2024 Web3 Sector Analysis, a perspective from Asia's largest Crypto VC Hashkey

As one of the most active crypto VCs, HashKey Capital regularly analyzes every Web3 sector internally. In the New Year of 2024, we “open source” our internal sector analysis and insights as our contribution to the industry.

The authors of this article include (in alphabetical order): Arnav Pagidyala, Harper Li, Jack Ratkovich, Jeffrey Hu, Junbo Yang, Stanley Wu, Sunny He, Xiao Xiao, Yerui Zhang, Zeqing Guo.

In 2023, the ZK (Zero-Knowledge) track underwent significant expansion compared to previous years. It extended its scope beyond its traditional applications, such as scaling and cross-chain functionality, and ventured into various specialized scenarios while diversifying into distinct tracks.

zkEVM

In the context of zkEVM (Zero-Knowledge Ethereum Virtual Machine), several significant developments have occurred, categorized into type0, type1, and type2. Each category presents its unique set of characteristics and challenges:

Type0: This category remains closely aligned with Ethereum, sharing full equivalence. However, it faces technical hurdles, particularly in terms of block production speed, release schedules, and state verification.

Type1: Type1 has made notable enhancements and compromises based on the Ethereum Virtual Machine (EVM) foundation. It stands out for its improved overall application experience and compatibility with EVM opcodes.

Type2 and others: These categories have launched their mainnets earlier and are actively developing their respective application ecosystems. The specific progress and project dynamics vary and require individual analysis. Notable projects include Polygon’s CDK and StarkNet’s full-chain game, each following its unique development trajectory.

zkVM

In the realm of zkVM (Zero-Knowledge Virtual Machine), the predominant technical route currently being pursued is zkWASM, known for its highly scalable architecture. This architecture lends itself well to collaborations with exchanges for the creation of high-performance Decentralized Exchanges (DEX). Notable players in the zkWASM field include Delphinus Labs, ICME, and wasm0, among others.

Moving towards the RISC V architecture direction, RISC0 is exploring new possibilities and stands out for its friendliness towards front-end languages and back-end hardware compared to WASM. However, potential challenges exist in terms of efficiency and proof time. The scope of application scenarios continues to expand, with instances such as Reth simulating the Ethereum execution environment, the utilization of Fully Homomorphic Encryption (FHE) for running environments, and Bitcoin Rollup, to name a few.

Additionally, there’s zkLLVM, where =nil; recently introduced a Type-1 zkEVM based on this technology. This advancement allows for the swift compilation of high-level programming languages into zkSNARK circuits.

ZK Mining

In the ZK Mining sector, the current efficiency levels of GPUs (Graphics Processing Units) and FPGAs (Field-Programmable Gate Arrays) are relatively comparable. However, it’s worth noting that GPUs tend to be more expensive, while FPGAs often serve as prototypes for validation purposes. The landscape may witness increasing distinctions with the emergence of specialized ASIC (Application-Specific Integrated Circuit) chips and evolving requirements related to Fully Homomorphic Encryption (FHE) and other specialized needs.

Furthermore, there has been a significant rise in Prover DAOs, with computational power emerging as the core competitive advantage. Prover DAOs formed by mining machine teams are particularly well-positioned to leverage their combined computational capabilities, giving them a competitive edge in the field.

ZK Middleware

Within the realm of ZK middleware, there exists a diverse array of components, including zkBridge, zkPoS, ZK Coprocessor, zkML, zk Trusted Computing, and more, all serving various scenarios of verifiable computation. Among these:

ZK Coprocessor scenarios stand out with notable clarity, with a majority of projects advancing to the testnet stage.

The zkML track continues to generate significant interest, characterized by ongoing projects that demonstrate diversification in their development progress and competitive landscape.

Furthermore, a novel track in the ZK ecosystem has recently emerged: ZK proof sharing. This involves the collective sharing of proofs within a network, with income distribution occurring after batch processing. This development adds an exciting dimension to the ZK landscape.

MEV

MEV Takeaways:

The focus can be mainly on earlier stages in the txn supply chain. i.e. intents.
Rise in next-gen DEX design/infra that solves LVR and improves LPing, attracting more capital.
Private auctions/mempools will drastically improve the txn supply chain if done efficiently. We will be following developments in FHE, MPC and ZKPs
Most systems today use centralized relays, permissioned solvers, trusted builders etc. We believe the end game will be permisionless to enable the most competitive markets
Where could the MEV supply chain be changed? APS, Execution tickets, PEPC etc.

