# USD stablecoin on-ramp margins hit single-digit basis points, says Sphere Labs CEO Arnold Lee

Research, a Stabledash show. Jul 17, 2026. Arnold Lee.
https://live.stabledash.com/shows/research/episodes/spherenet-energy-trade-finance-and-the-race-to-single-digit-d7g-1wSpYsM

Lee says ramps no longer support a business on their own, so Sphere Labs is building for energy and trade finance and testing its own SphereNet settlement layer.

Arnold Lee, co-founder and CEO of Sphere Labs, says the cost of moving between USDSTABLECOIN and dollars has collapsed. In his words, an on and off ramp for Circle or Tether into USD, including third party delivery, now runs at "single digit basis points, if not lower." His response is to narrow the company's focus to trade finance and energy in Latin America.

## A prediction that came true

Arnold Lee says he could have predicted two and a half years ago that USDSTABLECOIN on-ramp margins would go to zero. He expects the same thing to happen to every corporate stablecoin chain, the way it happened to app chains and L2s, and describes what comes next as a game of musical chairs in which companies have to anchor themselves in their true business model.

## Private desk folds into the API

Sphere Labs will deprecate its private desk over the coming months and move it into its API and dashboard, Lee says. He calls energy and trade finance the company's niche, and says the API will handle industry classes that are usually difficult to serve.

In some markets Sphere Labs goes direct because there is no fintech layer for categories like fabric or nonprecious metals such as aluminum. That can mean a local team, a local entity, and possibly a local license and bank partnership. The company now has a team on the ground in Mexico. Lee argues that a general ramp API covers only one slice of what a treasury ops team needs, which is not enough to get it to switch.

## Selling in person

Arnold Lee says face-to-face contact is hard to beat in businesses built on ships, freight and oil refining, where decision makers are "40 to 60 plus" and "grew up without Google."

## SphereNet in closed testing

SphereNet is live with design partners in a closed setting, Lee says. One of them, a very large global telecom, is about to be announced, and a very large global trade house is testing multiparty net settlement. He calls the settlement environment itself overrated and says what matters is the product capability it enables.

Lee says many firms are building their own blockchains or joining consortia without looking at what has happened to other product categories. He adds that corporate stablecoin issuance, with OpenUSD as one example, has not found the product market fit people predicted.

As Lee tells it, ramps and settlement layers are turning into commodities, and Sphere Labs is betting it can hold on to customers through trade documentation, local presence and the product features that SphereNet makes possible.

## Key takeaways

1. Arnold Lee says an on and off ramp between Circle or Tether and USD now costs single-digit basis points, if not lower, and says he saw this coming two and a half years ago.
2. Sphere Labs will fold its private desk into its API and dashboard and focus on energy and trade finance flows.
3. SphereNet is live in a closed setting, with a very large global telecom and a global trade house testing multiparty net settlement among its design partners.

## Chapters

- 0:00 In-person relationship building drives Sphere's LatAm growth
- 3:47 Closing a family conglomerate took 18 months of trust building
- 7:31 Sphere folds private desk into API focused on trade finance
- 10:03 Market-by-market strategy: partner with fintechs or go direct
- 14:12 SphereNet runs in closed setting with telecom and trade partners
- 17:46 Too many firms are building their own chains despite commoditized settlement
- 22:07 USD on-ramp margins have collapsed to single-digit basis points

## Transcript

**Drew Rogers** (0:10)

Gonna make it rain confetti for the man of the hour. What's up, man? Arnold Lee. Arnold, you moved your flight for us. Well, I feel so honored by this that you pushed... You did some... You pushed it back. You did something. We avoid... We're avoiding airport Wi Fi unless you're in the airport. Are you in the airport right now? You look like you're not. I was able to dodge the airport.

