On Friday, Bill Pulte (Director of the Federal Housing Finance Agency and chairman of both Fannie Mae and Freddie Mac) announced that America’s two government sponsored housing finance giants are exploring equity stakes in technology companies.

Speaking at the ResiDay conference in New York, Pulte described potential partnerships where tech firms would offer Fannie and Freddie equity positions, explicitly citing the Trump administration’s recent investment in Intel as a model.

I’m reminded of the American baseball legend Yogi Berra’s famous quip: “It’s like déjà vu all over again.”

What we’re witnessing here is the American government (through Fannie Mae and Freddie Mac, state owned mortgage institutions) once again heading down a path that bears a disturbing resemblance to the past.

The 2008 playbook

Leading up to the 2008 financial crisis, Fannie and Freddie played a central, though unfortunate, role.

These two institutions were designed to make home loans cheaper and more accessible by purchasing loans from banks and reselling them as securities to investors.

However, to keep pace with private competitors, they began taking on increasingly risky mortgages.

When housing prices fell, the losses were enormous, and in 2008 the American government had to take control to prevent a total collapse in the housing and financial sectors.

And now we see Bill Pulte (Trump’s handpicked chief of the powerful regulatory body, the Federal Housing Finance Agency) openly discussing how these state backed mortgage institutions should invest money in American technology companies.

Pulte stated that he views Fannie and Freddie somewhat differently because they operate as actual businesses, albeit private ones. He indicated that the GSEs will likely take ownership stakes in various companies as those firms offer equity in exchange for business partnerships with Fannie and Freddie.

More tellingly, Pulte acknowledged the coercive nature of these arrangements, explaining that major technology and public companies are offering equity to Fannie and Freddie in exchange for business partnerships. He noted that the GSEs are considering taking these equity stakes because of the substantial power Fannie and Freddie wield over the entire housing finance ecosystem.

Crony capitalism comes to Silicon Valley

I must be frank: this is beginning to stink.

Over the past week, we’ve received a series of signals that the American government now intends to directly support American technology companies through precisely the kind of arrangements that characterise crony capitalism, where political connections and government favour, rather than market competition, determine winners and losers.

On Monday, the Wall Street Journal reported that OpenAI’s CFO, Sarah Friar, hinted that government guarantees could cut AI funding costs – later backtracking to say they weren’t seeking a bailout. But the message was clear: tech giants now assume taxpayer backing, just like banks did before 2008.

So if you’re asking yourself why American tech giants (particularly the so-called “Magnificent Seven”) are trading at the extraordinary valuations we see today, you likely have part of the explanation here: investors aren’t just pricing in technological progress or profit growth. They’re pricing in implicit government guarantees.

They’re pricing in the expectation that the Trump administration will support these companies, that losses will be socialised whilst gains remain private, that these firms have effectively become “too big to fail.”

This is moral hazard on a grand scale, but now applied to an entire sector rather than just individual institutions. The market isn’t valuing these companies based on discounted future cash flows. It’s valuing them based on the anticipated probability of state intervention to prevent failure.

And the connections between political donations and government favour are becoming impossible to ignore.

Elon Musk, who contributed hundreds of millions of dollars to Trump’s campaign (making him the largest individual political donor in the 2024 election cycle), now wields extraordinary influence over government policy.

Marc Andreessen and Ben Horowitz, whose venture capital firm has invested heavily in OpenAI, together donated over $5 million to Trump aligned political action committees.

Peter Thiel, the co-founder of Palantir, spent $15 million to elect JD Vance to the Senate in 2022. Vance is now vice president.

These aren’t coincidences. They’re investments with expected returns.

But to my mind, this is also a very clear signal that shouldn’t be ignored: all is not as it should be with the tech giants. Over the past few weeks, we’ve seen turbulence surrounding share prices in the sector, but perhaps more worryingly, also rising tensions in credit and money markets.

