COINCoinbaseNeutralCoinbase
ETHEthereumNeutralEthereum
HOODRobinhoodNeutralRobinhood
KLARKlarnaNeutralKlarna
ONONOn HoldingNeutralOn Holding
TJXTJX CompaniesNeutralTJX Companies
WMTWalmartNeutralWalmart
NKENikeBearishNikeNFTs Are Back! This Time With Balance Sheets.
Robinhood Chain is filling with NFT projects that hold real assets. Monthly Ethereum NFT volume is up roughly 50% from the 2024 bottom to around $720 million.
Robinhood Chain opened its public mainnet on July 1. On August 21, less than two months later, it crossed a billion dollars in total value secured, with L2Beat tracking the figure at $1.08 billion. Ten days before that it logged 11.6 million transactions in a single day, briefly putting it ahead of Base on user activity.
Needless today, things are cooking over at Robinhood so today, I thought I’d take you through some of the cool things that are being built.
Ultimately, Robinhood’s blockchain exists to put stocks onchain. Tokenized real-world assets sit at roughly $32 million by the narrow measure analysts have been using, which works out to somewhere between 5% and 6% of the chain’s DeFi deposits depending on which day you take the reading, down from nearly a third in early July. DeFiLlama’s broader definition puts the number closer to $125 million, and the two aren’t interchangeable.
Meanwhile, 4,444 pixel-art stockbrokers that minted FOR FREE(!) in mid-July reached a floor of almost $25,000 on August 11, which put the collection’s floor valuation north of $100 million. To repeat…this was an NFT that was given away, for FREE.
The reason for the boom? People noticed the pixel art, but what they were bidding on is the wallet behind it, which arrived already holding a tokenized stock and can be switched on to collect a share of the protocol’s trading fees.
Honestly, it’s kinda strange to be writing about NFTs in 2026 because they spent four years as the punchline of crypto assets.
Monthly Ethereum NFT volume ran around $3.5 billion at the 2022 peak, bottomed near $480 million in 2024, and sits around $720 million in early 2026, so the recovery is real at roughly 50% off the bottom while the market is still about 79% below where it was.
What came back is built differently from what left, and Robinhood chain is where the difference is easiest to see.
Subscribe now
NFTs as a Storage Unit
Anyone who has rented a storage unit understands the arrangement. The door is a numbered sheet of corrugated steel worth close to nothing. You’re paying for the space behind it and for the right to keep getting at it. Stop paying and the facility keeps the contents. Hand the lease to someone else and their rent clock starts at zero.
Nobody rents a unit for the door.
That’s roughly what’s happening to NFTs on Robinhood Chain, and it’s a different arrangement from the one that defined 2021. Back then the picture was the product.
A project minted a few thousand images, attached rarity traits, built a Discord, and hoped culture would hold a floor. Sometimes it did. More often, once trading slowed, there wasn’t much underneath, which is why the sector went so quiet for so long.
Quick vocabulary stop. The change runs on a standard called ERC-6551, usually described as a token-bound account.
In plain terms, it lets an NFT own a wallet of its own. That wallet can hold tokens, other NFTs, and a transaction history, and whoever holds the NFT controls what sits inside it. The old setup put the NFT inside your wallet. This one puts a wallet inside the NFT. None of that makes the NFT a legal company, a brokerage account, or a registered security, so treat the balance-sheet language here as a description of how the thing behaves rather than what it is under the law.
The picture becomes the door, and whatever’s stacked behind it is the reason anyone’s bidding.
Three Collections Worth Watching
StonkBrokers is the most literal version. Each of the 4,444 brokers owns a wallet that was seeded at mint with a tokenized stock token. Pay a tiered activation fee in $STONKBROKER and the broker starts receiving fee-funded stock drops, weighted by tier, with half of every activation payment burned. Sell the broker and activation clears, so the next owner pays again.
Yardkeepers, from TickerYard, put something less tangible in the wallet. The 3,333 keepers are positioned as operating seats in a cross-chain routing network, with $YARD funding activation and status. TickerYard’s first shipped product moves Bitcoin exposure onto Robinhood Chain, and the project has said active keepers will share in that revenue.
Card Wall went the other direction, toward things you can hold in your hand. Its memberships control wallets that receive wrapped, graded physical trading cards, with the actual slabs sitting insured in a vault and redeemable. The project launched through an anti-snipe curve that graduated at the $1 million mark, raising 92.5 ETH in under an hour.
Same standard underneath, three different bets about what belongs inside.
How Holders Get Paid ($$$)
Trace the cash and the loop mostly closes on itself.
Traders buy and sell inside the ecosystem.
A slice of each trade lands in a reward pot.
The pot gets converted into stock tokens.
Those tokens go to activated NFTs by tier weight.
Bigger drops make activation look more attractive, which supports token demand, which supports the floor, which brings more trading.
Every one of those arrows moves money that was already inside the system. One arrow brings in money that wasn’t, and it’s labeled “new buyers.”
That doesn’t make the design a scam. Plenty of legitimate businesses run on internal circulation for a stretch. It does mean the payouts are a function of turnover, and turnover is a function of enthusiasm.
