Why a Perp Exchange Asked Carmen Li to Make a Market in Her Own Index

Transcript

When S&P came out with their rating last year for strategy, they stated that they gave zero credit for the Bitcoin. And so, I know some of the people on your show have talked about why strategy has a junk rating from S&P. But, when you actually read the methodology, they say they did not give literally a single penny of credit for the Bitcoin. Strategy has over 800,000 Bitcoin. Like, when I talk to to actual institutional buyers, trillion-dollar asset managers that have positions in strategy and strive, they think that that's a joke. They actually just completely write off S&P's rating methodology because they don't give credit to the Bitcoin. Now, to S&P's credit, when you look at Basel frameworks and and, you know, the the financial kind of, whatever you would call it, our regulatory bodies, it makes

sense to me why S&P today might say we give zero credit to the Bitcoin. But, when you put it in the perspective of an asset manager that's actually tasked with underwriting the credit risks of these agencies, like, I can literally go on a call with a trillion-dollar asset manager and say, "S&P gave strategy a B- rating. Look at it, they gave zero credit to the Bitcoin. Do you agree with me that there should be some credit to the Bitcoin?" Every single person says, "Absolutely. The rating is too low." And our view on on digital credit actually is that it solves one of the biggest problems [music] in at least America, probably the world right now, which is the retirement crisis. And And so, my background, I came from a pension. I've been all in on helping solve the retirement crisis for everyday Americans. And what [music] we see is

that there is an income problem. And we've seen this for years, the 60/40 portfolio being dead. My view is that 40% of portfolios in most people, the income portfolios, there's just not a lot of attractive options. >> The number one question to understand the direction of markets is this: Is the data center cancellation rate going to increase or not? That's it. That's the whole question. The demand for chips, the demand for infrastructure, the demand for energy is there. [music] The hyperscalers are trying to hand the money out the door to Micron and Nvidia and TSMC and Broadcom. But, you are seeing higher cancellation rates, an increasing [music] public backlash, including from politicians. And by my account, I believe all the presidential contenders in the [music] midterms,

which will come into focus on the other side of the bicentennial, are voicing some concerns around this. The other constraint is at the energy layer. We saw this deal that was announced today in the Wall Street Journal. Chevron's providing [music] a two-plus gigawatt data center for OpenAI, which is tremendous, right? There's no There's no data center that consumes that much energy on the planet today. Put that in perspective. >> My other piece of advice to the CME would be you should be real careful what you wish for on this one. Cuz as somebody also familiar with the guts of Dodd-Frank, one, everybody remember the distinction between a future and a swap is at the discretion of the CFTC based on certain factors, which are not the

payoff profile [music] or strictly having an expiration date. If you really poke the bear, they have a huge amount of discretion looking at these things. They could easily say rolling a futures position is actually a swap cuz now you have eliminated the expiration date, and it needs to trade like a swap, right? And you're not going to like how that works compared to futures treatment. They can also go the other way and say as long as you construct your swap as a rolling sequence of indefinite futures, this is a futures contract. >> If you hold crypto on your phone, your biggest vulnerability isn't your wallet. It's your carrier. AT&T, Verizon, and T-Mobile have been breached again and again. And [music] SIM swaps are still one of the easiest ways for attackers to drain accounts.

That's where Cape comes in, America's privacy-first mobile carrier. Same premium service, but Cape rotates the identifier on your SIM every 24 hours, deletes your call and text metadata after a day, and protects against SIM swaps with a 24-word [music] recovery phrase that only you control. You also get two middle-to-end encrypted secondary numbers for banking and sign-ups, so you stop handing your real number to every app that asks. Go to cape.co/unchained and use code Unchained for 33% off your first 6 months.

Fidelity [music] has been researching and investing in blockchain since 2014, long before it was a headline. And they're hiring crypto and DeFi professionals to join their team and discover [music] what's next in finance. Fidelity is looking for people with fresh perspectives from different backgrounds, whether it's tech, UX, or product design. Whether you're crypto savvy or crypto curious, as long as you have the passion to make a real impact at a company striving to make finance accessible to all. Explore crypto careers at Fidelity today and make the decision that could change your future for the better. Visit crypto.fidelitycareers.com to learn more. That's crypto.fidelitycareers.com.

Fidelity is an equal opportunity employer. Before we dive back in, a quick word about something we make ourselves. If you like the way we cover crypto on this show, clear-eyed, no shilling, just what actually matters, you'll like the Unchained Daily. It's our free newsletter that lands in your inbox every weekday with the most important stories in crypto written by journalists, not marketers. No jargon, [music] no hype, no number go up cheerleading. Just the news you need to stay ahead in the time it takes to drink your coffee. Subscribe for free at unchainedcrypto.com/newsletters.

Heads up everyone. We've now fully transitioned Bits N' Bips [music] to its new dedicated channels. So, if you're not yet subscribed to the new Bits N' Bips channel on X, YouTube, [music] Spotify, Apple Podcasts, or wherever you get your podcasts, go there and subscribe now. [music] For the next few weeks, we'll play segments of the full interviews on Unchained as a reminder, but for now, you'll only hear those short segments of Bits N' Bips on Unchained. No more full episodes. You can get links to all of the dedicated channels with the full episodes at unchainedcrypto.com/bitsnbips, [music] spelled b i p s. Thanks again.

