavy mission. SpaceX’s next-generation launch vehicle, still in flight testing, promises to carry payloads to low Earth orbit measured not in tens of metric tons but in hundreds — in a fully reusable configuration. When Starship reaches operational status, it will not merely compete with Falcon Heavy; it will displace it for most missions, while simultaneously enabling satellite constellation architectures of a scale and cost structure that will make today’s GEO programs look like the previous generation of space infrastructure — necessary, valuable, and eventually superseded.

The timing of ViaSat-3 F3 thus acquires a particular resonance. This spacecraft will likely remain in commercial operation for fifteen years or longer. By the time it retires from service in the early 2040s, the satellite broadband market will look almost unrecognizable compared to what we see today. The operators that survive will be those who have built the most flexible, multi-orbit, software-defined network architectures — and who have done so without betting so heavily on a single generation of hardware that they cannot pivot when the next generation arrives.

The Geopolitics of Coverage: Who Gets Connected, and Who Decides

Zoom out one more level, and the ViaSat-3 F3 launch carries implications that extend beyond corporate strategy into international relations and development economics.

The Asia-Pacific region is the world’s most economically dynamic. It is also the region with some of the most pronounced disparities in connectivity. The aviation market — Viasat’s primary immediate revenue target in the region — connects the affluent and the mobile. But the underlying capacity infrastructure that ViaSat-3 F3 provides will also serve maritime vessels, island communities, remote enterprise sites, and eventually, through service expansion, populations in some of the world’s most connectivity-starved areas.

This is not altruism on Viasat’s part; it is market expansion. But the geopolitical dimension is real. When U.S.-headquartered satellite operators extend high-throughput, high-reliability broadband coverage across the South China Sea, the Pacific Islands, and the maritime corridors of Southeast Asia, they are making infrastructure decisions that have strategic implications. The race between American and Chinese satellite operators for coverage of the Indo-Pacific region is not merely commercial — it is a contest over which country’s technical standards, legal frameworks, and network architectures become the default infrastructure for an economically and militarily critical region.

China’s own ambitions in this domain are serious and well-funded. China Satellite Network Group, the state-owned entity overseeing the Guowang LEO constellation, has filed for orbital slots that would place it in direct competition with Starlink and other western operators for limited spectrum resources. The completion of Viasat’s GEO coverage over the Asia-Pacific, combined with ongoing LEO buildout by U.S. operators, represents a concrete broadening of American-aligned connectivity infrastructure across a region where that presence matters.

Conclusion: The Weight of a Rare Launch

Eighteen months of quiet, and then: twenty-seven engines, 5.1 million pounds of thrust, a spectacular double booster landing, and a six-ton spacecraft on its way to geostationary orbit above the Pacific. There is something fitting about the rarity of Falcon Heavy’s flight pace. Each launch carries more weight — literal and figurative — than the routine. Each one lands in a market landscape that has shifted since the last, and must be interpreted against that shifting context.

Today’s mission completes what Viasat set out to build. Whether that completion arrives soon enough, at sufficient capacity, and at competitive enough terms to hold meaningful market share against the LEO operators is the question that will determine the company’s next decade. The honest answer is: probably, in some segments; probably not, in others. The in-flight connectivity and government markets will sustain meaningful GEO operators for the foreseeable future. The mass consumer broadband market — where Starlink and eventually Kuiper will compete on price and latency — is likely beyond recovery for GEO-only strategies.

But the more durable insight from watching Falcon Heavy lift off today is about the infrastructure of ambition. The rocket that launched a Tesla Roadster toward Mars for a demo flight in 2018 has, in twelve missions, launched classified military satellites, a spacecraft headed for Jupiter, weather observation platforms critical for hurricane forecasting, and now the final piece of the first commercially deployed global multi-terabit broadband constellation. It has done so at a fraction of what its predecessors cost, with a booster recovery system that turns what used to be expensive expendable stages into reusable assets.

That is the story the launch market keeps telling, in different configurations and with different payloads: that the economics of access to space have been permanently disrupted, that the disruption is still accelerating, and that the satellites we put up today will operate in a world the launch industry of a decade ago could not have anticipated. ViaSat-3 F3 will look down from 35,786 kilometers at a world connected in ways its designers planned for, and ways they did not. That is, perhaps, the most precise definition of infrastructure worth building.

The call came, as these things often do, without warning. A seasoned equity partner at one of the Big Four — two decades of late nights, cross-border engagements, client dinners, and carefully cultivated relationships distilled into a six-figure “units” allocation — was summoned for what was framed as a career conversation. The language was collegial, the room was quiet. And then, politely but unmistakably, the message landed: you will no longer share in the firm’s profits. We are moving you to a salaried partner role.