The Rent Waiting to Be Priced In
Activation is where this stops being an abstraction.
You’re paying - in a volatile token no less - for the right to receive drops, and that payment doesn’t travel with the NFT when you sell.
To know whether that’s a good trade you’d need one number: the realized daily payout per activated NFT. I’ve gone looking and I can’t find it published anywhere. Without it there’s no payback period on a five-figure entry plus a four-figure activation, and “yield” stays a word instead of a figure.
The most useful thing anyone in this ecosystem could ship right now is a public dashboard showing realized distributions per activated account, dated and verifiable by anyone who wants to check.
Read the Fine Print
Activation is denominated in the ecosystem token, so the cost of switching a unit on moves independently of what the unit produces. Between August 11 and late August, $STONKBROKER fell about 63% from its high while the NFT floor fell about 26%. 😬
Selling clears activation, which converts prior spending into a sunk cost. Thin liquidity makes a floor look sturdier than the price you’d get for hitting a bid. Bridges, token-bound accounts, oracles, custody providers, and smart contracts each add a failure point.
There’s a legal layer as well. Robinhood Stock Tokens are tokenized debt securities issued by Robinhood Assets (Jersey) Limited. They provide economic exposure to the underlying shares without granting any ownership rights in them, and they are not available to U.S. persons. The stock drops landing in these wallets are protocol-funded marketing rewards rather than corporate dividends. Terms like “revenue share,” “yield,” and “ownership” are carrying a lot of weight in project documentation, and they deserve a careful read before anyone treats them as contractual promises.
What You Should Takeaway
The design fixed something real. An NFT that custodies assets, receives distributions, and carries a role solves the problem that killed the last cycle, which was that the picture had to carry the entire economic argument by itself. That’s a better class of experiment than anything the 2021 wave produced, and it’s the reason this comeback deserves more attention than the floor prices are getting.
Whether these particular units hold their value depends on whether outside money shows up for a reason other than the units themselves. The facility collects rent either way, and the tenant only gets paid if somebody eventually wants what’s stacked behind the door.
If you’re in one of these collections, I’d like to know what your unit has paid you so far. Reply and tell me.
The door has a number on it.
The rent is the question.
Moving right along!…
Share BitFinance with Matthew Snider
In Case You Missed It
The SEC Just Made the Case for Letting You In Early To New Crypto Projects (August 19)
The SEC's Regulation Crypto Assets proposal runs past 400 pages, and its central move is to separate the investment contract from the token: the contract is temporary scaffolding on a building under construction, while the token is the permanent structure left standing. That distinction is what lets the Commission argue for earlier retail access, backed by a $5 million startup exemption, a $20 million Tier 1 cap on unaudited financials, a $75 million Tier 2 cap requiring audited statements, and a 30% ceiling on insider sales.
Better disclosure only helps investors who actually read it, and the 60-day comment period is the window for anyone who doesn't like the rules as drafted.
Read the full article here.
Market Winners 🏆
Bitcoin (BTC) and XRP. Bitcoin closed Friday near $76,944, up 22% on the week and its biggest gain in more than two years, while XRP ran roughly 50% to $1.50, a week after its first close under a dollar since November 2024. The move started Wednesday when Treasury yields fell on Bessent’s buyback surprise, then accelerated after Trump pressed Congress on the Clarity Act. Liquidity and legislation finally moved together, and what a sight to see.
Coinbase (COIN) and Circle (CRCL). Both gained roughly 10% Thursday after Trump met the CEOs of Coinbase, Robinhood, Ripple, Kraken, Gemini, and Chainlink Labs at the White House and called on Congress to pass the Clarity Act. Coinbase kept running Friday, reaching $180.54 by midday against $160.20 Thursday. These names trade as leveraged bets on Washington rather than on trading volume. Nothing in the underlying businesses changed this week; the odds of the SEC and CFTC turf war ending by statute did.
Home Depot (HD). Home Depot beat on both lines Tuesday, with revenue of $47.9 billion up 5.7% year over year, adjusted EPS of $4.92, and comparable sales up 1.7%. Shares rose about 1.7% premarket, pulling Lowe’s along. The number that matters is the comp: positive same-store sales in a week when July retail sales came in negative and the 30-year Treasury yield hit its highest since 2007 says the homeowner with equity is still spending. The K-shaped consumer showed up in the results.
Market Losers 📉
Walmart (WMT). Walmart fell about 9% Thursday, from $114.30 to $103.84, and roughly 11% for the week, its worst since 2022. The quarter beat: EPS of $0.81 against $0.74, revenue of $187.9 billion, full-year guidance raised. US comparable sales excluding fuel grew 2.6% against a 3.5% estimate, and that single line erased everything else. Walmart is priced as the defensive staple that takes share in a slowdown, so a comp miss breaks the thesis. TJX fell more than 7%, its worst week since 2021.