>> When S&P came out with their rating last year for strategy, they stated that they gave zero credit for the Bitcoin. And so, I know some of the people on your show have talked about why strategy has a junk rating from S&P, but when you actually read the methodology and they say they did not give literally a single penny of credit for the Bitcoin. Strategy has over 800,000 Bitcoin. Like, when I talk to to actual institutional buyers, trillion-dollar asset managers that have positions in strategy and strive, they think that that's a joke. They actually just completely write off S&P's rating methodology because they don't give credit to the Bitcoin. Now, to S&P's credit, when you look at Basel frameworks and and, you know, the the financial kind of, what do you want to call it, a regulatory bodies, it makes

sense to me why S&P today might say we give zero credit to the Bitcoin. But, when you put it in the perspective of an asset manager that's actually passed with underwriting the credit risks of these agencies, like, I can literally go on a call with a trillion-dollar asset manager and say S&P gave strategy a B- minus rating. Uh look at it, they gave zero credit to the Bitcoin. Do you agree with me that there should be some credit to the Bitcoin? Every single person says absolutely, the rating is too low. In our view on on digital credit actually is that it solves one of the biggest problems in at least America, probably the world right now, which is the retirement crisis. And and so my background, I came from a pension. I've been uh all in on helping solve the retirement crisis for everyday Americans. And what we see is that there

is an income problem, and we've seen this for years. The 60/40 portfolio being dead. My view is that 40% of portfolios of most people, the income portfolios, there's just not a lot of attractive options. >> The number one question to understand the direction of markets is this. Is the data center cancellation rate going to increase or not? That's it. That's the whole question. The demand for chips, the demand for infrastructure, the demand for energy is there. The hyperscalers are trying to hand the money out the door to Micron and Nvidia and TSMC and Broadcom. But you are seeing higher cancellation rates, an increasing public backlash, including from politicians. And by my account, I believe all the presidential contenders uh in the midterms, which will come into

focus on the other side of the bicentennial, are voicing [music] some concerns around this. The other constraint is at the energy layer. We saw this deal that was announced today in the Wall Street Journal. Chevron's providing a 2-plus gigawatt data center for uh OpenAI, which is tremendous, right? There's no There's no data center that consumes that much energy on the planet today to put that in perspective. >> My other piece of advice to the CME would be you should be real careful what you wish for on this one. Cuz as somebody also familiar with the guts of Dodd-Frank, one, everybody remember the distinction between a future and a swap is at the discretion of the CFTC based on certain factors which are not the payoff profile or strictly having an

expiration date. If you really poke the bear, they have a huge amount of discretion looking at these things. They could easily say rolling a futures position is actually a swap cuz now you have eliminated the expiration date and it needs to trade like a swap, right? And you're not going to like how that works compared to futures treatment. They could also go the other way and say as long as you construct your swap as a rolling sequence of indefinite futures, this is a futures contract. >> If you hold crypto on your phone, your biggest vulnerability isn't your [music] wallet. It's your carrier. AT&T, Verizon, and T-Mobile have been breached again and again. And SIM swaps are still one of the easiest ways for attackers to drain accounts.

That's where Cape comes in, America's privacy-first mobile carrier. Same premium service but Cape rotates the identifier on your SIM every 24 hours, deletes your call and text metadata after a day, and protects against SIM swaps with a 24-word recovery phrase that only you control. You also get two >> Hi everyone, welcome to another episode of Bits and Bips the interview. Uh my

name is Steve Ehrlich. I am the head of research at Shapelink and also your host. We've got a really exciting show today, but as always before we begin, uh let's [music] nothing that you hear or see on the show should be considered financial or investment advice. For full disclosures, please see unchainedcrypto.com/bitsandbips. And let's take a brief pause to hear from one of our sponsors [music] uh who make the show possible. >> This episode is brought to you by Cape, America's privacy-first mobile carrier. Same premium service you'd expect from any other carrier, but designed so your number, your location, and your data actually stay yours. Get 33% off 6 months at cape.co/unchained.

Fidelity has been investing in blockchain since 2014. They're not wondering if digital assets will shape the future. They're hiring the talent to help ensure they do. Explore opportunities today at crypto.fidelitycareers.com. Fidelity is an equal opportunity employer. >> All right, welcome back. So, really exciting guest today. I have Carmen Lee. She is the CEO of two different companies, Silicon Data Silicon Data and Compute Exchange. And she's basically created an entire industry focused on tracking the price of GPUs during this just exponential increase in demand. So, welcome Carmen.

>> Thank you. Glad to be here. >> Really excited to have Carmen you as a guest and we've been friends for a long time ever since our days at City. Really impressed with everything that you've been doing since you left City. I guess since we both left I became a journalist and now I'm a charlatan. You went to Bloomberg and I think led like all enterprise data partnerships at Bloomberg which anyone who ever used a Bloomberg terminal knows is a pretty big job. And then you got involved in sort of building an index tracking GPUs for the first time. Excellent foresight given where everything has gone and you're building out your entire uh I guess enterprise to not just track the prices, but also build the financial

plumbing and infrastructure that will underpin all of it. So, really excited to have you on today. >> Thank you. It's always nice to uh speak to a friend. >> Yeah. So, I mean I mean look we're going to talk a lot about potential areas of overlap um similarities and differences between what you're doing and and areas of commonality with crypto. But just uh to make sure everyone understands what you do, what your two different companies offer, can you please just give me the two-minute elevator pitch on Silicon data and Compute Exchange?

