a16z Crypto: Four Key Capabilities for Blockchain to Evolve into Financial Infrastructure

Author: a16z Crypto

Compiled by: Jiahua, ChainCatcher

Finance is one of the most obvious application scenarios for blockchain and also one of the scenarios with the highest infrastructure requirements. When discussing the financial uses of blockchain, people often mention one of its advantages: it can incorporate assets, ownership records, and execution rules into a shared system that is not controlled by any single counterparty. However, this does not mean that blockchain is inherently capable of meeting the vast scale and complex operational requirements of global financial markets.

For many years, discussions about whether blockchain is suitable for large-scale applications have mainly revolved around cost and throughput. Throughput refers to how many transactions a chain can process per second; without sufficient processing capacity, large-scale implementation of financial applications is out of the question. However, as the gap in throughput among many formally operating systems gradually narrows, the focus of the discussion has begun to shift.

As financial institutions start executing transactions on-chain and issuing stablecoins and tokenized assets, they need to assess whether these networks can meet the operational requirements of real financial markets. This includes not only stable and reliable transaction access and predictable transaction processing rules but also the ability to decide whether and when to disclose sensitive information. Even in the face of network congestion, system interruptions, or attacks, these guarantees must not fail.

Blockchain designers have been working on these issues for years. This article will discuss which problems have made progress, which remain unresolved, and why today's blockchains are more capable of supporting financial markets than those of the past.

Predictability: Ensuring Transactions Are Included and Ordered According to Rules

For many years, processing capacity has been the most obvious bottleneck for blockchain. Fundamental issues such as insufficient transaction processing capacity, network congestion, and high fees have made it impossible to implement many real financial scenarios. Without sufficient processing capacity, routine operations like placing or updating orders become costly and unstable once network activity increases.

In the past few years, especially during the last round of infrastructure upgrades, this limitation has been significantly alleviated. Over five years, the total throughput of the blockchain industry has increased by more than 100 times, and some formally operating blockchain systems can now process tens of thousands of transactions per second. This is just the beginning; as stress tests progress and new architectures are implemented, this number may continue to rise.

But the new problem is that throughput alone does not indicate when a transaction will actually be included in a block. This is precisely the issue that predictability needs to address.

In financial markets, many operations are only valid within specific time windows, so transactions must be included in blocks on time and processed according to rules that participants are clearly aware of in advance. For everyday payments, a second of uncertainty may be inconsequential; but in financial markets, even millisecond-level uncertainty can shake participants' trust in the trading venue.

Take on-chain order books as an example: if a trader's cancellation instruction does not take effect in time, others may execute trades based on outdated quotes. Market makers will factor this risk into their quotes, widening the bid-ask spread to protect themselves, ultimately forcing all participants to face worse execution prices.

In other words, predictability requires two types of guarantees: first, the guarantee that transactions can be included in a timely manner, which is anti-censorship; second, a clear guarantee of transaction ordering.

Resilience: Ensuring Participants Can Reliably Submit Transactions

Even if the ordering rules are well-designed, what happens if a single entity can still control who can submit and execute transactions? Imagine if the New York Stock Exchange randomly selected one node from global participants every second to control all transactions entering the market. Clearly, these nodes would gain enormous power over transaction execution.

Transaction ordering determines the sequence of orders once they are on-chain, while resilience focuses on the earlier step: can participants actually submit orders? Must they rely on a single entry point or operator, such that the latter's decisions are sufficient to influence prices, risks, or transaction outcomes?

Therefore, blockchain developers are seeking stronger guarantees, shifting the focus from "transactions will eventually be included" to "can transactions be included in a timely manner." The goal is that as long as a valid transaction reaches the network in time, it should be included immediately, rather than merely receiving a promise of "it will be processed soon." This is much stricter than "transactions will eventually be processed," and it is this certainty that financial markets require.

Researchers at a16z Crypto have proposed the concept and measurement method of Strong Chain Quality, which centers on ensuring that participants' chances of entering the next round of execution no longer depend on a single queue controlled by an operator. Specifically, blockchains can reserve a portion of space in each block to process transactions submitted from other paths in the network, providing multiple entry points for valid transactions and thus reducing a single operator's control over which transactions will be prioritized.

