The Dilution Delusion
Markets are concentrated in fewer companies. Capital is concentrated in fewer managers. Power and wealth is concentrated in fewer people.

Rage, fear, jealousy, envy—the primal forces of human psychology are themselves concentrated on the have-compute, the have-revenues, the have-portfolios, the have-networks and the have-influence.

In chemistry, a solution with too high a concentration can be diluted—a delusion that’s spreading. New York and other states are proposing moratoriums on new data centers, writing themselves out of future innovation and economic growth. California is proposing a wealth tax on billionaires, forcing its most successful entrepreneurs to flee to freedom elsewhere. America is raising barriers against the smartest minds coming to the world’s greatest economy. Investors are advised to buy a balanced basket of ETFs even as the Mag 7 runs away with the market.

Concentration is capitalism’s secret success. It goads asset allocators to place the heaviest bets on the brightest prospects. It prods founders to seek the strongest partners in the most fertile niches ready for returns. It cajoles employees to headhunt to emerging winners. That concentration is also where risk emerges. Returns, like reality, obey the laws of physics. As Richard Feynman noted, physics—unlike people—can’t be fooled. The dilution delusion must be ignored in favor of confident concentration in the ceaseless competition known as civilization.

Our last letter was retrospective—a 25-year origin story built on gratitude, luck and some reasonably good decision-making. This one looks forward, with what we hope is concentrated clarity where others tend toward dilutive doubts.

On the Concentration of Public Markets
Every January delivers the same lesson, and every January most forget it. The forecasters who called 2025 with such precision are nowhere to be found—because they don't exist. We mistake pattern recognition for prophecy, confuse narrative coherence for causal understanding, and leave intact one certain truth: nobody knows nothing.

What's available to anyone willing to trade prediction for preparation is a landscape dense with both opportunity and risk. The known risks array across familiar axes: geopolitical fracture accelerating, social cohesion fraying in ways that mock prior polarization, economic structures straining under debt loads once considered unsustainable, and technological disruption escalating faster than institutions can metabolize. Had we written that sentence to open any letter of the last decade, it would have been equally true. What matters instead are the unknown and unexpected surprises (are there any other kind?) that arrive within weeks of each new year to re-concentrate us after the dilutive doldrums of December.

Physics teaches that systems in disequilibrium eventually correct and the question is never whether, but how—and how violently. Excess energy must dissipate, whether by gradual cooling or explosive release.

At present, highs abound—perhaps much higher than most appreciate: concentration risk in public markets, debt levels for companies and governments alike, and a cheery consensus that the year ahead will deliver only positive returns. What a setup for negative surprise and disappointment. We say "failure comes from a failure to imagine failure" not out of pessimism but because its inverse is the secret to both happiness and sanity: positive surprise and expectations exceeded. Our doubts shouldn’t dilute our actions, but concentrate them.

On market concentration: five companies now represent roughly 30% of the S&P 500’s market cap. The top 10 exceed 40%—the highest concentration in 50 years—while contributing over 50% of the index's return. Even as the Fed cut rates at the short end, the 10-year Treasury yield resisted.

Consider the concentration of risk and structural fragility beneath that surface. In 2008, nearly 40% of foreign portfolio investment into the U.S. came from central banks and sovereign wealth managers. Today it is 13%. Meanwhile, the share of U.S. equities held by foreigners has nearly tripled to a record high—from just over 20% to just under 60%. When the marginal foreign holder of U.S. assets is an institution chasing returns rather than a central bank storing reserves, the structure is both more concentrated and more fragile.

U.S. equities underperformed after the dot-com bust. Foreign buying of Treasuries went up, so yields went down. Since investors chased yield, the markets produced a glut of suspect structured products promising it, and we crescendoed to the crisis of 2008. Today, if equity concentration unwinds and the dollar continues its decline (already down 10%, though it fell 40% after the dot-com bust), then foreign investors may repatriate and the dollar may drop even further. On the other hand, they may turn skittish and rotate into Treasuries, handing Trump, Warsh and Bessent lower rates without major Fed intervention. Concentration risks demand watching.

Macro conditions set the terms for micro decisions. Rising public markets invite privately-held companies to list, and in biotech the invitation is getting warmer: the XBI was up over 35% last year, M&A has reawakened, and an IPO window appears to be opening amid the thaw of another cyclical biotech winter. Biotech companies produced more than 70% of newly approved drugs in recent years, yet biotech investment as a share of overall private funding hit a 20-year low.

