The arithmetic is what stopped me.
Six point four kilometers of tunnel. Four stations. Roughly four million cumulative passengers since the first Loop segment opened beneath Las Vegas. And, according to the disclosure that crossed my desk this week, a valuation of twenty-three billion dollars.
That last figure is a four-fold increase over the $5.675 billion mark The Boring Company carried in April 2022. Across the four years in between, the company closed exactly one announced round — a $675 million Series C — and published no audited revenue, no EBITDA, no cash-flow statement, and no per-kilometer cost figure for its tunneling platform. The re-rating did not come from operations. It came from somewhere else, and locating that somewhere is the only part of this story that matters to anyone who reads a crypto publication.
Because that is how the story reached most of you: repackaged, tagged with Web3 keywords, and placed beside a sovereign wealth fund's spot Bitcoin ETF position. Two facts, one headline, and a conclusion that does not follow from either.
Truth over hype. Always. So let us begin with what the disclosure actually says — and, more importantly, with what it conspicuously leaves out.
The Company, the Round, and the Tag That Was Sewn On
The Boring Company digs tunnels. It has never claimed to be anything else. Its technical centerpiece is a boring machine platform called Prufrock, which the company describes as capable of launching and recovering without a surface shaft and without a crane. Its commercial centerpiece is the Vegas Loop, an underground dedicated roadway in which Tesla vehicles carry passengers between stations. By its own count it has twenty-five tunnels complete or under construction, fourteen of them in Las Vegas, and it holds a pilot contract with Dubai's Roads and Transport Authority.
The Series D was led by an Abu Dhabi-linked entity — the disclosure does not name the vehicle — and followed by a syndicate that would be strange to find behind a fraud: Human Capital, Vy Capital, Valor Equity Partners, Sequoia, a16z, Temasek, Shamal, and Baron. That list is genuinely Tier 1, and it is the strongest single piece of evidence that this is a real company doing real engineering in the physical world.
The crypto adjacency is thinner than the coverage implied. An Abu Dhabi sovereign entity holds spot Bitcoin ETF exposure — a fully regulated product, unremarkable in 2026, and increasingly the default institutional wrapper for digital-asset exposure. Separately, reporting elsewhere describes roughly two billion dollars of institutional support for Binance from a state-linked source the article declines to name. Neither fact involves The Boring Company. Neither fact was caused by The Boring Company. Both were placed in the same paragraph as a $23 billion tunnel valuation, and the proximity performed the rhetorical work that the evidence could not.
I have a specific allergy to that construction, and it was formed the hard way. When I built our publication's regulatory-literacy column in 2025 and worked through the EU's MiCA framework alongside counsel, the most frequent error I had to correct in readers was adjacency — the assumption that because two things are regulated within the same jurisdiction, they share a risk profile. They do not. A sovereign buying a Bitcoin ETF and a sovereign leading a tunnel round are two different balance-sheet decisions that happen to be signed by the same hand, for two different reasons, with two different time horizons.
So let me be precise, because precision is the entire job: this is not a crypto story that happens to involve tunnels. It is a capital-allocation story that crypto media borrowed without permission and returned with a new cover.
When a Document's Silences Carry the Argument
In 2017, I spent an entire season auditing token distributions rather than chasing allocations, which at the time made me the least interesting person in the newsroom. Two of the offerings I examined were EOS and Golem, both of them the most oversubscribed events of their moment. My editors wanted speed. I wanted to know why certain things were absent from the paperwork.
I found three structural vulnerabilities in those distributions, and none of them required reading a line of code. They required noticing what the whitepapers never quantified: how much supply sat with insiders, on what unlock schedule, and under whose unilateral discretion that schedule could be revised. Those documents were not lies. They were omissions arranged so that the omissions read as simplicity. I wrote them up, took some hostility for it, and eventually earned a grudging respect from colleagues who preferred a fast story to a correct one. That season is why every trend piece I have written since opens with a structural vulnerability before it discusses upside.
The TBC disclosure has the same texture as those whitepapers had. Here is what I could not find inside it.
No revenue, and no usable proxy for revenue. Passenger counts are not revenue. A free or subsidized Loop ride produces a number that looks like traction and functions as marketing.
No cost per kilometer of finished tunnel, which is the only metric that would validate the Prufrock thesis.
No advance rate — meters per day, against what geology, with what machine class — which is the only metric that would validate the schedule.
No cap table, no allocation across the Series D, no lockup terms, no liquidation preference, no disclosure of how the new round sits relative to the 2022 entry.