It’s simply about nervousness that some AI companies potentially face liquidity problems, and ultimately whether this might infect the banking sector.

The market is already showing cracks

And these aren’t just theoretical concerns. Over the past fortnight, we’ve seen concrete signs of stress in American credit and money markets.

The Federal Reserve has been forced to inject substantial liquidity into the system to prevent tensions in the repo market from escalating.

On 31 October, the Fed injected $50.35 billion into the system (the largest single day operation since 2021), followed by an additional $22 billion on 3 November.

The Secured Overnight Financing Rate has shown unusual volatility, with repo rates spiking to their highest levels relative to the Fed funds rate since 2020, precisely the kind of pressure that typically precedes broader credit events.

Major Wall Street banks, including JPMorgan and Deutsche Bank, have warned that money market stress could flare up again.

Dallas Fed President Lorie Logan has suggested the central bank might need to begin purchasing assets if the rise in rates proves more than temporary, echoing the kind of emergency measures deployed during previous crises.

More tellingly, major international banks are taking defensive positions. Deutsche Bank, which has extended billions in loans to data centre firms powering the AI boom, is now reportedly exploring hedging strategies including shorting baskets of AI related stocks and purchasing credit protection to offset potential losses if the current pricing regime collapses.

When one of Europe’s largest banks starts hedging against its own AI lending book, that tells you something important about where sophisticated risk managers think we are in the cycle.

JPMorgan’s Jamie Dimon warned in October about a probable market decline of 10 to 20% within the next 6 to 24 months, citing concerns about overheating in the technology sector that could trigger a correction similar to the dotcom crash of 2000.

Goldman Sachs’ David Solomon echoed these concerns this week, stating that a 10 to 20% drawdown in equity markets within two years appears likely due to AI related risks and trade tensions.

Even the Bank of England’s Andrew Bailey has warned of “growing risk of a sharp correction” if AI expectations falter, with “alarm bells ringing” over private credit and AI concentration in major indices.

Regional American banks have already begun reporting losses on loans to distressed investment funds, and credit default swap spreads on major banks have risen to elevated levels.

And I’m hardly the only one concerned. I’m completely convinced that credit and risk management departments in all the major global banks are worried about precisely this right now. The parallels to 2007 (when credit markets began showing stress months before the broader crisis became apparent) are uncomfortably clear.

The Intel precedent: when subsidies become equity

To understand what Pulte is proposing, we must first examine the model he explicitly referenced.

In August 2025, the Trump administration took a 9.9% equity stake in Intel Corporation worth $8.9 billion, converting previously awarded CHIPS Act grants and Defence Department funds into common stock ownership. The government purchased 433.3 million shares at $20.47 per share.

Commerce Secretary Howard Lutnick justified this approach by arguing that the government should receive equity stakes in exchange for the funding that had already been committed under the previous administration.

But Intel was only the beginning.

The Trump administration has also taken equity stakes in MP Materials (rare earth mining, 15%), Lithium Americas (lithium production, 5-10%), and Trilogy Metals (copper zinc mining, 10%). Reports emerged in October that the administration was exploring similar arrangements with quantum computing firms including IonQ, Rigetti Computing, and D-Wave Quantum, though the Commerce Department subsequently denied “currently negotiating” such stakes (language that leaves ample room for future deals).

This represents an extraordinary shift in American economic policy and a troubling embrace of what can only be described as crony capitalism.

Unlike previous government equity stakes during financial crises (TARP in 2008 or airline support during COVID-19), the Trump administration has taken these positions without any financial emergency.

The ideological precedent is clear: converting government grants and subsidies into equity ownership across strategically important industries, creating an intertwined relationship between state power and private profit that undermines market discipline.

Now Pulte wants to extend this model to Fannie and Freddie. According to Pulte’s statements from May, Fannie Mae has approximately $4.3 trillion on its balance sheet, whilst Freddie Mac holds over $3 trillion.