That gap matters more than usual. Over 90% of FDA-approved drugs in the last half century had some form of NIH funding, and public-sector R&D now faces serious cuts across nearly every major federal agency. A retreat in public funding puts a premium on private ventures with pathbreaking labs producing cutting-edge compounds. The dilution of public funding must be matched by the concentration of private investment.

Capital markets have a capricious capacity for compartmentalization. The sickening in SaaS has sustained, stripping value from software companies deemed prey for AI predators—even as AI itself has evolved from speculative mania into hardened infrastructure investment, concentrated now on enormous (and possibly unsustainable) capex for compute and power generation.

Hyperscalers have hyperscaled their spending to over $600 billion in the year ahead. Only the Louisiana Purchase of 1803 has exceeded their spend as a percentage of GDP (3% vs 2.1%). The current wave of tech capex (now over $1 trillion) exceeds the Manhattan Project, the Apollo program, the interstate highway system and rural electrification combined, even as revenues remain in the tens of billions. Until recently, debt was mostly absent, dampening risk and keeping things comparably sober. It has now arrived at the party and even brought a plus-one in the form of off-balance-sheet funding structures. Debt is just dilution in disguise.

In public markets, AI will flatten alpha to beta, eroding much of the edge that most investors possess or believe they do. Of the three kinds of edge—informational, analytical and behavioral—the analytical erodes first and fastest. Models can find anomalies and analyze cash flows and call transcripts better than humans, and do so more quickly and with superior pattern-matching. Behavioral edge will persist, especially as so many investors are passively herded into so few names. The surprise is the informational edge. In public markets, regulation has largely flattened it. In private markets, it widens. Why? Because there is no publicly listed universe of traded equities for companies that have not yet been created. This is the addictive nature of what Lux does: like the unreported discovery of a scientist, we seek to know something the rest of the world doesn't and won't until we tell them.

On the Concentration of Private Capital
In a few weeks, a new company will launch that only a handful have heard of, taking an approach few have seen, with unfair access to talent, technology and secrets its founders won't speak of publicly. There’s no Bloomberg terminal for that. No query, no quant model, no sum-of-the-parts, no high-frequency trade to front-run it. It’s just this: we knew a guy who knew a guy, and both those guys knew and liked us all at Lux.

Dilutive doubts abound in venture: right team, right idea, right time? Risk is nothing more than the economic expression of doubt. Confident concentration comes not from heedless hubris but rather through careful combinatorial construction of peerless networks, deep partnerships, exclusive intelligence, execution agility and an ever-evolving flexibility of mind toward the world as it is than as it was.

We enter the year with a lot of dry powder. We closed our ninth Lux fund at $1.5 billion, with substantial reserves from prior funds, and now manage $7 billion for some of the greatest institutions and missions globally. Our raise took three months from launch to close—notable because overall venture fundraising dollars are down roughly 20% to the lowest level since 2019 and the average GP now spends nearly two years raising, double the timeframe of 2021. Credit belongs to our team, but even more so to the Lux-family founders who give us the right to win and the track record to earn it.

With heightened scale comes heightened responsibility—not only for the capital we manage and those for whom we manage it, but for the gravity of what our companies do. Venture is often described as a game of optionality and upside. True financially, but incomplete morally. Our portfolio companies don’t build e-commerce sites or mobile apps. They build sensors that precisely guide munitions and therapeutics that extend human life—matter that matters.

More than ever, our work at Lux finds us as the earliest believers and backers, including starting companies de novo and spinning founders out of large companies into new ones. We treat investing as the bespoke craft that venture was at its roots—not as “asset management.” In a frenzied venture environment, it’s easy to dilute attention, and ever more valuable to concentrate it.

It’s a Lux mantra to believe before others understand, something we only get to do if we realize (returns) because others also realize. In the past quarter, we realized returns with distributions from Lux companies including Nozomi (acquired by Mitsubishi), Kyruus (acquired by RevSpring), Chronosphere (acquired by Palo Alto Networks) and EvolutionaryScale (acquired by Chan Zuckerberg Initiative). This is against a generally lackluster backdrop for exits. Less than 10% of 2021 vintage funds have returned any capital (compared to 25% for 2017 vintages). Median hold periods have stretched from just over four years in 2010 to just under seven.