>> Yeah, of course. So, Silicon data is the Bloomberg market intelligence layer of Compute. So, think about as a Bloomberg S&P for Compute, meaning we have indices, we're alternative data, we're benchmarking, we have everything around intelligence layer, right? So, for example, our GPU indices, token indices, things can actually very soon can be tradeable at different exchanges, right? So, you need data, you need information, you need financial products, that's Silicon data. Compute Exchange, the other side, is the marketplace for compute resources procurement. You could actually get compute resources. So, forward contracts, reserved contracts, um refurbished GPU, um you actually can

get servers from us, and token packages, which we'll be announcing very soon. So, you can get actual physically delivered compute resources at Compute Exchange. >> Great. And I mean, your progress has has really been um has really been tremendous. I mean, maybe talk a little bit about that and uh and in particular, I I uh anyone that has been kind of watching your progress, I mean, you've been on CNBC and Bloomberg, you were just on Odd Lots. Uh I mean, you're launching futures contracts that are going to be traded on CME. I believe they're they're cash-settled at at some point in the second half of this year. Would love to hear more about that. And then just in in particular, in in a world where there's so much demand for for GPU resources, uh how do you build an index

that that that tracks it? I I'm not a tech person by by any stretch of the imagination, but but I know enough to know that there's all different types of setups. Like, this is not a homogeneous commodity as much as GPUs perhaps are becoming commoditized. So, how do you integrate that into an index that can just track the price, so people can um can hedge exposure or speculate. >> Yeah, so give you a little history, right? So, you imagine a little bit as I said in my background experience. Before that, I was at the I W um doing option market making for different commodity products and also at different index products. So, uh purely from a market participant's point of view, as a trader point of view, I really want a market that's liquid, trustworthy, and have a excellent price discovery mechanism,

right? Um I moved to Bloomberg, obviously got a ton of exposure from the data business itself, including indices. The way I look at compute in 2024, right? When the whole post ChatGPT, when I start in the 2024, you I saw tons of uh AI um software companies starting already, right? They're leveraging GPUs back then, leveraging potentially other uh large language models. They're creating services and products for hedge funds, right? So, what was interesting back then is they have great business traction, which was good for them.

The same time, they were complaining to me about their margins. So, translate to they, even though their top line was fantastic, they could not control their cost component, which translate to their margin was all over the place. They could be negative. They have no control over that. It's almost like it's in my mind it's like, yeah, of course, like American Airlines cannot hedge your oil prices or any airline, right? It's a It's a big problem for them to manage their operations, right? For any any businesses have exposure to any source of natural resources, right? Or resources in general, it would be nice for them to hedge either from revenue point of view or from cost point of view.

So, in my mind, compute will be the largest human resource very soon. If that's the case, we need benchmark price discovery for this particular asset class. And to your point, no commodity is the same, right? Even oil, there are different grades of oil. You can't really shifting oil from one place to different refineries, right? There's all nuance to that. Similar to GPUs. That's the reason why we start uh indices in that, "Hey, let's separate out A100 from H100 from B200. They actually have very different architecture behind that." Even within the H100 family, for example, there's 20, 30, 40 configurations. Translate to they have different CPU, different RAM, different interconnect, different geo location, for example, right? We have to

normalize it even within the sub class before you can start doing calculation. Or you're just comparing apple to oranges, right? That's a very important process to normalize to a base case, and then you do settlement process on top of that. >> And And talk to me just a little bit about volumes that that you're seeing. Uh in particular, I I know that you're seeing uh some significant growth on compute exchange. Uh And And And I believe that focus is more on on GPU rental, so it's not the same thing as as people buying directly from from Nvidia or or any of the other fabs, but but just contextualize that a little bit.

>> Yeah, there's a few layers to this, right? Um when Nvidia with their partner Dell, Microsoft, Super Micro, the world, when they sell servers, right? So, that's people, you know, hyper scalers, new cloud, they consume servers from from those um uh OEMs. And then they start rental process, right? They rent it out to people like enterprises. So, once the server is out out of Dell and the other, you know, OEMs, the rental price is a complete uh supply demand curves. They don't control the rental prices. It's purely demand supply driven, right?

So, that's a very interesting factor. That's what we really looking at is the on-demand uh compute per GPU power rate. So, there's a lot of numbers being thrown out how big the um AI infrastructure market is. Some numbers have, you know, been around for a long time, you know, we talk about $600 billion expenditure for for the server build-outs. Obviously, Nvidia recently mentioned in the future they're looking at $3 trillion. The derivatives market usually 10 to 15 X this spot market transaction. So, you're looking at a range of 6 trillion to 30 trillion derivatives market easily, right? So, that's the that particular market alone. For Compute Change wise, right? Obviously, we don't target every single thing out there as

an exchange. We're targeting people coming off queue different number of nodes through auctions. Um and then to your point, we don't do on demand at this point. We do reserve longer term contracts and token. We definitely see high volume growth at Compute Change. We see clients from traditional AI startups to enterprises to infusing platform even to cloud provider themselves. Um some of the months I look at the growth is like 1,000% month over month. I was like, this is crazy, right? The market as a whole, the notional traded it just I'm getting used to numbers, right? Like billions is just the way the market is. Everything is million to billion. It's the size of market. That translate to why the index and derivatives layer is so critical. All

the notional traded is so huge. You have to have a really good very honest with base trading going out, right? For you to hedge accurately cuz again, the size is just really dwarf other asset class. >> Yeah. And uh I want to That's pretty fascinating and I know in crypto a lot of times we're we're talking about like different asset classes. People have been trying to do the AI trade. People have uh in all sorts of ways trying to kind of appropriate onto what they're doing. Um but but your product and what you're building, you're seeing real value real volumes and um and it's really um satisfying a a true need which I think is something that product market fit

that we're always searching for. I want to get a lot more into it, but I think a nice place of convergence is just to talk a little bit more about your experience dealing with the CFTC. You're partnering with with the CME in terms of launching those cash settled futures contracts which will happen at some point in the year. My audience in particular is very familiar with the CFTC because people in crypto they strongly believe rightfully so that most digital assets are commodities. They're not securities and therefore they fall under the CFTC's purview not not the SEC and especially when it comes to launching different types of futures contracts on regulated futures on CME.

