BTC $64,566.21, ETH $1,856.62, SOL $75.31, XRP $1.093, HYPE $59.97, DOGE $0.072337, AAVE $89.74.
Sunday’s useful signal is account displacement.
July 17 was about access control: crypto ETFs, market-maker capital, payment standards, brokerage APIs, agent wallets, and institutional vaults. July 18 moved into routing and gatekeeping: Stripe versus Swift, SBI’s Asian corridor, Robinhood’s DeFi front door, France blocking Polymarket, MiCA conversion paths, FTX payouts, and supervised digital-securities rails.
Today’s cleaner read moves away from that distribution-heavy lane.
The next fight is over where the account lives and who keeps the market usable once assets move there. Wallets are starting to look like bank, brokerage, payment, and yield accounts in one place. Brazil’s Pix fight shows Washington treating payment systems as dollar-policy terrain. Bitcoin’s DOG Mode fight asks whether mempool policy becomes private routing or open propagation. DeFi’s idle-liquidity problem shows that capital can be onchain and still useless. Options traders are targeting a Fed-window BTC move, while Uniswap is trying to make protocol revenue less theoretical.
That is a different Sunday mix. Less “who gets licensed?” More “where does the account live, who routes the transaction, and does the capital actually work?”
Price snapshot via Coinbase spot prices for BTC and ETH, plus CoinGecko simple-price data for SOL, XRP, HYPE, DOGE, and AAVE around 07:00 HKT.
1. Digital Wallets Are Starting To Replace The Account Mental Model
CoinDesk reported that crypto executives and bankers increasingly expect younger users to hold payments, savings, tokenized assets, and crypto inside wallets rather than separate bank and brokerage accounts.
That is the most useful Sunday framing.
Steakhouse Financial manages more than $4B in blockchain vaults. Visa’s stablecoin tracker recorded $6.6B in retail-sized stablecoin volume across 132.4M transactions in the latest 30-day period. Standard Chartered expects stablecoin circulation to rise roughly sevenfold to about $2T by 2028, while agent-led purchases could grow from 1% of e-commerce in 2025 to 12% in 2029.
The bank account is not disappearing tomorrow. The shape of the account is changing.
If users hold stablecoins, tokenized deposits, money-market funds, crypto, cards, savings, and trading access inside one identity-linked wallet, the old product categories blur. The bank becomes infrastructure. The exchange becomes payments. The wallet becomes the daily account.
The hard question is custody. Self-custody is elegant until a lost key means no recovery, no recourse, and no insurance. That is why the likely winner is not pure bank replacement. It is a hybrid account model where wallets carry the user interface and regulated institutions still provide parts of the money, compliance, recovery, and risk stack.
2. Brazil’s Pix Fight Turned Payments Into Dollar Policy
CoinDesk reported that the U.S. will impose a 25% Section 301 tariff on most Brazilian goods starting July 22, partly targeting Brazil’s Pix instant-payment system.
That is a payments story with a geopolitical edge.
Pix is not a crypto rail. It is a state-run instant-payment system used by more than 90% of Brazilian adults. In June, it processed nearly 7B transactions worth about R$3T, or roughly $590B. In the second half of 2025, Pix handled 42.9B transactions versus 23.8B across credit, debit, and prepaid cards.
Washington’s complaint is that Pix gives local payment infrastructure an unfair advantage over U.S. card networks. The deeper issue is dollar routing.
Brazil has promoted local-currency settlement and non-dollar payment channels through BRICS policy work. Yet dollar-linked stablecoins already account for about 90% of Brazil’s crypto transactions, according to CoinDesk’s report.
That makes the politics awkward. The U.S. is attacking a non-dollar payment rail while dollar stablecoins are quietly spreading through the same economy. If stablecoins become the private-sector dollar export, Washington may find that the dollar’s strongest digital path is not Visa, Mastercard, or SWIFT. It is USDT, USDC, and wallets.
3. DOG Mode Reopened Bitcoin’s Relay-Governance Fight
CoinDesk reported that DOG Mode, an alternative Bitcoin client introduced by developer Leonidas, relaxes default relay policies affecting Ordinals and Runes transactions without changing consensus rules.
This is a better Bitcoin story than another price take.
Consensus rules define what can enter a valid block. Relay policy defines which valid transactions nodes will forward before miners include them. DOG Mode targets the second layer.
That distinction matters. If a transaction is consensus-valid but default node policy refuses to relay it, users may need private channels, direct miner relationships, or specialist brokers to get the transaction mined. DOG Mode tries to make those transactions propagate through the peer-to-peer network instead.
