I wanted to pay tribute to Perry Mehrling, one of my great mentors, by translating my vision of his Money View applied to Tokenized Money into a full-length article I'm on it But in the meantime, I wanted to share the hierarchy of Tokenized Money and how it has shifted over the years In the article, I will explain how this could potentially shift even further in the future, for instance, if regulated payment stablecoins end up permeating the central bank layer and entering systemic territory Thank you, Perry, for every minute of your insight

Settlement in a tokenized system is instant, though there's plenty to argue about what that actually means before we scare the treasury desk that needs $500m for a repo. And its ledgers never close, which scares the hedge funds who enjoy a beer with their families on a Saturday. Both can raise liquidity needs at first, depending on how settlement is set up. In current systems a large share of obligations is netted before settling, and only the net balance moves. That doesn't mean you can't settle net on a blockchain. But netting needs a window in which obligations build up, and the faster things move the less there is left to net, so intraday funding and settlement peaks go up. That extra requirement eases if the same balance gets reused several times during the day, although reuse depends on incoming payments arriving on time and shrinks in periods of stress, when everyone holds back their payments. The other problem is that a balance sitting on one ledger is no use for paying on another until someone moves it. And the weekend has to be covered with whatever you already hold, because the money market is closed (I hope that changes soon). The most lasting change will be that liquidity gets a price within the day. Today the shortest tenor that trades as a matter of course is one day, and even if that's just a standard by social convention and going shorter is theoretically viable, tell that to an industry that would have to change its whole operating model and every model built around it. What happens between the open and the close is covered with firms' own buffers and with intraday lines that the custodian bank or the prime broker extends as part of the relationship. If collateral can move in minutes, funding for a matter of hours becomes viable, and the moment you pay takes on a cost or a return. The collateral that today sits spread across custodians and clearing houses just in case also comes down, because it can be sent when it's needed. Shortening the settlement cycle reduces counterparty exposure and margin requirements, and shifts that reduction into a greater need for immediate cash. It's an exciting time to see whose balance sheets will hold up the system in the future.

The price of a stock is the purest expression of information asymmetry in markets I’m not a big fan of the initial design of TSVs. For perfectly valid regulatory reasons, they leave participants highly exposed, and I don’t think they can serve as a real-world test of how listed equity markets should work, even if they sidestep the NMS rules. Revisiting those rules will take many years and won’t be easy. Still, the fundamental tension is clear: Risk management versus accessibility I don’t deny that the LPs supplying liquidity to AMM pools can be sophisticated. But an LP in an AMM is, almost by design, candy for arbitrageurs: it can’t reprice its quote when the stock moves elsewhere. On a pure capital-efficiency basis, I don’t believe this is the right model for the future of capital markets. I could be wrong Mutualization among sophisticated traders might create new incentives that make the model viable. If that happens, though, it would effectively mean retail is subsidizing the cost of that lack of sophistication, because a pool charges uninformed flow the same fee as informed flow That said, it doesn’t mean these venues shouldn’t exist. Mutualization can produce environments that are less sensitive to information, more accessible, and more elastic relative to the traditional stock-pricing system.

It's fascinating to watch compute behavior because it has literally become the strategic asset that gold is for central banks, but for the so-called "central banks" of AI technological sovereignty (aka hyperscalers) If you're a neocloud, understand that your position is quite vulnerable because the market sees you as necessary yet replaceable

I have blind faith that one of the greatest use cases for the AI and monetary/financial tokenization pairing will be in the treasury management of major digital corporations (and eventually non-digital ones too) That capacity to usher in a quantum era of money where, acknowledging other factors like regulation, capital controls, and liquidity buffers for black swan contingencies, money can move completely frictionlessly between tokenized deposits for banking relationships, tokenized money market funds to yield returns on idle capital while freeing up dealer balance sheets, and stablecoins for transactional efficiency

The end-state for on-chain corporate treasury is a reality where the treasury is a vault and the Treasurer its curator TDs will manage bank relationships for revolvers, asset-backed lending, and core facilities, while seamlessly shifting between stablecoins, tokenized reserves, and tokenized assets like MMFs

We're seeing significant traction in stablecoin use cases for a much better delivery of both USD PnL from NDFs and currency settlement in FX forwards Happy to chat with anyone looking to optimize capital efficiency thanks to T+0 settlement

In Brazil, card sales settle 30 days later. That’s a real cash flow gap, so many merchants sell their card receivables early. Funding is mostly from local capital which is expensive. Onchain lenders can bring a much broader pool of capital, which should mean better pricing over time.