The goal is to make block space a shared infrastructure with multiple entry points, rather than a queue controlled by a single entity. This requires more complex protocol designs than currently operating systems, but the added complexity may be worth it: even under significant network pressure, participants can trust that their transactions will be processed normally.

Ordering: Allowing the Market to Anticipate Transaction Sequence

Predictability also depends on how the trading venue determines transaction order. Once transactions enter a pending state, the market needs a clear set of ordering rules.

On many blockchains, transactions are typically not completed one by one for final confirmation; instead, they are first aggregated into batches, known as "blocks," which are then written into the ledger by block builders or proposers. In many current systems, block builders or proposers can largely decide which transactions to include and how to order them. This design simplifies the protocol and may be acceptable for certain applications. However, in financial markets, transaction order directly determines who gets priority, at what price trades occur, and whether the execution process is fair.

More importantly, if one party controls the next block, they may see transactions earlier than others and take action. In the crypto market, this is often referred to as MEV, or the value gained from including, excluding, or reordering transactions in a block beyond the regular block rewards.

When explaining MEV, people often use the example of a "sandwich attack": a trader sees a pending transaction and executes a trade before it to push the price, then immediately completes another trade after its execution to profit from the price change. For institutional markets, the issue is even more direct: whoever controls the order flow may act before other participants, using that information to make decisions.

This is very similar to the flaws in existing financial market structures. In traditional markets, centralized intermediaries may profit from the privilege of receiving order flow and trading intentions first. If on-chain markets cannot provide strong guarantees for transaction ordering and execution rules, they may recreate the same information asymmetries.

Blockchain protocol designers are exploring clearer ordering rules, mainly including deterministic rules based on priority fees and rules set by specific trading venues according to their applications. The specific mechanisms remain an active area of research, but their basic principles are not unfamiliar to traditional finance: market participants should be able to accurately anticipate how their transactions will be ordered and executed, similar to the price priority and time priority principles used by traditional trading venues.

Transaction inclusion and ordering must work in tandem. If a bidding quote or clearing transaction is selectively delayed before entering the ordering phase, then even the fairest ordering rules are of no use. Similarly, if participants cannot determine how their transactions will be handled after inclusion, then the significance of timely inclusion itself will be greatly diminished.

High throughput makes on-chain finance possible; only predictable transaction inclusion and clear ordering rules can ensure it is truly usable under time-sensitive and market-pressure conditions.

Privacy: Protecting Transaction Intent Before Execution

A transaction may expose information or intentions held by participants. For example, the size and direction of an order may indicate whether a fund is building or reducing a position. If this information is disclosed before the order is fully executed, other traders may trade ahead and influence the price.

This is especially important in an on-chain environment, as many blockchains will disclose pending transactions before final confirmation. Therefore, even if post-execution information cannot be kept completely confidential, financial markets still need to protect privacy before execution.

Current protocol designers aim to keep the specific contents of transaction information confidential during the phase when it may still be exploited by others, while supporting more diverse privacy models after transaction execution. After a transaction is completed, the rules for information disclosure can change accordingly: public markets can disclose complete records, while banks can only disclose relevant records to counterparties and regulators. Even if others cannot see all the details, they should still be able to verify whether the transaction complies with the rules.

To this end, protocol designers are exploring different types of cryptographic technologies, such as timelock encryption and threshold encryption. These technologies can keep transaction contents hidden until a predetermined deadline ends or a committee confirms that the transaction's position in the block has been established. This way, even if the market sees the transaction at that time, it can no longer be front-run.

The market can remain transparent without requiring every participant to view others' unexecuted orders in real-time. On-chain finance also needs to draw this line: auditable after execution, but confidential before execution.

If blockchain is to become core financial infrastructure, it must be able to support a market that participants can trust.

Throughput is just the first test. Without sufficient processing capacity, on-chain finance will struggle to move beyond niche applications; but once speed has become a fundamental capability, the real challenge is whether blockchain can meet financial institutions' existing requirements for market infrastructure, including predictable transaction execution, clear and unambiguous priority rules, stable and resilient transaction access, and protecting information before execution without sacrificing post-execution transparency.