No independent confirmation of any kind. The disclosure cites no Form D, no audited financials, no data provider. Several of its most load-bearing figures carry no source attribution whatsoever.
And then there is the timestamp. The embedded post carrying the primary details is dated September 10, 2026 — a date I could not reconcile against any wire record, any filing, or any subsequent confirmation in the public record. I have spent twenty-five years reading disclosures for a living, and the first thing I do is check whether a document agrees with itself about when it exists. This one does not, at least not on any record I can access.
That does not make the company fake. It makes the document unusable for decision-making until somebody independent verifies it. Those are two very different claims, and conflating them is precisely how people get hurt in this industry — once in the direction of believing a fraud is real, and once in the direction of dismissing a real company because its press release was sloppy.
Prufrock Is the Only Asset With Technical Content, and It Is Unquantified
Strip away the branding and TBC possesses exactly one differentiated technical claim: that Prufrock can be launched and recovered without a surface shaft and without a crane.
If that is true, it is not a small thing. Surface launch shafts are among the largest cost and permitting burdens in conventional tunneling. Removing them collapses the fixed cost of starting a project, and fixed starting cost is the barrier that makes urban tunneling uneconomic at small scale. Nearly every serious tunnel project in the world begins with a hole in a city that did not want a hole in it, followed by a decade of litigation. A machine that eliminates that step changes the unit economics of the entire category.
But "if that is true" is carrying an enormous amount of weight in that sentence, and the disclosure gives me nothing to test it against. No meters per day. No cost per kilometer against a Herrenknecht-class baseline. No specification of the geological conditions under which the claim holds. No peer review, no third-party engineering assessment, no delivered-project data broken out by machine class.
This is the exact shape of a pattern I have audited across DeFi for years: the one number that would settle the argument is the number left out. When a lending protocol promises capital efficiency and omits its liquidation parameters, that is not an oversight. When a rollup promises throughput and omits its proving costs, that is not an oversight. When a tunneling company promises a cost collapse and omits cost, the omission is the finding. Not the fraud — the finding.
Noise filtered. Signal preserved. And the signal here is that the single most load-bearing technical claim inside a $23 billion valuation is currently unfalsifiable by anyone outside the company and its lead investor.
The Capacity Question Nobody Wants to Price
The Vegas Loop is routinely described as transit. It is more accurately described as an underground dedicated roadway carrying Teslas. That distinction is not pedantic. It determines what the asset can ever be worth.
A conventional metro line moves tens of thousands of passengers per hour per direction, on fixed headways, with no dependency on a third party's vehicle production. The Loop, at pilot scale, moves a fraction of that, with capacity bounded by how many vehicles Tesla chooses to make available and how many drivers — or, eventually, how much autonomy — can be mustered to move them. Six point four kilometers and four stations is a demonstration, not a network. Four million cumulative passengers over several years is a respectable activation number and a very poor utilization number once divided by hours of operation.
Then there is the phrase "150 kilometers of tunnel partnerships." Read against the actual operating footprint, that figure almost certainly describes approved, permitted, or intended mileage rather than completed bore. I have watched that linguistic move for a decade in token launches: total addressable market, written in the grammar of current capability, with the tense left deliberately ambiguous. Notice also that the stated use of proceeds includes hiring for production roles. If Prufrock were a mature platform in volume production, you would not need a capital raise to staff the line.
Dubai is the real test. A pilot with the RTA, in a jurisdiction whose sovereign capital is simultaneously the lead investor in the company, is the cleanest natural experiment available. It will either produce delivered kilometers or it will produce a very expensive press release, and either outcome will be legible within eighteen months.
The Valuation Arithmetic Does Not Originate in Operations
Here is the sequence, laid out plainly.
April 2022: $5.675 billion, on the back of a $675 million Series C.
Now: $23 billion, on the back of a Series D with no disclosed size.
Between those two points, the disclosed operating footprint moved from a single Loop to a handful of tunnels and one small international pilot. Passenger counts accumulated. No revenue was disclosed. No path to profitability was disclosed. A four-fold valuation increase on that operational delta is not a cash-flow story, and anyone presenting it as one is either misreading the document or hoping you will.
It is a capital-supply story. The marginal buyer for this class of asset in 2026 is Gulf sovereign capital with an explicit mandate to convert hydrocarbon balance sheets into durable real assets, and with a demonstrated appetite for founder-brand premium inside the Musk orbit. SpaceX and Starlink established that precedent, and the premium is now being priced as a category, not an exception. The valuation reflects what a strategic buyer is willing to pay for exposure, not what the asset earns.