I mean some of those it's a different it's a different type of environment where unlike the SEC where you need kind of formal approval to everything before launching the ETF the CME kind of provide some some additional difference to issuers and lets them innovate a little bit more with the ability to sort of step in if if if deemed necessary to protect market integrity etc. So I'd love to just hear a little bit about your experience experience dealing with the CFTC especially if you can at all under the new chairman Mike Sellick and sort of what it's going to take or or what are the the steps necessary to get these new futures contracts to begin trading.

>> So personally I haven't met Mike yet. Um what what we have done in the past two years right we started 2014 we've been very transparent to all market participants. Hey this is what we're looking to do. This is the indices we think it'll tremendously help with price discovery for hedging for risk management for the whole entire ecosystem. It's going to be critical. We've been very transparent every step along the way. Obviously my stakeholders some of them have been communicated to CFTC and and other regulators. The good thing about CM our relationship with CME as well as the future products it is very traditional better words right it's not it's not extremely different right if you think about any

other futures launch it's on the line as a class cash settled or following exactly same CME process. So, if if you're if the future contract with CME today, you don't have to do anything different, right? Your bank, you have anyone of CMs of the world for you, it's just adding another contract versus, right, doing something anything different. So, that's a good thing. Um obviously, we will be meeting up meetings and reviews with CFTC very soon, which I'm really excited about. Um CME has been tremendously helpful throughout the process. >> All right. So, what really got me excited about bringing Elon, um is talking about the areas of overlap and convergence with crypto.

Uh because there's just so much demand to get involved in the AI trade. Uh I mean, there's even actually I I think it's on Objective. There's one DEX already that has um some degree of of perp futures uh related to uh related to GPUs that I don't think it's doing a a ton of volume. It's on it's on Helix. But I I I mean, there's so many different um protocols and projects that are offering like tokenized um compute resources, tokenized agents. Uh there was so much excitement surrounding the the SpaceX IPO, ways to uh like perp futures looking at like uh to trade it before the it even um before it even actually IPO'd.

Um there are I guess like synthetic opportunities to get tokenized access to like the likes of Anthropic and OpenAI, and then actually became a whole other issue that that kind of blew up. But when we were talking, I mean, you've sort of mentioned how you've been inundated with requests from some of these platforms to perhaps bring your product, your index in particular, on chain. And it's really interesting, too, cuz I know Hyperliquid in particular, they license the S&P 500 for uh for one particular perp market that they have now. So, I'd really love for you to just kind of share some of those stories, share some of the the what those experiences were were like

for my audience. >> Yeah, so um I'm very new to the perpetuity markets and anything uh traditional capital markets focused. So, my experience actually two-folds. It's pretty interesting. One is from conduction side. We have a lot providers. They were traditionally the mining um facilities, right? Um data centers first. They're converting um to AI uh data centers. So, we've been working with them, right? Helping them think through, "Hey, you know, how do you think about migrating to different servers? And some servers you're not doing dynamic air cooling." Translate to you can't support some of the servers.

Or you don't forget about liquid cooling, right? So, we're doing a lot of infrastructure trans- transformation with I'm sorry. >> Go ahead. >> I'm sorry. Your your audio cut out a little bit in the very beginning of that answer. Did you say that you you've been working with some of the Bitcoin miners that I have been looking to transition from Bitcoin mining into uh just like long-term HPC computing contracts? >> That's right. AI-centered um compute. So, facility-wise people are trying some of them at least choose to um pivot, I guess, to more AI-centric compute providers.

>> At this point, I think it's pretty much all of them. I actually did a show on this uh I think a few months ago. Uh and I mean, virtually they're they're all pivoting because the the money economics is it's so adversarial. And HPC, you can sign like a 10-year with Google or some other hyperscaler and you just you're you're just set because there's no way to replicate all of the the power generation and and the infrastructure. You can't just recreate it uh like that. So, that makes a lot >> It is sounds very different. And I also always coaching them, you know, you might not be your facility itself might not be set up to sort of the bare ovens because of, you know, coolings and the your whole infrastructure. Do you can this the floor supporting the new server at weights, right? For example, that's a

simple physics or for math. Um and then the other the flip side is if financially you know innovation with second data, right? So I'm always pro financial innovation. To me is what is the the core purposes of that, right? Take any products, my limited understanding of perps, right? The the the additional value you can bring is really the 24/7 angle, the weekend trading, right? And then, you know, some other um the nature of design perpetuity futures versus um the term structure, right? The rolling everything else. So there's a different value proposition for perps. But same time, to me is always to S&P, for example.