The risk is mempool fragmentation. Different node policies can mean different views of pending transactions, different fee estimates, and uneven propagation. Consensus survives, but the path to inclusion gets less uniform.
The governance lesson is sharper than the Ordinals fight itself. Bitcoin does not need a hard fork for power to shift. Defaults, relay policy, miner relationships, and client diversity can all decide who gets practical access to blockspace.
4. DeFi Has $1.6B Of Capital Sitting In The Wrong Place
CoinDesk reported that Dune research commissioned by 1inch found $1.6B in DeFi liquidity was underutilized in the first half of 2026.
This is a capital-efficiency problem, not a TVL problem.
The study tracked $1.84B across concentrated-liquidity pools on Uniswap, PancakeSwap, and Aerodrome. Roughly 85% was underused, and about $542M sat fully outside active trading ranges in an average week. Those out-of-range positions earned no fees and provided no market depth.
The annual fee loss is estimated around $150M.
That matters because DeFi often sells liquidity as its core advantage. But a liquidity position only helps traders when it sits inside the live price range. If capital is posted too high or too low, it looks good in TVL charts while doing nothing for execution.
This is where automation can be useful without sounding like a pitch. Active range management, cheaper rebalancing, smarter pool design, and better intent routing all become more important as onchain markets mature. DeFi’s next improvement may come from making existing liquidity work harder, not from attracting another billion of passive capital.
5. BTC Options Traders Are Targeting The Fed Window
CoinDesk reported that traders bought $2.5B in notional bitcoin call spreads on Deribit, targeting $72K by July 31.
The timing is the point.
The July 31 expiry lands two days after the Fed’s July 29 interest-rate decision. Markets are currently leaning toward a hold in the 3.5%-3.75% range, with futures pricing roughly 75%-80% odds of no change.
The trade is not a moonshot. A bull call spread buys upside around $70K while selling the $72K call to reduce cost and cap gains. That is a structured bet on a moderate move, not a blank-check long.
Why does this matter after a week of risk-off AI and Gulf headlines? Because large traders are separating event risk from direction. Oil and Iran risk can still hit the tape. But some options desks are willing to pay for the possibility that softer inflation and a Fed hold reopen the path toward $70K.
Spot demand still has to confirm it. Options flow can show positioning. It can’t create durable demand by itself.
6. Tokenization Is Becoming A Hybrid-Market Budget Item
CoinDesk reported that a Broadridge survey found 84% of financial institutions now see tokenization as strategically important.
The number is loud, but the implementation detail is better.
Broadridge surveyed 200 North American financial-services executives. Sixty-eight percent said tokenization will at least partly reshape markets within three to five years. Almost one-third plan to increase tokenization investment by 26% to 50% or more over the next two years.
Yet the market is not preparing for everything to move onchain at once.
Ninety-two percent expect digital and traditional assets to coexist, and 69% plan to integrate tokenization into existing systems rather than build separate blockchain-native infrastructure. Capital-markets firms are ahead, with 44% already running tokenization initiatives in production or at scale, versus 20% of asset managers and 9% of wealth managers.
That makes tokenization less exciting and more real. The winning version probably looks like hybrid custody, hybrid settlement, tokenized funds, money-market products, collateral movement, and back-office compatibility before it looks like every equity trading 24/7 in a retail wallet.
7. Uniswap Is Trying To Turn V4 Volume Into UNI Burn
The Block reported that two Uniswap governance proposals are heading to final onchain votes from July 19 through July 26.
One proposal activates protocol fees for selected v4 pools across seven chains. The other extends v2 and v3 fee collection to Robinhood Chain.
That matters because Uniswap has long had the paradox of massive usage and uncertain token value capture. December’s UNIfication overhaul created a UNI burn mechanism, turned on fees for v2 and v3 Ethereum mainnet pools, and burned 100M UNI from the treasury. V4 fees were deferred.
Now v4 may enter the revenue path.
Robinhood Chain makes the vote more interesting. Uniswap deployed all three protocol versions there at mainnet launch, and those deployments passed $6B in cumulative swap volume by July 10. Early volume was memecoin-heavy, but fee capture does not care whether the flow came from a perfect institutional use case.
The useful question is whether UNI can graduate from governance claim to cash-flow claim. This vote is a small but concrete test.
8. Bitcoin Japan Shows The Treasury Copycat Trade Can Dilute Fast
The Block reported that Bitcoin Japan, formerly apparel wholesaler Marusho Hotta, approved a financing package with EVO Fund that could raise about 9.66B yen, or $59.5M.
The headline sounds like another corporate BTC treasury story. The details are rough.
