What Kind of Blockchain Does Finance Really Need When Throughput Is No Longer a Bottleneck?

Today, transaction speed has become the industry baseline, and more stringent tests are on the horizon: how can blockchain meet the mature standards of transaction infrastructure required by financial institutions?

Written by: a16z Crypto

Compiled by: Luffy, Foresight News

Finance is the most intuitive application scenario for blockchain, and it is also one of the most demanding. In the context of financial applications, a major advantage often cited for blockchain is that it incorporates assets, ownership records, and execution rules into a shared system not controlled by a single trading counterparty. However, this does not mean that blockchain can naturally adapt to the scale and constraints of the global financial market.

For many years, discussions around the implementation capabilities of blockchain have focused on cost and throughput, specifically the number of transactions a single public chain can process per second. Without sufficient performance limits, financial applications cannot achieve scalable implementation. However, as the throughput gap between many commercial systems continues to narrow, the focus of industry discussions is shifting.

As financial institutions begin to execute transactions on-chain, issue stablecoins, and tokenize assets, they need to confirm whether these networks can truly meet the business needs and expectations of real financial markets. This includes stable and reliable access capabilities, predictable transaction execution rules, and controllable permissions regarding when sensitive information is disclosed. Even in extreme situations of network congestion, service interruptions, or attacks, these safeguards must continue to function.

Blockchain developers have been working on these challenges for many years. This article will outline the technical difficulties that have been overcome, the unresolved legacy issues, and why today’s blockchain is one step closer to supporting the financial market.

Predictability: Ensuring Certainty in Transaction Onboarding and Ordering

For a long time, performance limits have been the most intuitive development bottleneck. Insufficient transaction processing capacity, frequent network congestion, and high fees have made it difficult for many real financial scenarios to be implemented. Once market activity increases, routine operations like placing and canceling orders become costly or the execution results become unreliable.

In recent years, especially during the last infrastructure upgrade cycle, this constraint has improved. Across the industry, the overall throughput of blockchain has significantly increased over five years, with some commercial networks now capable of processing tens of thousands of transactions per second. However, this is just the beginning; current stress tests and new architectural designs will push performance metrics to even higher levels.

But new problems arise: relying solely on throughput does not determine when a transaction is ultimately completed on-chain. The predictability of transaction execution has become a new core issue.

Transactions need to be confirmed within a predetermined time window and adhere to rules that market participants can know in advance. The value of many operations in financial scenarios exists only within a very narrow time frame. For everyday payments, a one-second execution delay may be inconsequential; but for trading markets, a millisecond difference can determine whether participants trust this transaction infrastructure. For example, in an on-chain order book, if a trader’s cancellation instruction is confirmed too late, other market participants can complete transactions based on outdated prices. Market makers will factor this risk into pricing, widening the bid-ask spread, ultimately leading all traders to face worse execution prices.

In simple terms, predictability requires two layers of assurance: first, transaction admission assurance, which is the ability to resist censorship; second, transaction ordering assurance.

Resilience: Stable and Reliable Transaction Execution Channels

Even with a complete ordering rule, if a single entity can still control the execution permissions of transactions, risks remain. Imagine if every second at the New York Stock Exchange, a randomly staked node decides transaction admission. Clearly, such nodes would have a strong market manipulation ability over traders within the exchange.

The ordering mechanism determines the sequence of a transaction on-chain relative to other orders; while the interference resistance needs to address the preceding link: whether participants can successfully submit orders without relying on a single gatekeeper or operator—the latter’s decisions could affect pricing, trading risks, and even the final execution results.

For this reason, blockchain developers are striving to build stronger certainty guarantees, pursuing immediate confirmation of transactions rather than “final confirmation.” The goal is that if a compliant transaction arrives at the network on time, it can be immediately packaged on-chain. This standard is far stricter than “your transaction will eventually be confirmed,” and this is precisely the capability needed by financial markets.

Our researchers have proposed and defined the concept and measurement standard of “strong chain quality,” ensuring that the execution permissions for the next round of transactions are not monopolized by a single operator’s queue. The underlying logic is that blockchain can reserve a portion of space in each block to accept transactions submitted by other nodes in the network, providing multiple on-chain paths for compliant transactions. Ultimately, this reduces a single operator’s control over the transaction packaging order.