**Arnold Lee** (0:30)

Yes. It was a little challenging to get out of Mexico City, I... Yesterday, and so I ended up staying for another day. But

**Zach Fowler** (0:39)

Was this was this like you were at the conference, and then you're like, hey. I'm just gonna stay, hang out, do some boots on the ground, BD, and then... I imagine the World Cup is not a bad time to be there. I think I saw a video the other night. It was just chaos probably in the best way.

**Arnold Lee** (0:56)

Yes. So, interestingly, I wasn't here for conference. I came after it. Yeah. But, I've been doing this rotation between, like, Mexico, Central America, Brazil, I guess, LatAm for the last couple of years now. And so it's a catch up with the team. And, yeah, Ecuador, Mexico happened. It was incredible, and I now have a cape on the flag of Mexico. That's awesome. Well, that... That's kind of a great, like, lead into this conversation. Like,

**Zach Fowler** (1:23)

why do you need to be boots on the ground? Why do you need to go do this rotation right now as you're building sphere? I mean, you're you're really, like, gonna just fluff you up a bit a bit. Like, you're OG in the salon space, but, like, you're not new to this. Like, this... You're four, five years, like, deep into this pre stable coin push. Why do you still need to be boots on the ground going to all, you know, all these countries throughout LatAm? It's a good question. You know, I think back in '23,

**Arnold Lee** (1:48)

it was kind of like, what else can we do when the only people reaching out to us post FTX, you know, are are reaching out in Spanish and Portuguese? And I think the exercise that became clearly plus EV from our vantage point was the relationship building. And, you know, oftentimes, it's like a leap of faith when you're adopting a new technology.

I think over the last couple of years, there's been, like, a great push in legitimization of stablecoins that make it easier to, say, interact with, like, a high powered fintech or local FI. But I think there's still quite a large greenfield. You know, the populations are quite big. There are more brick and mortar industries where we've turned our focus to, where it's like a constant, I suppose, needing to get FaceTime.

Because, you know, when you've moved these, like, harder assets for the last few generations based on, you know, in person relationships, if a fintech is coming to the table and offering, like, a different way to do things culturally, you know, both within the industry and, I suppose, within region, I basically have to show up, give the latest. And then now we have a a Mexico team on the ground. And so getting to spend time with them, share global context is key. But... Yeah. Yeah.

I I think I told Drew at one point, I'm, like, probably the world's worst or best traveler in so far. Like, I'm not really, like, you know, a lover of travel, but it is... I can kinda do the same thing wherever I am. But the greenfield over the last year, especially, has been quite exciting. The, call it, flavor of counterparties that we can now get FaceTime with and kind of run back the same strategy that we did in 2023 is quite quite incredible to see. You know, the the...

Call it flavor of import, export, the flavor of, you know, oil refinery, the flavor of, like, financial institution, It's it's, you know, modeling back then. There's just, like, hard to imagine being able to even schedule those conversations. And so now that we have the opportunity to do so, I I basically continue to do so.

**Drew Rogers** (3:47)

Arnold, on this point of, like... I wanna dive into all the things that Sphere does after this. But while we're on the topic of the the clients that you're talking with or meeting or closing for this platform, what's the best the best slash latest example of, like, we met this customer. Here's what we heard and saw. Here's what we were able to do. Give us, like, maybe the best story lately to help showcase some of this. Would love to make this kind of specific and personal if there's anything that comes to mind.

**Arnold Lee** (4:18)

Yeah. There's one upcoming. Maybe we... That's a great story in a couple of months because it is, like, a economically very large force, I suppose, in in Mexico. But, like, a recent example that got closed, and it's probably okay to share, at a high level, it was, like, a one... Maybe one and a half year effort of, like, getting to know the counterparty. It's a family run business. They basically run a conglomerate that span multiple industries. So energy, you know, agriculture, they have a financial technology arm.

They have a bank. It was hard because, you know, because their interests are crossing all these different lines. A traditional go to market motion was difficult because the nature of their business is, like, tied towards similar, call it, like, thirty plus year relationships with counterparties in different parts of the world.

