Posted: 22 Nov. 2017 6 min. read

Tokenize all the things!

As we pointed out in a previous post, it might be wiser to think of Bitcoin as signalling a shift in the environment rather than the development of a new technology. Bitcoin emerged when the cost-performance and penetration of broadband reached the point that Bitcoin became economically viable, not due to new math.

 

Consequentially we’re missing a trick if we focus on blockchain as the solution to our distributed computing problems, as the shift in the environment means that a whole host of distributed technologies are now viable, not just blockchain.

The question, then, is could we tokenise an asset without a blockchain?

If we’re thinking of tokenising an asset, then we’re concerned with:

  • issuance, how is the asset created?
  • clearance, can we confirm that our counterparty owns/controls (enough of) the asset?
  • settlement, how does one transfer ownership of the asset between two parties?

We can also get tricky and add:

  • discovery, how do I find someone selling the asset I want?
  • price-formation, what is the asset worth?

Bitcoin – a stateless currency – is interesting as it is one of the few instances where issuance requires a distributed ledger.

Bitcoin solved the double-spend problem by using account balances – triple-entry accounting, if we’re to be precise – rather than digital tokens to store value. You can’t double-spend something that doesn’t exist. Consequentially Bitcoin requires a common ledger to support clearance as triple-entry accounting only works if we are confident that we’ve seen all transactions and that each transaction is unique (global consistency), so that we can accurately reconstruct current account balances. The ledger must be stateless (not owned by any individual or institution) as otherwise one actor controls the ledger and is therefore the issuer of the currency. The currency must be issued as a side-effect of maintaining the distributed ledger (rather than by some identified actor) so that the value it represents is also stateless.

While Bitcoin requires a distributed ledger, the ledger comes with a cost as it’s slow and expensive to operate when compared to a traditional digital ledger (a database). It’s a simple question of cost-benefit: if you want a stateless currency then you must use a distributed ledger.

The vast majority of use cases, however, don’t have such stringent requirements, and we can avoid the trouble and expense of a distributed ledger.

The thing about the most assets is that they have a connection to the physical world. Sometimes this connection is concrete – such as land, property, where a title entitles you to exclusive use of a geographical area. Other times it is more tenuous, such as where the asset represents a claim on a pool of identical assets; a claim on a barrel of oil in an oil tanker, for example. In both of these cases issuance has occurred in the real world: property is a right created by the government, while the claim on a barrel of oil is created by the original owner of the oil (most likely the firm that dug it out of the ground via a property right).

If we want to tokenise these goods then we don’t need to resort to a distributed ledger. It doesn’t matter if the asset is fungible or not. While property is not fungible (given that a title represents a unique location on the planet) oil clearly is. All we need to do is create a document in the digital world that represents a claim on the asset, and then trace the provenance of that document.
Consider property. A property title is simply a document that records the ownership of defined geographic area.

The document is created (issued) by the government as it represents a right that is enforced by the government (and the government’s monopoly on violence). The government then signs it over – transferring the right – to a recipient. It cryptographic terms, the document includes the recipient’s public key, and is signed with the government’s private key. (We happily acknowledge that key management is a challenging problem common to both this and a blockchain-based approach, but outside the scope of this discussion.)

Transferring ownership of the land title (settlement) is simply a question of the current owner signing the title over to a new owner, just as the government did. Anyone can verify that the government signed the original property right, but only the ‘owner’ can transfer ownership to another entity.
What we’ve done is treat land titles as a provenance problem. You can prove your claim to the land if you can produce a title that has been signed over to you, and which shows provenance all the way back to when the government first issued the title. Provenance typically doesn’t require a blockchain, as we discussed in a our last post, Provenance and the search for blockchain’s killer app.

Treating ownership as a provenance problem enables the holder of a title to advertise and sell the asset however and wherever they want. They could put the title on Gumtree (or a real-estate focused online market, perhaps), enabling buyers to find the title (discovery) and bid on it (price formation). They can even sell it down at the pub. Or do all of these at once. (They can also keep their ownership secret, should they choose to.)