I hold a related view that I have argued for years in a different context, and it applies cleanly here. Most of the "problems" our industry is told it has are manufactured by the parties who want to sell the solution. Liquidity fragmentation is the canonical example: nobody was actually confused about where liquidity lived, but somebody needed a reason to launch another chain, so the confusion was manufactured and then monetized. The $23 billion figure is the same maneuver at a larger scale — a valuation produced by capital availability and then presented as a market judgment, with the supply side of that judgment left off the page.
What Sovereign Capital Is Actually Doing, and Why Crypto Keeps Misreading It
This is the part worth your attention, and it has almost nothing to do with tunnels.
The same capital pool that led this round holds spot Bitcoin ETF exposure. That is the actual information. Not "crypto wins because a sovereign funded a tunnel," but "a sovereign balance sheet is now allocating across hard infrastructure and regulated digital assets inside a single strategic framework."
Read that carefully, because it inverts the standard institutional-adoption narrative. The industry spent a decade insisting it would be validated when institutions arrived. Institutions have arrived. What they purchased was an ETF wrapper, a custody arrangement, and a settlement rail. They did not purchase governance. They did not purchase community. They did not purchase the ideological content of the asset class. Trust is the only currency that matters, and sovereign allocators buy it pre-certified, with a legal opinion attached.
I find a clarifying parallel in the rollup wars. The widely repeated claim is that OP Stack and ZK Stack compete on cryptography. They largely do not. They compete on who can persuade more projects to deploy a chain — distribution, political alignment, grant programs, founder relationships, and the patience to court teams for eighteen months. The technical differences are real and are routinely overstated relative to the commercial ones. TBC occupies the same position inside its own market. Its differentiator against Herrenknecht and against municipal transit authorities is not primarily engineering superiority. It is the demonstrated ability to get a government to sign. That is a distribution moat, and distribution moats are genuine. They are simply not the thing the technical narrative claims to be selling.
The competitive picture, stated honestly, looks like this. Incumbent tunnel-boring manufacturers and municipal rail authorities hold the absolute majority of the market on capacity, maturity, and regulatory qualification. TBC holds a brand, a machine platform with an unquantified cost claim, and a demonstrated capacity to close sovereign and municipal deals. Other Loop-style startups hold neither. Market-share comparison is impossible because the disclosed data does not support it, and I would rather say so than construct a table that implies precision I do not have.
The Bridge Problem, Applied to Capital
There is a structural parallel here that I think about constantly, and it is worth stating plainly.
Cross-chain bridges have been drained of more than two and a half billion dollars cumulatively, across every architecture, every audit tier, and every security model the industry has produced. And the industry still depends on them, because the alternative is not connecting to anything. That is a fundamental security paradox, and it persists not because nobody noticed but because the dependency is more powerful than the risk.
Sovereign capital dependency has the same shape. A company that raises its Series D from a single state-linked lead, in a jurisdiction that is simultaneously awarding it public contracts, has not eliminated its financing risk. It has concentrated it, and placed it inside a relationship that is political before it is commercial. Investment-for-localization is a stable model right up until the local politics change, at which point renegotiation is not a legal event but a diplomatic one.
I am not predicting that outcome. I am observing that the disclosure does not let me price it, and that any analysis treating the sovereign lead as unqualified validation is skipping the only risk that actually scales with the valuation.
Governance and the Concentration Problem
Two governance facts deserve more attention than they received.
First, key-man risk is extreme and entirely unhedged. The founder's attention is distributed across an electric vehicle manufacturer, a launch provider, an AI lab, and a social platform. The disclosure itself notes that his wealth derives principally from the first two, which tells you where the strategic priority sits. A company whose $23 billion valuation is anchored in founder brand is a company whose valuation is exposed to that founder's unrelated decisions, unrelated litigation, and unrelated reputation events.
Second, disclosure asymmetry. A syndicate of this quality — Sequoia, a16z, Temasek, Valor — receives information at a granularity the public will never see. That is normal for private markets, and it is also the precise condition under which outside observers should stop behaving as though they are inside. The investor list is an endorsement of the company. It is not an endorsement of the number, and it is emphatically not an endorsement of any crypto thesis layered on top of it.
The Regulatory Frame, Briefly
For completeness: there is no token, so there is no Howey question and no securities analysis to perform on the asset itself. The Series D sits in the private-placement world. The Bitcoin ETF exposure held by a sovereign entity is a regulated product and a compliance non-event in 2026.
The genuinely interesting regulatory fact is structural rather than legal. Dubai pairs a government contract with sovereign investment, which is the standard investment-for-localization template. Capital enters, industry lands, contracts emerge. That model is stable and legible, and it is also politically contingent in a way a purely commercial financing would not be — which is precisely the trade-off a state-linked lead buys you.