No performance improvement plan. No transparent benchmark they had failed to meet. Just the quiet arithmetic of a partnership that needed fewer people at the table.

This is not an isolated anecdote. According to reporting by the Financial Times, both KPMG and EY have in recent years removed members of their UK equity partnerships and instead offered them “salaried partner” roles — a demotion wrapped in the same title, drained of its financial substance. And on April 23, 2026, the story took on transatlantic dimensions: KPMG announced it was cutting roughly 10% of its US audit partners — approximately 100 individuals — after years of failed voluntary retirement programmes. The message to the profession has never been louder: the partnership is no longer a destination. It is, increasingly, a temporary assignment.

The Golden Ticket, Tarnished

For generations, making partner at a Big Four firm was the legal and financial world’s closest equivalent to a tenured professorship. You had, in the popular imagination and in contractual reality, arrived. The equity partnership conferred ownership, profit-sharing, prestige, and an implicit understanding that barring catastrophic misconduct, your position was secure until mandatory retirement. It was, in the language of another era, a job for life.

That compact is dissolving — not with a dramatic rupture, but through a series of quiet institutional manoeuvres that, taken together, signal a structural reorientation of how these firms are governed, whom they reward, and what professional excellence is now expected to deliver.

The statistics are unambiguous. Big Four partner promotions across the UK fell to just 179 in 2025, a five-year low and a sharp retreat from the 276 promoted at the peak of the post-pandemic boom in 2022, according to analysis by the Financial Times of Companies House filings, press releases, and LinkedIn data. EY elevated only 34 equity partners, down from 74 in 2022. Deloitte made just 60 promotions, against 124 in 2022. Overall, the total number of equity partners across the four firms fell for the first time in five years, dropping by roughly 80 to approximately 3,050.

The belt-tightening is deliberate, and its beneficiaries are the incumbents. KPMG’s average UK partner pay reached £880,000 in 2025 — an 11% year-on-year increase — putting it ahead of both PwC (£865,000) and EY (£787,000) for the first time since 2014. Deloitte partners crossed the £1 million threshold. Revenue, meanwhile, has barely moved: EY reported 2% growth in what it called a “challenging market”, while KPMG posted just 1% growth after 9% in 2023, and Deloitte suffered its first annual revenue decline in 15 years.

The mechanism is elementary. When you constrain the denominator — fewer equity partners sharing the profit pool — the numerator rises for those who remain. Profit-per-equity-partner (PEP) is the prestige metric in professional services, the figure that determines lateral hire competitiveness, graduate recruitment marketing, and the partner’s own sense of institutional worth. And right now, the Big Four are protecting it with considerable ruthlessness.

Demotion Without Firing: A New Instrument of Control

What distinguishes the current moment from previous cycles of partner attrition is not the reduction in numbers per se — firms have always managed their equity pools — but the instrument being used. The introduction of a salaried or “non-equity” partner tier creates a new, lower rung on the ladder that can be used not merely as a holding pen for promising directors, but as a landing zone for underperforming incumbents.

Deloitte, EY, and KPMG have all introduced this salaried partner tier, widely regarded in the industry as a mechanism for retaining senior staff without sharing profits. PwC, the only firm still operating an equity-only partnership, has created a “managing director” grade as its structural equivalent. The title is preserved; the economics are fundamentally altered.

In the case of KPMG’s UK operation, multiple people with knowledge of the matter told the Financial Times that partners were called into rooms for what were “positioned as career conversations” but were in reality mechanisms to reduce equity partner headcount. Some received the news with little warning, having been given positive performance feedback until the conversation itself. Several chose to leave rather than accept what they experienced as a demotion, describing the process as blindsiding.

EY, meanwhile, has demoted a small number of equity partners to salaried roles since introducing the tier in 2022, according to three people familiar with the matter. The firm declined to comment.

To be clear, “departnering” is not unique to accountancy. Goldman Sachs has long managed partner membership with clinical precision; law firms regularly de-equitise underperforming partners, particularly in mid-tier practices. But the cultural signal from the Big Four is significant precisely because of the scale, the prestige mythology, and the professional pipeline implications. These are the firms that recruit tens of thousands of graduates annually on the implicit promise of a meritocratic climb toward a life-altering outcome.