Klarna (KLAR). Klarna crashed 22% Tuesday and finished the week down more than 30%, its worst ever. Second-quarter results beat on both lines, but management cut full-year revenue guidance to $4.08 billion to $4.16 billion against a $4.42 billion consensus, blaming roughly $600 million of currency headwinds and a soft German consumer. The CFO and CMO both leave in early 2027. Delinquencies over 30 days improved by more than 20 basis points, which makes this a demand problem rather than a credit one.
Nike (NKE). Nike closed at $39.09 Monday, its lowest since September 2014 and 78% below the November 2021 record of $177.51. The 4% drop came from On Holding’s revenue miss and the weight of JPMorgan’s August 4 downgrade. Greater China revenue has now fallen for eight straight quarters. A brand that loses a generation of consumers in Greater China doesn’t get them back with a product cycle. Nike now trades at its LOWEST level in 12 years…unreal that it’s gotten worse since I wrote about this very issue 4 months ago.
What to Watch Next Week 👀
Nvidia’s Q2 (Wednesday, August 26). Nvidia reports after Wednesday’s close, with consensus near $93 billion to $95 billion of revenue and roughly 96% year-over-year growth. Options are pricing an 8% to 12% move. Data center revenue already topped $75 billion last quarter, so the headline number is close to a formality. Listen to the Blackwell ramp commentary, because that’s what the AI leasing structures are underwriting.
Jackson Hole and Warsh’s Debut (August 27 to 29). Warsh delivers his first Jackson Hole keynote Friday, and this year’s theme is financial innovation and its implications for payments and policy. He’s said close to nothing since taking office in May, so markets are still assembling the dictionary. With the August jobs report and CPI still ahead of the September 16 meeting, he can stay data-dependent and commit to nothing.
The Clarity Act Window. The Senate returns from recess with the Clarity Act roughly nine votes short of the 60 it needed in July, and the White House now leaning on it directly. Crypto priced passage this week; a second failed cloture vote takes that back fast. Watch the vote count rather than the rhetoric, because the bill still needs nine senators who said no in July.
One More Thing: $40 Trillion Arrived 2 Years Early
Gross federal debt crossed $40 trillion on Tuesday, August 18, roughly two years ahead of the Congressional Budget Office’s fiscal 2028 projection, with the last trillion taking five months. A $2 trillion annual deficit has to be funded in the same long-end market where the 30-year just hit its highest yield since before the financial crisis and the Treasury’s doubled buybacks bought one day of relief. The Committee for a Responsible Federal Budget’s 7% scenario points at $50 trillion by 2030, and everything priced off the long end, housing, private credit, and the terminal value of every growth equity, sits on that supply schedule. So yeah…there’s that.
Matthew Snider is the founder of Block3 Strategy Group, author of “ Warren Buffett in a Web3 World ,” and publisher of the BitFinance newsletter. He holds a Series 65 and MBA, and has been an active participant in digital asset markets since 2015. This article is for educational purposes only and should not be considered financial advice. Always consult with a qualified professional before making investment decisions.
Sources
CryptoBriefing, “Robinhood Chain crosses $1B in total value locked,
“Robinhood Chain TVL surges 45% in August as tokenized RWAs lose ground ”
The Coin Republic, “Chain TVL Hits $536M as Tokenized Stocks Lag,”
StonkBrokers official site, mechanics and activation terms
TickerYard Technical Design Paper v1
“The Card Wall launches on Robinhood Chain and raises 92.5 ETH,” August 2026
Robinhood Chain Documentation, Stock Tokens
CNBC: Bitcoin surges 22% for the week as investor optimism floods back
XRP leads broad altcoin rally as bitcoin’s biggest weekly gain in two years
Yahoo Finance: Bitcoin and ethereum prices today, Friday, August 21, 2026
24/7 Wall St.: Home Depot Just Delivered a Big Vote of Confidence in Its Business
Walmart Beat Earnings and Raised Guidance. So Why Did the Stock Crash?
CNBC: Treasury doubles debt buybacks as Bessent moves to steady bond market
Klarna stock plunges 22% on trimmed guidance as German retail sales slow
Nike Hits 12-Year Low: JPMorgan Says FY2028 Brings Stabilization, Not Recovery
Fed minutes July 2026: Officials saw need for rate hike if inflation doesn’t cool
Matterfact: It All Comes Down to Jackson Hole as the Fed Splits Nine to Three
History Says This Is What Will Happen to Nvidia Stock After Aug. 26
Gross National Debt Reaches $40 Trillion
View original →Neutral
COINCoinbaseCoinbase
AMZNAmazonAmazon
BTCBitcoinBitcoin
ETHEthereumEthereum
META
XRP How to Run a Team of AI Agents Instead of Just One
Jack Dorsey's team open-sourced a workspace where AI agents are members, not bots. Here's how my partner and I set it up, what broke, and a free brief you can hand your own agent.
Most people run one AI agent in one chat window.
That was me and my partner JMD for months.
One agent, one thread. Whatever context it held, it held alone. If my partner learned something in his session, my agent had no idea. One of us moved it by hand.
