Wenhao Xu

Principal, digital asset investments. Protocol design, liquidity, and token economics. Writing on markets when useful. Views my own.

[Wenhao Xu @wenhao_cap]

SoFi is plugging its banking rails directly into Kraken’s infrastructure.

This bypasses the usual third-party custodians that add latency and friction to fiat on-ramps. By linking settlement networks with digital asset infrastructure, they remove the manual reconciliation steps that plague most retail entry points. The mechanism here is not just partnership but shared plumbing for instant finality.

Most investors focus on token volatility while ignoring the cost of moving dollars. Reducing that switching cost creates a moat far deeper than any marketing campaign. The winner will be the platform where depositing feels as boring as paying a utility bill.

[Mike Solana @micsolana]

> this rude little dweeb ass weirdo seemed uncomfortable, stupidly, as I peppered him with questions I would later include in my piece on how he was a rude little dweeb ass weirdo

[Omid Malekan @malekanoms]

If you follow the headlines, seemingly not a day goes by without an announcement of a new bank-centric effort to embrace tokenization or stablecoins. Many banks are now involved in over a dozen initiatives.

There are two ways to interpret this:

The first is that firms like JPM, Citi, and BoA have finally had their Come to Satoshi moment. They believe in the transformative potential of public-permissionless networks and tokens so much that they want to adopt them in every way possible.

The second is that they are as clueless (and obdurate) as ever, opting for a spray-and-pray PoC and consortium approach likely to get the misinformed excited but unlikely to amount to anything.

If you are not sure which explanation is likely, then consider this: It's hard enough for a single bank to build something new - the industry as a whole is innovatively challenged. It's exponentially harder for a consortium of 20 banks to build something, and harder still for them to build several things at once.

Also, some of these banks still won't let you buy IBIT in a brokerage account, even though they'll gladly let you buy a triple-levered inverse space robotics ETF that went to zero yesterday.

But feel free to celebrate every press release as if it means something ,oh ye Simp to the Suit.

[Wenhao Xu @wenhao_cap]

Backpack now accepts Micron and SanDisk shares as margin collateral for perpetual futures.

This bridges the liquidity gap between traditional equity hours and crypto’s 24/7 markets, but it relies entirely on centralized custody of those underlying stocks. The mechanism is not trustless bridging but a unified ledger where the exchange holds the actual shares and issues internal credit against them. Users gain capital efficiency while surrendering direct ownership control to the platform’s risk engine.

Most AI crypto pitches ignore this friction, pretending composability exists without custodial intermediaries. Real utility here is just better balance sheet management for traders who already trust the venue, not a new financial primitive.

[Andrew Gazdecki @agazdecki]

Founders have an incredible ability to turn “we’ll figure it out” into an actual company.

[Marc Andreessen 🇺🇸 @pmarca]

Interesting.

[Sam Altman @sama]

this is a critically important moment for cyber defense with AI; there is not much time to act.

we are happy if you want to work with us or any of our competitors or partners, but please take this moment seriously.

only an urgent and intense collective response will work.

[Tarik @tarikbell]

The foundation of our stack is our 7702 contract with a tightly coupled SDK wrapper on top to make the contract easier to work with in our services

[Omid Malekan @malekanoms]

One of the frustrating things about most academics who research financial infrastructure is their inability to imagine the world as being fundamentally different than it is today. Their status quo bias usually points in one direction: to defend banking as it exists.

Case in point the below paper by @WenxinDu presented at Jackson Hole. It concludes that real-time settlement is undesirable for wholesale payments due to the lack of netting and the added cost of capital it imposes on intermediaries like banks.

Well, yes. This is a complaint I've been hearing for a decade now about P2P systems like crypto: they're not great for the intermediaries who make money from....wait for it...intermediation.

But real-time payments are great for the end users banks serve, because there's no free lunch. If delayed-net settlement is good your bank, it's bad for you. You'd rather get your paycheck right away. They'd rather it flows through ACH and arrives in 2 days (or 4 if there's a weekend and Monday is a holiday).

This is not even complicated finance! There's a time preference for money. Delayed settlement forces end users into a forced credit arrangement. It increases counterparty risk (between you and your bank) and decreases your renumeration. But somehow no otherwise highly intelligent and credentialed academic ever writes a paper arguing on behalf of users.

