Comparing the Circle IPO public filing to the Chia Network IPO confidential filing (but now available on the crypto task force site), and particularly the glossaries contained in each set of Form S-1 filings, it becomes quickly apparent that investors should care about the technological details.
Blockchains are Blockchains
Circle starts with their definition of “blockchain” or “blockchain networks”. They choose to use the neutral “blockchain networks” rather than the generic term, which is “distributed ledger technology”. Why? If they said “DLT” or “distributed ledger technology,” people in the industry would’ve booed at it, hooted and hollered, and wrinkled their noses at it. But this is how they describe it:
They also list in their business section the “blockchain networks” they support:
Whereas Chia describes the term “blockchain” in detail. In deep, technical detail:
I point this out because the level of detail matters a lot.
The Challenge of the False Equivalence
Investors and consumers often think that the same word means the same thing or at least an equivalent. But in this case around the word “blockchain”, these definitions are NOT.
A blockchain is literally a sequential set of transactions. Timestamped blocks that contain spends, one after another. Quite literally, this is a chain of blocks. But a “blockchain network” as defined by Circle can include DLT like distributed acyclic graphs (DAG), hashgraphs, application chains (“app chains”), and shards. Think the Provenance blockchain that Figure Markets builds on (a DAG), Hedera (a permissioned hashgraph), Cosmos (a set of app chains), or where Ethereum (proto dank sharding) wants to go.
These graphs and app chains enable parallel chains that enable parallel processing. That sounds great, but what happens if you double spend across shards/graph segments? Or you didn’t intentionally double spend, but because the graphs or shards actually record different sequences that only later get reconciled or settled at L1 (if at all that is), then imagine how that could affect a transaction.
This is all nice and cute when we’re playing with funny money. But it isn’t so cute when the blockchain becomes real financial rails and assets like your home, your retirement account, and more are held on chain. Imagine the chaos if the timestamps were wrong, or worse, the sequences got messed up. Transactions could be invalidated altogether, people could double spend an asset to you that then gets settled incorrectly, or you become the victim of a complex web of MEV. At Chia, we are insistent that blockchains are blockchains, and anything else that doesn’t record transactions on a single ledger with accurate timestamps is not. As Gene Hoffman, CEO of Chia Network Inc, says, they are “mockchains.”
Keys
Circle describes a cryptographic key:
And they describe private keys as:
At Chia Network, we define the term private key to include the details of public and private keys:,
And this distinction matters because public keys are good identifiers for wallets and users, while private keys are what you have to protect. The corresponding risks around cybersecurity and privacy are different for each. Which means that Circle’s risk factors may not be granular enough to truly describe the risks.
Interoperability and Security Status
Circle provides their definition of interoperability, but they don’t describe how interoperability works.
But the definition matters.
- Is it a layer 0?
- Is it a bridge, which often is a centralized/multisig smart contract (and may be an unlicensed broker)?
- Is it a messaging protocol?
This matters where a “peer to peer” transaction is between USDC and a security as the swap. What if the bridge is centralized for example? An unlicensed broker-dealer is one way to describe that bridge. Or what if an AMM or centralized exchange allows USDC to be swapped with securities? The AMM, like a centralized exchange, may need to register as an exchange.
To their credit, Circle drafted the following risk factor with the assumption that USDC isn’t a security (which I agree with). However, what happens if USDC is made available in centralized systems that also list securities for trading?
That couldn’t possibly happen, right? Well, they support Polygon as a native L2 blockchain, which is controlled by a multisig administrative set of keys. So, that blockchain system is centralized, controlled, and possibly allows securities to trade there.
Circle is just passing the problem and liability downstream.
Consensus Mechanism
And of course, since Circle is primarily issued on Proof of Stake (PoS) chains, they use this definition:
Customers and investors are asked to make the assumption that PoS is okay, without a description of how it works, social slashing, and how PoS lacks much of the necessary security capabilities and technology.
Circle completely omits other consensus mechanisms. They are generally irrelevant to Circle issuances and users, but they might be relevant to Circle investors.
Compared to Chia’s definition of its consensus mechanisms in great detail:
The Chia blockchain isn’t even a PoS chain, yet we accurately describe the risks inherent in issuing assets on such a blockchain. For example, how finality works which leads to the long range attack:
It’s almost as if Circle doesn’t want investors to understand how insecure PoS actually is. And while that can pass muster for poorly executed marketing materials, it’s not acceptable for an S-1, where a prospectus is a CYA document and insulates from legal liability.
In fact, the elephant in the room is just how centralized some of these “decentralized” systems are.
This is marketing speak–a new internet financial system “reducing society’s dependence on centralized intermediaries.” But the issuance of USDC or EURC by Circle, which are revocable assets means you have to trust Circle, further centralizing the blockchain. This is not much different than simply trusting your bank. But you now also have to trust that your smart contracts are actually decentralized. (Spoiler: most are not)
Not to mention that a lot of EVM based systems aren’t very decentralized because they rely on centralized infrastructure like Fireblocks and AWS, relayers which are VERY centralized, and more.
Am I nitpicking? Yes, I am. But I’m nitpicking on specific details that matter. The average investor will read the Circle S-1 and think “This is Great!” But they may not realize where there are landmines and risks. There is an actual information asymmetry here, so caveat emptor.
And this is why we did such a thorough job on our S-1 to describe the technology (and risks) in full–so investors could truly understand what’s underlying the network.