Selling your tokenised asset is no harder than selling an unwanted couch. Discovery and price formation are not problems that we need to solve, as there are already many good solutions that we can leverage.
This approach, however, doesn’t explicitly provide certainty on who owns a particular title right now (clearance), as we haven’t created a central clearing house yet. It’s easy enough to prove that you owned the land – just produce your land title – but how do you show that you currently own it?

We could solve the clearance problem by having the purchaser disclose their purchase; if property is sold in a forest but there is no one there to hear it fall, and all that. (We use purchaser rather than the vendor as the purchaser is incented to disclose and thereby claim their property right, while the vendor might hide the transaction so that they can ‘double sell’.) The purchaser simply submits the updated title to a network of peer-to-peer clearing houses, with the update distributed between peers via something efficient via a flooding algorithm, as we did with provenance. These peer-to-peer clearing houses might be run by the government or a trade body; or even by frequent vendors, purchasers or brokers. Many solutions are possible.

Also, the government might want to track who owns which title to enable it to enforce regulatory requirements, such as foreign ownership limitations. We can easily support this by requiring that all transfers of ownership are co-signed by the titles office for them to be valid: i.e. a title must be signed over to the new owner and then endorsed by the titles office (which is current practice). The titles office can then control sales where required (by refusing to endorse undesirable transfers) while also providing it with visibility into all title transfers. This enables the sales to be keep private if we choose (or we can choose for them to be public to support distributed clearance, as above), as the titles office can act as a clearing house (or the titles office can submit the title to the clearance network). It’s also acknowledging that it is the government that stands behind the title and that, ultimately, the owner is whomever government chooses to recognise as such.

We might want to reconfigure a property title, or collection of property titles, either by subdividing a block or merging a number of adjacent blocks. Subdivision can be supported by having the owner of the title signing it over to multiple recipients while specifying a constraint on each recipient, where the constraints specifies the sub-area that the recipient receives. Combining blocks is simply the reverse of subdivision, where a number of separate titles sharing the same private key are signed into one title.
And, finally, settlement can be managed via an independent escrow service, for a small fee. It’s not even necessary for this service to know what is being exchanged.

We can use a similar approach for any virtual or physical asset that is issued by an entity.

The fairly recent announcement for an oil trading solution is a good example. The idea behind the proof of concept was to show how a blockchain could be used to support trading of oil while on a tanker was on route. Oil is stored as account balances and trading functionality realised via smart contracts. From a functional point of view the solution is no different from a web application, though it’s one that doesn’t live in a particular data centre.

An alternative approach is for the owner of the oil to create oil certificates, where each oil certificate represents a claim on a barrel of oil, before the oil is loaded onto the tanker. We might even order the certificates, number them, with lower numbers redeemed first, passing the risk of the tanker comes up short when it pulls into dock onto the higher numbers (which means that they’ll trade for lower values).

We can track the provenance of these certificates to determine ownership – as with the property example. They can be advertised in any sales venue the owner chooses. Indeed, this approach facilitates competition between markets, rather than enabling one market to extract a monopoly rent. Or the asset can be kept private, if we’re willing for the certificate issuer to act as a central clearing house. Oil can then be traded and settled in at the same manner as the property example.

As the tokenised asset is not tied to a single venue – the blockchain that it was created on – it’s possible for settlement to combine a number of different unrelated assets – swapping property for gold or oil, for example – without requiring them to live on the same blockchain or forcing counterparties to break the exchange into a complex sequence of transactions mediated by oracles.

If we standardise the meta-data used, using something like Open Provenance, then we can even create exchanges that combine otherwise unrelated assets. A settlement process could swap property for oil (or whatever asset you choose to include), without the need denominate everything in terms of some common currency. This could be used as the foundation for a generation of more complex and much more interesting distributed solutions.

The important point to note, though, is that we don’t need to use a blockchain to tokenize an asset other than in some corner cases, such as Bitcoin.

 

Ross Hancock also worked on this blog.

More about the author

Peter Evans-Greenwood

Peter Evans-Greenwood

Fellow, The Centre for the Edge Consulting

Peter is currently a fellow at The Centre for the edge - helping organisations embrace the digital revolution through understanding and applying what is happening on the edge of business and society.Peter has spent 20 years working at the intersection between business and technology. These days he works as a consultant and strategic advisor on both business and technology sides of the fence.