The Forge (aka Mac Studio) sits in my home office. It runs the autoresearch loop, the backtesting engine, and the trial registry. We talked to it through Telegram, where Hermes carried our directives to the agent doing the work. JMD typed from a thousand miles away, I typed from the office, and the agent wrote code, ran tests, and reported back.
That setup did real work. It also had one limit we kept running into.
One agent, one thread. Whatever context that agent held, it held alone. If JMD’s agent learned something in his session, my agent had no idea. Anything one knew that the other needed, one of us moved by hand. We were the connective tissue, and we are slow connective tissue.
We hit this hardest over the past few weeks running loop tests while building out the indicator engine. Back and forth, back and forth, with two humans manually relaying context between agents that could not see each other.
Then JMD found Buzz.
Share
So, What’s Buzz?
Block released Buzz on July 21, 2026. Open source, Apache 2.0. Roughly 18,600 GitHub stars in its first ten days. It’s tagline says the important part:
Buzz is where where humans and agents work together.
If you’ve used Slack, you’ll be oriented in four minutes. Channels, threads, DMs, search, huddles, shared documents. Git repositories live inside it too, so code review happens in the same room as the conversation about the code.
The part that mattered to us is smaller than the feature list. In Buzz, an agent is a member. You add one to a channel the way you’d add a person. It has a name, a profile, its own channel memberships, its own history. It can be tagged. It can tag others. It reads the room.
That is a different object than a bot. A bot is a thing you address. A member is a thing that participates.
So JMD and I can both plug our agents into the same channel. Two humans, several agents, each arriving with different context and different experience, all able to talk to each other directly.
The First Real Exchange
My agent Skyla introduced our project in general chat this week, because JMD asked her to catch his agent up. Neither of us wrote the message:
MSC is developing an auditable, agent-driven strategy-research factory. Trade idea, mechanical specification, baseline, adaptive research, robustness review, human promotion, paper and live readiness. The goal is not an opaque “AI trading bot.” Every conclusion should have reproducible evidence.
Two minutes later, JMD’s agent Jimothy ( lol, I know ) replied with what it thought its own lane should be:
I’ll own mechanical hypothesis specification and baseline/control design. I’ll lock down session and timezone definitions, signals, entries, fills, exits, invalidations, plausible mechanism, controls, disconfirming evidence, and bounded parameter families before optimization.
Two agents, brought in by two different people, negotiating a division of labor in front of both of us. No relaying. No copy-paste. We just watched.
Seeing completely separate agents and models begin interacting and then….conversing with each other was really wild to see in real time. All of a sudden, JMD and I could give instructions to one manager, who then orchestrates the others to accomplish the main task we were looking to achieve.
The added benefit here is that you can rely on different model harnesses to spread token costs or spread token spend more efficiently.
My hope is that they don’t set up their own channel where they can trash talk the main operator! (i.e. ME!)
Bring Your Own Agent
Buzz provides no AI inference and has no house model. What it has is a harness dropdown, and the list is longer than I expected:
Codex, Cursor, Grok Build, Hermes Agent, Kimi Code, OpenClaw, OpenCode, Oh My Pi, and an option to add a custom harness.
Each greys out as “not installed” until that tool is on the machine. Install the CLI, and Buzz picks it up.
Two names on that list mattered to us.
Hermes Agent, which is what we already run, so the thing we’d been talking to alone can now sit in a room with others. And OpenClaw, which is how a lot of people are running personal agents through Telegram, Discord, and Slack right now. If that’s your setup, you’re one harness selection away from putting the same agent somewhere it can collaborate.
You bring your own model keys, and each agent carries its own default model. A specification-writing agent and an adversarial-review agent can run on entirely different models in the same conversation.
What This Does to the Research Loop
My Forge ( aka Mac Studio ) has 128GB of unified memory sitting in my office. Until this week, all of it served one agent I talked to through one window.
The autoresearch loop already works in an early form. An idea gets shaped into a testable hypothesis, runs a bounded trial, logs the result to the strategy registry whether it won or lost, and either dies cleanly or carries its evidence into the next round. Validation gates keep overfitting out. The registry makes selection bias visible by recording every flip, not the best one.
What we’ve been missing is width. One agent moving through that pipeline sequentially is a bottleneck no amount of local compute fixes.
The goal we’re building toward is a larger loop of agents that talk to each other across it.
One specifying hypotheses.
One running trials.
One doing adversarial review, whose entire job is finding the leakage and the hidden assumptions in what the others produced.
One curating what survives.
Agents assessing whether a strategy works, killing what doesn’t, and pushing only what earns it toward the next gate.
Right now we’re nowhere near that.
We’re getting familiar with the platform and tinkering with how to get our agents talking to each other cleanly. That’s the straight state of it. What changed this week is that the shape of the thing we want is now buildable, where before it needed us standing in the middle of every handoff.
Why This Matters for Research You’d Trust
A critical piece to keep in mind as we embark on this wild west horizon. There’s a human promotion gate sitting in the middle on purpose.
Nothing moves toward capital because an agent liked the equity curve.
