Red Flags with Project Vault

Yesterday, I argued that the Cramer-Warner proposal to reauthorize the Export-Import Bank for a decade and raise its lending cap by $70 billion fails on its own terms. The Bank already sits on $100 billion in unused capacity. Its last major strategic mandate, the China and Transformational Exports Program, spent only 22 cents of every dollar Congress allocated to it. And the institution’s defining feature, across nine decades, is not strategic vision but the reliable funneling of benefits to a small group of politically connected corporations.

But what about Project Vault? As some see it, it is a $12 billion critical-minerals stockpile, $10 billion of it backed by the Ex-Im Bank. Many build their hope on the belief that the Bank has reinvented itself as a serious instrument of industrial policy. Meanwhile, other people, who are maybe not as familiar with Ex-Im, support the idea because they like the end goal of creating a rare-earth stockpile to minimize the impact of China using its dominance of rare-earth processing as leverage against the U.S.

I don’t agree. Project Vault doesn’t redeem Ex-Im’s poor record. It extends it, at the highest stakes the Bank has ever taken on, and for the first time, without an Inspector General watching.

Further, project Vault perfect illustrates the wrong thinking exhibited by many scholars who conclude that just because something is desirable, it should be done by the government – in this case through Ex-Im.

Red Flags All Over Project Vault

Project Vault is a government-backed commodity stockpile covering all 60 minerals on the U.S. Geological Survey’s critical-minerals list for private manufacturers. The federal government, through a $10 billion Ex-Im loan and roughly $2 billion in private capital, will purchase and store critical minerals in facilities across the United States. The $10 billion is a direct loan, not a guarantee, not insurance. Further, the loan is for 15 years, more than double the length of the bank’s previous largest deal.

Three commodity trading firms (Hartree Partners, Mercuria, and Traxys) will procure the minerals. Large manufacturers (Boeing, GM, GE Vernova, Google, and others) pay commitment fees and lock in a fixed purchase price, giving them the right to draw from the stockpile during supply disruptions. If they draw down materials during normal times, they must replenish them. If a major disruption hits, they can withdraw their full allotment at once.

I would like to flag several issues with this project.

New project, Old Patterns and Same Cronyism:

The architecture of Project Vault will be familiar to anyone who has studied agricultural price supports: the government finances and stores a commodity, private participants get guaranteed access at stable prices, and the taxpayer absorbs the losses when markets move the wrong way. Expect the same here.

Farm programs that began as emergency measures in the 1930s are still with us nearly a century later and form the backbone of one of Washington’s most entrenched lobbying complexes. Project Vault is being sold as a temporary response to a supply-chain crisis. So too were farm subsidies.

Those who follow Ex-Im closely will also recognize several of these names as perennial beneficiaries of the Bank’s financing, including Boeing (it’s as if the Bank doesn’t do anything if it doesn’t benefit Boeing). The cast of characters at the trough has barely changed. The only thing that’s changed is the justification on the placard.

Worth noting: Ex-Im Chairman John Jovanovic was previously an Investment Director at Mercuria Energy Group, where he managed investment and business building across North and South America. His former employer is now one of the direct beneficiaries of the largest loan his agency has ever issued. This doesn’t have to be an issue, but we may never know if it is as the person whose job it was to flag exactly this kind of concern if it arises is gone. In October 2025, Trump removed Ex-Im’s Senate-confirmed Inspector General, Parisa Salehi, who at the time was overseeing 15 open investigations into possible violations of federal law. Project Vault was announced weeks later. The Bank is now operating without independent oversight at the precise moment when it is taking on the largest and most complex commitment in its history.

The SPR Analogy Cuts the Wrong Way

Many advocates are comparing Project Vault to the Strategic Petroleum Reserve (SPR). It is meant to confer legitimacy by association. But it deserves closer examination because the actual history of the SPR does not flatter the analogy.

The SPR was created in 1975 in response to the 1973 Arab oil embargo, with a clear and narrow purpose: to buffer the United States against severe supply interruptions in a commodity on which the entire economy depended. It was a government asset, controlled by the President, releasable only upon a finding of severe energy-supply disruption, and managed by the Department of Energy since 1977. With no role for private commercial participants. Whatever its flaws, the architecture was coherent: a public reserve for a public emergency.