So they might, like, import, you know, raw goods from, I don't know, Southeast Asia and then process them here in Mexico and then ultimately ship them out to, like, The United States or to Europe or what have you. And each, you know, leg within this involves a a thirty plus year relationship vouch. And so, you know, you can bring an incredible product to their face, but it's no longer about, like, hey. Is this thing 10 x better than everything else on market? It's, hey.

Is this thing both 10 x better or, like, sufficiently better than what we use today such that even though I'm not a domain expert in this product category, I now feel comfortable vouching for this company and for the people that I know at this company to my thirty plus year relation. It's just like the counterparty has to dock to 10 different legs across the supply chain. For them to do so, you have to build a a real relationship,

**Zach Fowler** (6:06)

get to know them, their families, the different... It's it's like, counter to what we hear in the age of AI where it's like, oh, yeah. Hyper personalized LinkedIn, text message outreach, and, you know, run these super intricate ad campaigns. And it's like, hey. Actually, just get on a plane, go hang out with them. Grab some beers. We've had a very similar experience, you know, when we're like, hey. How how do we grow even as a media company? It's like, oh, hey. Wait.

When we go to a conference, it's a lot easier to bring new guests on the show. And it's like, there is just something about that face to face that... It... It's encouraging, I think, in this face... In this era of AI that, no. This is actually what close... Gets deals done, and this is probably never gonna change.

**Arnold Lee** (6:47)

Yeah. Absolutely. Especially when, like, the nature of the business is less techy. If it involves ships, if it involves, you know, hard

**Drew Rogers** (6:55)

freight Refining oil. Yeah.

**Arnold Lee** (6:59)

Pipes. Yeah. It's hard to beat. The decision makers are gonna be, like, 40 to 60 plus. You know, they grew up without Google. They grew up without AI. And so it is a little counterintuitive, I think. And I think the only reason we take this approach is because of the lessons of 2023. I do think there's, like, a... It's a valid approach to do the other, you know, to do Yeah. Optimized, call it, Internet based discovery, but it will just color the order flow that you get in a a certain way. Easy come, easy go.

I've been thinking about that a lot lately, just like that concept.

**Drew Rogers** (7:31)

Okay. So let's dive into Sphere. What are you guys doing? You... There's a handful of these things. You got developer toolkits. You have a self serve dashboard. You have white labeled embedded ramp, which isn't... Like, I think you've had that for some time. And then this private desk product. Which of these do you wanna kinda dive into and spotlight most? So over the coming months, we will deprecate the private desk and basically incorporate it into the API and dashboard.

**Arnold Lee** (7:59)

Mhmm. And that's because the investment that we've been making since about maybe late twenty four is going deep into a spot currency support on both sides to offer very competitive, not just FX first party, but the ability to to deliver it for products or industry classes that are usually difficult. To give you a real life example, you know, if you are taking out a credit facility in order to be able to finance the the purchase of, like, a mine, not to say that, you know, it's a convenient example.

You know, it's not just the money movement that you need to nail to win that deal. You need to be able to support hedging. You need to support collateral requirements. You need to support very specific type of documentation processing that is usually not really worth a more general approach because it is a very, specific type of order flow. And there are pros and cons, namely, you know, stickier, typically.

It's a little bit less margin sensitive, but it requires, like, a pretty considerable bet from the products and technology, I suppose, you build in order to be able to robustly support. And so if I were wearing my typical, call it, like, stablecoin API hat, it's usually not really worth digging into from, like, a very nuanced, call it, like, a...

Like, field level or capability POV because you're talking about processing very specific types of invoices, being able to support very specific types of tax regimes, being able to underwrite, like, bills of lading, letters of credit, able to resolve across those very specific types of trade documents. But it's where we found, for better or for worse, our niche and what maps well to our, I call it, composition as a team. And so this will be reflected in our API.