Why Now? Three Interlocking Forces

1. The Consulting Hangover

The pandemic generated an extraordinary and, in retrospect, unsustainable surge in demand for advisory services. Governments needed economic modelling, corporations needed digital transformation, boards needed risk assessment. The Big Four expanded headcount aggressively. By 2022, PwC was promising to add 100,000 staff globally; KPMG was promoting equity partners at a rate it could not sustain.

The hangover has been severe. PwC’s revenue growth slowed to 2.9% in fiscal 2025, down from 9.9% in 2023. Consulting revenues have contracted across the sector as clients, now operating in a tighter macro environment, question the value of expensive advisory mandates. James O’Dowd, managing partner at Patrick Morgan, told City AM that the firms are “cutting jobs to protect partner profits and rebalance bloated teams” after years of aggressive post-pandemic hiring.

2. AI Restructuring the Audit Architecture

Perhaps more structurally significant than the revenue cycle is the accelerating role of artificial intelligence in reshaping what partners actually do. KPMG launched its Workbench multi-agent AI platform in June 2025, developed with Microsoft, connecting 50 AI agents with nearly 1,000 more in development. EY granted 80,000 tax staff access to 150 AI agents through its EY.ai platform, investing more than $1 billion annually in AI platforms and products. Deloitte struck a deal with Anthropic to deploy Claude AI to its 470,000 employees worldwide.

The point is not that AI will replace partners tomorrow. It is, rather, that the work historically required to justify a partner’s existence — managing audit workflows, overseeing large teams of junior staff performing repetitive compliance tasks, supervising structured data review — is increasingly automated. KPMG acknowledged as much in its US announcement, noting that artificial intelligence is “increasingly handling key steps of audits, spurring firms to rethink staffing and delivery”. At PwC, leadership has indicated that new hires will be doing the work of managers within three years, supervising AI rather than performing the audit tasks themselves.

This compression of the value chain has a direct implication for partner economics. If AI can execute the audit procedures that previously required six team members, you need fewer partners to supervise them. The case for a large partnership structure becomes harder to make.

3. The Future-Revenue Problem

Laura Empson, professor of management at Bayes Business School, has articulated the third driver with particular precision. The question being asked of potential partners has shifted from “can you generate enough business this year?” to something more existential: “Will this person generate a substantial stream of income for the foreseeable future — and right now the future is particularly hard to foresee?” A director with a strong practice in regulatory compliance was, five years ago, a safe bet. Today, as AI takes on compliance automation and regulatory technology firms encroach on traditional advisory turf, the projection is far murkier. The firms are not just managing the present — they are hedging against futures they cannot yet model.

Winners, Losers, and the Long Game

The winners in this restructuring are, in the near term, the incumbent equity partners who remain. By shrinking the pool and reweighting units toward rainmakers — under KPMG’s current leadership, the firm has reallocated profit units to place less weight on tenure and more on business generation — the firms are concentrating extraordinary wealth among a smaller group. KPMG’s UK partners, who were earning £816,000 on average in 2025’s reporting cycle and £880,000 in the most recent period, now out-earn their counterparts at EY for the first time in a decade.

The losers are harder to count but easier to identify. The most acute damage falls on the cohort of ambitious directors and senior managers who have spent a decade or more building toward equity partnership as their defining professional objective. James O’Dowd of Patrick Morgan noted that whereas 20 years ago, Big Four employees could make equity partner by around 35, they are now looking at their early 40s — if they get there at all. The salaried partner tier is, for many, not a staging post but a terminus.

There is also a diversity dimension that deserves sharper scrutiny than it typically receives. Research consistently shows that informal sponsorship, visibility networks, and the “cultural fit” judgements that govern partnership decisions tend to replicate existing demographic profiles. When promotion cycles compress and the bar rises, historically underrepresented groups — women, minorities, first-generation professionals — disproportionately absorb the attrition. The firms publish annual diversity data with admirable transparency; whether that transparency translates into accountability when the pressure is on remains a live and uncomfortable question.

More troubling still is the impact on institutional knowledge. Partnership models, whatever their flaws, created an incentive for long-term relationship stewardship. A partner who owned the firm had reasons to invest in client relationships, mentorship, and institutional culture that extended well beyond the quarterly cycle. When you strip equity from people who have spent twenty years building domain expertise, you create a class of high-skilled employees with diminished loyalty and a market incentive to take their networks elsewhere — to boutiques, to in-house roles, to competitors offering better economics. The knowledge transfer implications are real.

The Contrarian View: Are They Trading Resilience for Returns?

Here is the question the managing partners are not asking loudly enough: does concentrating profits in fewer hands make these firms better, or merely more profitable in the short term?