The scale of what’s at stake

Fannie Mae and Freddie Mac don’t just participate in America’s housing market; they effectively are the housing market. They guarantee roughly half of all outstanding U.S. residential mortgages (the largest share of the approximately 70% of American home loans that receive some form of federal backing, including FHA, VA, and other programmes).

When institutions of this size and systemic importance start deviating from their core mission, the consequences ripple through the entire financial system.

My friend and Richmond Fed veteran economist Bob Hetzel wrote in his seminal 2009 analysis of the financial crisis about how financial safety nets inevitably create moral hazard by increasing incentives for risk-taking.

The critical mechanism he identified: financial institutions receive implicit subsidies from safety nets that grow larger as their portfolios become riskier, as they increase leverage, and as their capital buffers decline.

2008: when mission drift became catastrophic

The last time Fannie and Freddie strayed from their mandate, it ended catastrophically. In September 2008, both institutions collapsed under the weight of massive losses and required a government takeover that has lasted 17 years. At the time of conservatorship, they held or guaranteed about $5.2 trillion of home mortgage debt.

The scale of the losses was staggering. About 80% of Fannie and Freddie’s combined $213 billion in credit losses between 2008 and 2011 involved mortgages that were either Alt-A, interest only, or both. These were loans made to borrowers with relatively high credit scores but featuring riskier structural characteristics. The critical error wasn’t the specific loan types. It was the strategic decision to chase market share by expanding into riskier segments well outside their traditional remit.

Starting in 2006 and 2007, just as the housing market reached its peak, Fannie and Freddie increased their leverage and began investing heavily in subprime securities and Alt-A loans in an ill-fated effort to win back market share from private competitors. This is textbook mission drift: institutions designed for one purpose (providing liquidity to mortgage markets) taking on unrelated risks with predictably disastrous results.

Hetzel understood the fundamental problem.

Writing specifically about the GSEs, he noted that understanding the subprime crisis required grasping how Fannie and Freddie had increased demand for housing stock, pushed homeownership rates to unsustainable levels, and thereby contributed to sharp rises in housing prices given the relatively inelastic supply of housing due to land constraints.

Perhaps most damningly, Treasury Secretary Timothy Geithner told the Financial Crisis Inquiry Commission in a private interview that moral hazard was pervasive throughout the system, with the GSEs representing the single largest source of this problem.

The new mission drift: from mortgages to tech equity

Now we’re watching a remarkably similar pattern emerge, only this time the target isn’t housing. It’s technology.

Pulte indicated at the conference that “one of many companies” seeking equity arrangements with Fannie and Freddie is a firm whose involvement would leave observers “blown away with how much money is involved,” though he declined to name it.

Fannie Mae has already signed a partnership agreement with Palantir for fraud detection efforts, though financial terms weren’t disclosed (precisely the kind of opacity that should alarm anyone concerned with accountability).

The Palantir connection is particularly revealing. Since Trump took office, Palantir has secured large federal government contracts. Peter Thiel, Palantir’s co-founder, spent $15 million electing JD Vance to the Senate.

Vance is now vice president. Joe Lonsdale, another Palantir co-founder, contributed $1 million to Elon Musk’s America PAC supporting Trump. This isn’t a market economy. This is a system where government contracts and partnerships flow to companies whose founders financed the campaigns of those now in power.

The proposed model is seductive in its simplicity: tech companies offer Fannie and Freddie equity stakes in exchange for access to the GSEs’ enormous housing finance ecosystem.

But consider Pulte’s own reasoning: the GSEs are considering taking equity stakes in companies specifically because of the substantial power Fannie and Freddie exercise over the entire ecosystem.

This isn’t market allocation. This is using control over critical infrastructure to extract equity positions from private companies. This is crony capitalism at its most transparent.

Moral hazard at scale: the implicit guarantee premium

The fundamental problem here isn’t just about Fannie and Freddie taking equity stakes in a few tech companies.