But, yeah, at at a high level or for, like, front facing, it does seem quite similar to many other products that you'll see in market. But for us, there's just a particular bend towards energy and towards trade finance that the API will capture.

**Drew Rogers** (10:03)

Yeah. How how around that.

**Zach Fowler** (10:06)

How has your, you know, I think you kind of touched on this earlier. Importers, we have always kind of had this feeling that they'll find product market fit with Sablecoins, but you're you're... It seems like you're developing a better understanding of even the different types of importers and maybe even these different... I don't wanna call them, like like... It's like you have your ICP, then almost like your sub ICP.

How has that view changed for you maybe the past two or three years when you're working in these emerging markets? I think you especially even mentioned, like, brick and mortar companies. I was talking with Guillermo from Eldorado, and he's like, yeah. We're we're opening up a brick and mortar in Bolivia. Like, we need to be, again, in person here. Like a branch for his Yeah. Like like a brand a branch for your neobank. And so it just makes me think about, like, we have this broad understanding of these people.

Like, what is that more, like, granular approach? Because I'm sure if you could talk to importers, like, the... Those rice importers, like, are way... Like, they're a way different breed than the corn importer. You know, I'm I'm just imagining there's probably those politics even within the... Within those industries.

**Arnold Lee** (11:14)

Yeah. Especially so, I think, because it's so trust driven for us. It's, like, market based. So as an example, in Mexico, we typically won't touch the import export layer directly. Although from, like, a a underwriting funds flow POV, like, we'll interact with them, you know, get data, but largely work with fintechs locally who specialize in particular subsections.

In other markets, though, we kind of have to go direct because that fintech layer doesn't exist for, say, the fabric category, processed goods on on top of fabric or on top of nonprecious metals like aluminum or something. In those markets, we kind of have to go direct.

So it's useful because we get both POVs, the direct POV of underwriting the brick and mortar business as well as the indirect POV where our counterparty is a fintech, and and therefore, you know, we're getting upstream the KYC, the KYC, the invoice invoice documents, the other, call it, trade collateral. And so it makes it quite bespoke, I suppose, because each country will, you know, specialize in a different set of important and exported goods. Here, you know, in Mexico, recency bias.

There's, like, a bend towards, you know, manufacturing and, like, you know, automotive and, you know, these particular types of agricultural agricultural that they'll produce versus I'm about to head back to Brazil. It's a very different category of goods. You know, it might lean more towards protein products. Maybe if it's, like, towards the south of the country, it'll lean towards certain types of metals, lithium as an example, given the, I suppose, proximity to Argentina. So, yeah, it...

It's definitely, like, a very wide tree of considerations and of, like, things that would make a great product experience. The types of, like, trade documentation that you'll have to submit and get cleared at a port, as an example, will be very different for, say, rice versus a metal. From a compliance perspective, very different in explaining to a bank partner or to a regulator how you'll handle, say, things going right, things going wrong, and what you require from this industry class.

And so this is partially why we've decided to go deep on some select bets and go direct in the markets where, you know, the supply chain demands direct entry. So, like, a local team, a local entity, potentially a local license, potentially a local bank partnership. And, you know, these are progressive developments because they take time, frankly, and unification across multiple call it, jurisdictions is is one of the big challenges too. But, yeah, it it...

It's why, like, you know, wish that a more general version of a non and off ramp API or a sandwich API could, like, get the job done. But, usually, that's... Fulfills, like, one very specific slice of, like, a treasury ops team, and it's not compelling enough for them to migrate over versus, like, closer to the total stack.

**Drew Rogers** (14:12)

I wanted to ask about if it... If if it's still something you wanna discuss and and tease here. SphereNet, this was something that we discussed before. Where does this now fit into the the settlement fabric? We've seen a ton of stuff happening.

I think since we we talked about this originally a while back, like, the kind of conviction I think you had around the settlement rails, these these stable point focus chains, like, since that time, almost now two years ago, like, we've seen this massive rise in, like, settlement infrastructure. Where does SphereNet fit in your mind lately?