Only the 1.5B yen bond tranche is committed upfront. The plan sets aside 662M yen, or about $4.1M, for the company’s first bitcoin purchase. A prior December warrant program raised only 54% of its target and allocated nothing to bitcoin, instead funding stakes tied to SpaceX and Figure AI.
The dilution math is the warning.
At the initial price, full conversion and warrant exercise would create new shares equal to 95.33% of the existing share count. At the floor price, potential new shares would equal 110.08%. The stock closed down 26.7% Friday after touching a year-to-date low.
This is why the public-company bitcoin-treasury playbook is not portable. Strategy built a capital-markets machine around BTC accumulation. A microcap borrowing the brand language without committed capital, scale, or clean terms can turn the treasury story into dilution first and bitcoin exposure second.
9. GitHub Trending - Fresh Picks After The Repeat Filter
The repeat tracker ruled out yesterday’s picks github/copilot-sdk, tirth8205/code-review-graph, and RyanCodrai/turbovec, plus recent repeats like apache/ossie, PostHog/posthog, codecrafters-io/build-your-own-x, jamiepine/voicebox, and microsoft/markitdown.
bojieli/ai-agent-book is leading Trendshift’s daily list with 1.3K stars and 120 daily momentum points. It is an open Chinese-language book on AI-agent design principles and engineering practice, with chapters, PDFs, and companion code. The signal is that agent work is moving from demo prompts into engineering curriculum.
xai-org/grok-build has about 2K stars and 328 daily momentum points on Trendshift. It is a coding-agent harness and TUI described as fullscreen, mouse-interactive, and extensible. The useful read is interface pressure: coding agents are becoming development environments, not only chat panes beside an editor.
ibelick/ui-skills has 4.9K stars on GitHub Trending and gained 242 stars today. It packages design guidance as skills for design engineers. That fits the broader turn toward reusable instruction files: teams are trying to make judgment portable enough for agents to apply during real interface work.
Morning Read
Read CoinDesk’s wallet-account story, then read CoinDesk’s DOG Mode explainer.
The number to remember is 132.4M.
That is Visa’s latest 30-day count of retail-sized stablecoin transactions. It explains why the wallet-account debate matters more than another app launch.
The second number is $542M. That is the average weekly capital sitting fully out of range in concentrated-liquidity pools. DeFi already has capital. The next job is making more of it useful.
Sunday’s read is account displacement. Wallets are pushing into bank-account territory. Stablecoins are quietly exporting dollars through private rails. Bitcoin is debating who gets practical access to blockspace. Uniswap is trying to convert usage into burn. BTC options traders are targeting the Fed window.
The market is not only asking which asset goes up. It is asking where money lives, how transactions propagate, and whether onchain capital earns its keep.
Evening Update - 18:50 HKT
BTC $64,526.32, ETH $1,865.08, SOL $75.96, XRP $1.095, HYPE $61.08, DOGE $0.072449, AAVE $90.14.
The evening update is deliberately not another pass over wallet accounts, Pix geopolitics, DOG Mode, idle liquidity, BTC call spreads, Broadridge tokenization budgets, Uniswap fees, Bitcoin Japan dilution, or the morning GitHub picks.
The cleaner late-day signal is implementation risk.
Zcash is trying to scale private payments without losing verifiability. U.S. stablecoin regulators missed the GENIUS Act rulemaking deadline, so issuers now face a shorter runway before January 2027. Security researchers found malware moving through fake GitHub apps and recruiter workflows. Tokenized Pokemon-card gacha hit $324M in monthly spending while the wider crypto market was weak. Centralized exchanges finally broke a five-month volume slide, but the rebound was uneven. Tokenized equity and RWA perp flow kept showing up in market data before the legal rails are fully settled. Balaji’s Malaysia fight showed that network-state projects still answer to passports, borders, and local politics.
That is a different Sunday evening mix. The market already has ambitious stories. The hard part is making them survive real nodes, real rules, real attackers, real liquidity, and real jurisdictions.
Price snapshot via Coinbase spot prices for BTC and ETH, plus CoinGecko simple-price data for SOL, XRP, HYPE, DOGE, and AAVE around 18:50 HKT.
11. Zcash Shipped A Node For Privacy At Payment Scale
CoinDesk reported that Zakura, a new Zcash full node, launched as a pruned fork of Zebra with zcashd compatibility.
This is the first fresh story worth leading with tonight.
Zakura is not a marketing wallet. It is node software aimed at the ugly bottlenecks that keep private payments small. The team says a ready-made pruned chain snapshot can get a new node running in under two minutes, and the client supports the Ironwood upgrade scheduled for July 28.