Our goal is to make the block space a public infrastructure with multiple admission channels, rather than a single queuing line. Achieving this requires more complex underlying logic than current commercial protocols, but the benefits are sufficient to match the R&D investment. Even under market pressure, participants can be assured that their transactions will be executed smoothly.

Ordering: Clear Rules Understandable by Market Participants

Predictability also depends on how the trading platform decides the order of transactions. When multiple transactions are ready to be executed, the market needs a clear set of ordering rules.

The vast majority of blockchains do not complete the final confirmation of transactions one by one but package transactions into “blocks,” which are written into the ledger by block builders or proposers. In many current systems, this entity has significant autonomy over the selection list and ordering of transactions. This design aims to simplify protocol logic and is acceptable for some application scenarios. However, in financial markets, transaction ordering directly determines execution priority, final execution price, and the fairness of the entire trading process.

Furthermore, if a single entity controls the packaging permissions of the next block, it may see transaction information ahead of others, seizing market opportunities. In the crypto industry, this phenomenon is often referred to as Maximum Extractable Value (MEV): operators extract additional profits beyond basic block rewards by selecting, excluding, or reordering transactions within blocks. A sandwich attack is the most typical case, where a trader monitors pending transactions and places orders to drive prices before completing a reverse transaction for profit. In institutional trading scenarios, the logic is even more straightforward; the party controlling the flow of transactions may make trading decisions based on order information earlier than everyone else.

This mechanism flaw is highly similar to existing problems in traditional financial markets, where centralized intermediaries profit from privileged information regarding order flow and trading intentions. If on-chain markets cannot provide strong guarantees on transaction ordering and execution rules, they will ultimately replicate the same information asymmetry issues.

Blockchain protocol designers are developing clearer ordering mechanisms, mainly including deterministic rules based on priority fees and business rules customized by trading platforms. The exact implementation plans are still in the academic research stage, but their core logic is consistent with traditional finance: market participants have clear expectations for transaction ordering and execution, aligning with the price-time priority execution mechanism used by traditional exchanges.

Transaction admission assurance and ordering mechanisms must work in tandem. If bidding orders and clearing instructions can be artificially delayed before the ordering rules take effect, then even the fairest ordering logic is meaningless. Similarly, if users cannot predict the processing rules after a transaction is on-chain, the value of immediate confirmation will also be greatly diminished.

High throughput is merely a foundational condition for on-chain finance; only when paired with predictable transaction admission mechanisms and clear ordering rules can on-chain finance be deployed in time-sensitive, real market pressure environments.

Privacy: Locking Transaction Intent Before Execution

The order information of traders exposes their trading judgments and operational intentions. For example, the order size and direction can reflect whether a fund is building a position or exiting. If the order is publicly disclosed before it is fully executed, other traders have the opportunity to manipulate prices for profit.

This is particularly prominent in on-chain scenarios. Many blockchains directly disclose the data of pending transaction pools before the final confirmation of transactions. For financial businesses, even if privacy protection cannot be achieved after a transaction is completed, privacy protection before execution is a necessity.

The market can remain transparent and open, but there is no need for all participants to view others’ unexecuted orders in real-time.

Current protocol designers are focusing on protecting data privacy during the window period when transaction information can still be arbitraged, while supporting diversified privacy disclosure modes after execution. After a transaction is completed, the disclosure rules can be flexibly adjusted: the market can publicly display complete transaction records; interbank transactions can only open permissions to counterparties and regulatory agencies. Third parties that cannot view all details can still verify that the entire transaction complies with business rules.

To this end, protocol designers are exploring various encryption solutions, such as fully homomorphic encryption and encrypted transaction pools. These solutions can hide transaction content until the deadline arrives or the committee confirms the final ordering position of the transaction within the block. By the time the market can see this transaction, the window for arbitrage has already closed.

The market can balance transparency without having to open unexecuted orders to everyone in real-time. On-chain finance also needs to implement this differentiated logic, supporting audit verification after execution while strictly protecting transaction confidentiality before execution.

If blockchain wants to become the core infrastructure of the financial market, it must build a transaction system that market participants can trust sufficiently.

Throughput is just the first test. When performance is insufficient, on-chain financial applications will always be limited to niche scenarios. Today, transaction speed has become the industry baseline, and more stringent tests are on the horizon: how can blockchain meet the mature standards of transaction infrastructure required by financial institutions? These standards include predictable execution mechanisms, unambiguous execution priority rules, interference-resistant transaction access channels, and privacy protection before execution, while retaining audit transparency after execution.