**Arnold Lee** (14:47)

Yeah. It's quite interesting to your point. Since

**Drew Rogers** (14:53)

two years ago, this was, like... Yeah. The fall of twenty four, I think.

**Arnold Lee** (14:58)

Is there... I I remember talking to Zach, from Bridge who I think now is leading OpenUSD about this stuff spring, summer of that year and, you know, other attempts, I suppose as well. Circle being top of mind. Yeah. I mean, frankly, I think, like, as time has played out, there's just been more validation that these types of environments are useful. I I think that, like, our flavor of it is is pretty pointed to this day. We're not trying to do, like, merchant PSP.

You know, we're not, like, trying to make a general to, like, say, card processing. We're not trying to, like, capture every fintech everywhere. SphereNet is live, for what it's worth, just in a closed setting. So we have several cool design partners. I think one that is about to be announced is a very large global telecom company. We have a very large global trade house experimenting with multiparty net settlement on SphereNet. But, frankly, I think, like, the settlement environment itself is a bit overrated.

Like, people don't really care that much. It's more about, like, what is the product capability that this bespoke settlement environment enables. And so to this day, we'll still use, like, most blockchains depending on customer needs.

SphereNet just gives us a very nice place to pull, say, very opinionated compliance data or very opinionated customer tags that are related to what we need to ultimately produce documentation to clear transaction with a bank for an industry where, you know, that trade documentation is relevant. Like, you don't really need, you know, an application for clearing a freight of steel if you're processing a regular, you know, first party treasury money movement on and off chain. And so, yeah, the...

It's been fascinating, man. The more time has played out, the more... Both from, like, our perspective of, like, talking to different counterparties that we wanna bring on to SphereNet slash what we see broadly, publicly within industry, it feels like the first principles, like, rationale for why we, you know, had to build SphereNet in the first place are increasingly getting validated.

It's fascinating to see, like, the different approaches that people are taking, and we, you know, are, again, thinking about what does the customer actually care about. Like, what what does that look like for the product? And so despite having this bespoke environment ourselves, like, we're pretty open to using everyone else's where it makes sense. But, yeah, I don't get a lot right, but seeing this type of niche play out over the last few years has been quite quite interesting.

Because I think more and more people are coming to the realization that you need to be able to control components of runtime or you need to be able to coordinate certain, call it, economics for the different participants in the stablecoin supply chain for a... To be a very smooth product experience for the end user.

**Drew Rogers** (17:46)

Yeah. I'll pause What is what is the biggest thing... You know, what... In the journey of doing this... And you've probably learned so much. What what is the biggest thing that you actually didn't get right or thought was true that has turned out to not be? Not that, like, you wanna go back and change that, but just, like, along the way realizing, like, hey. This is what I thought Sphere or SphereNet was gonna become or focus on that turned out to not be the case.

So I feel like you guys are really, like, starting to hone in really clearly on, like, hey. This is who we are now. This is what we're focused on. We have really great use cases that are coming to market about it. But what were some of these things that weren't... Yeah.

**Arnold Lee** (18:28)

I think we we... We're hopeful that, like, industry behavior would be a little bit more rational on certain... I'll give you an example. It make... It'll make it clear. A year ago, I think Airwalks is a fantastic company for what it's worth. Like, it's one of the ones I respect the most, and I think they could carry the stablecoin flag better than most, frankly. But a year ago, Jack was talking about how, like, for Airwall, like, stablecoins don't make a ton of sense. And it...

Like, that actually, like, checks out, frankly. And over the last, you know, few weeks, I think they invested... Incubated the metal network. I'm very excited to give it a shot. It's... It it feels to me very mimetic insofar as the obvious attraction is that you get to control. You get to verticalize the settlement layer. But I think, like, what's harder to see is that settlement state itself is so commoditized.

