It’s common nowadays to hear people describe Bitcoin as “digital gold.” The term “mining” itself suggests — quite intentionally — the image of old-time prospectors, clawing through mounds of rock in the hopes of discovering a new vein of precious metal. Then there’s the near-mythical status of gold as a reserve currency (i.e., a store of value) and unit of account, mirroring the fact that Bitcoin is the blockchain universe’s dominant asset (and one side of most trading pairs on exchanges).
A large part of Bitcoin’s appeal as “digital gold” stems from its fixed supply and — until recently — the absence of complex financial products built on it. For perspective, trust in governments and banks was at an all time low when Bitcoin was first released, as evidenced by the Genesis Block’s reference to “The Times 03/Jan/2009 Chancellor on brink of second bailout for banks.”
For Satoshi, Bitcoin was a political statement. It represented the cypherpunk’s technological answer to the increasingly political control of money and central bankers’ ability to debase the currency through inflation.
Satoshi likely drew inspiration for Bitcoin’s monetary policy from gold itself, which has maintained long-term price stability quite effectively. Although we will likely never know, it’s probable that Satoshi believed that this long-term stability resulted from the inherently finite amount of gold in existence, and that he sought to emulate that scarcity. Satoshi constructed the protocol to produce a steady (but decreasing) supply of new assets, eventually capping out at a total value of 21 million coins. Presumably, this finite supply would thus impose an artificial scarcity that would allow markets to establish a price based on demand for money, enabling its use as “a peer-to-peer electronic cash system.”
This monetary policy is based on what I call the Resource Scarcity Fallacy, which is exactly what it sounds like: the intuitive, but mistaken, belief that resources are limited.
The late economist Julian Simon differentiated between raw materials (i.e., stuff in the ground) and resources (i.e., stuff in usable form). While the former is finite, the latter is not: as resources get more expensive, humans, guided by markets, seek out alternatives and new sources.
When the price of gold rises, people invest in mining equipment and processing technologies, increasing supply and lowering the price. In the long run, the marginal rate of return on gold production should approximately match that of the market in general. If the rate of return is too high, capital will flow in as investors seek economic profits; if the rate of return is too low, capital will flow out as people seek better opportunities elsewhere. Of course, these effects aren’t immediate; gold markets have short-term volatility, but the net result is that the purchasing power of gold remains relatively stable in the long run.
In reality, the “monetary policy” of gold looks nothing like the monetary policy of Bitcoin. Bitcoin has no market dynamic governing its supply. Rather, the supply of bitcoin is fixed, and will end at 21 million tokens in the 2040’s, come hell or high water.
This monetary policy has a few notable and interrelated problems which should be apparent for even a casual observer of markets. At the most basic level, the supply of new assets is not responsive to demand for money. Rather, new coins are created on a fully predictable basis. As a result, when the value of the network increases (such as when a major retailer begins accepting Bitcoin), the new demand for money is not matched by a new supply. That means only the price, and not quantity, will increase with new demand.
Because increases in network value accretes to tokenholders rather than active users or validators, Bitcoin strongly rewards “hodlers” and discourages spending. Without significant sell pressure, markets have a thumb on the scale, driving up the price ever further. This also invites speculation, which tends to amplify price volatility.
In turn, this makes producers of goods and services unwilling to build the infrastructure necessary to accept payment in Bitcoin. Why take payments in the form of an asset no one wants to spend, if it means extra overhead and security costs? And how can prices even be established when the price of Bitcoin fluctuates by 10% or more in a matter of minutes?
The greatest barrier to widespread adoption (aside from the pesky “scaling debates”) is likely rooted in Bitcoin’s monetary policy; if people are to use it, it must have a reasonably stable long-term value — not one that increases endlessly. Without such stability, it will be unattractive for general use and suitable for speculation only. Consumers must not believe that their assets are producing a return on investment, and producers must not believe that the asset’s value could suddenly collapse.
But what if we could make a token that had the market dynamics of gold? Essentially, we would need a way for the market — rather than the protocol — to determine what the supply of money should be. If tokens got too expensive, network participants could simply create new tokens (after incurring a real economic cost, such as contributing hash power, of course). This would tend to dampen upward price movements, thereby suppressing unproductive speculation and fostering the stability necessary for both consumer and producer adoption.
One option for having a “rational” monetary policy with gold-like properties might be to separate out the processes for transaction validation and those for asset creation. In this conception, a validator would be responsible (and rewarded) for verifying transactions, but any participant could also create new assets by broadcasting a transaction backed by an appropriate proof-of-work — much like Bitcoin’s current coinbase transactions.
If block sizes are capped, these coinbase transactions would compete with value-transferring transactions for block space. Each transaction would come with fees, of course, so miners (i.e., those creating new coins, not validators) would only do so if the rate of return exceeds the market rate of interest. The seigniorage value in this system accretes to miners (and to a lesser extent, validators), not existing tokenholders. As a result, it doesn’t reward hodling or speculation, and is a more stable platform for commerce.
An interesting consequence of this approach is the ability to create hybrid blockchains — for example, using proof-of-stake for consensus, and proof-of-work for token generation. The network can thus process transactions without ridiculous energy use, but if the price of the asset rises too much it will incentivize the (costly) generation of new assets. Thus, the network’s energy use will be high only while the network is growing. (This energy use can also be mitigated by establishing some baseline rate of money creation, although the mechanisms for that are beyond the scope of this post.)
Moreover, such a design can greatly improve the decentralization of the blockchain, across multiple different dimensions. First, new users can acquire tokens without relying on accessing centralized gateways such as exchanges. This would improve privacy and censorship-resistance. Second, it enables drawing on the best security features of both proof-of-work and proof-of-stake. Centralization of the validator set could be undermined by the ability of users to add new assets, as an attack would require both the energy and physical infrastructure necessary to create sufficient stakes to attack the network, and sacrificing the value of those stakes.
It remains to be seen if such a protocol will be successful, if its ever even built. One obvious challenge is the radical disparity in electricity prices between regions and across time. For the time being, though, the analysis of monetary theory and its implications for blockchain market stability, user adoption, governance, and security promises to be interesting and fruitful.