It’s about what those stakes signal to the broader market about the existence and scope of implicit government guarantees.

Hetzel identified the core mechanism with precision: financial safety nets (including deposit insurance, too-big-to-fail protections, Federal Home Loan Banks, and the Fed’s discount window) allow banks to access funding at costs that don’t rise with the riskiness of their portfolios.

The same logic applies when government backed entities like Fannie and Freddie take equity positions in private companies.

That government guarantee (now explicit after 17 years of conservatorship) means US taxpayers ultimately backstop losses whilst any gains accrue to whom exactly? The Treasury? Tech companies?

This creates a fiscal time bomb where downside risk is socialised whilst upside is privatised.

But the effects extend far beyond the specific companies receiving these investments.

When the market observes the government taking equity stakes in Intel, exploring partnerships with OpenAI, and now planning to inject Fannie and Freddie capital into technology firms, investors rationally update their beliefs about which companies enjoy implicit state backing.

The extraordinary valuations we observe across the Magnificent Seven and related AI companies aren’t just about technological optimism. They reflect the market pricing in an implicit guarantee that these firms are “too big to fail.”

This is moral hazard pricing on a sectoral scale. And it raises a troubling question: if these companies are indeed “too big to fail,” are they also becoming “too big to save”?

Too big to fail or too big to save?

Consider the arithmetic. The combined market capitalisation of the Magnificent Seven alone now exceeds $20 trillion. Add in the broader ecosystem of AI related companies, data centre operators, and semiconductor firms, and you’re looking at market value that’s at least partially predicated on the assumption of government support.

Now place that against America’s fiscal position. U.S. national debt stands at $38 trillion (more than 100% of GDP). Interest payments reached $841 billion in just the first ten months of fiscal year 2025, already exceeding Medicaid.

The Congressional Budget Office projects debt will reach 156% of GDP by 2055, with interest payments hitting $1.8 trillion annually by 2035.

Here’s the uncomfortable arithmetic: when the next crisis comes (and it will come), can the American government actually afford to bail out a technology sector whose market capitalisation approaches half the entire national debt? The fiscal buffer that existed in 2007, problematic as it was, has completely evaporated. We may be creating a class of companies that markets believe are too big to fail, but which the American government is quite possibly too indebted to save.

And this is before considering the international dimension.

When Fannie and Freddie (institutions with explicit government backing) take equity positions in tech companies, it sends a signal to foreign holders of U.S. Treasury securities that America is extending its contingent liabilities even further into speculative territory.

For a Chinese central bank holding a trillion dollars in Treasuries, or a Japanese or Scandinavian pension fund with massive exposure to American debt, this doesn’t look like prudent fiscal management.

It looks like the American government is systematically increasing the risk that it will face multiple, simultaneous calls on its financial resources that it cannot meet without inflating away its debts or defaulting.

The conservatorship paradox and regulatory capture

Pulte confirmed on Friday that Fannie and Freddie will remain in government conservatorship whilst potentially conducting an IPO of up to 5% of their shares this quarter or early next year.

He indicated that he anticipates the president will make a decision on the IPO timing either this quarter or in early 2026.

Think carefully about what this means: these institutions remain under explicit government control because they’re deemed too important and too risky to operate independently in housing markets, yet they’re simultaneously being encouraged to speculate in technology equity markets.

The conflicts of interest are staggering. Pulte runs the agency that regulates Fannie and Freddie whilst simultaneously chairing both companies.

If those companies become equity investors in major tech firms, he’ll effectively be regulating entities in which his institutions have direct financial stakes.

This is regulatory capture taken to its logical extreme: the regulator, the regulated entities, and the companies receiving investment all bound together in a web of mutual dependency that eliminates any possibility of arms length oversight or genuine market discipline.

If Fannie and Freddie aren’t trustworthy enough to exit conservatorship after 17 years of profitability in their core business, what possible justification exists for expanding their mandate into tech investment?

