Markets Sending, Just Buy Consensus, Meta's AI Moment & Who Wins Tokenization? | Weekly Roundup

Bitcoin and a broad range of crypto assets rallied, including projects and protocols speakers described as profitable. They took the move as evidence that markets can rise without regulatory clarity, while still preferring passage of the CLARITY bill. They also pointed to SEC and CFTC signals supporting innovation, including tokenization guidance, Project Crypto and a possible developer no-action letter. One speaker credited part of the positive week to the Fed hiking rates and preserving its credibility amid concern it might be pressured not to raise them. The 10-year Treasury yield appeared headed toward 7%, with jitters and the odds of another hike rising, though speakers cautioned that macro forecasting was not their expertise.

Since the market bottom and year to date, sustained momentum had been concentrated in Venice, Hyperliquid, Lighter, Derive and Zcash, which speakers saw as a flight to quality. Some traders interpreted rallies in Bitcoin Cash and other older coins as a possible market top; another speaker thought a top was less likely if more assets were beginning to rally. A broader investing thesis was that consensus names have outperformed in both public markets and crypto as momentum and attention concentrate returns. The speaker favored quality and a long-term view, re-underwriting positions and avoiding rotation unless a thesis is invalidated. Hyperliquid, once a consensus investment near $20, later traded in the $40–$60 range; Nvidia was another example, and Meta was becoming more widely accepted.

The firm’s third fund, a 2022 vintage, expected about 70% of invested capital to accrue value through tokens, while its fourth fund had shifted to roughly 70% equity. Polymarket was the third fund’s largest position, with Lighter among its other large holdings. The firm sees fewer on-chain, token-specific opportunities and lower founder quality in that area, but remains open to investing. Speakers disagreed about liquid-token prospects: one said Derive merited a fresh underwriting because of its remarkable options-market efficiency and could be a 10x opportunity; another said such opportunities were scarce and not what their LPs sought. The firm’s average investment is $10 million, an amount that could move Derive’s market at a $100 million valuation, while $250,000 to $1 million can matter to a smaller fund. Speakers noted that funds often separate venture and liquid assets into sleeves because LPs have different return expectations and may find sharp quarterly drops harder to defend than steadier venture marks.

Meta’s Muse was described as under-recognized and poised to “rip,” with one speaker saying they had recommended Meta without first weighing the risks. They compared the idea to their Robinhood trade: a business moving into an area the market had not yet appreciated. Speakers said Muse was growing faster than ChatGPT and that subway ads reflected its widening reach; one had known about it for two to three weeks before broader public attention. Alexander Wang said Muse was built by a small skunkworks team operating more like a separate startup than through Meta’s usual processes. One speaker found Muse less feature-complete than Instinct but faster and more responsive, attributing that to greater compute and expecting it to catch up on features quickly. That speaker planned to switch fully from the slower Instinct, while relaying uncertainly and secondhand that Instinct’s burn rate was about $900 million a year.

Speakers said Opus 5.5 was 30% faster and cheaper than Opus 5, and that AI tools had become about 50% cheaper and faster over roughly 60 days. One compared AI progress to blockchains moving from scarce, expensive block space to abundant capacity, while another stressed that AI faces different physical constraints and may require trillions of dollars in compute infrastructure. More capacity could make consumer applications more important, but AI remains compute-constrained. On tokenization, speakers saw potential gains in access, user experience, securities lending and 24/7 on-chain trading, but no obvious winner in the middle of the stack. They questioned how liquidity would fare if assets were split across ATSs and who would capture the value, while allowing that the market could expand to benefit issuers, consumers, exchanges, crypto and surrounding protocols.

Potential infrastructure beneficiaries included Alpaca, Superstate and Securitize, though no company’s outsized value capture was clear. One speaker estimated Alpaca might custody 97% of tokenized securities, and speakers said early infrastructure and systems of record could charge substantially higher fees than traditional markets; DeFi could also benefit. The landscape included Blockchain.com users gaining access to tokenized stocks and ETFs through Nizzy’s digital ATS, although one speaker was unsure whether the planned ATS existed yet; ICE was distributing data to Blockchain.com, Nizzy was working with Securitize and tZERO, MoonPay had bought a license from North Capital, and Coinbase had begun in-app IPO allocations that week. Speakers also questioned the economics: tokenization deals are often paid, trade costs vary across offerings, and authorized participants behind Robinhood Chain and similar efforts—often the same market makers—could earn substantial returns.