OFA

2023 marked the rise of orderflow auctions. No longer will high value txns be routed to the public mempool, rather to an OFA which aims to return the value users create. There are various OFA implementations that serve a spectrum of needs around price discovery and execution quality, from an RFQ auction to a blockspace aggregator. Moving forward, we expect to see a growing % of ETH transactions moving through OFAs.

Blockbuilder

As we’ve seen via relayscan, the builder market has centralized around only a few builders, some of which are notorious HFT firms that first serve their own trading needs.

Moving forward, we expect the advantages of HFT firms to reduce as the amount of CEX/DEX arbs will reduce.

Relay

The relay market faces two foundation problems. (1) The market is centralized around a few players, namely BloXroute and Flashbots (2) There is no incentivization mechanism for relays making it an unprofitable business.

Moving forward, we expect to see the rapid development and implementation of optimistic relays. And possible proposals for relay incentivization.

The AA (Account Abstraction) track is mainly divided into two categories: smart contract wallets and modular services.

In the smart contract wallet sector, companies in the AA wallet track are already aligning with the overall wallet market trend. Gaining traffic solely through features is increasingly challenging, and instead, the wallet factory aspect is worth attention.

In modular services, Bundler and Paymaster are essential functions that all basic service providers must offer. These two services have already become standard offerings.

Current trends in this track include:

Most infrastructure is already in place, and the development is relatively stable. Looking at the track’s data, it has entered a phase of rapid growth, with user wallet numbers starting to increase from June, reaching over 6 million Userops by November and approximately 200K MAUs.

AA development on L2 is better than on L1, with EF considering native support for L2.

The issue of DApps not supporting AA remains significant, including challenges in cross-chain and cross-rollup account implementation, requiring new solutions.

Private mempools will converge with MEV and intent, optimizing user experience.

Intents

After gaining attention this year, Intents have developed rapidly, facing issues like solver misconduct and order flow trust, but feasible solutions exist.

For Intents to progress further, considerations around order flow and user acquisition are essential. Architecturally and business-wise, Intents fit well with the MEV and AA frameworks. Builders and Searchers are ideally suited for matching and solver roles.

Telegram Bots are likely to evolve towards Intent, their advantage in order flow giving them significant bargaining power over builders and SUAVE, even exceeding larger wallets.

The DA (Data Availability) track has limited participants, including Ethereum, Celestia, Eigenlayer, and Avail, with varied project progress. The head effect is obvious, leaving little opportunity for the waist and tail. DA projects mainly focus on security (including data integrity, network consensus), customizability and interoperability, and costs. With Celestia’s launch and rising prices, the overall valuation of the DA track has increased. However, DA is essentially a B2B business, closely tied to the quantity and quality of ecosystem projects.

From a client perspective, launching DA on Ethereum is the safest but most expensive solution. Ethereum’s fees have significantly decreased after protodanksharding, so major Rollup projects still prefer Ethereum as their DA layer. Current DA project clients, apart from EigenDA, are mainly Cosmos ecosystem projects, RaaS projects, etc. EigenDA’s unique position, related but not directly tied to Ethereum, may attract more middle-ground clients. Additionally, some early-stage DA projects and specific scenario DAs, like Bitcoin DA, might gain substantial market share in niche sectors.

Rollup Frameworks & RaaS

Rollup market is saturated and needs to evolve. 30+ VC backed RaaS projects + infra providers like Gelato are entering the market. We need to see what use cases are successful on RaaS and which interoperability solutions will work.

Some L2/L3 frameworks, like OP Stacks have already got lots of traction, with strong public goods fundings, and developer adoptions.

Some specific applications, like DePIN, could potentially use Ethereum rollup by using customized execution environments.

There’re still lots of new tech about Rollups, for example: Risc Zero Zeth/other projects could change the way rollups verify state without needing to to relying on the validator or sync committees; FHE Rollup can offer fully generalizable, private DeFi when used alongside ZKP and MPC primitives, etc.

