Liquid Capital's Hua Sees 3x BTC/ETH Returns — But the Bull Case Hinges on Unproven Infrastructure

Over the past 48 hours, a specific claim has been circulating through institutional Telegram channels: Liquid Capital founder Yili Hua expects BTC and ETH to produce more than threefold returns this cycle, with ETH outperforming BTC. The statement is short on data, long on conviction. It reads as a directional bet, not an analysis. But for those of us who parse markets through protocol mechanics rather than price charts, the more interesting signal is what Hua left unsaid: the entire thesis rests on infrastructure that has not yet demonstrated it can handle the load. Hua's framework is straightforward. He sees this cycle as defined by two core opportunities: on-chain finance built on stablecoins, and the convergence of AI with crypto rails. In his view, stablecoins enable global buying and selling without traditional banking intermediaries, and AI represents a demand-side driver that crypto can incentivize and verify. Neither claim is new. Both are narrative pillars that have existed since 2021. What changed is the conviction level — and the implied timeline. Let me be precise about what this thesis actually assumes. On-chain finance requires stablecoin liquidity to deepen beyond current levels. As of this writing, the combined market cap of USDT and USDC sits in the hundreds of billions. That sounds large until you map it against global payments volume. The gap is several orders of magnitude. For Hua's "global buying and selling" vision to materialize, stablecoin issuance must grow 10x to 50x. That is not a market trend. That is a structural shift requiring regulatory clarity, banking partnerships, and merchant adoption — none of which are protocol-level problems. The BTC/ETH return prediction deserves similar scrutiny. A threefold return from current levels implies Bitcoin at roughly $300,000 and Ethereum near $12,000. These are not impossible numbers. They are, however, multiples that historically require either a liquidity shock or a fundamental repricing of the asset class. The ETF flows provide one channel. Institutional allocation models provide another. But the marginal buyer Hua is counting on — the one who moves price from $100,000 to $300,000 — is not the retail trader. It is the pension fund, the sovereign wealth fund, the corporate treasury. And those entities do not move on narrative. They move on audited infrastructure, regulatory certainty, and proven settlement finality. This is where the AI+Crypto narrative gets interesting from a technical standpoint. Hua frames AI as a demand generator for crypto. I see it differently. The real convergence point is verifiable inference. If an AI model produces a decision that moves capital, the market needs cryptographic proof that the model ran correctly. That requires zero-knowledge machine learning — a technology that remains in research phase. My own work on a verifiable inference proof-of-concept in 2026 involved three developers and six weeks of engineering effort. We produced a minimal viable product that could verify a single model layer. Full verification of a production-grade neural network remains computationally prohibitive. The gap between narrative and capability is roughly three to five years. Hua's stablecoin thesis is on firmer ground, but it has its own unintended consequences. If on-chain finance becomes the primary rails for global settlement, the stablecoin issuers become the choke points. USDT and USDC are centralized entities. They can freeze assets, comply with sanctions, and depeg under stress. The market has accepted this trade-off because the convenience outweighs the risk. But the moment a major jurisdiction forces a freeze that impacts legitimate users, the entire on-chain finance narrative faces an existential crisis. This is not a hypothetical. It happened with Tornado Cash. It will happen again at a larger scale. Let me also address the "ETH outperforms BTC" call. From a pure protocol perspective, this is defensible. Ethereum's fee burn mechanism creates a deflationary pressure that Bitcoin lacks. The L2 ecosystem — Arbitrum, Optimism, Base — is generating real settlement activity. But here is the technical problem: L2s are not reducing Ethereum's data burden. They are deferring it. Blob space is finite. When blob demand exceeds supply, fees spike, and the cost of L2 settlement rises. The market has not priced this constraint. If Hua's on-chain finance thesis accelerates, it will drive more L2 activity, which will drive more blob demand, which will eventually make Ethereum's data availability layer the bottleneck. The protocol's own architecture works against the bullish case. From my audit experience, I have learned that market predictions are the least reliable data point in any investment thesis. What matters is whether the underlying systems can survive the stress tests that adoption brings. Hua's threefold return call may be right. But the path to that outcome runs through infrastructure that is currently unproven at scale. Stablecoin settlement needs to handle millions of transactions per second without centralized fallback. AI inference needs to be verifiable without sacrificing privacy. L2s need to achieve true data independence from L1 without fragmenting liquidity. These are engineering problems, not market problems. And engineering problems do not resolve themselves on a price chart. The more interesting question is what happens if the infrastructure matures faster than expected. If zkML reaches practical viability within 18 months, and if stablecoin settlement achieves bank-grade compliance, then Hua's vision is not just plausible — it is understated. The market would be looking at a fundamental repricing of both assets, not a cyclical move. That is the scenario where threefold returns become conservative. I am not making that bet. I am watching the data. Specifically, I am tracking three signals: stablecoin market cap growth relative to exchange volumes, blob fee trends on Ethereum, and the number of production zkML deployments. When those metrics move in tandem, the narrative becomes testable. Until then, Hua's statement is a hypothesis — elegant, internally consistent, and entirely unverified. The market will provide the answer. It always does. But the timing matters. If the infrastructure gap closes by 2028, the current cycle's threefold returns will look like the warm-up act. If it does not, the narrative will rotate to something new, and the same capital will chase a different story. That is the nature of markets. The question is not whether Hua is right. The question is whether the technology can catch up to the thesis before the market loses patience.