Bitcoin's security is the last unpriced layer of the asset. The venue to price it already exists. What is missing is the buyers.

Two letters ago I argued that Bitcoin's largest holders have remarkably little exposure to the infrastructure securing their wealth. In the second I argued that this infrastructure is now being outbid for its inputs by AI, and that the price of Bitcoin's security is a formula containing no term for what security is worth to anyone. That left one question standing: if security matters this much to holders, why has nobody been paying for it?

The answer is not negligence. It is rationality. And once you see why, the shape of the solution follows.

Why the foundation went unfunded

Bitcoin's security has the economics of a public good: every holder benefits from the network's proof-of-work, and every individual holder is rational to let someone else pay for it. A treasury company that spends on security strengthens the network for every free-riding competitor alongside itself. Public goods stay underfunded until a mechanism converts the collective benefit into an allocatable private decision, a way for one participant to buy a defined amount of the good, at a price, for themselves.

Bitcoin has never had that mechanism. The market built extraordinary sophistication for pricing the asset: spot, custody, lending, structured access of every kind. It built almost nothing for pricing the security underneath it. A large Bitcoin position today carries embedded exposure to hashrate health, template concentration, and miner economics, exposure that few institutional balance sheets explicitly measure, let alone manage, because there has been no instrument through which to manage it. Institutions can, of course, buy mining companies or build facilities. But there has been no simple, open-market instrument through which a Bitcoin holder can allocate capital directly to independently directed proof-of-work, without owning the machines, the power contracts, or the pool.

Decentralization is a stack, not a product

The industry is not standing still, and honesty requires saying so.

Bitcoin's security has layers, and each concentrates differently: hashrate protects consensus, mining infrastructure determines who exercises it, and block-template construction determines who decides what enters the ledger. Protocol work is making real progress on that last layer. The Stratum V2 standard lets individual miners rather than pool operators construct block templates. Many major pools have announced support for it; production use of miner-built templates remains early, and the two are not the same thing. That work matters. But it redistributes decision-making. It does not change who funds the hashrate, how mining economics survive the subsidy's decline and the hyperscaler bid, or whether the owners of Bitcoin's economic value have any way to participate. Protocol upgrades fix the layers they can reach. The funding and participation layer, the allocation layer, is not a protocol problem. It is a market problem.

Bitcoin's quiet advantage over AI

And here the AI comparison from my last letter turns around, because on one dimension Bitcoin is far ahead of the technology outbidding it.

Hashrate is already a commodity: standardized, measurable, deliverable, trading continuously on open markets around its reference value. AI compute, for all the capital chasing it, still has nothing comparable in maturity. Much of it changes hands in bespoke, bilateral, multi-year contracts, with no liquid spot market, no transparent reference price, no mature trading layer. Bitcoin may be losing the beauty contest for capital, but its security input is the more financially mature asset of the two.

Think about what that means. The infrastructure race everyone is watching still transacts its core resource like industrial real estate. Bitcoin's core resource trades like a commodity, today, with continuous price discovery. The rails for allocating capital toward Bitcoin's security already exist. What has been missing is not the market. It is the participation of the asset's largest beneficiaries in it.

The wrong answer: buying the stack

Before the right answer, name the wrong one, because it is the one institutions reach for by instinct.

The instinctive response to "you are exposed to mining" is vertical integration: acquire the exposure. Follow that path to its end (ETF issuer acquires custodian acquires mining company acquires pool) and you get a vertically integrated security apparatus that would make Bitcoin more concentrated, not less, and would recreate precisely the trusted-intermediary structure Bitcoin exists to avoid. Replacing one concentrated owner with another is not decentralization with better branding. It is the same disease in institutional clothing.

The goal is to make participation itself open: capital should be able to enter and exit the security market without owning the infrastructure. The answer to institutionalization is not institutional ownership of mining. It is institutional access to an open market for mining.

Demand is the other half of decentralization

This is where open hashrate marketplaces enter, not as the solution, but as a missing piece of the stack alongside independent miners and protocol upgrades.

At NiceHash, we operate infrastructure, under Swiss regulatory supervision, through which any participant can acquire real SHA256 hashrate at open market prices and decide where it works. The mechanism matters. Purchased hashrate does not have to flow back into the concentrated pool structure. It can be directed to solo block production outside the dominant coordination points: blocks produced under the buyer's mandate, with rewards delivered straight from the chain. That is the difference between renting computation and adding an independent source of block production.

It is also the mechanism the AI era makes urgent. When the energy, capital, and machines securing your asset all have a better-paying alternative use, security stays only if the demand side is willing to pay for it. Mining's flexibility lets it survive on the grid's margins; whether it thrives there is decided by demand for hashrate, and that demand is now something any holder can be.

Put it in the language of a CFO. Ignoring the security layer was a defensible position when the subsidy was rich and the mine had nowhere else to go. Neither is true anymore. An allocation to hashrate buys three things no custody arrangement offers: independent block production at the margin of a concentrating system, direct participation in the infrastructure securing the asset, and optionality on a market that grows more important with every halving. This is an emerging use of capital, not yet a conventional one, which is precisely the point: the institutions that move first will define its terms. And the deeper logic is this: precisely because a Bitcoin mine is a data center with options, Bitcoin's security needs a liquid market that can bid to keep it. The alternative is not the status quo. The alternative is watching the infrastructure leave for the industry that pays.

I will be equally honest about scale. An institution allocating a fraction of a percent of a large Bitcoin position to hashrate does not, by itself, move a Nakamoto coefficient. What it does is add independent block production at the margin, establish that the demand side of mining is a place capital can be allocated, and set a precedent that makes the same allocation legible for every institution that follows. Decentralization was never going to arrive as a single decision. It arrives as an allocation habit.

And for those who make it, the contribution need not be invisible. Blocks produced this way can carry the participant's own signature and message, written permanently into the chain: the on-chain manifestation of a philosophy worth stating plainly. We don't just custody Bitcoin; we allocate capital to the infrastructure that secures it. Not a press release. A block.

The last background condition

Every mature financial system eventually builds markets for its own infrastructure. Liquidity became a product. Custody became a product. Credit, settlement, insurance: all began as background conditions and ended as priced, allocatable markets. Bitcoin's security is the last background condition left, and AI has ended the era in which it was free to ignore.

In my first letter I said Bitcoin holders are standing on the mine. They still are. The difference is that the mine now has somewhere else to go, and for the first time, so does the capital that could keep it here.

The market learned to price the asset. The next phase of Bitcoin's maturity is pricing what keeps it alive: the emergence of a market for security itself.