Cosmos

Cosmos Hub will continue to try to strengthen its position in the ecosystem from various aspects in the future. For example, Partial Set Security (PSS) can be more flexible to allow some verifiers to provide inter-chain shared security (ICS) without forcing all Cosmos Hub’s validators to join, which can reduce the pressure on a lot of validators and make it easier to promote; On the other hand, Cosmos Hub plans to enable multi-hop IBC to enhance UX; in terms of protocol implementation, it plans to add the functions of Megablocks and Atomic IBC to provide atomic cross-chain transactions, which can form a unified MEV market, similar to the shared sequencer and SUAVE in the Ethereum ecosystem.

As for the Cosmos ecosystem, the application chain development thesis has been affected by L2 and other development frameworks to a certain extent recently, and the number of new app-chain projects has declined. However, due to its highly customizable underlying framework, it is more flexible, so it is possible to look for new public chain deals with customized modifications as the mainstream narrative continues to evolve.

Security

Projects in the security track have made progress at various levels, offering products that include tools and protocols for on-chain detection and interception, on-chain tracking tools, manual audit and bounty services, development environment tools, and applications of various technical methodologies (such as fuzz testing).

Each tool is more suitable for detecting specific types of vulnerabilities and has specific methods to check for vulnerabilities in smart contracts (static analysis, symbolic execution, fuzz testing, etc.). However, the combination of tools still cannot replace a complete audit.

Besides the different positioning mentioned above, other dimensions to evaluate projects include maintenance and update speed, size of the vulnerability database, carriers, and the actual needs of partners.

The current direction of the Crypto and AI integration mainly includes underlying computational infrastructure, training based on specific data sources, chat tools, and data labeling platforms.

In the computational infrastructure and power network field, various projects innovate at different points but are generally in early stages, and they need to consider sustainable commercial expansion paths beyond creating different types of agents.

Data labeling platforms convert traditional manual annotation services into web3 formats, with the ability to acquire orders being crucial. Also, because low-threshold data labeling is easily replaceable by AI in the future, it is necessary to focus on high-value, high-threshold data to obtain more orders.

Moreover, many new projects combining AI are developing To-C chat tools and others.

As a crypto fund, we will focus more on ZKML, projects with data strengths in crypto sectors, or to-C products that are tightly integrated with AI, rather than infrastructures such as large language model, which we don’t specialize in.

DeFi

One notable change we observed in 2023 is a shift in investor preferences towards seeking real yield products, as opposed to those reliant on emission-based yields. Investors are increasingly interested in products sourced from Liquid Staking (or re-staking) Derivative finance (LSDfi) or Real-World Assets (RWA).

Meanwhile, with centralized exchanges facing significant regulatory pressure and assets needing to find liquidity, there is a tremendous opportunity for DEX, especially with L2 coming online bringing the potential for high-performance applications, which could focus on DEX opportunities on L2.

We also noticed big opportunities in projects that enable non-native users, including institutions, to access Web3 yields. Projects that abstract away the on-chain element and provide a safe environment for non-native users could bring massive amounts of capital into the space.

Perp DEXs, including famous projects like dYdX, GMX, Drift and Jupiter, generate the highest ratio of fees by sector, according to Messari’s report.

Liquid staking has seen consistent growth throughout the year. Nearly 22% of all ETH is staked, with Lido taking 32% of the staked ETH market share. In fact, liquid staking remains the largest DeFi sector, with $20B TVL.

Gaming & Entertainment

Studio

Studio-type projects, based on their own categories, have different characteristics and points of focus, as shown in the table below：

The overall gaming industry can be succinctly categorized as mentioned above, with a significant improvement in overall game quality and the professional level of teams than those in previous cycles. We continue to look forward to studios entering the web3 space, equipped with experience in mature development and operational products, and continuously seeking game producers and KOL founders who are keen learners, sensitive to crypto and communities, and willing to speak out.

In terms of Web3, because of the short history of blockchain gaming at the moment, there is not much mature experience to learn from, so we will put more emphasis on whether the team’s ideas and thoughts are align with web3, as well as learning fast, rather than having rich experience in web3.