S&P is a fairly liquid product in traditional finance, right? Um and then we know the product itself is very solid and provide a very true economic value to the general public. So the underlying core S&P is a solid product. And then in the same time, you know, I think for any perps to make sense, perps to the wrapper, right? What's the product itself? Can this product the indices itself, this product itself is always can it actually do the prices covered the hedging needs in the real world, right? If your product is something that's manipulative manipulatable by others, right? Or it doesn't have much, you know, um really trading attractions, then whatever

wrapper you put out is just not going to be helpful. >> Yeah, I'm I'm curious. I want to double click on that a little bit because it it cuts to and and we were talking about this a little bit yesterday. It cuts to like the core value proposition of of perps and and derivatives in in general. And it's sort of in my mind a little bit of a cousin to the idea of having cash settled futures. Um because how what's the utility of that as a hedging tool to to someone just that's not intimately involved in these industries? If I'm an orange farmer and I want to do like physically settled futures, I understand the value of that and and protecting uh protecting sort of my revenue etc. But but I mean just like um traditional markets, derivatives are

such so much bigger than spot. Um GPUs right now it's it's the inverse. I think we're going to get there and I am curious your thoughts on how like the products you're going to be launching with CME and um and I think I I think ICE is actually launching a competitor, too, um because there's there's so much demand for this. But what what what you're launching with CME and how that might smooth out price discovery, how that might just be beneficial um in order to kind of help um companies involved in this industry hedge and uh to the extent you think perps can be useful. Um I'm very much on record even back going back to my time as a journalist at saying that like at some point I I think sometimes derivatives can just be detached from from reality.

I I mean I I interviewed Sam Bankman-Fried a million times, but one of the first conversations I ever had with him long before uh long before like FTX blew up and on all of that, I asked him why he was offering 100x margin on derivatives at FTX. Like like what is the actual utility of that? And and it it created such a volatile swings that I mean people couldn't put on very big positions cuz just like a very minimal swing would would just liquidate and and he said, "Well, people wanted it." And uh So I'm I'm curious like how you reconcile all of that because having a liquid derivatives market can certainly smooth over a lot of the nuance that you

laid out earlier in um in your answer. But at the same time, we want to make sure that it's not just only speculation for speculation's sake. >> 100%. So the way we design contract, we have our natural hedging mind, right? If you are building a data center, there's you know, there's number of hours you're looking rent it out. You have your spot exposure. For example, most of time you're doing 1-year, 2-year, 3-year long contracts, right? Your financing can be 5 years versus your your customer contract can be 2 to 3 years. What's going to happen in the next 3, 4, 5 years, right? What kind of volatility?

You need the money to pay down your loan and your revenue that is uncertain. Can you hedge that for that particular duration? What's interesting about GPU does have a term structure, right? Versus let's say other things. You need to hedge your term structure efficiently. You need a data future products. That's why I think GPU future GPU by itself naturally lend [clears throat] it to data future as a better fit product. However, large language model token, for example, we have a for better words, now very famous large language model token indices.

It is continuous. People use that on the weekends. It maybe is a better approach spot, but I don't know, right? So, I think the depends on underlying assets. The better product can be different. Right? So, that's one thing. The second thing is about liquidity. The way I think about liquidity, obviously to your point, is, you know, leverage is one thing. The way I think about it is bid-ask spread, right? If a product has a really wide bid-ask, it's not very efficient for you. It's expensive to hedge, right? That translates to. I like, selfishly, a very narrow bid-ask spread for my products. That translates to liquidity, translates to cheaper to hedge for every more than market participants. Right? That's what I'm looking for. The efficiency of the product really drive down the really

drive up the liquidity and the market participants. And the whole natural hedge will going to come to you. The only reason they want to come to you is you can actually help them to hedge. And then to your point about competitors, the way I think about it is I I actually believe in capital market efficiency. I believe people will gravitate gravitate to towards to the best product for their their needs. Your product design is not relevant. It's easily manipulated by others, Uh the basis trading is all over the place, then maybe less people want to participate because of the flaw of your indices versus others, right? Once people that the trust has been built, right? The financial institutions have to trust your indices to put trillions dollars swaps on your indices, right?

How do we get that, right? So that's going to be a gradual process. At the same time, it takes a lot of rigorousness and methodology, data breadth, depths, right? All those things come into play. >> All right. Well, we're going to get a lot more into that, but before we do, we have to take one more break so we can hear from another one of our sponsors. >> If you hold crypto on your phone, your biggest vulnerability isn't your wallet, it's your carrier. AT&T, Verizon, and T-Mobile have been breached again and again. And SIM swaps are still one of the easiest ways for attackers to drain accounts.

That's where Cape comes in, America's privacy-first mobile carrier. Same premium service, but Cape rotates the identifier on your SIM every 24 hours, deletes your call and text metadata after a day, and protects against SIM swaps with a 24-word recovery phrase that only you control. You also get two middle-to-end encrypted secondary numbers for banking and sign-ups, so you stop handing your real number to every app that asks. Go to cape.co/unchained and use code Unchained for 33% off your first 6 months.

Fidelity has been researching and investing in blockchain since 2014, long before it was a headline, and they're hiring crypto and DeFi professionals to join their team and discover what's next in finance. Fidelity is looking for people with fresh perspectives from different backgrounds, whether it's tech, UX, or product design. Whether you're crypto savvy or crypto curious, as long as you have the passion to make a real impact at a company striving to make finance accessible to all. Explore crypto careers at Fidelity today and make the decision that could change your future for the better. Visit crypto.fidelitycareers.com to learn more. That's crypto.fidelitycareers.com.

Fidelity is an equal opportunity employer. Before we dive back in, a quick word about something we make ourselves. If you like the way we cover crypto on this show, clear-eyed, no shill filling, just what actually matters, you'll like the Unchained daily. It's our free newsletter that lands in your inbox every weekday with the most [music] important stories in crypto written by journalists, not marketers. No jargon, no hype, no number go up cheerleader, just the news you need to stay ahead in the time it takes to drink your coffee. Subscribe for free at unchainedcrypto.com/newsletters.