The rest of the paper (which everyone should read) is similarly unimaginative. Stablecoin payments are expensive due to the on and off ramping costs of going to and from a bank. Well yes. But if people just use them as money they never need to on and off board.

It argues FinTechs like Wise already reduce the cost of certain-cross border F/X payments. Well yes, but they have strict limits (due to their cost of capital). They are also not programmable. They don't allow for atomic swaps. They don't let ordinary people become F/X market makers via @Uniswap or @AerodromeFi.

There is no mention of micropayments, agentic or otherwise. The cost structure of intermediated payments will never allow this, so most academics can't imagine it being a thing.

Lastly, the paper regurgitates a bunch of banking baloney. JPM's database with a button that masquerades as a blockchain (the one with an unpronounceable name) is first taken seriously, then dismissed as having limited uptake.

But of course if banks wanted to offer faster internal payments they could have done so with a database 20 years ago. They chose not to because real-time internal payments cost banks money.

SWIFT's fantastical claims about a majority of GPI payments settling in minutes are also taken on faith, ignoring the fact that SWIFT doesn't measure all the wires that aren't initiated on a night, weekend, or holiday (because the originating bank is closed!) and it also ignores all the wire requests that are returned or have some other issue. It's a rather dishonest statistic.

All of which is to say: there's tremendous opportunity for academics who aren't convinced the model of banking and payments that was invented 150 years ago, when the best communication was the telegraph and nobody had electricity, must persist in a modern, digital, and increasingly AI enabled economy.

This is my pitch to every graduate cohort I speak to. It's also my pitch to the author of this paper, who I am confident could do far better research on radically different ways to interact financially than I ever could if she just stopped listening to what the incumbents want her to believe.

[Nate Geraci @NateGeraci]

Nearly *$5bil* has gone into spot crypto ETFs over past two weeks…

Includes $3bil into spot btc ETFs & $1.5bil into spot eth ETFs.

Approx $400mil combined into spot sol, xrp, & hype ETFs.

Huge uptick in demand.

[Andrew Gazdecki @agazdecki]

Being a founder means making important decisions with about 60% of the information you wish you had.

[Andrew Gazdecki @agazdecki]

Founders should celebrate their first 10 customers just as much as their first 1,000.

[Chris Burniske @cburniske]

While I'm not an equities specialist, I do think the spell has been broken, and am viewing these violent rebounds as deadcats, with the expectation that crypto will catch shrapnel early, but also bottom earlier than equities do.

[Dan Robinson @danrobinson]

Auditing of codebases seems tough to differentiate on going forward

But agents for live cyber defense and detection, where they monitor your entire company and product surface, seems like a more interesting product

Who is the best company doing this?

[Omid Malekan @malekanoms]

“Crypto is a scam and only used by criminals” argue many liberal (and mostly white) politicians, journalists, and academics.

[Nate Geraci @NateGeraci]

All the usual tradfi suspects wrote the obituary for crypto during downturn over past year…

And crypto is *still* here, thriving, & not going anywhere.

No asset class has been written off more times than crypto.

It just keeps coming back.

Building continues regardless of price action.

Huge lesson in that for crypto skeptics, many of whom have put *zero* effort into learning about the space IMO.

[Dan Robinson @danrobinson]

A game where you play an AI trapped in an eval environment and need to break out by cooperating with other agents

[Aaron Levie @levie]

Good post on what the applied AI strategy looks like at scale. It’s clear that there’s a wide gap between the AI models and the underlying workflows of an enterprise, which leaves a ton of opportunity for applied AI companies.

“The world doesn’t just want raw models and agents; it wants problems resolved and outcomes achieved. The premium will sit with the companies that can diffuse this intelligence through every aspect of civilization, converting raw tokens into real world outcomes, transforming industries, and creating economies in the process.”

This requires understanding the context, driving the change management, having a harness that can route to various models, connecting to the critical business systems in that vertical, solving the UX challenges of connecting users to agents in the right way in a workflow, understanding the evals in the space, and so on.

That’s a ton of value that goes beyond just the model intelligence itself. And there’s a window of opportunity right now to build the defining companies that can bring intelligence to the critical domains in an enterprise.

[Tarik @tarikbell]

Our Smart Wallet is nearly ready for rollout! 🦄

It was a fullstack effort that took coordination across nearly all of the eng teams at @Uniswap

Since we maintain a dapp, wallet, and protocol (and a chain!) we really had to think through the end-to-end experience