D.L. White

@pmarca wrote something about Bitcoin in 2014 that I think has aged remarkably well. Not the micropayments, not the merchant-adoption curve, and not the expectation that rising usage would naturally turn Bitcoin into a global payment network.

The part that still looks exactly right is much more fundamental: Bitcoin created, for the first time, a way for one Internet user to transfer a unique piece of digital property to another without relying on the normal institutional machinery that determines ownership and transfer.

That is an extraordinary claim, and I think it is correct in its entirety.

The mistake comes in what tends to be inferred from it. Andreessen moves quickly from digital property to digital money, and Bitcoin discourse has spent much of the subsequent decade treating that transition as though it were contained in the technology itself. Bitcoin is scarce and transferable, therefore Bitcoin is money; because it is money, it should become a global medium of exchange; because the base layer cannot support that kind of transaction volume, whatever additional infrastructure is required can be built around it while the resulting system continues to be described as Bitcoin.

The problem is that none of those later steps follows automatically from the first.

This is why Michael Saylor’s characterization of Bitcoin as “digital capital” strikes me as much closer to the thing actually sitting in front of us. The appeal of the phrase is not that @saylor has some privileged ability to define Bitcoin, but that the description requires very little future projection.

Bitcoin is valuable property.

I can exclude others from it, transfer it without asking a bank, broker, issuer, or government agency for permission, and verify ownership through a public network. If I secure the keys properly, there is no title office where someone can simply amend the record, no corporate issuer that can cancel the asset, and no bank account that must remain open for my claim to continue existing. Someone can still compel me physically, steal credentials, or exploit poor security, but the property itself is extraordinarily resistant to administrative seizure.

That is not a theoretical future use case. It is what Bitcoin demonstrably does now. And it is already a strange, incredible, and historically important property-rights arrangement.

The phrase “Bitcoin is money” usually carries much more baggage. It is often treated as axiomatic because Bitcoin is scarce, divisible, durable, portable, and transferable. From there the rest is quite literally imagineered into place: wages will be paid in it, groceries priced in it, debts denominated in it, governments disciplined by it, fiat displaced by it, and global economic life eventually reorganized around a fixed supply. Yet nearly every meaningful component of that description refers to an institutional state Bitcoin does not presently occupy at monetary scale.

The usual answer is that the missing functions will be supplied by layers. Lightning can handle payments, custodians can aggregate transactions, banks can hold reserves, credit can be issued against Bitcoin, stable instruments can settle against it, and new financial infrastructure can extend its reach.

That may happen.

But the more infrastructure that must be layered on top to make Bitcoin perform the functions of a monetary system, the more obvious it becomes that those functions are being supplied by the surrounding institutions rather than by the base asset itself.

That distinction matters because the features that make Bitcoin unusual become diluted as you move away from the base layer. Once I am operating through custodians, routing networks, legal claims, credit systems, software intermediaries, and settlement hierarchies, I have reintroduced trust, counterparties, failure modes, and institutional dependence. Those systems may still be useful and may ultimately settle against Bitcoin, but that is conceptually different from saying that Bitcoin itself directly performs all of the functions attributed to “money.”

The digital-capital framing handles this much more cleanly. Bitcoin can sit at the base as scarce property. Claims can be built against it. Institutions can hold it. Credit can reference it. Payment systems can move obligations whose ultimate value is tied to it. People can save in Bitcoin while transacting in something else entirely. None of that diminishes Bitcoin. It simply stops pretending that every layer above it is economically identical to the thing underneath.

More importantly, this framing fits actual observed behavior much better. People acquire Bitcoin and hold it. They celebrate long holding periods, lower time preference, shrinking liquid supply, lost coins, and unwillingness to sell. Companies construct treasury strategies around accumulating it. Holders increasingly prefer to borrow against it rather than spend it where possible. These are perfectly coherent behaviors around scarce capital. They are much stranger if the ultimate thesis is that Bitcoin should become the mundane transactional denominator of daily life.

The fixed-supply issue makes the distinction harder to avoid. As capital, absolute scarcity is easy to understand. If demand for an asset rises while its supply does not, appreciation is expected. That strengthens the property thesis. As the monetary unit organizing an entire economy, however, the same feature creates harder problems.

If a state makes Bitcoin strongly monetary by requiring taxes in it, then it begins collecting the fixed stock. If it persistently retains more than it returns, its share necessarily rises until the private sector is progressively starved of the units required to satisfy the next round of obligations. The obvious solution is for the state to spend the units back into circulation, but at that point fixed supply has not eliminated monetary management. It has merely transformed the management problem from issuance into circulation.

The opposite case is also awkward for the monetary thesis. Suppose taxes, wages, debts, rents, and ordinary prices remain denominated in dollars while Bitcoin continues appreciating against them. Under those conditions, rational actors have a strong incentive to spend the weaker unit and retain the stronger one. If I can buy groceries with dollars that I expect to lose purchasing power while preserving Bitcoin that I expect to gain purchasing power, the decision is not mysterious. The appreciating asset becomes the thing I prefer to dispose of last. Again, that is exactly what one might expect from capital.

It is less obviously what one wants from the universal circulating unit.

This is where Andreessen’s original paragraph remains so important. He identified the actual technological novelty before the argument raced ahead to assign Bitcoin a much larger historical role. Bitcoin created a new form of digital property with unusually strong exclusion and transfer properties. That is observable, already real, and does not depend on theoretical hyperbitcoinization, theoretical global merchant adoption, theoretical universal Bitcoin-denominated wages, or an imagineered future in which every existing monetary institution has been replaced by a Bitcoin-native analogue.

On that view of Bitcoin, the less “digital capital” sounds like a retreat from the monetary thesis and the more it sounds like a description disciplined by what the system actually does. Bitcoin may become reserve collateral for large financial structures, a preferred long-term savings asset, an important sovereign balance-sheet instrument, or a way for individuals to retain property in environments where ordinary ownership claims are precarious.

Any one of those would be historically consequential.

What seems much less necessary is the insistence that Bitcoin must also become wages, rent, taxes, coffee, credit, accounting, payment rails, commercial banking, and every other function that happens to sit beneath the loose heading of “money.” That begins to look less like observation than theoretical annexation: a new thing appeared, and instead of asking what it was uniquely good at, we immediately assigned it the entire job description of an old thing.

Andreessen may have had the more important insight before doing the same himself. Bitcoin gave us a new form of digital property. It’s very unclear why that should be regarded as an insufficient outcome.