Concentration of attention is turning into concentration of capital. Several Lux companies trade on secondary markets at premiums—not discounts—including Databricks, Ramp and Anduril, all of which carry high demand and high expectations for future IPOs.

We see concentration across the venture landscape. Our long-expressed view of the coming extinction and involuntary exit of small VC funds (“minnows”) is empirically underway. Active VC firms declined 25% from around 8,000 in 2021 to 6,000 last year. First-time funds averaged just $7 million—less suited to venture and more a restaurant opening. The minnow extinction will accelerate and further concentration will follow. At the other end, the “megas” are gathering assets at an increasing pace as predicted, expanding with foreign country initiatives, launching multiple products, and moving toward voluntary exits with several planning to go public. This follows the path PE firms charted a decade ago.

The surviving firms—including Lux—will enjoy less competition, better pricing power and most importantly, compounding relationships with coveted founders whom allocators will pursue for exposure and direct co-invests.

The concentration of private capital is only matched by the concentration in equities. Nearly $340 billion flowed into U.S. deals, the second-highest amount ever, yet it was packed into the fewest deals of the decade, with nearly half the capital concentrated in a few dozen deals over $500 million. The top 1% of companies by valuation now absorb a third of all venture capital deployed. The bottom half get 7%, technically “venture-backed” but also kind of like sitting in the parking lot at a concert and claiming you saw the show.

Half of all venture money went to 0.05% of deals, while half of all LP capital went to a handful of funds. And where did the money ultimately go? Those two vowels on our collective keyboards are as worn from striking them as our ears are from hearing them.

Five AI companies are already worth more than every venture-backed IPO of the dot-com era combined. Whether they represent prophecy or hubris depends entirely on the next two years. “This time is different” are dangerous words that may yet be true. There’s a vast gap between what investors feel and what prices imply. Investors are “hedgespeaking.” In public, nobody wants to be on the record throwing shade if the future turns out to be bright. Privately, so many doubts are harbored they can’t get a dock slip.

Lux’s concentration remains invariant: matter that matters behind the most ambitious people who turn “this time is different” into “this time will be different.” We spend our days evaluating risk: sizing founders, technologies and markets to determine whether capital should flow in or out. To take a first meeting or a second. To convert concentrated conviction into focused energy from the entire Lux team, and to earn the right to win. We make many small decisions every day and a few very consequential ones every year—and to be honest, it’s often not clear until later which was which.

The financial calculus of committing is straightforward: uncertain asymmetric upside if we are right (or lucky), certain loss of what we risk if we’re wrong (or unlucky)—along with some reputational risk, mostly recoverable. But throughout Lux’s history, our errors of omission have proven far more painful than those of commission. The deals we did that missed (eventually realizing losses) haunt far less than the deals we didn't and dismissed (eventually realizing the profits we missed). Our never-ending task is to find the balance: minimizing the permanence of regret while maximizing a highly selective bar.

An Aside: Who Watches the Watchers
Lux concentrates on the gap between sci-fi and sci-fact, but sometimes the most potent alpha comes from watching what the world is watching. Every major market cycle produces a cultural artifact that captures—and shapes—its zeitgeist, often signaling the top just as the herd piles in. In markets, narratives don't just follow prices: they prepare them, justify them and prosecute them. It’s an arc of romanticizing, rationalizing, and then regretting.

Fitzgerald's The Great Gatsby didn't merely describe the Roaring Twenties; it was a cultural tombstone for an era of desperate excess, published just four years before the crash of 1929. Gatsby reaches for the green light just as the market reaches for the exit. In 1987, Oliver Stone gave us Wall Street with Gordon Gekko and brick-sized car phones—intended as indictment, taken as invitation, arriving just as the LBO craze crested.

We’ve previously chronicled the cycle: The Social Network canonized the founder-king before Facebook's IPO, romanticizing a decade of code in the aftermath of the dot-com crash. Silicon Valley satirized peak techno-optimism while feeding it oxygen and normalizing its absurdity. Billions hyped hedge funds and equities at their apex. Then came the prosecution in the form of a hangover trilogy: The Dropout, WeCrashed, and Super Pumped, marking the death of growth-at-all-costs for retail investors who watched as the bill came due and their portfolios emptied. Hollywood is a delayed but surprisingly reliable signal, sensing what capital is about to reward and what it’s already stopped forgiving.