>> [music] >> All right, welcome back. So, we we were talking about sort of the applicability of of Carmen's businesses to to crypto, some of the the nuances and challenges associated with bringing her products and her index on chain. And I I want to go a little bit deeper because I know and you've talked about how you've been courted by by many of the without naming names, many of the major like DEXes and and and perp issuers or or perp marketplaces to sort of license your index to them similar to what the S&P 500 did, etc. And uh

I'm sure um that it's interesting, potentially appealing to you, but but you did mention how you were also asked a few they made a few requests of you that you found very unusual. And dare I say concerning. So, I'd love for you to just kind of walk us through a couple of those experiences. >> Um yes, so again, I'm extremely new to the purpose market. So, um it was great. We received that kind of enthusiasm from perpetuity marketplaces. So, I almost say everybody almost everybody, right, reached out and have conversations. So, my question to them is always, "Hey, love to this love to explore. Tell me

what should I do? What kind of product that will be interesting enough for your audiences, right? Um what kind of relationship you're looking for? A few things I at least ought to me from traditional again market participants point of view from capital markets is um one of the things is is so bizarre is they ask me, "Hey, Carmen, aside from licensing indices, will you consider become a market maker of your own indices?" And so, I thought they were joking, right? Cuz clearly there's a huge conflict of interest by me trading my own indices that completely rule the neutrality from index provider point of view, right? By them asking, I thought they were completely joking and then

realized they were not joking. And then I was like, "You do realize a big conflict of interest. Are you doing that with any other indices?" They were like, "Yeah, we do that with other indices as well as well." And I said, "Did anyone say no because it's by design they not you're not supposed to do so, right? So, it's fascinating to me that was even a discussion to begin with, right? That is not something I will feel comfortable as a trader, right, to trade a product I know the index provider also can trade their own products.

>> That's number one. >> Yeah. And and maybe just to explain a little bit more again, we have a sophisticated audience here and they're quote unquote like bilingual in terms of crypto and and tradfi. But but maybe just give like one or two examples for for people who perhaps may not understand the significance of you market making on your own index, what are the risks? >> Um How do I explain this? Um it's almost when I tell my daughter, "Hey, you know, you can go play if um if I if I decided that um my work can be done by 4:00 p.m." right? And then I I'm I'm the only person to decide if I can

work by my 4:00 p.m. by 5:00 p.m. Then her ability to play or not completely on my preferences, right? So for example, if I have a open interest about a position on my own indices that's long right? It'd be really hard is the natural incentive for me to set up my indices at a higher price, right? Cuz I will benefit from this particular position versus others, right? Such a conflict of interest. And everybody can say, "Hey, you know, we the trading side is different than my index side at a lower price I set." But doesn't matter. Your incentive as a company to maximize your P&L, that's a huge It's a huge conflict of interest.

You can't >> Yeah. >> give [laughter] the power of deciding set up prices to the provider while you're trading at the same time. >> Yeah. Yeah, and I appreciate that. And and it's >> It's a conflict of interest. It wasn't even a question, right? >> Yeah. I mean it really is I mean clearly it it's in your interest for your next to keep going up and up and up. And and as you're looking to I mean potentially license to some of these perp offerings and many of which I think the vast vast majority of which are quote unquote like unregulated and probably geo-fenced around um offering or servicing US clients.

Uh at the same point I I know that there's I think there's some ETF issuers that are potentially looking to launch products based on your futures contracts at at CME. And I I only imagine the SEC and CFTC having some real big questions about the the fairness of those of those products if like theoretically you have the ability to manipulate the price of of your index that everything is based off of. >> So it's incredibly important for us to make sure that our neutrality stands, right? That's why we're going through external audit our on second data through ISKO compliance third party lens, right?

We have very strict internal trading policies, right? Not only you cannot trade my indices obviously, but you have disclose your other positions, right? For example, you're long on the video. Right? That can be a potential conflict of interest. Everything taken extremely seriously at second data as index provider. As as we said, everything in our world the scale is just unparalleled compared to other other product asset class. So that's number one. Number two is I think it's interesting when I talk to the um the uh perpetuity market issuers and market participants. Another thing I was bringing up is how do you make sure the index, you know, obviously they have their choices, right? How do you make sure it's it's not being manipulated? It's very

important, right? When you pick a pick partners. Um and then how do you make sure the trader can actually make sense? Some of them suggest that hey, by the way, we're going to pipe into pipe your indices and then mix with other indices. And to me it's all right, so when you're going to move from one index to the other, they have different methodologies and they actually essentially different products, right? How can anyone trade if you just anytime can switch from one to the other? How are you going to keep your position open and close? This sounds like again another box of worm just came up, right? Um I think the lack of understanding of the GPO markets the index index per se itself is is a big problem for me to pick the right

partners. Uh essentially, when my public my product become tradeable at CME. We have a suite of institutional player. They are can't wait to use our indices to hedging. We're talking about all the banks, hyper scalers, neo clouds, big AI companies. We're talking about billions dollar traded at at traditional indices traditional fusion options products, right? The energy traders and then all the basis trading going to be happening. And then if they going to I'm sure some of them want to have a purpose venue, right? I don't blame them for that. How can we make sure they're in they're they're they're trading in a purpose markets on my indices and also in a reg- regulated in a very um

similar rigorousness as traditional markets. That's some that's something I'm looking for, right? I don't I can't just publish my indices and let people trade without being able to sort of guardrail that particular venue. >> Yeah, and it's really interesting the point you mentioned about like how you can't just swap indices. I don't have a ton of experience with this, but um one of my prior roles in in crypto was at Kraken. Um and if you don't know who they are, Carmen, I mean they're they're they're they're one of the larger Okay, you do know who they are. Um and and they have a dedicated index business called CF Benchmarks. And I used to work with their team to to understand and they're actually partnered with CME. And they work with their team to understand like how they build these indices, the types of like redundancies and resiliencies built in,

