"Strategy sells more shares for cash reserves" Jul 20, 2026

Strategy sold 2.7 million MSTR shares between July 13 and July 19, raising $225 million in cash and bringing its total cash reserve to $3.2 billion. The company made no Bitcoin purchases, sales, or share repurchases during that period, signaling a deliberate shift toward rebuilding liquidity rather than accumulating more Bitcoin. Strategy currently holds just over 843,000 BTC valued at more than $54 billion, representing roughly 4 percent of Bitcoin's maximum supply, though the position carries approximately $9 billion in paper losses. Michael Saylor hinted at a possible next acquisition move with another tracker post. JP Morgan cited the larger cash position and improving institutional demand in Bitcoin futures as encouraging signs, and CryptoQuant said the new capital management approach eases liquidity concerns. Despite those endorsements, Strategy's stock fell 4 percent last week.

Vitalik Buterin released an early working demo of an anonymous message board built on Aztec, where users deposit ETH on Ethereum and post privately on Layer 2 with no public link to their wallet. Buterin described the project as a vibe-coded toy based on an idea he first proposed in 2022. The system includes an on-chain moderation policy with a designated censor role that can flag posts and delay the offending poster's next message, and a locally run large language model that automatically checks posts against that policy before they go out. Posting frequency scales with the size of a user's deposit. The project includes formal verification in Lean 4 with 70 completed proofs covering privacy guarantees, rate limits, moderation checks, and deposit security, giving the lightweight demo a more rigorous technical foundation than its toy framing suggests.

Japanese logistics company AZ-COM Marua Holdings plans to use the JPYC stablecoin to pay business partners and contractors including truck drivers, which would represent the first large-scale corporate adoption of JPYC. The company's stated motivation is that faster payments could help attract workers as Japan's logistics sector contends with an aging workforce and tightening overtime regulations. The move is notable less for the stablecoin itself than for the specific labor-market problem it is being deployed to solve.

Hyperliquid plans to enable permissionless creation of HIP prediction markets, launching first on Testnet before moving to Mainnet. Deployers must stake 500,000 HYPE to create markets, and that stake is subject to partial or full slashing for unclear market rules, incorrect settlement, or failure to settle within one week. The stake remains locked for six months, and all markets under a deployer's account must be settled before any withdrawal is permitted. Each deployer starts with capacity for 100 outcomes, and settled markets free up slots for reuse. Validator-created canonical Hyperliquid markets are expected to remain rare, with fewer than 10 outcomes per year anticipated from that channel, meaning the permissionless deployer route is intended to carry most of the volume.

The Hyperliquid design places meaningful financial accountability on market creators through the slashing and lockup mechanics, which distinguishes it from prediction market platforms where bad or unresolved markets carry little direct cost to the deployer. Whether the 500,000 HYPE stake requirement filters for quality or simply limits participation to well-capitalized actors is a trade-off the Testnet phase will likely surface before the Mainnet rollout.

The Strategy share sale and the Hyperliquid prediction market structure share a common thread in that both reflect institutions and protocols choosing to hold larger capital buffers as a condition of operating, whether to manage liquidity risk on a leveraged Bitcoin treasury or to enforce accountability in a decentralized market system. The AZ-COM stablecoin adoption and Buterin's Aztec demo sit at the other end of the maturity spectrum, representing early-stage deployments where the primary question is whether the use case proves out at scale rather than how to manage risks already well understood.