There is a credible argument that what looks like strategic discipline is actually a structural fragility in the making. The Big Four derive much of their value not from capital but from trust — the trust that a client places in an auditor’s independence, the trust that a regulator places in a firm’s quality controls, the trust that markets place in a signed opinion. That trust is accumulated slowly, through relationships, through institutional memory, through the kind of deep sectoral expertise that takes years to develop.

When you compress the partner class aggressively, you signal to the broader professional pipeline that the implicit social contract has changed. Junior auditors at KPMG UK, earning around £32,500 as new graduates while partners take home nearly £880,000, are already observing a ratio that strains credulity as a meritocratic proposition. Removing overtime pay for busy season, shrinking the equity pool, and quietly demoting long-tenured partners does not create the conditions for the recruitment and retention of the next generation of exceptional audit professionals.

There is also the audit independence question. The Financial Reporting Council and its international equivalents have long expressed concern that commercial pressures on audit firms compromise the independence of judgment that audits require. A partnership model explicitly oriented toward protecting PEP — where the primary signal of success is partner compensation rather than audit quality — does not obviously serve the public interest that audit is meant to protect.

What Comes Next: Three Scenarios for the Profession

The optimistic scenario holds that these are rational adjustments to a structural oversupply of partners accumulated during an anomalous boom period, and that AI will simultaneously create new value — in AI assurance, ESG verification, regulatory technology — that supports a leaner but higher-margin partnership in the medium term. EY’s vision of a “service-as-a-software” commercial model, where clients pay by outcome rather than hour, might indeed generate the next platform for partnership growth.

The bearish scenario holds that compression of the talent pipeline, combined with AI-driven commoditisation of core services, will accelerate the fragmentation of the Big Four’s market position. Boutique advisory firms, technology-native audit platforms, and specialist consultancies are already capturing the mid-market segments where the Big Four’s scale is a disadvantage rather than an asset. If the firms price themselves out of the talent market by narrowing the partnership pathway, the talent goes elsewhere — and so, eventually, do the clients.

The structural scenario — and the one with the most historical precedent — is that this marks not a temporary adjustment but a permanent restructuring of what professional partnership means. The partnership model of the 20th century was predicated on human capital scarcity: expertise was concentrated in senior people, and those people needed to be economically incentivised to stay. AI erodes that logic. The next model may look less like a traditional partnership and more like a technology firm with a professional services overlay — equity concentrated at the top, a salaried technical workforce in the middle, and an AI infrastructure doing much of the work below.

For Aspiring Partners, Directors, and Regulators

If you are a director or senior manager at a Big Four firm reading this, the strategic implication is uncomfortable but clear: the pathway to equity partnership is narrower, later, and more uncertain than at any point in the past two decades. The hedge is diversification — cultivating expertise in areas where AI augments rather than replaces human judgment (regulatory navigation, complex cross-border transactions, AI assurance itself), and building client relationships that are genuinely portable. The salaried partner tier may, for some, represent a viable and well-remunerated alternative. For others, the boutique and in-house markets have never been more attractive.

For regulators, the questions are structural. Does the concentration of equity in fewer, higher-paid partners improve or compromise audit quality? Do the oversight frameworks that govern partnership conduct need updating to reflect the new realities of AI-assisted audit and performance-managed equity pools? The FRC and PCAOB have the tools to ask these questions. The political will to pursue them publicly is another matter.

For the firms themselves, the most important question may be one they are reluctant to examine: is the protection of partner compensation a strategy, or a symptom? A strategy would involve investing in the next generation of talent and expertise with the same vigour applied to protecting the equity pool. A symptom would be the short-term extraction of value from a franchise whose long-term competitive position is quietly eroding.

The Covenant, Rewritten

There is a moment, in the mythology of professional services, when a young accountant or consultant first allows themselves to imagine making partner. It is a moment of ambition and delayed gratification — the belief that if you are good enough, disciplined enough, client-focused enough, the institution will eventually reward your investment with a share in its future.

What KPMG and EY are doing — quietly, through human resource conversations in unremarkable meeting rooms — is rewriting that covenant. The reward is no longer guaranteed by longevity or even by excellence across a career. It is contingent, performance-managed, and revocable. In that sense, they are asking their most senior professionals to accept an employment relationship that the most junior associates have always known.

That may be a more honest model. It is certainly a more anxious one. And whether the profession that emerges from this restructuring will be better equipped to serve the public interest — or merely better equipped to serve the interests of those already at the top — is the defining question for the decade ahead.