What actually happened to the SPR over the following decades is a preview of what happens when a national security asset gets anywhere near a congressional appropriations process. In 1996, Congress authorized the sale of roughly 28 million barrels, partly to cover the cost of decommissioning a damaged storage site, partly to reduce the federal deficit. Then, starting in 2015, it happened again on a far larger scale: through a series of budget acts Congress embedded SPR sales directly into spending legislation as offsets, mandating the sale of well over 100 million barrels to fund programs that had nothing to do with energy security. The reserve had become a piggy bank with a “national security” label on the outside.

Then came the most dramatic chapter. Between late 2021 and late 2022, the Biden administration released roughly 217 million barrels. That represents the largest withdrawal in the Reserve’s history. Some of this release was defensible, a response to the disruption caused by Russia’s invasion of Ukraine. But much of it is hard to defend. Critics argue that the release timing was calibrated at least in part to decrease retail gasoline prices ahead of the 2022 midterm elections. By mid-2023, the SPR had fallen to its lowest level since 1983, under 350 million barrels, less than half its capacity. The drawdowns also caused structural damage to the salt cavern facilities, which will require an estimated $100 million in repairs, while refilling to near-full capacity is projected to cost around $20 billion. A reserve built to protect against emergencies had been spent down to manage political ones.

Proponents of the SPR will argue that the lesson here is better governance, not that the concept is flawed. Perhaps. But that argument requires you to believe that Project Vault, managed by an institution without an Inspector General, structured to serve commercial participants, will somehow resist the same political and fiscal pressures that hollowed out a much simpler program.

National Security Excuse vs The Market Process

None of this has received adequate scrutiny, in part because the national security framing makes it politically radioactive to raise serious questions about it. Who wants to be the person arguing against securing America’s mineral supply? But that framing is exactly what makes the risk so large. An assertion of national security is the one justification that short-circuits every normal check on government action.

The United States already maintains a National Defense Stockpile, managed by the Pentagon, for materials deemed essential for military readiness. If the concern is fighter jets and guided missiles, the mechanism already exists. Expanding it might be wise, but that is a different question. Project Vault is something different. It is a physical mineral stockpile for civilian use. Its participants are commercial manufacturers, not defense contractors fulfilling military procurement orders. The national security argument is being stretched here to cover an entirely different claim: that any change in the price of minerals, or any supply shock affecting civilian industry, is itself a national security emergency.

We are told this is justified because China dominates the processing of rare earths, and that if Beijing decided to restrict access, as it has in the past, we would face shortages and sky-high prices. It is a serious-sounding argument. But the private-sector case for a government-backed stockpile is weaker than it appears.

First, rare earths aren’t rare at all. Tim Worstall, a rare-earths specialist and economist, reminds us that “rare earths are neither rare nor earths, and they are nearly everywhere.” He also notes that “The biggest restriction on being able to process them is the light radioactivity the easiest ores (so easy they are a waste product of other industrial processes — monazite say) contain. If we had rational and sensible rules about light radioactivity — alas, we don’t — then that end of the process would already be done.” That explains the China dominance in processing.

However, he argues, the dominance in rare-earth processing is, in economists’ terms, a contestable monopoly: stable only as long as it is not weaponized. Beijing is now weaponizing it with its licensing scheme. Under Beijing’s current rules, every buyer of Chinese rare-earths must disclose entire production chains and obtain approval from China’s foreign trade ministry for any change in supplier, volume, or end use, all the way down to products for which rare earths constitute as little as 0.1 percent of the final value.

Worstall’s point is that this is an inherently self-defeating strategy. Western industry is simply not going to accept permanent dependence on a Chinese ministry’s license, especially since rare earths are everywhere. Non-Chinese materials will command a premium precisely because they come without the bureaucratic leash, and that premium is all the incentive the market needs to build alternatives. This isn’t just theoretical. That’s what happened in 2010, the last time we had such a rare earth scare.

In other words, China’s own export-licensing regime is doing much more to break its processing dominance than any government program could. As Worstall puts it, the desire of end users to be free of China’s system “doesn’t need any guidance from other governments to make it happen. Nor financial support nor investments nor buying rare earth companies and the rest.” The contesting has begun.

