At a very high level, one of the basic design concepts for a CBDC is the decision to design a purely token-based solution or should the central bank take on a broader mandate to offer an account-based solution that would support digital currency tokens and potentially a variety of other assets as well. It needs to be clear that regardless of the choice here, tokenization of the digital currency is in play either way. A central bank is going to leverage the tokenization of value to operate a digital currency. Is there value in going to the next level and offering accounts? This may depend on the level of disintermediation a central bank wants to employ within their economy.
N.B.: One theme that I will continually return to in this series is the need for designers and architects of CBDCs to not get caught up in the old concepts of money and finance. Trying to make technology mimic or reinforce the models of physical currency will fail. Disruption and disintermediation of an economic system are inevitable with any digital currency solution. Trying to avoid these disruptions in a CBDC design will be economically catastrophic. Solutions need to be approached with critical thought and the designs will need to identify and account for the changes and impacts that they create. The constructs, processes and models of physical fiat are useful to understand but not applicable to a CBDC and also the constructs, processes and models of cryptocurrencies and existing digital currencies are not as broadly applicable to a CBDC as many try to imply. A successful CBDC will require stepping outside of many of the thought boxes we’ve created thus far. I will incorporate concepts from existing solutions as a reference point, but it does not mean an endorsement for using the concept as exactly represented by those solutions within a design for a CBDC.
On Tokens and Tokenization
Tokenization is a concept that has been around for a long time. We’ve learned that by utilizing tokenization, we can create a tokenized representation of any asset, be it digital or physical. Looking to the bitcoin model, this was implemented as an exact proof that you can tokenize value within digital currency and allow users of that currency to split and combine values from various tokens to make commerce work.
Some earlier concepts of digital money had tried to reuse serial numbering style concepts from physical currency, but those solutions immediately become very cumbersome because you needed a serial number on every element of the currency. Moving a dollar meant aggregating 100 serial numbers for 100 cents to be moved. Larger transactions became almost impossible as there was no effective model to aggregate and assign groups of serial numbers to a larger single serialized value and then break that down again later. So, a big part of the tokenization revolution within the bitcoin model is the proof that we could have a single source of value in the system – the mining rewards – and tokenize that value to a large number of users who then transact by generating more tokens of higher or lower value. Because the ledger that logs those transactions is immutable and cannot be changed, the value can move around the system without the need for a specific serial number on every penny to track it. The fact that bitcoin offers a single method of value creation within the system and no method for value destruction, means that it is also very simple to ensure that no other sources of value creation exist and the nodes in the system ensure that this is always true.
Not all tokens are created equal. It is not recommended to apply the tokenization model of currency value such as those found in Bitcoin or Ethereum to the tokenization of other assets. You should never mix different purposes into a single token. If you plan to create a tokenization model for representing and transacting in the value of the digital currency – then the tokens should only be used for that purpose. The tokenization models that we use to apply to physical assets and other instruments usually look very different. As an example, the model we used to tokenize an encrypted digital master recording of a song in Ujo music to then spawn tokenized watermarked versions of that master for use in movies, commercials or streaming channels is very different from a token that is used to combine and split currency value. The tokenization model that we use to register a physical piece of art linked to a digital encrypted representation of that art, acquire 3rd party valuations of the art and then track the sales/ownership of that piece of art is also very different from tokenizing currency value.
Designing a token to be multi-functional creates a drag on a system because it needs to determine the specific use of the token and which contracts apply to that token. For efficiency, a token should be created and used for a specific purpose to reduce the loads on the tokenization engine and the systems that need to use or process those tokens for the specific function it was designed for.
The Central Token in Account
Central banks already need to take on a lot in creating and managing fiat currency, creating a digital currency and then transitioning an economy between the two. There is no need to exacerbate the solution with a leap to provisioning a centralized accounts platform for commercial and retail use of the digital currency. There should be plenty of opportunities for FinTechs and banks to build out the account-based solutions and wallets that enable the utilization of the digital currency tokens and also offer the broader capabilities of account-based solutions. The value of an account-based platform is that it can be an open wallet for a variety of tokenized and untokenized assets. I can have my digital currency tokens, my real-estate property tokens, my digital art tokens, and better aggregate the value of my assets in an account-based solution. There is little value that a central bank can offer outside of the operations and management of the digital currency.
Disintermediation will be at the core of this argument and the appetite for disintermediation will vary among jurisdictions. Central banks that feel they are being overrun or overruled by their current banking environment may opt to significantly change the playing field by centralized account solutions in either a wholesale and/or retail setting. More centrally controlled economies can certainly create even more centralized control by cutting banks, financers and other intermediaries out of the picture with a centralized account-based banking solution. This would significantly change the face of banking in regions that choose this model.
While the concepts and implementation of tokens are fairly simple, the tools and smart contract solutions for a central bank to manage seigniorage, creation and destruction of currency within an economy, manage liquidity, generate interest, monitor valuation of the currency domestically vs internationally, manage central bank-owned assets and other aspects will all need to be considered in the delivery, but the token is the core method that will enable these controls. Adding a central bank-based account solution at the wholesale level may have some value, but an offering of accounts by a central bank at the retail level would likely be a lot for a central bank of any sizable nation to take on just from the infrastructure, management, operations and risk perspectives. We will deep dive into the retail vs wholesale models in a future post.