Yesterday I had the great pleasure of speaking to the Securities Services team at BNP Paribas in Madrid. Many thanks to David Calabaza for the invitation. We covered a bit of everything regarding the tokenization of financial markets, from how it impacts money markets to how it connects with capital markets, and I wanted to share a few thoughts here. Someone asked how I would define tokenization for a person who has no idea what a blockchain is. I didn't have a rehearsed pitch ready, so my mind went to Marc Levinson's book, The Box, and how shipping containers transformed global trade. Before containers, everything was shipped loose as breakbulk cargo, in boxes, sacks, and barrels, and every load had to be handled differently: loading a ship took days, fixed infrastructure costs were sky-high, and shrinkage and theft were constant, even if the system seemed efficient at the time. Once the container was standardized, you could suddenly move goods from truck to ship to train without ever touching what was inside. That is essentially what tokenization does for assets: instead of every FMI, bond, equity, or form of money operating in its own operational and intermediation silo, you put it inside a standard digital "container." Now, the usual pushback is completely logical: "Sure, but a tokenized financial asset still carries all its underlying legal complexity, remains subject to local regulations, and the underlying asset is still what actually matters." I completely agree. But this reminds me of the corner fittings on shipping containers: standardization didn't eliminate customs at every port or make the cargo inside stop mattering. Previously, every shipping line had its own rigging system, meaning a crane at one port couldn't lift another company's box. Once corner fittings were standardized and the patents were released, any crane could handle any container. At the time, everyone said it was impossible, that it required too much investment, that it would clash with regulations, and that there would never be consensus. We cannot evaluate a structural shift by assuming current variables are static, or we will fall into severe myopia. Systems naturally reorganize themselves; initial friction shapes the path, but as soon as incentives shift, the regulator ultimately adapts to the new environment. We also spoke at length about tokenized money. I shared the same view I always stand by: all formats will coexist, and each will serve its own purpose. From tokenized central bank reserves acting as the anchor of the system (wCBDC), to tokenized deposits (TD) preserving the banking relationship and credit creation, all the way to stablecoins opening up corridors with greater agility, including natively on-chain settlement systems like Tempo or Ethereum to tap into their native global programmability and composability. Today they are viewed as competitors because, let's be honest, they are, especially when one has a multi-year operational head start over the other. The problem is that many banks are jumping into stablecoin and tokenized deposit initiatives without a clear understanding of the "why." It is the equivalent of buying AI licenses for a company without changing its culture or its processes. With tokenized deposits in particular, you need a deep modernization of the core banking stack to operate in real time; in fact, once you reach that point, you realize that the token itself is the most irrelevant part of the entire internal transformation you had to execute. If you allow me to lean into a crypto analogy and oversimplify what I have in my head, the real magic of all this lies in a multinational treasurer's ability to move capital without friction: the corporate treasury starts functioning like a vault, the CFO acts as its curator, and the flow across Tokenized Deposits <> Stablecoins <> MMFs moves atomically without cut-offs, without liquidity trapped in precautionary buffers, and without today's heavy operational overhead. Finally, we touched on my favorite topic: liquidity and the false myth that T+0 is free. Today's financial system runs on a constant recycling of liquidity (netting), and that is possible precisely because settlement is not instantaneous and because we have indirect legal ownership that facilitates the fungibility of instruments. To me, the efficiency of the system lives in a trilemma across three variables: settlement speed, the collateral required to cover counterparty risk (margins at the intermediary), and the liquidity needed to fund that settlement. If you accelerate settlement toward real time, you reduce counterparty risk and margin at the intermediary, but you drastically increase the need for pre-funding and gross liquidity. If you slow down the settlement cycle, you save a massive amount of liquidity thanks to netting, but you increase counterparty risk exposure throughout the cycle. The market lives in that constant balancing act. If the destination is T+0, market structure will reorganize itself to solve for that liquidity cost, just as central banks continue evolving their facilities to maintain the stability of the system. Once again, thank you to the BNP Paribas team for the debate and the opportunity.