We will keep an eye on UGC (User-Generated Content) in the future. Currently, the centralization problems of web2 UGC remain unsolvable. Beyond providing tools for users to create content, there should also be a fully transparent reward mechanism and the freedom to trade assets. Decentralization could be an excellent solution to these problems, offering additional value. Teams with web2 creator resources are promising, as they can attract these creators to web3 UGC platforms with transparency and higher earnings.

Game User Acquisition (UA)

Projects in the game UA category are all centered around establishing user profiles, combined with on-chain, off-chain, and social dimensions. They are divided into customer acquisition (Carv) and operational strategy directions (Helika). However, all acquisition platforms often face retention challenges. Nevertheless, it’s undeniable that player data is valuable, and this value increases with the number of users. If there is confidence in massive adoption brought by applications, game data analysis can capture part of this value.

Regarding game publishing platforms, those that focused on infrastructure and tools as core products have gradually lost competitiveness in the last cycle. Now, the ability to succeed in publishing relies on core game products that can attract many users.

User/Fan Engagement

Projects in this vertical primarily revolve around the entertainment, sports events, and film and television industries. Depending on their collaboration with Intellectual Property (IP) rights holders, these projects can generally be categorized into two forms:

Direct Operation by IP Rights Holders: In this approach, IP rights holders take a more hands-on role in the platform’s operation. This approach places a higher emphasis on connecting IP events and content with the end product. Additionally, it allows for the creation of NFTs (Non-Fungible Tokens) that grant users subsidiary rights. This strategy often results in better fan incentives and feedback.
Licensed Cooperation: Licensed cooperation, on the other hand, places less operational pressure on the platform but relies heavily on the resources provided by the IP rights holder. While it may require fewer operational efforts, its effectiveness hinges on the level of support and resources offered by the IP rights holder.

In addition, through the tracking and observation of several projects, it was found that platforms with a certain base of their own community (such as Karate combat), turning existing users towards IP fans, tend to be more successful than starting from 0 and doing fan engagement based entirely on the IP’s fan base.

Moving forward, our focus will be on entities with high IP value, audiences/fans that overlap with the gaming and betting demographics, and those engaged in more direct IP collaborations.

Institutional Service

The institutional service track can be categorized into the following sub-tracks:

Trading/Brokerage Services: This encompasses activities such as exchanges, liquidity providers, brokers/traders, clearing/settlement, and related services.

Asset Management: This includes services like fund management, high-frequency trading, arbitrage, custody, and more.

Banking/Payments: This sub-track involves payment processors, deposits and withdrawals, card issuing, and other banking-related services.

Other Services: This category includes providers of trading technology and similar services.

Key trends within this track include:

Continued Steady Growth: The institutional service track is expected to experience steady growth in the coming years.
Emphasis on Compliance: Compliance is a significant focus within the institutional services sector, with companies actively working on building compliance frameworks.
Precise Division of Responsibilities: Service providers are becoming more specialized, with each participant focusing on their core responsibilities. This specialization, along with checks and balances and industry supervision, contributes to a more transparent and efficient market.
Rise of Prime Brokerage (PB) Services: The market share for Prime Brokerage service companies is expected to gradually increase. Attention should be given to companies with a traditional financial background (ECNs, fully regulated clearinghouses, Cross Margin capabilities) entering the crypto space.
ETF Applications Impact: The ongoing consideration of ETF applications in the crypto space is creating competition between crypto-native service companies and traditional financial firms. This competition is driving demand for compliant-grade products and reshaping the market structure.
Geographic Focus: Europe is a hub for developing the institutional service track, and emerging markets like South America also show promising potential.

Bitcoin

While inscriptions have recently drawn attention to the Bitcoin ecosystem, it’s essential to recognize that, unlike Ethereum, the Bitcoin blockchain does not share a global state, and its entire approach, including its state, accounts, and computation model, varies significantly. Therefore, constructing infrastructure and applications for Bitcoin requires a different mindset in the medium to long term.

In light of this, it’s worthwhile to focus on key developments such as Taproot Assets, Rollups, and the Lightning Network. Additionally, emerging technological paths like Statechains are worth watching.

Sidechains

Technological paths, including Stacks, have long dominated the narrative of Bitcoin’s second-layer network due to their minimal technical overhead (which allows high programmability off-chain) and the minimization of negative environmental impact. However, sidechains mainly rely on cross-chain and main-chain anchoring technologies and may lose out to other new technological paths that attract more traffic and attention.