So what is today’s signal? Landman. The zeitgeist is shifting from people who code bits to people who extract atoms. From equity investment to debt financing. From cap tables to mineral rights. From Palo Alto garages to West Texas lease agreements. From "growth solves all" to "cash flow pays interest." It’s less nostalgic and more anticipatory cultural positioning. 

Audiences grok what many sector analysts haven’t yet comprehended: the next decade of wealth creation will look more like Landman than The Social Network: capital-intensive, physically constrained, and far more dependent on legal permitting and infrastructure access than genius coding. If capital goes where it’s welcome and stays where it’s well-treated, then this cultural moment—of pickup trucks and long drives to oil rigs with massive machinery that must be manned and maintained—anticipates a gusher of cash to physical assets (beyond just data centers) that are airgapped from software systems, isolated from AI disruption, and immune to automation. The hero has changed: from the coder nerd moving fast and breaking things to the politically connected engineers navigating the brutal physics of the real world.

On the Concentration of Power
From the Middle East to Venezuela, the past year has seen the most muscular manifestation of American power and realized foreign policy results in recent memory. The increasing concentration of power among the hegemons—America, Russia and China—has led to widespread bemoaning that the world is more confusing or chaotic or fragile than ever before. 

Power abhors a vacuum, and the reality is this: the United States has intervened in over 400 foreign conflicts since its founding—not out of charity, but because intervention serves four core interests: geopolitics (denying rivals like Russia and China strategic territory), economics (protecting trade routes, the petrodollar, and the predictability commerce requires), values (serving as a humanitarian beacon to deter dictators and aid those who suffer), and domestic security (neutralizing threats abroad before they reach our homeland). Every foreign entanglement maps to this framework. We’ll cover a few geopolitical trends that directly guide and impact a few Lux portfolio companies, which in turn may impact these trends.

In aggregate, we have several billion dollars worth of investments across aerospace and defense. Total U.S. venture investments in the sector hit $21.4 billion in 2025 across over 300 deals, nearly doubling its prior peak. Domestic defense spending may climb to 5% of GDP by next year, a figure that would have seemed indecent to many allocators who, not so long ago, were inserting ethics clauses to avoid exposure to precisely the kind of work they now seek to fund.

We have long held that there is a moral imperative for good actors to produce superior technologies that deter and disincentivize bad actors, so as to prevent war and suffering. Alas, patriotism and profitability are not synonyms. Too many companies will be funded, bubbles will form, and most will fail (and will have been wrongly comped to the few that succeed). Many investors will shortly learn that they have diluted their defense dollars across too many almost-primes.

We have invested across U.S. and allied defense, with Lux companies actively engaged in deterring conflict in the UK, Australia, Israel, Bulgaria, Japan and Nigeria. On the latter, we’ve funded efforts against the destabilization flowing from the Sahel and Maghreb, regions where violent extremists, Russian mercenaries and Chinese debt traps risk producing a continent of chaos. The West may not concentrate (positively) on Africa today, but some groups in these regions concentrate (negatively) on the West. Lux companies are involved in providing technologies—hardware and software, submarines and satellites, drones and counter-drones, AI and surveillance—to allied warfighters and defenders of liberal democracies across the globe.

Consider China. While we in the West debate whether AI is a bubble, China signed 550 Belt and Road deals worth over $130 billion, up 75% in just a year. The cumulative tab since 2013 is over $1.4 trillion. What Beijing is building isn’t merely infrastructure, but a parallel operating system for the global economy. China ships systems to Pakistan, Russia and Iran, not because they share Beijing’s worldview, but because they undermine ours. China doesn’t need to pick sides when they control the parts bin. Their components sit inside drones on both sides of the battlefield. When Beijing decides to close the supply spigot in one direction, we won’t see it fire a shot. Instead, they’ll stop shipping motors. A motivation for us to find and fund American and allied alternatives.