The Honest Bridge to Crypto Is Slower Than the Headline
There is a legitimate Web3 connection here, and I want to name it precisely so that it does not get inflated into something it is not.
Infrastructure capital is beginning to accept regulated digital rails for settlement, custody, and — increasingly — for the tokenization of real-asset cash flows. That trend is real, it is proceeding, and its direction runs toward institutional plumbing rather than retail participation. A sovereign allocator already comfortable holding a spot Bitcoin ETF has crossed the threshold where a tokenized infrastructure receivable is not conceptually alien. Real-world asset tokenization is not a meme. It is the natural extension of what these balance sheets are already doing in the analogue world.
But that connection moves at the speed of procurement cycles, not news cycles. It will appear in custody agreements, fund structures, and legal opinions long before it appears in a token price. Anyone using this week's tunnel valuation as evidence of near-term crypto demand is reading a document about a different asset class and finding in it what they already wanted to find.
The Counterargument I Will Not Make
Before the contrarian turn, one thing I refuse to do: I will not tell you this is a fraud. I have no evidence of that, and I have watched too many careful people damage their credibility by reaching for the most dramatic available conclusion when the boring one was sufficient. Twenty-five tunnels, four million passengers, a Tier 1 syndicate, and a named municipal contract are not the profile of a fabrication. They are the profile of a real company with an unverifiable valuation, an unfalsifiable technical claim, a document that disagrees with itself about its own date, and a set of crypto tags that do not belong on it.
The boring conclusion is the correct one. Carry it.
The Part Where Both Sides Are Wrong
Here is where I part ways with the bulls and the bears alike.
The bears are running the wrong critique. "Overvalued" assumes a discounted-cash-flow framework in which $23 billion is a claim on future profits that a pilot cannot justify. That is the right analysis for a public equity and the wrong analysis for a strategic allocation. Sovereign capital is not purchasing a cash-flow stream. It is purchasing optionality on a relationship, a technology import, and a counterparty who returns calls. You cannot call an option overpriced by comparing it to an annuity. The valuation is not wrong. It is answering a different question than the one being asked about it, and both parties in the argument are failing to notice that they are holding different instruments.
The bulls are running a worse one. The crypto coverage treated this as validation — a sovereign buying ETF exposure, a sovereign leading a Musk round, read together as institutional arrival. That is narrative grafting, and it is the same rhetorical move that has separated retail readers from their capital in every cycle I have covered. In 2021, I spent months interviewing Bored Ape collectors and artists, and concluded that the value driver was social credentialing rather than art. The collectors who understood that distinction were fine. The ones who mistook a status market for a fundamentals market were not. The Web3 tag on this tunnel story is a credential. It is not a connection.
And here is the genuinely contrarian position, which I hold with some discomfort. The uncomfortable lesson is not that crypto is being validated by sovereign capital. It is that sovereign capital has quietly shown us what it considers a mature asset class — and none of it is the thing our industry spends its energy arguing about. Not governance tokens. Not community treasuries. Not decentralization theater. Regulated wrappers, legal certainty, identifiable counterparties, and contractual claims on real cash flows. If that is the bar, then most of this sector is not being adopted. It is being politely skipped, and the skipping is happening without a single headline about it.
I learned a version of this during the 2022 collapse, when the loudest voices in the industry went silent and the readers who stayed with us were the ones who wanted fundamentals rather than forecasts. I restructured our coverage toward resilience and education that year, mentored three junior analysts through their own panic, and kept our output steady while competitors published nothing. What I took from it is simple: in a bull market, the audience's tolerance for risk exceeds its tolerance for being told about risk, and that gap is where the damage accumulates. We are in a bull market now. The gap is open again.
Forward, Not Backward
Watch three things, in this order.
Whether a Form D or equivalent filing surfaces to corroborate the round — because right now the primary document does not agree with itself about its own date, and that has to be resolved before anything built on it can be trusted.
Whether the Dubai pilot converts into delivered kilometers, because that is the only data point that prices the Prufrock claim and the only one the company cannot narrate its way around.
And whether the same sovereign capital continues allocating across hard infrastructure and regulated digital assets in parallel, because that pattern — not this valuation — is the signal with a five-year horizon, and it will still be legible long after this round's headline is forgotten.
One question to sit with, and I mean it as a genuine question rather than a rhetorical flourish. If sovereign balance sheets are now the marginal buyer for tunnels and for Bitcoin alike, what happens to both prices on the day they decide they are finished buying?