like which trading platforms are allowed to feed in, which one like what what's the criteria for inclusion or then perhaps being like excluded or or relegated. And I mean just looking at I'm blanking on the name of your one other competitor with the the New York Stock Exchange. Their index I think is based on slightly different methodology than than yours. It's more based on like just like the actual um transactions. And I'm not necessarily necessarily saying one is better than the other. I'm sure you think yours is better, but I'm I'm also guessing that you probably say that theirs is not wrong. It's it's just it's just different, but it means you can't swap you can't just swap them out because like they're they're not they might be cousins, but they're not identical twins. And like that's a

nuance that professional uh money managers will understand. Retail traders may not. >> So, one of the things I do want to um sort of that's the correct but that is a you know, some people's marketing and PR just went all over the place. I want to correct that. We are building on transaction prices plus other of the real prices. We have 200 data provider information and then we have the most well covered indices which translates to with at least easily manipulated manipulatable. Meaning you have to manipulate 200 data sources of ours which is is very hard to do versus you have two three data sources. Just big because your transactional price doesn't mean it's not manipulatable, right? If you know there's two data sources and

you do one trade can change your prices, you can do so. >> And yeah, we're very familiar with the term of wash trading and like uneconomic trading which can certainly be used to to to push up a price of something. So, so we've got >> a transaction price doesn't translate to you cannot be manipulated, right? The best way to hedge that right risk is having a variety of data sources. And again, there's a huge pipeline rigorousness you have to through what what to include what not to include is over standard deviation. It's a real trade but hey, the price is two standard deviation against your prices, right? You probably have to throw that out, right? How rigorous is with the data pipeline? So, all those into it's very very rigorous a very complicated process to create indices that's tradeable. And to your point, I think the capital

markets will vote um by trading volumes, right? Which one they think will most helpful for their hedging needs, speculation needs, you know. >> [snorts] >> What do you think of the CFTC um like quote unquote approving US regulated perps? I know that there's some some litigation on ongoing um uh I think the CME's taking exception to it but uh I mean you're launching cash-settled futures. Have you paid attention to this at all? >> Um so, obviously, from market observer point of view, I've been I've been following the news. Um from my point of view, um you know, I think there's a time place for everything. For GPU futures right now, I think data futures is best use cases for hedging term

structures for GPUs. So, I do believe that's a best venue to start and let the natural hedger come and actually hedge the exposures with the other with the other counterparties, right? And then going forward, obviously, we're we're totally open um to do different things um as the as the volume of liquidity start building up. >> Okay. Uh one other aspect of uh crypto I wanted to touch on before we kind of wrap up by by taking a a broader picture of everything is uh that you also told me that you're very open to accepting stablecoins as as payments and and had conversations with um had conversations with one major industry player if you feel comfortable talking about. And uh

it really would just come down to I think you said customer demand, but could you just sort of walk us through uh your thought process there? >> Yep. So, for Compute Exchange, we had discussed here should we take stablecoin or any uh currencies, right? Uh obviously, we open whatever facilitate as an exchange, right? We have to facilitate any transaction that both party in the market demands. So, we completely open to do so. We had to your point, we had discussion last year talking about wall integrations and you know, both ends. Obviously, as a marketplace, you have providers and sellers and buyers, right? So, they have to want it on both sides. We we don't really push either way, right? So, if they both want it, we have to support whatever form they want to they want to do. The nuances is, again, we're talking about huge dollar amount for almost all

transactions right? Um and sometimes it triggers different form of payment, a deposit, and monthly recurrings, and things can the billing process can get complicated and then we do clear trades. We sometimes we are in the intermediate taking payment from one and then to the other. So, there's a complication involved with that. But yeah, as a as a company from computing point of view, we have to take her take any form of payments as both parties demand. >> Yeah. I I just imagine I I I mean maybe just to level set real numbers like like what is a typical size of a transaction?

I mean this this stablecoin market cap right now is about 300 billion dollars that it fluctuates. Um US Treasury Secretary Scott Bessent has suggested it could get to three or four trillion dollars by the end of the decade. I mean at that point who knows? Uh what what are the dollar transactions that you're seeing? I mean people are familiar with the headline numbers of hundreds of billions of dollars even trillions of dollars of investment but that those are like the big big big guys. Like what are you seeing? >> So, we see a which is the beauty of exchange, right? We don't only cater to the large player which we do. If you think about per gigahash power at the same state $5 per gigahash power, every node has eight, right? And then every hour of day, right? Once you enter a

contract, you're paying every single hour, right? So, if you need 2,000 nodes, right? And then you times eight times um $5 and then times 24 hours times the duration, quickly becomes billions dollars, right? It's it's very the math just adds up. So, as exchange, we do provide liquidity to people want a window, two nodes, right? For for year all the way to people needs thousands, 20, 30, 40, 50, 60,000 nodes easily. So, we have big wide range of interval contract. So, interval transaction size is a big variety in terms of obviously notional traded volume distribution 20/80 rule, right? It's just there's people concentrating in the high end but

we love to support from smaller nodes point of view. For me it doesn't matter, right? Liquidity is liquidity. >> Okay. All right. So, let's start to wrap up. Uh Uh crypto and AI, as I kind of mentioned in the beginning of the interview, they like to see themselves, at least from the crypto side, we like to are like to see ourselves as cousins and see us rise together. Um but I as I'm I know you don't follow crypto as much as I do, but uh there has been a a bit of a divergence as AI keeps going up and up and and up. Uh crypto has has been struggling.