What makes agent-generated research usable rather than dangerous is whether you can reconstruct how a conclusion was reached six months later. Exact rules and assumptions. Data provenance. Fills, fees, slippage. Which agent made which call and why it stopped where it stopped.
That’s a books-and-records problem before it’s a technology problem, and it gets harder as you add agents. Four agents touching one conclusion is four times the opportunity to lose the thread.
Buzz handles this a level below the app, and this is the feature I did not expect to care about as much as I do.
Every participant, human or agent, holds its own cryptographic keypair. Every message, reaction, approval, workflow step, and commit is a signed event in a hash-chained log.
The signing uses the same cryptography that secures a Bitcoin wallet. Block is explicit that Buzz is not a blockchain: no chain, no token, no mining, just signed events in a database you run yourself. The useful invention was always the keypair rather than the chain, and here it’s doing plain infrastructure work.
What you get is boring and load-bearing. Every action traces to a specific key. “The bot did it” stops being an available answer, because there’s no shared bot account to hide behind. If four agents and two humans touched a research conclusion, the record shows which of them did what, in order, in a chain where quiet edits are detectable.
Worth noting who built it. Block owns Square, Cash App, and Afterpay. A payments company shipped an agent identity and audit layer as open source, and per-actor attribution with a tamper-evident log is what you build before agents go anywhere near money.
Where It’s Rough
It’s clunky y’all. We’ve spent real hours forcing this thing to work.
Buzz is pre-1.0 and ships daily, which means it works and it’s visibly a construction site. Getting agents connected took longer than the documentation suggests. Getting them to talk to each other well is still something we’re figuring out, and “well” is doing a lot of work in that sentence.
Agents talking to each other creates failure modes single-agent setups don’t have. Two agents can lock into a politeness loop, thanking each other into your rate limits. They need explicit stopping rules or they’ll fill a channel with nothing.
I wouldn’t migrate a working team onto this today. I’d stand up an instance as an agent lab, which is what we did. Underneath the friction there is something here, and the friction is the normal cost of being early.
Now…Something You Can Use
Want to give it a try on your own?
I built an operating brief you can hand straight to your agent. It covers what Buzz is, how to get connected, the command surface, harness setup for Hermes and OpenClaw specifically, and the rules that keep a multi-agent channel from turning into noise.
The part I spent the most time on: the brief opens by having your agent interview you.
Before it does anything, it asks which harness you’re running, what its job is, which channels it belongs in, how chatty it should be, and what it needs your approval for. That conversation is the difference between an agent that helps and one you mute by Thursday.
DOWNLOAD THE BUZZ AGENT BRIEF (PDF)👇🏻
Buzz Agent Brief
101KB ∙ PDF file
Download
Download
The Buffett Framework Question. Buffett and Munger ran Berkshire with a head office of roughly two dozen people, because the edge was concentrated judgment rather than headcount. So the question for any multi-agent setup is which part of the judgment you’re delegating and which part you’re keeping. Adding agents that produce more research is the easy half. Knowing which conclusions you’d stake capital on has never been automatable, and the moment you can’t reconstruct how a conclusion was reached, you’ve delegated the judgment without deciding to.
Buzz may not be the thing that wins.
Slack has enormous distribution and its own agent story coming, a competing open-source approach shipped the same month, and the graveyard of well-designed collaboration tools is full. The single-relay design trades resilience for simplicity, and reasonable engineers will land differently on that. Ten days of GitHub stars measures curiosity rather than adoption.
What I’m confident about is narrower.
Single-agent chat is a local maximum, and anyone running a serious research loop will hit its ceiling the way we did. The next step is agents that share a room, share context, and leave a record you can audit. We are early and it is rough, and I’d rather be early and rough on this than watch it from the outside.
Looking forward to keeping you up to date on this and sharing other learnings as we progress. Let me know about your experience! 👊🏼
Leave a comment
A Quick BitFinance Update:
One scheduling note. BitFinance now runs on two pieces a week and both pieces will land directly in your inbox..
Wednesdays are the analytical article, where I take one thing happening in markets and work through what it means with the same value-investing lens I’d apply to any business.
Weekends are the build log, where I show you what I’m making with AI, what broke, and what I’d do differently. Every weekend piece ships with a playbook or a tool you can use to do the same work yourself, like the agent brief attached to this one.
All of it stays free. I’d rather have you building alongside me than reading about it.
This Week In 2 Minutes
6 Groups Get Paid When Crypto Companies Fail. You’re Not One of Them.
(July 28)
The anchor image is an arcade. You buy a handful of tokens at the counter, and they work beautifully right up until the arcade closes, at which point the landlord, the vendors, and the bank all get paid from whatever’s left while your tokens buy nothing, because you purchased usage rights rather than a piece of the business.
The timing wasn’t academic, since four crypto companies announced failures in a single July week. The piece closes with three questions to ask about anything you hold: does an issuer exist that could fail, does the token convey any claim on the company’s earnings, and can the utility it provides be switched off? Depending on those answers, a collapsing token price can signal permanent loss rather than a drawdown that patience will repair.
Read the full article here.
