Antony Lewis is a bitcoin and blockchain consultant, who previously served as the director of business development at bitcoin exchange itBit.
In this article, Lewis attempts to describe two very different uses for blockchain technology and why the division of these two functions is increasingly of interest to enterprise firms.
For 90 minutes, we heard from an all-star panel: Blythe Masters, CEO of Digital Asset; David Gurle, CEO of Symphony; Daniel Gallancy, CEO of SolidX; and Tim Grant, head of R3's Lab and Research Centre.
One of the panelists said something that got me thinking – using blockchains for what she called 'proof of satisfaction', ie evidence of agreement or assertion that something happened to the satisfaction of the parties involved.
It seems that two types of blockchain* or distributed ledger are emerging: blockchains as digital token ledgers, and blockchains as registers where proofs of activities (transactions, bets, etc) are recorded.
These ledgers record ownership and changes to ownership of digital tokens.
There are two distinct types of token:
Digital assets: The token is the asset. The classic example of this is bitcoin. You own a bitcoin, that's it – you can’t take that to anyone and claim the 'underlying' thing. Satoshi Nakamoto, the creator of bitcoin, doesn’t own a bunch of gold he promises to give you if you wave a bitcoin at him. (Neither do central banks for fiat currencies, but that’s for a different post).
A bitcoin is your asset, and there is no corresponding liability owed by someone else.
Digital claims: The digital token is a claim for something against someone else who promises to give you the actual asset upon presentation of the token. It's a debt token, or an IOU. It’s your asset, but their liability.
The classic example is user-issued tokens on Ripple, where "gateways" (eg banks) issue tokens representing underlying assets (eg gold) or issue them in place of existing liabilities (eg currency deposits).
These tokens can be passed around to different owners, where changes to ownership are recorded on the shared ledger, and where owners ultimately need to come back to the issuer to claim the underlying thing, rather like claiming your coat by presenting your cloakroom ticket at the theatre.
(Note: user-issued tokens on Ripple are not to be confused with Ripple’s underlying token XRP, which actually is an asset).
2. Activity registers
This second class of blockchain contains entries that feature data, usually fingerprints of data called hashes, possibly stored with some other non-sensitive metadata. The data can represent anything, but in the financial services context, they might be trading facts about transactions negotiated and agreed "off-chain" via another channel.
This keeps the commercially sensitive secrets off the industry-shared ledger, while leaving an indelible mark on it that can be referred to later if needed.
- Two banks agree on a trade, with various trade facts (dates, price, amounts, etc). They can agree via any mechanism such as telephone, email, click and trade, or carrier pigeon.
- They both record the trade details in some very standardized way (standards are important here as pedantry is key when creating and agreeing hashes).
- They calculate and agree the hash of the trade facts.
- The hash gets submitted onto some sort of timestamped industry blockchain, cryptographically signed or agreed by one or more parties.
Note that the hashes themselves can't be backwards-engineered into the trade fact data if they are well constructed. But in the future, either party can take their copy of the trade facts and show it to anyone interested and point to the hash on the blockchain which proves:
- These are the original trade facts – or at least these trade facts existed at the timestamp time.
- Whoever signed the hashes, must have also agreed to the trade facts (if there was some some sort of requirement for both parties to put a signature against the hash in the first place).
This concept can also be used to store data such as images, newspaper headlines, identity information, etc.
Which blockchain should activities be registered on?
Private blockchain supporters prefer to store proofs of activity in databases visible to only known participants, with the comfort that all writers are known entities, and censorship may be possible, if required, and it is also cheap to write on a private blockchain.
Public blockchain supporters would argue that if all you’re doing is putting hashes somewhere without giving away any trade facts, then why not use bitcoin’s blockchain as it is the most tamper-resistant and battle-tested chain, and the cost of transactions can be kept low by bundling data and committing the bundle at once – using techniques like Merkle trees.
Personally, I think the industry needs more differentiated terminology, or to at least be clear on what we are talking about. I would suggest "distributed ledgers with asset tokens", "distributed ledgers with claim tokens" and "distributed activity registers".
Not as sexy as ‘blockchain’, but probably more clear.
Blockchains seem to be being used for very different uses: recording ownership changes of digital tokens, and recording activity with timestamped proofs of existence of data or agreements about data.
Bitcoin has a blockchain which contains:
- Asset tokens (BTC) – this was the intended use
- Claim tokens in the form of volored coins etc – not really the intended use
- Hashes of data – not really the intended use.
Ethereum has a blockchain which contains:
- Asset tokens (ETH or ETC depending on which ethereum!)
- A multitude of 'user generated' tokens that are created and managed through smart contracts
- Hashes stored in smart contracts.
Ripple is a distributed ledger which contains:
- Asset tokens (XRP) used mainly as an anti-spam mechanism to make each transaction cost a tiny amount
- Claim tokens (gateway-issued tokens) whose transmission from user to user is the main purpose of Ripple
- Hashes of data – not really the intended use
Private Industry Activity Registers
These may be distributed or not, but they are databases that contain:
- Immutable(ish) records of hashes of private activities between participants
- Usually cryptographically signed by known participants for non-repudiation.
The distributed ledgers may not include records of real final settlement of actual digital assets eg stock ownership unless the token is the asset a la bitcoin or a real central bank-issued digital currency.
In the private blockchain space, there are still privacy issues around letting your competitors know what you own.
This article originally appeared on Lewis's blog Bits on Blocks, and has been republished here with his permission.
Chicken and egg image via Shutterstock
Disclaimer: The views expressed in this article are those of the author and do not necessarily represent the views of, and should not be attributed to, CoinDesk.
'Arrival' Trailer: Amy Adams Can Talk to Aliens
Remaining Time -0:00
This is a modal window.
Foreground --- White Black Red Green Blue Yellow Magenta Cyan --- Opaque Semi-Opaque
Background --- White Black Red Green Blue Yellow Magenta Cyan --- Opaque Semi-Transparent Transparent
Window --- White Black Red Green Blue Yellow Magenta Cyan --- Opaque Semi-Transparent Transparent
Font Size 50% 75% 100% 125% 150% 175% 200% 300% 400%
Text Edge Style None Raised Depressed Uniform Dropshadow
Font Family Default Monospace Serif Proportional Serif Monospace Sans-Serif Proportional Sans-Serif Casual Script Small Caps