If defense needs are already covered by the existing Pentagon stockpile, and commercial diversification is already being driven by the market incentives that China itself has created, the question becomes unavoidable: What exactly is that stockpiling for? What exactly is the $10 billion for? The answer could very well be as simple as the creation of a private sector’s commodity risk management underwritten at public expense.

Not Everything Requires the Government

That leads me to my final point. Project Vault is genuinely appealing, not just to the Boeing lobbyists and the connected trading firms, but to serious, intellectually rigorous people who have spent years studying supply-chain vulnerability and have no particular interest in corporate welfare.

The appeal rests on a chain of reasoning that feels airtight at each step. China dominates rare-earth processing. That dominance is a real vulnerability. A stockpile would buffer against disruption. Therefore, the government should build one. The logic seems to flow without flaw. But that’s a misimpression. This ‘logic’ contains three separate breaks, each worth naming, because conflating them is how well-intentioned analysis produces bad policy.

The first is the leap from a problem exists to government should solve it. This is perhaps the most common error in policy thinking, and it is not limited to one ideological tradition. Scholars who would never dream of calling themselves interventionists nonetheless tend, when they’ve identified a genuine market vulnerability, to move almost automatically to government remedy.

The identification of a problem is intellectually satisfying. It generates papers, testimony, task forces, and eventually legislation. What gets far less attention is the prior question: Is this actually a problem that government intervention will improve, or is it one that government intervention will redirect, distort, and ultimately make more durable? A problem that markets are imperfectly solving is not the same as one that requires a federal program. The bar for intervention should be that it is the government’s role and that it can do better, not merely that a gap exists.

The second break is the move from a private stockpile would be beneficial to therefore a government stockpile is justified. These are not the same proposition. If the logic is that manufacturers would benefit from having stable access to critical mineral inventories, the relevant question is why they aren’t building that inventory themselves. If the answer is that the economics don’t support it at current risk levels, that is important information. It means that the market is pricing the risk differently than the hawks are. A government program that overrides that pricing doesn’t resolve the underlying disagreement about risk; it just makes taxpayers bear the cost of one side’s assumptions. If, on the other hand, the economics do support a private stockpile and companies simply haven’t gotten around to building one, then what’s needed is not a government program. What’s needed is for someone to point this out, which the current threat environment is doing rather effectively.

The third break concerns this specific institution. Even granting everything — that the problem is severe, that government involvement is warranted, that a government-engineered stockpile is the right form of intervention — none of that implies that the Export-Import Bank is the right vehicle. An institution’s fitness for a task is a separate question from whether the task is worth doing. Ex-Im brings to Project Vault a ninety-year pattern of benefit concentration, a demonstrated inability to deploy capital toward strategic mandates, a chairman whose former employer is a direct program beneficiary, and, as of October 2025, no Inspector General. The vessel matters. Pouring a sound policy idea into a compromised institution does not produce sound policy. It produces the policy’s vocabulary attached to the institution’s incentives, which is, in practice, corporate welfare with a supply-chain press release.

Taking all these points under consideration in addition to the fact that whatever problems we have due to China dominance might already be in the process of solving itself should make us all skeptical of Project Vault.

Conclusion

Reasonable people can disagree about the severity of China’s rare-earth dominance and the speed at which markets will correct it. But you don’t need to resolve that debate to see that Project Vault is the wrong answer. Even if the threat is exactly as dire as the hawks claim, the response should look nothing like what is being proposed. It should fix the permitting system that makes the United States the second-slowest country on earth to open a mine. It should rationalize the radioactivity rules that lock away the most accessible rare-earth feedstocks behind irrational regulatory barriers. It should establish a dedicated federal mining office, as Canada and Australia have done, so that companies face a coordinated approval process rather than a decade-long obstacle course. It should feature narrowly targeted defense procurement under the Defense Production Act for the handful of materials where military readiness genuinely cannot wait.

None of that requires a $70 billion increase in Ex-Im’s lending cap. None of it requires a ten-year reauthorization designed primarily to insulate an institution from accountability. And none of it requires routing public money to stockpile minerals that still need to be processed through a bank that has spent nine decades proving it will spend that money on the connected few.

Project Vault is being sold as a solution to a supply-chain crisis. What it actually is, is a test of whether the national security frame has become powerful enough to make Congress stop asking questions. So far, the answer looks like yes.