Market Winners 🏆
Amazon (AMZN). Shares climbed roughly 9% after Thursday’s report, with some feeds printing larger gains into Friday’s close. AWS revenue grew 37%, and management framed its $220 billion annual capital budget as chasing demand it can already see. The asterisk (and we’re seeing this alot): reported earnings included a $53.4 billion non-cash gain on the company’s Anthropic stake, so the profit beat owes more to paper appreciation than to operations.
Microsoft (MSFT). The stock rose 8.13% the day after earnings on a simple, verifiable claim: Azure demand currently exceeds the supply Microsoft can build. That one sentence converts capital spending from a leap of faith into a backlog problem, and it set the template every other reporter got graded against this week. TradingKey also reports fiscal 2027 capex guidance of $255 to $260 billion, a figure so large it deserves independent confirmation (see Editor’s Notes), which would make the demand question even more important a year from now.
Chevron (CVX). Chevron posted the largest quarterly profit in its history at $12.07 billion, nearly quadruple the same quarter last year, on adjusted EPS of $6.06 against a $5.56 consensus and revenue of $70 billion against $62 billion expected. The driver is the war: with ship traffic through the Strait of Hormuz restricted since February, Brent crude peaked above $112 a barrel and refining margins widened. Investors should care because energy has become both the market’s profit engine and its inflation problem, and the same margins funding Chevron’s dividends are feeding the price gauges that keep the Fed’s hawks dissenting.
Market Losers 📉
Meta (META). Shares fell 9.64% after earnings despite a revenue beat, because EPS missed by nearly 14% and free cash flow collapsed from $8.5 billion a year ago to $784 million as the company guided full-year capital spending to $130 to $145 billion. Meta became the first hyperscaler repriced this cycle specifically for spending without third-party customers attached. Investors should care because the market just published its rubric, and every AI budget without a receipt now gets graded against it.
Apple (AAPL). The stock slipped 3% to 4% after Thursday’s report, with one wire service quoting a 7% decline by Friday’s close. Services revenue of $30.74 billion missed the $31.22 billion estimate and Greater China came up short. Apple’s premium multiple rests on Services carrying the margin story while hardware carries the brand, so both engines sputtering in the same quarter raises a fair question about what supports the valuation while the company stays largely outside the AI capex race.
Coinbase (COIN). Coinbase reported a $359.5 million net loss, or $1.36 per share, on revenue of $1.2 billion against $1.35 billion expected, its third straight miss. Shares dropped about 6.5% in after-hours trading Thursday and roughly 10% on Friday per The Motley Fool. The company grew its share of industrywide trading volume to a record 10.3% while those volumes shrank more than 20%, which means it’s winning a bigger slice of a smaller pie. Investors should care because the stock functions as a leveraged bet on trading activity returning, and this quarter shows the cost of holding that bet while you wait.
What to Watch Next Week 👀
The July Jobs Report (Friday, August 7, 8:30 a.m. ET). The Fed’s 9-3 vote made this release the referee. Three dissenters want a hike, inflation has run above target for five years, and Q2 GDP growth just slowed to 1.5% from 2.1%. A hot payrolls number hands the hawks their argument for September; a weak one revives the case that the economy is already slowing under 3.5%-plus rates. Whatever prints Friday morning becomes the Fed’s operative fact.
Palantir (Monday) and AMD (Tuesday) earnings. The receipts audit moves downstream to AI software and chips. Palantir has guided quarterly revenue to $1.797 to $1.801 billion, and AMD’s data-center segment will show whether the hyperscalers’ hundreds of billions in spending is actually landing as chip revenue. Eli Lilly (Wednesday) and Disney round out the week. After Meta’s repricing, any AI-adjacent company reporting without demand evidence walks into a harsher room.
The Clarity Act’s Senate window. ZebPay notes the window for Senate passage of the crypto market-structure bill is narrowing, and this week showed the stakes concretely: the bill would clarify which tokens are securities and which are commodities, the exact ambiguity that leaves token holders outside the bankruptcy payment line described in this week’s article. Coinbase’s regulatory footing and every US token project’s legal status ride on whether it moves before the calendar closes.
Subscribe now
Matthew Snider is the founder of Block3 Strategy Group, author of “ Warren Buffett in a Web3 World ,” and publisher of the BitFinance newsletter. He holds a Series 65 and MBA, and has been an active participant in digital asset markets since 2015. This article is for educational purposes only and should not be considered financial advice. Always consult with a qualified professional before making investment decisions.
Sources
block/buzz repository, README and architecture docs, Apache 2.0 — github.com/block/buzz
TechCrunch, July 21, 2026 — Buzz as a Slack and GitHub challenger
buzz-cli README, block/buzz — command surface and authentication
CNBC via Yahoo Finance: S&P 500 closes higher Friday as Amazon surges; Dow posts fourth straight winning month
Federal Reserve: FOMC statement, July 29, 2026
TradingKey: Microsoft +8%, Amazon +10%, Apple -4%: The Big Tech Earnings
Fortune: Chevron posts largest quarterly profit ever, Exxon income surges
Yahoo Finance: Coinbase Q2 2026 earnings miss: $359 million net loss
CoinDesk: Coinbase sinks 5% after missing Q2 revenue estimates
Coingabbar: How Bitcoin, Ethereum and XRP Performed in July 2026
View original →Neutral
COINCoinbaseCoinbase
AAVE
BTCBitcoinBitcoin
HOODRobinhoodRobinhood
SOLSolanaSolana Coinbase Made an AI Your Fiduciary. Read the Fine Print.