The scale target is aggressive: Visa and Mastercard-level throughput above 50,000 transactions per second. Current Zcash privacy tech would demand huge verification bandwidth at that level, so the broader plan leans on Project Tachyon’s recursive proofs and Valar’s private information retrieval work.
The risk is deeper than speed.
Zcash’s Orchard pool had a long-running soundness bug that forced an emergency response in June. Ironwood adds a turnstile mechanism to cap what can leave the pool and limit any counterfeit supply that might have been created.
That makes Zakura a useful privacy case study. Privacy chains have to solve scale, wallet UX, node operation, and supply assurance at the same time. One missing piece can break trust.
12. GENIUS Act Rules Missed The Deadline
The Block reported that U.S. regulators missed the GENIUS Act’s July 18 deadline for final stablecoin implementing rules.
That is a stronger stablecoin story than another market-share update.
The law was signed on July 18, 2025. It told the primary federal payment-stablecoin regulators, Treasury, and state regulators to finish implementing rules within one year. As of the deadline, major rule packages from the OCC, FDIC, NCUA, Treasury, the Fed, FinCEN, and OFAC were still proposals or still open for comment.
The important date did not move.
The law still takes effect on Jan. 18, 2027, unless final rules trigger an earlier 120-day path. That means issuers, banks, state regulators, custodians, and reserve managers may have to prepare against draft rules that can still change.
This is what implementation risk looks like. Congress can pass the framework. The market still needs final answers on reserves, tokenized reserve assets, customer identification, state certification, AML, sanctions, custody, and disclosures before stablecoins become boring financial plumbing.
13. Crypto Malware Moved Into GitHub And Hiring Flows
Cointelegraph reported that Kaspersky identified OkoBot, a malware framework targeting crypto investors through social engineering and trojanized GitHub apps.
The security story is no longer only malicious links in a Telegram DM.
OkoBot can harvest wallet files, browser data, credentials, and wallet application windows. Kaspersky said it has seen multiple attacks involving the family since January 2026, with payloads coordinated through an SSH tunnel.
SlowMist flagged a related developer attack path: fake LinkedIn recruiters send blockchain developers GitHub repositories to run as part of a supposed interview. The workflow looks normal enough to be dangerous. Pull a repo, install dependencies, launch an app, and the attacker gets a path into project keys, cloud credentials, wallet extensions, or Telegram sessions.
This matters because crypto teams live in GitHub, npm, Discord, Telegram, and contractor pipelines.
Smart-contract audits do not protect a team that runs attacker-controlled code during hiring. Wallet hygiene does not help much if the developer workstation is compromised first. The next security budget has to cover people, repos, dependencies, and recruiting rituals, not only protocol code.
14. Onchain Card Gacha Hit $324M While Crypto Was Weak
Cointelegraph Magazine reported that users spent a record $324M on onchain gacha in June, up from roughly $50M a year earlier.
This is the oddest RWA story of the day, which is why it is useful.
Projects such as Collector Crypt and Courtyard vault graded physical trading cards, issue NFTs tied to specific cards, and let users open randomized digital packs. The token can be held, sold, bought back by the platform, or redeemed for the physical card.
The demand did not wait for a bull tape. June was rough for crypto, but tokenized card spending hit a record. Pokemon’s broader collectible market helped: Cointelegraph cited Circana data showing the brand reached $2.5B in U.S. sales in 2025, up 87% from a year earlier.
The risk is the loop.
Randomized packs plus instant buyback can feel closer to gambling than investing. The asset is real, but the user still depends on vault integrity, grading accuracy, redemption promises, and platform solvency.
Tokenization works best when it removes friction from a market people already want. This one proves the point, then immediately asks whether speed makes the behavior healthier or more addictive.
15. Centralized Exchange Volumes Finally Rose Again
CoinDesk Research reported that combined spot and derivatives volume on centralized exchanges rose 13.0% to $4.99T in June.
That ended a five-month decline.
Spot volume climbed 15.3% to $1.11T, while derivatives volume rose 12.1% to $3.88T. Gate had the largest spot-share gain among tracked exchanges, with volume up 50.8% to $66.1B and spot share rising to 5.95%.
The interesting part is where the rebound happened. Bitcoin fell 20.5% in June, but spot market share still rose. That suggests some users were buying the drawdown rather than only adding leverage.
DEX activity improved too, but CEXs recovered faster. DEX spot volume rose 9.10% to $195B, while DEX futures rose 13.2% to $676B. The split is a reminder that onchain venues are gaining real scale, but centralized venues still absorb the largest liquidity shock when volatility returns.
The better read is simple: volume came back, but venue power did not decentralize evenly.