And I would love to just kind of get your sense of how you're seeing like broader macro trends. Like we're we're still in I guess like a uh set uh I guess an environment of euphoria in light of the SpaceX IPO. We're going to have Anthropic and OpenAI go sooner. I think SpaceX just issued more debt right after the largest IPO in history. I mean, that's a sign of frothiness. Uh I if that's not a sign of frothiness, I don't know what else is. But but like how are you seeing demand for for GPUs and AI um in the next like 6, 12, 18 months? And and to the extent you feel comfortable talking about how it might relate to crypto, I'd love to get your thoughts.

>> Uh so, our world So, I think the world, not just US, is bifurcated in a way where you see tons of um productivity gain, um you know, even some dollar-wise, company valuation. So, just in general, spend and you know, everything, right? In the in the AI semi industries, crypto, you know, that particular segment, a lot of capital goes in. And you see tons of productivity gain. You see stock pro- pri- prices overall across all goes up, right? Imports, exports, right? People in you know, exporting semis and the Koreans and you know, they they've been experiencing tons of growth.

And then all the other other segment of industry, right? Different bifurcated, right? Is not doing as well. A lot of traditional businesses, right? Even in America, even in China, even in Korea, right? They're not They're just static for the most part, right? Traditional finance traditional marketplaces, traditional industries, um you know, to a a point where they will start enjoying that kind of uh um growth because of AI, ideally, right? Other things that would that would be interesting to see. So, and crypto again is is a one of the what they call the industry verticals in in this world, right? So, obviously, I think right now they try to link say, "Hey, AI is going bonkers. Let me just tag along, right?"

From AI infrastructure point of view, some of them are doing so, right? Um maybe they can tag along or they try to figure out a way to grow themselves in different ways. Obviously, payment is a big deal. Obviously, you know, tokenize other asset classes within the financial industry is a big deal. So, all those things can can be helpful, but in general, I think the world is pretty bifurcated, right? The traditional industries versus AI semi world of what we're doing here. Um and then next 12 months, um I'm not smart enough to tell you next 12 months, but what I can see from my term my forward curve, right? Um still huge demand for AI. Um and then we'll see a proliferation of obviously LLM

uh LLM uh models in general, right? Open source, open weights, closed source, state keeping, you know, which is good. We want the competitiveness of the industry. Um and token prices obviously partially driven by GPU prices, partially driven by competitive nature of the industry itself, and go up and down, and the mixture of models go go, you know, they they change all the time. So, things have been pretty pretty interesting, you know, lately. >> Yeah. What do you think is the biggest risk to your to your business um over the next like 12 to 18 months? >> Yeah, so I think where we stand from both um company point of view, we're in a interesting stage where both company agnostic, right? Silicon data, we benchmark all different GPUs, not just the the the one to design houses, right?

With we and then we are completely agnostic in terms of um indices calculation. We do have a AMD indices and my 300, right? So, compute inside, again, we're completely agnostic of chips, right? We have done a lot of AMD deals, right? We have to do any design house, any independent design houses. We work with all new cloud providers, all different clients, right? So, in a way, we are hedged against individual trends, right? TPU or particular chips doing well, particular design house doing well, we hedged in that sense.

So, the only the only thing can happen is the outcome worse is just no one use compute, no one use AI, no one care about the market, right? So, then, if no one use compute or AI, no one need to hedge, right? Obviously, and then no one need to do transactions in spot markets. So, unless that happened, those two companies are well positioned to grow with the industry. >> Got you. Okay. And just again, cuz we are primarily a crypto show. I just want to before I give you a chance to share any other final thoughts, looking through your crystal ball and and maybe a year from now, how big a part of your business do you think crypto or like blockchain-based financial rails will be as how big of a part of your business do you think it

might be? >> Um so, from the payment point of view, right? From compute change, we can support variety variety of support any payment types. So, if there's a need from both sides, happy to do that right away, right? We're we're we're very unbiased in that sense. Um so, it depends on the market adoption, again, right? From Silicon data point of view, obviously, issuing perps on some of products in different venues is something we're actively engaged in discussion. Um for me, it's just find the right partner, find the right venue, right product constructs, right? And the right timing. That's that's the critical thing. So, yeah, maybe I don't the liquidity can come and obviously I kind of think once the traditional CME future launched, we will

see a lot more institutional participants centered around our indices. And then those folks may be more willing or happy to trading other products, similar products at in perpetual forms, right? So, I think the industry will will take up pretty nicely next year. >> Okay. Great. Um well, we're at time, but before I let you go, um any other final thoughts that that you like to share? Any any comments >> Yeah. So, I keep telling people I'm not cool. I wasn't, you know, I'm very close to crypto all the time, but never cool enough to be in the industry, right? So, as a full-time market participants, so I love financial innovation. I think it's something I'm really really excited about. I just It's for me it's all find

the best product and bring institutional player with me, right? That's who are supporting us so far in traditional markets. I think people are exploring with us together. Um so, yeah, excited to to learn more and then do more. >> All right, great. Well, I've never been called cool in my my life either. I think we have kids around the same age as and one one time I don't know if your kids use the term goated, but I think my oldest daughter used the term goated to describe when I opened a jar of jelly. So, I guess I'll I'll take that. Um but Carmen, um great to have you on. Um we'll have to monitor your progress and then and see what happens. And uh yeah, thanks for joining us. And yeah, and so that wraps up wraps up our show um for our this episode of

uh Bits N' Bips the interview. Thank you everybody for for watching and joining and um tune back next week for another exciting episode. Have a good day.