Israel, described in the U.S. National Defense Strategy as a “model ally,” has pushed its defense budget to 8.3% of GDP, doubling in the last two years. Its defense manufacturers are globally sought and the government will now sell public stakes of its major primes. IAI's net income rose 55%, Rafael's 64%, and the country’s defense exports hit a record of nearly $15 billion. Israeli defense startups attracted over $1 billion in financing and M&A in 2025—more than all previous years combined. Lux has invested over $100 million in several companies (some still in stealth) covering air, land, subterranean, long-range power projection and cyber. We believe these companies—founded by what may be Israel’s greatest generation, forged in fire and existential fight—will emerge as the next major defense primes whose technologies will be sought by global allies.
On Iran, the entire world watches the standoff—negotiations serving as either pretext or prelude to kinetic military action. American warships sit off the Persian Gulf. Sanctions have crashed the rial to 1.6 million per dollar. The IRGC, which possesses the largest ballistic missile arsenal in the Middle East, is fracturing under pressure. All tools are being deployed at once: economic strangulation, soft-power rhetoric to embolden protesters, and military hardware “locked and loaded.” We have long and vocally held the view that ending this regime—its creation of and support for global terror proxies, its oppression of tens of millions of brilliant Iranian women and men who yearn to be free—may yield the greatest peace dividend the world has seen since the end of World War II. The known risks are real: roughly 20% of the world's oil flows daily through the Strait of Hormuz, a 21-mile-wide chokepoint Iran can squeeze and sabotage. Our own Saildrone provides the U.S. Navy and its allies with a fleet of solar-powered, persistent and low-cost ships that conduct intelligence, surveillance and reconnaissance in those waters. The known upsides are also real: a peaceful Middle East and the roll up of anti-American and anti-democratic networks throughout the world.

In short, this is matter that matters concentrated in the hotspots where the hinge of history hangs in the balance. We are proud to fulsomely back this new arsenal of democracy.

On the Concentration of Wealth
The link between Ayatollah Khamenei and American kitchen tables is direct: the Strait of Hormuz connects to the price of oil, which connects to the politics of affordability—the single word that dominates messaging from last year’s NYC mayoral race to this year’s midterms. Higher oil prices mean less affordability and more political finger-pointing. And that finger already points to the final concentration we will discuss: the concentration of wealth.

The discontent around wealth accumulation and the perceived failures of capitalism—with no distinction made for the crony kind—is only matched by the gauzy promise of socialism peddled by TikTok influencers for those not knowing and thus condemned to repeat its history. This perception divide creates a social risk that may be the next great movement diluting American power and growth in the decade ahead.

What decides and defines a movement is the concentration of attention. Where are people concentrating? Every few years, a new outrage arises and captures the zeitgeist, arriving like a cultural cloud raining down with a thunder clap of legitimacy before building to a torrent of too-much. After a few years, fatigue sets in, indifference gusts through and attention dilutes. The old cause becomes too hard to sustain while the new one becomes too hard to resist.

The sequence has run: trade, housing, guns, gender, race and religion. What comes next is class. Each prior movement—Occupy, Parkland, #MeToo and BLM—built up infrastructure but ultimately hit a ceiling. The identity axes that organized them produced passion but imposed limits; coalitions fractured along their own seams. Class is different. Rich versus poor does not divide households; it unites them against a common target. It recruits across racial lines. It reframes every prior grievance—police violence, immigration, healthcare and housing—as symptoms of a single disease: capital's capture of the state.

Mamdani's New York victory was won on “affordability” and rent control, not identity. Much of the enthusiasm is organic. Much may also be fomented by foreign foes skilled at finding fissures in the American fabric, as clever in the tearing as we are naive in being torn. That’s most frustrating in a year—our 250th—when our States should be more United than ever.

Confident Concentration
Let’s return to a clarifying reality: concentration is not destructive, and dilution is a delusion. The technologies that made us more productive and prosperous—the electric lights, radios, telephones, televisions, personal computers, cars, planes, refrigerators and microwaves—were all once unaffordable to even the wealthiest and most powerful for a simple reason: they didn’t exist. Once they did, capitalist competition drove prices lower, quality higher and access wider. Things begin not unaffordable but impossible, become a luxury, then a commodity, then an inevitable background assumption of modern life.

AI follows this arc, as do the emerging frontiers we are funding where physical-world applications meet computational scale. They will displace jobs while creating capabilities that never existed at any price: automated scientific discovery, accelerated drug development, specialized always-on tutors teaching future generations that eclipse even the elite pedagogy of aristocrats past. The "concentration" that critics lament is more rightly understood as what creation looks like at the moment of inception: an idea in one mind, then ten, then ten million. Our job at Lux is finding and funding those minds first. Fiat Lux.