Layer2

Many Bitcoin Layer2 solutions, while resembling sidechains in core technical principles, have established complete frameworks for execution, settlement, verification/challenge, and Data Availability (DA), mirroring Ethereum’s model. Differences among these projects primarily hinge on technology stack choices at various levels, such as Cosmos SDK, OP Stack, Polygon zkEVM, Taiko, etc., at the execution layer. Additionally, there’s a trend of “account abstraction” and multi-chain wallet integration for user convenience, accommodating both Ethereum and Bitcoin address formats.

Client-side validation

Technologies like RGB and Taproot Assets, which enable client-side validation, allow for asset issuance and trading with minimal on-chain footprint and are worth continued attention.

Lightning Network

Lightning Labs plans to launch stablecoins and other assets on Taproot Assets next year. Furthermore, the promotion of native asset yield products by Liquidity Service Providers (LSPs) is also anticipated.

BRC20-like Assets

BRC20-like assets heavily depend on specific infrastructures, such as indexers. It’s important to monitor these infrastructures and new asset types like ARC20 while being mindful of the technical implementation risks involved.

DLC

Discreet Log Contracts (DLCs), although proposed early on, have been difficult to promote due to previously low demand. With the large-scale expansion of ecosystem construction, the application of DLC technology may become more widespread, especially following collaborations with certain oracles. However, it is necessary to be cautious of the centralization risks introduced during the implementation of DLCs.

DePIN

DePIN is a sector that can easily take off in a bull market. Similar to games, DePIN is also a sector that is easy to break the barrier and convert traditional users, so it has attracted a lot of attention from the industry. To invest in DePIN, it has a few key points: 1) decentralization, as well as the mechanism, which can be said to be the critical points of DePIN, and the investment in DePIN projects needs to look at the mechanism first.2) timing, good mechanism needs to be coupled with good timing. It is definitely easier to get customers in the early stage of the bull market, so the team is required to have sensitivity to the web3 market.3) Industry fundamentals, the type of hardware chosen by the project, as well as the characteristics of the users, may determine the success or failure of the project. The following is a categorization based on hardware types:

Focusing on toC scenarios, and markets where similar hardware is not yet popularized, there may be revolutionary changes: for ToC high-frequency-use hardware (e.g., wearables), Web3’s gameplay and mechanisms actually provide project parties with better and more efficient product crowdfunding channels, lowering barriers for both users and merchants. In the world of DePIN, because there is a clear token incentive, so the user’s motivation to buy hardware will be stronger (quickly get return on the investment), the merchant can even pre-sale and then production; with a flexible cash flow, the merchant can be postponed to do the fundamentals, such as how to enrich the software ecosystem, how to do the linkage with other hardware, how to empower the token in the ecosystem, and so on. Especially for underdeveloped areas, if there is no DePIN may never buy certain hardware, but early participation in mining speculative heart may bring large-scale real hardware popularization.

Cautious about improving hardware: For ToC placeables or low-frequency-use but just-in-time hardware (e.g., routers) with a large user base and retention, DePIN could be an opportunity to improve the experience. Theoretically, DePIN realizes the redistribution of costs and revenues by redistributing resources and demand between the supply side and the users, resulting in a more reasonable unit economic efficiency and cheaper services on the user side.

However, there are also difficulties: 1) whether the technology can achieve decentralization over centralized solutions, for example, many decentralized computing or storage solutions, or even more expensive and inefficient than the centralized solutions; 2) whether the business is offensive to the interests of the centralized big manufacturers, just demand and high retention means that there are many big manufacturers in this direction, they have many years of users, brands, capital deposits, DePIN If DePIN can’t make a revolutionary breakthrough in fundamentals, or can’t unite the forces beyond the influence of big brands, it’s actually hard to outperform web2’s competitors.

Wait-and-see by holding mining machines: for other low-frequency use, while the daily life of dispensable hardware, or even specifically purchased for the purpose of mining directed miners, DePIN may be able to bring short-term returns, but not necessarily able to form user stickiness. It is not excluded that DePIN really cultivates new user habits, but there is a certain degree of chance, it is more difficult to predict, we need to analyze case by case.