16. Tokenized Equity And RWA Perps Kept Pulling Flow
CoinDesk Research reported that monthly trading volume for RWA perpetual instruments on centralized exchanges rose 57.0% to a record $311B in June.
The driver was not another crypto-native token.
CoinDesk tied the move to SpaceX IPO activity and retail demand for trading traditional financial assets through crypto exchange rails. Binance led the category with $245B in volume, or 78.6% share. OKX followed with $33.3B, and Gate printed $12.1B.
This sits next to the week’s tokenization stories, but the angle is different.
Institutions are building supervised digital securities. Retail traders are already chasing synthetic and perp exposure to traditional assets on crypto venues. Those two tracks can coexist for a while, but they create different risks. One cares about custody, legal title, and settlement finality. The other cares about leverage, funding, liquidation, and venue solvency.
RWA perps are the fast path. Tokenized securities are the formal path. The gap between them is where regulators will spend the next few years.
17. Balaji’s Malaysia Fight Put Network States Back On Earth
Cointelegraph reported that Balaji Srinivasan sought legal assurances from Malaysia after authorities probed Network School’s Forest City community.
The allegation was that the community hosted Israeli citizens using second passports. Malaysia’s Home Affairs Ministry said initial checks found all 266 foreigners held valid documents.
Balaji’s response was to seek a memorandum of understanding that would give Network School legal certainty. Without it, he said the community could move capital elsewhere.
This belongs in a crypto digest because network-state theory keeps colliding with state power.
Crypto people like portable capital, online communities, parallel institutions, and city-scale experiments. Governments still control entry, residency, immigration rules, policing, telecoms, land use, banks, and political tolerance.
The useful lesson is not whether Network School stays in Malaysia. It is that even well-funded post-national communities need boring local certainty. A founder can move capital. A community still needs a jurisdiction that says the rules will hold.
18. GitHub Trending - Deepsec Turns Agent Security Into A Scan
The repeat tracker ruled out today’s morning picks bojieli/ai-agent-book, xai-org/grok-build, and ibelick/ui-skills, plus recent repeats like github/copilot-sdk, tirth8205/code-review-graph, shadcn/improve, diffusionstudio/lottie, makerspet/oomwoo, and moonshine-ai/moonshine.
vercel-labs/deepsec is leading Trendshift’s daily list with 5.5K GitHub stars and 336 daily momentum points. It is an agent-powered vulnerability scanner for large repos, with resume support, PR-diff mode, revalidation, exports, and optional Vercel Sandbox execution.
The signal is direct: coding agents are moving into security work where the output must be reviewed, rechecked, classified, and traced. The interesting part is not that an agent can read code. It is that the tool wraps the agent in a repeatable scan and validation workflow.
19. GitHub Trending - Castor Makes Web Video Locally Useful
stupside/castor is second on Trendshift’s daily list with 283 daily momentum points. It points at a web page, finds the video stream, transcodes it, casts it to a TV, and can handle subtitles.
This is a nice change from the agent-tooling pileup.
The read is local media control. Users keep getting boxed into apps, DRM surfaces, subscriptions, and brittle casting paths. A repo like Castor gets attention because it makes the web page itself the input, then handles extraction and playback from there.
That is not a crypto story. It is a builder story. Good tools often win by making an annoying everyday workflow direct again.
20. GitHub Trending - Harness Engineering Became A Repo
lopopolo/harness-engineering has 244 daily momentum points on Trendshift. It packages Ryan Lopopolo’s anthology, field guide, and agent context bundle for harness engineering.
That phrase is becoming less abstract.
As teams use agents for coding, research, security, writing, and ops, the job shifts from prompting one model to designing the wrapper around the work: instructions, checks, logs, review gates, state, recovery, evals, and human handoff.
That is why this repo fits tonight’s implementation theme. The winning agent stack will not be the loudest chat box. It will be the one with enough harness around it that the work can be trusted when nobody is staring at every token.
Evening Read
Read CoinDesk’s Zakura piece, then read The Block’s GENIUS Act deadline report.
The number to remember is Jan. 18, 2027.
That is the GENIUS Act effective date if regulators do not trigger an earlier path. Missing the July 18 rulemaking deadline did not give the market more time. It gave issuers less certainty.
The second number is 50,000. That is the transaction-per-second floor Zakura’s backers are using when they talk about payment-scale private money.
Sunday evening is about implementation. Privacy chains need faster nodes and better supply controls. Stablecoin issuers need final rules. Developers need safer hiring and repo habits. Tokenized collectibles need vault trust. Exchanges need real liquidity, not just product menus. Network-state projects need local legal certainty.
The ideas are big enough. Now the market has to prove they can survive contact with real operations.