What should happen instead

The solution is straightforward but politically difficult: complete the mission Fannie and Freddie were designed for, then either privatise them fully or wind them down entirely.

If conservatorship is necessary because these institutions require close government supervision, keep them focused exclusively on their housing mandate with strict limits on portfolio composition and leverage.

If they’re healthy enough to take tech equity positions, they’re healthy enough to exit conservatorship and face genuine market discipline.

What should not be accepted is this worst of all worlds hybrid: government guaranteed institutions with neither proper oversight nor proper market discipline, now venturing into speculative investments far beyond their expertise or mandate, at a time when the U.S. government’s own fiscal position is already unsustainable, all whilst creating a sectoral moral hazard problem that may prove impossible to resolve when the inevitable repricing occurs.

Hetzel proposed a radical but coherent alternative: eliminating the Fed’s legal authority to make discount window loans, suggesting instead that the central bank should flood markets with liquidity during panics through open market operations whilst maintaining its policy rate through interest on reserves.

The core principle: creditors and debtors will restrain financial system risk-taking only if they face genuine losses when financial institutions fail.

Every expansion of the safety net (whether through TBTF, discount window lending, government equity stakes in strategic companies, or now equity investments linking government backed housing finance to private tech firms) undermines this crucial disciplining mechanism.

Every implicit guarantee extended, every equity stake taken, every suggestion that politically connected companies will receive state support, moves us further from a market economy and deeper into crony capitalism where success depends not on serving customers but on securing government favour.

Conclusion: echoes from 2008

Seventeen years after Fannie and Freddie’s collapse nearly took down the global financial system, the same structural errors are being repeated, but this time with different assets, weaker fiscal foundations, and potentially far graver consequences.

The 2008 crisis taught us that Fannie and Freddie represent “entirely moral hazard” when they deviate from their core mission, as Geithner observed. Fannie and Freddie needed a $191 billion taxpayer bailout because they deviated from their core mission, driven by the same toxic combination of implicit government guarantees, inadequate oversight, and mission drift that is re-emerging today.

Now, under political pressure to “do something” about housing affordability and tech competitiveness, they’re being pushed down the same path again, only this time explicitly following the Trump administration’s model of crony capitalist equity stakes, at a time when U.S. federal debt stands at 100% of GDP, interest payments exceed defence spending, and the fiscal buffer to absorb another systemic crisis no longer exists.

But this time there’s an additional, more troubling dimension. This isn’t just recreating the moral hazard of 2008. This is creating it at sectoral scale across technology companies whose combined market capitalisation may literally be too large for the American government to backstop.

The stress in credit markets, the Federal Reserve’s emergency liquidity injections totalling over $70 billion in early November, the defensive hedging by major banks like Deutsche, the warnings from JPMorgan, Goldman Sachs, and the Bank of England—these aren’t abstract concerns. They’re happening right now, and they’re happening for a reason.

The market is beginning to ask the question that should terrify policymakers: what happens when companies that everyone believes are too big to fail turn out to be too big to save?

The next crisis won’t announce itself with sirens and flashing lights. It will begin, as the last one did, with seemingly reasonable people making seemingly reasonable arguments about expanding mandates, capturing growth opportunities, and using “power over the whole ecosystem” for strategic purposes. It will involve the gradual extension of implicit guarantees until the market prices them in as explicit. And this time it will happen in a context where U.S. federal debt is at historically high levels, the fiscal capacity to absorb losses has evaporated, and the companies that need saving may be orders of magnitude too large for the government’s balance sheet.

The lesson from Washington in 2008 was clear. Apparently, it needs to be learned once more, and this time, the tuition fees may be higher than ever, both for the financial system and for U.S. sovereign creditworthiness itself.

The only question is whether this administration is creating companies that are too big to fail, or too big to save.

I fear the answer will reveal itself soon enough.

Lars Christensen
LC@paice.io
+45 52 50 25 06

Related