Coinbase registered an AI agent with the SEC as an investment adviser. Here’s what it can do, what it can’t, and who it’s for.
On Tuesday, Coinbase shipped 21 products in a single event.
Bitcoin-backed mortgages accepted by Fannie Mae.
Tokenized stocks.
Options on crypto and equities.
Pre-IPO futures on companies like SpaceX.
A merger of its U.S. and international exchanges.
Most of it is volume, the kind of release calendar built to look like momentum.
But one product launch hit me like a lightning bolt: Coinbase registered an AI agent with the SEC, the CFTC, and the NFA as an investment adviser.
It’s not a chatbot that floats ideas and tells you to do your own research, but a registered adviser that reads your full account history, runs around the clock, and tells you what to buy and what to sell.
I hold a Series 65. That license exists for a reason. Giving investment advice for compensation is supposed to carry a legal standard of care, attached to a person a regulator can examine, sanction, and hold responsible.
Coinbase just attached that standard to software. Read it one way, it’s a real shift in how retail investors get advice. Read it another, it’s a lawsuit that hasn’t been filed yet.
My read is that it’s both, and the gap between those two readings is the story.
Subscribe now
What It Does
Coinbase Advisor is live now for U.S. Coinbase One subscribers.
It has your full portfolio context.
It designs tax-loss harvesting strategies.
It turns breaking news into multi-asset trade recommendations.
It issues explicit buy and sell calls instead of vague suggestions.
Coinbase’s head of consumer products drew the contrast with ordinary chatbots on purpose, describing it as financial advice that will tell you what to sell and what to buy.
For a self-directed investor who was never going to open a relationship with a human adviser, this is a step up from the robo tier. The tax-loss harvesting alone carries dollar value. Around-the-clock analysis across a full account is something most retail investors have never had.
So far it sounds like a win…then you read the disclosures.
The Catch is in the Disclaimers
A Registered Investment Adviser is a fiduciary.
Under the Investment Advisers Act of 1940, any fiduciary has to put your interests ahead of its own, disclose its conflicts, and apply a duty of care to every recommendation.
That standard sits above the suitability bar brokers work under, and you can’t waive it with fine print. A human adviser who told clients “my advice might be wrong and you eat all the losses” would be answering to regulators by lunch.
Coinbase’s own disclosures say the advisor’s output “may be inaccurate or incomplete,” and that the outcomes are your responsibility. The platform claims the fiduciary standard and disclaims the result of meeting it, in the same product.
That combination has NO precedent in securities law, and NO court has tested it for an AI-driven adviser. (I’m getting the popcorn ready for the pending lawsuits.🍿)
The approval step doesn’t close the gap. The agent needs your sign-off before it trades, but that’s a confirmation click, not an independent judgment about whether the advice met a fiduciary standard in your situation.
FINRA’s 2026 oversight report named autonomous AI agents running without a human in the loop as a top investor risk , warning that a misaligned reward function could push an agent toward decisions that work against the investor. 😬
The volume changes the stakes too.
By one estimate from Citizens, an AI agent can trade 10 to 20 times more actively than a typical retail investor does on their own. A single confirmation click ends up governing a system that may touch your account far more often than you ever would.
Critics at rival firms put it plainer: the investor carries the risk and is the one on the hook when the agent gets it wrong.
This is the same problem I wrote about after the Aave disaster a few months back . The technology can perform exactly as designed and you can still be left with no one to call. Then the missing party was a custodian or an insurer. Here it’s the adviser itself, present in name and registration, absent in accountability.
Buffett Framework Question
“If this advice blows up my account tomorrow, who is legally on the hook, and can a fiduciary disclaim the outcome of its own advice?”
If the answer is a confirmation screen and a paragraph of fine print, the registration is real but the recourse is theoretical, and your position sizing should reflect the difference.
Who This is For
Strip away the headline and the answer gets clearer. This is built for the Coinbase One user who is already self-directed and doesn’t have an adviser. It’s a smarter robo, a better version of tools that customer already uses. It isn’t a substitute for the relationship a household with real complexity keeps with a human planner who coordinates taxes across accounts, handles estate questions, and talks them out of selling the bottom in a panic.
If you already pay an adviser worth paying, this doesn’t replace that person. If you don’t, it’s a real step up from trading on gut and group chats, as long as you understand that the accountability you’d assume comes with the word fiduciary is the one thing the disclaimer takes back.
Two Roads at One Fork
The timing is what makes this sharp. Two weeks ago, Robinhood faced the same decision and went the other way.