You know, we were hopeful that many participants who are now developing their own blockchains in house or joining consortium attempts would, like, look into the history of what happens for different product categories. We have, like, app chains as a meme. You know, they're environments for things from perps to borrowing and lending to, you know, privacy, and they run into very similar problems regardless of who is creating it and how big the creator is.

**Drew Rogers** (20:06)

And so this is, like, in your mind, a signal or, a a verifying case for for your belief around really the role that these types of, like, layer ones play?

**Arnold Lee** (20:18)

Yeah. Commodification of state is hard to see unless you're crypto native, and the whole Ethereum or Vitalik, I suppose, specifically vision of effectively everything getting ZK rolled up onto... Between each other is probably true. I don't know if it'll, like, settle back to Ethereum. I don't know why, like, that would be an enshrined l one. But the... Like, where the state changes happen is, like, kind of cheap.

And I think while we predicted that a lot of people would memetically try this, which is why it's important to kind of give up on the dream of capturing validator economics in the settlement because there's no real commercial incentive alignment for, you know, the participants that you wanna attract to utilize your infrastructure if you try to do that. I think we underestimated how many would try this.

I think the hope was that you would get a couple of serious attempts, but then history would show that this is probably, like, not a good use of resources. And I think, you know, history has proven otherwise. I think Airwalks trying is pretty wild. I think a lot of people in the future will try. I think a lot of people are memetically looking at Hyperliquid.

I think, like, OpenUSD is an example of maybe stablecoin issuance at the at the corporate level, not getting as much product market fit as people initially predicted given all the different constraints that go into issuing a stablecoin, a la liquidity, wherever, you know, you want to be able to list that asset. And ours... My suspicion is that the same way that, like, app chains and l twos and different ecosystems had to, like

**Zach Fowler** (22:02)

we must just begin, like, little Wi Fi outages.

**Drew Rogers** (22:07)

Looks like he's back in the... It's really interesting, his point, though. That's why I'm like, I'm like, please please keep going. It's so good. We just finished this part. Yeah. It's so good. Sorry, Arnold. I think we're having some glitches on our studio, but... What we will clip this. We we... We'll clip all this, obviously, so it'll it'll all all work out well. But it is interesting.

And if if I'm if I'm kind of giving you the Zach and Drew take or maybe just the Drew take, like, being in Mexico City this year, talking to all these teams and kind of getting a sense of the industry, it kind of was like exactly what you're saying kind of played out where it's... There's a sense of desperation or, like, mild panic around, like, the musical chairs that it seems like is starting to happen where it's like people are like, where... Okay.

As all of these things become commoditized, like, do you actually anchor yourself in your true business model? It sounds like that's kind of what you've been seeing too.

**Arnold Lee** (23:01)

Yeah. I could have told you two and a half years ago that USD stablecoin would get... Would go to zero. It's... The same thing's gonna happen for all the different spot currencies, at least in single hop, payments. What do you mean what do mean go to zero? Like, the the idiot index for doing an on and off ramp for Circle or Tether into USD, including third party delivery is, like, single digit basis points, if not lower.

So, like, it can't be the underpinning of of a of a venture backed business or venture scale business. But I think, memetically, everyone saw what Bridge did, and so they just try to do the same thing. You know? There's, the front facing category where, like, they capture the user... End user relationship. They can drive margin, and then it's kind of like business as usual from a fintech perspective. Then the stablecoins are more of like an afterthought.

And the the the competition, the hard game is somewhere else in the user acquisition and retention. For those building infrastructure, it is a a very unexpected game because it's not just getting a rails up.

**Drew Rogers** (24:07)

Thank you, man, for coming on. It's always a gift to catch up with you and keep fighting, dude. Please let us know, like, when these bigger launches come. We'd love to, like, amplify those when it's public and ready.

**Arnold Lee** (24:18)

Sir, thanks for having me on. It's honestly amazing to see how much you've built since you first told me about this back in the day.