The Future of Fintech: Part 2

Part 1 can be found here if not read.

The Future of Payments

Public blockchains merge the rail layer into the vault layer, significantly simplifying the current complexity of payment networks. Current debit/credit transactions typically require four middlemen: the payment processor, the issuing bank, the payment network, and acquiring the bank (along with nine steps of interactions between these parties). International bank wires typically require four as well: the sender’s bank, the receiver’s bank, and two correspondent banks (along with four steps of interactions between these parties). Blockchains reduce the intermediaries in modern payment rails from approximately four to one, the chain itself. In turn, this reduces payment processes that previously took four to nine steps down to two: the payer signs the transaction and the blockchain executes the transaction.

Scalability

Scalability can be measured via two components: clearing/settlement speed and transaction throughput. Today, blockchains can clear and settle payments significantly faster than legacy rails. However, blockchains’ transaction throughput is currently limited by two parameters that blockchains optimize for security rather than scalability: block space and block time.

  • As block time increases, validators have more time to verify the correct state of the chain (more time allows for more communication among nodes), which increases a blockchains’ security. However, extended block times limit how frequently blocks (and therefore transactions) can be processed.

Layer-Two: A Brief Overview

The horizontal scaling provided by layer-two’s will enable public blockchains to process pseudo-anonymous transactions on par with current rails’ capabilities and private transactions more efficient than current rails’ international capabilities, while maintaining fees that are still orders of magnitude lower. Within two years, layer-two solutions will be able to process fully private transactions within the same realm as current rails’ capabilities, so long as innovation in the zero-knowledge proof space continues to improve at rate of at least 2x (in terms of both generation time and compactness of the proof). Additionally, if layer-one solutions successfully implement sharding, the combination of layer-twos on top of sharded layer-ones will allow for an order of magnitude improvement (or two) over today’s current system in both speed, fees, and privacy (within two years).

Image 1: A modern blockchain scaled through competing layer-two solutions

Sharding: A Brief Overview

Sharding is very similar in design to layer-twos: both horizontally scale blockchains, relying on an increase of validators rather than an increase in the size required to run validators; both are a set of individual blockchains that use a base chain to shuffle their validator sets and verify the correctness of their state. Layer-two solutions can be termed as ‘free market’ versions of sharding due to their implementation as smart contracts on their corresponding layer-one chains: no developer or validator is forced to use or validate a layer-two solution; they must explicitly choose to, whereas developers and validators will be forced to adopt to a sharding environment should one be implemented.

Image 2: A sharded public blockchain
Image 3: A neo-blockchain scaled through a sharded layer-one chain

Privacy

Layer-ones can currently generate ~2 private tx/sec through on-chain mixers. Mixer technology is still nascent, but numerous beta versions are live on public blockchains, one of which is under development by Ernst & Young, one of the largest global accounting firms.

Security

The security of blockchains can be analyzed via two sub-factors: the security of the chain itself (as well as its parts, such as layer-twos) and the security that, even if the chain operates correctly, users are not liable to the consequences of a fraudulent transaction.

Factor One: The Security of the Chain Itself

The security of blockchains rely on both technical security and social security. Technical security is achieved through three disciplines: cryptography, economics, and computer science. Cryptography ensures the validity of all transactions: users send transactions using cryptography; nodes package and verify blocks using cryptography. However, altruism is ineffective in ensuring that every node applies cryptography correctly. Thus, blockchains implement economic incentives to ensure that a decentralized community of actors append blocks to the chain that only contain valid transactions: validators receive block rewards for honest behavior and slashing penalties (directly or indirectly, depending on the consensus algorithm used) for dishonest behavior. All of a blockchain’s cryptographic and economic features are enforced through code, computer science, run by nodes of that blockchain.

Zk Rollups

Zk rollups garner their security properties from validity proofs, which leverage cryptography alone to prove correctness. This feature prevents zk rollup validators from posting invalid transactions, which in turn allows for users to exit from a zk rollup back to its correspondent layer-one in the layer-one’s next block (~13 seconds in Ethereum’s case). Some zk rollup teams have implemented economic incentives to enhance zk rollups’ scalability. Economic incentives in zk rollups allow for sub-second, guaranteed transactions (at the scale of 2000 tx/second). However, there is a strong difference between enhancement via incentives to perform optimally and a reliance on incentives to perform securely, as in the case of optimistic rollups.

Optimistic Rollups

Optimistic rollups rely on two blockchains for security: the rollup and its corresponding layer-one. To provide adequate security assurances through economic incentives, optimistic rollups implement challenge periods (via their smart contracts on layer-ones), where validators can challenge an optimistic block by locking funds in a smart contract. Challenges reward honest actors and punish bad actors (through distributing either the block proposer’s stake or the challenger’s stake to the honest party, similarly to layer-one proof-of-stake chains ‘in-protocol’).

Table 1: Comparison of Rail Solutions

Factor Two: Anti-Fraud Solutions

The other significant aspect of security remains the same regardless of which layer-one or layer-two solution is used to transact: public blockchain infrastructure must combat fraudulent payments.

Security of Centralized versus Decentralized Systems

Any system, whether centralized or decentralized, is composed of multiple parts, and each system type is only as secure as its least secure part. Decentralized systems inherently account for this in their designs. They build resilience into each part, with each part treating every other part as adversarial. Thus, if a hacker successfully attacks a part of a decentralized system, the hacker can only compromise that singular piece of the system. As decentralized systems grow in their number of parts, the security of each part remains the same.

Triggering the S-Curve

Blockchains will eventually deliver more scalable, more private, and more secure payments than today’s payment infrastructure, but it will take multiple years before these three features can be optimized simultaneously, as teams continue to innovate on horizontal scaling solutions and zero-knowledge proof efficiency.

The Future of the APIs

As blockchain-based banking/payment solutions proliferate, the APIs of today’s financial stack will be rendered irrelevant. The open-source ethos dominating public blockchain infrastructure renders obsolete both categories of current fintech APIs (rail APIs and data APIs).

  • If the user utilizes zero-knowledge proof based privacy solutions (such as on-chain mixers or a recursive zero-knowledge proofs), the user could store their encrypted transaction history on a decentralized storage protocol, such as Filecoin, and authorize access to digital wallets through a proxy re-encryption protocol, such as NuCypher.
Image 4: Future fintech infrastructure without sharding
Image 6: Future fintech infrastructure with sharding

Conclusion

There are three layers of fintech — the vault, the rails, and the APIs. Each of these layers will be disrupted by decentralized finance. While each of these three layers possess numerous moats, history serves as a reminder that all moats are temporary. For example, telecommunication companies also benefited from significant regulatory and network moats, and yet they could not do anything to prevent the advent of the internet. The largest open question that remains is whether layer-one chains will interface with traditional vaults (through stablecoin issuers’ DBSs), federal governments’ proprietary chains (through central banks’ CBDCs), or remain fully sovereign (through a decentralized communities’ FCOS). Blockchains leave little room for the incumbents of the payment and API layers, other than their becoming providers on top blockchain-based protocols (such as a validator of layer-two rollup or a ‘re-encryptor’ in a proxy re-encryption protocol).

Partner at Bizantine Capital