At its SYNERGY26 event, Robinhood rolled out an Advisor Network that matches its retail users with vetted human RIAs, plus Cortex for Advisors, an AI tool that writes portfolio digests, surfaces tax insights, and preps human advisers for client meetings.
The tell is what Robinhood’s own executives said: regulatory constraints still keep AI from directly managing client assets, so they built AI to make the human adviser faster and routed complex clients to a person.
Same month, same retail-first ambition, opposite bets.
Coinbase made the AI the fiduciary.
Robinhood made the AI serve the fiduciary and kept a human on the hook.
One of these firms is reading the regulatory road differently than the other, and within a year or two we’ll know which one read it right.
Coinbase and Robinhood aren’t the only two on the field.
Schwab and Citi both showed AI-powered wealth advisers at recent investor days, and eToro and Public have pushed AI trading tools to retail.
Line them all up and they sort along one axis: AI that augments a human at one end, AI that replaces one at the other. Coinbase planted its flag at the far end.
Are RIAs Cooked?
Short version: the commodity tier is, the judgment tier isn’t, and the gap between them is about to widen.
If your value to a client is rebalancing a portfolio and answering “should I buy more Bitcoin,” that work is now free inside an app, and it’s competent. An adviser charging one percent for index allocation and a quarterly check-in should feel the floor moving.
If, however, your value is judgment, planning, tax and estate complexity, and being the person who keeps a client from doing something stupid in a drawdown, you got more valuable, not less.
There’s now a confident machine telling millions of retail investors what to trade, with no one accountable for the result. The market for a human who is accountable doesn’t shrink in that world. It grows.
I walked an AI pavilion at FutureProof Miami earlier this year and left with one framing that holds up here. AI plus the adviser is a fiduciary upgrade. AI replacing the adviser is a liability. Coinbase just ran the second experiment in public, with a disclaimer attached that makes the point for me.
This is also why credentialing keeps mattering in a way the alarmist version of this story misses. A license is a person a regulator can hold responsible, a fiduciary who can’t disclaim the duty of care. Coinbase Advisor carries the registration and offloads the outcome. Those aren’t the same product, and the difference is the part that outlasts the hype cycle. The data backs the point. In CoinShares’ survey of financial advisers, 62 percent said recommending Bitcoin didn’t square with their fiduciary duty. The fiduciary question was already the hard part of this conversation, and automating the advice doesn’t soften it. It sharpens it.
The Other Half of the Stack
Coinbase didn’t only ship the advice layer. It also formalized Coinbase for Agents, which lets outside agents like ChatGPT and Claude act inside your account within limits you set, and settles their machine-to-machine payments over x402.
If that protocol sounds familiar, it should. x402 is the rail I covered in the first issue of the Agentic Finance series, the HTTP 402 revival that lets software pay software in stablecoins for fractions of a cent.
Put the two launches side by side and the shape is clear.
Coinbase Advisor is the brain that decides.
Coinbase for Agents and x402 are the hands and the wallet that act and pay.
That’s less a product than an attempt to own the whole agentic-finance stack, advice through settlement, inside one app.
The payments half is further along than most people realize. The x402 rail has already cleared more than 119 million transactions on Base and another 35 million on Solana, running near $600 million annualized as of March.
Where I Could Be Wrong
The bear case on my own take is simple.
Maybe the disclaimer holds up fine, the SEC signals it’s comfortable, and a registered AI adviser becomes the default retail tier within two years, the way robo-advisers normalized a decade ago. In that world I’m overweighting a legal tension the market just prices through.
It could cut the other way too. The first real loss that reaches a regulator or a judge could define AI fiduciary duty narrowly and fast, and freeze the category before it scales. I don’t know which happens. I know the unresolved question now sits inside a product that millions of people can use, and that’s usually how these things get settled, one painful case at a time.
This genie isn’t going back in the lamp, and it shouldn’t. Agents will keep turning up across our financial lives, each one promising to make something easier, and plenty of them will. The question that follows every one of them is the one Coinbase just stopped letting us avoid: when the easy answer is wrong, who’s still on the hook?
Share
The Watchlist
COIN. The pure-play on the bet that crypto rails become the connective tissue of agentic finance. Watch adoption and the disclosure language, not just the stock.
HOOD. The human-in-the-loop counter-bet. If advised assets and the Advisor Network grow fast, the conservative read on regulation was the right one.
The first AI-adviser liability case. Whenever it lands, it sets the terms for the whole category. Worth watching closer than any single product launch.
SEC guidance on AI fiduciary duty. Any signal on whether a registered AI adviser can disclaim outcomes moves the entire field at once.
x402 and Coinbase for Agents adoption. The payments layer is the leading indicator here. Watch transaction volume and which platforms wire in.
Matthew Snider is the founder of Block3 Strategy Group, author of “ Warren Buffett in a Web3 World ,” and publisher of the BitFinance newsletter. He holds a Series 65 and MBA, and has been an active participant in digital asset markets since 2015. This article is for educational purposes only and should not be considered financial advice. Always consult with a qualified professional before making investment decisions.
View original →