🧠 Circle Goes Full Layer 1. Coinbase Fights for Your Fees. Affirm's AI Knows You Better Than Your Ex.

0:00So this week, while basically everyone you know was probably doom scrolling their savings account yields. Oh, yeah. Just anxiously waiting for the Fed. Right. Waiting for the Federal Reserve to announce their first rate hike since 2023. It's just such a spectacle watching the whole market freeze up like that. It really is. The entire financial ecosystem essentially just, you know, holds its breath to see what borrowing costs are going to do. Exactly. We all sit there staring at the Fed trying to figure out if chasing like an extra quarter of a percent APY at some random online bank is actually worth the hassle of moving all our emergency funds.

0:37And all the oxygen in the room gets sucked up by that single macroeconomic narrative. It creates this collective paralysis. Which is exactly why we are taking a totally different approach for this deep dive. Because while the rest of the world was distracted by all that macro noise, we dug into a stack of sources you sent us. And the contrast is pretty wild, honestly. It really is. We found three major fintech giants who completely ignored the rate hike drama. I mean, they didn't pause their operations. They didn't panic about borrowing costs. They just put their heads down and decided to build.

1:11And historically, I mean, that is when the most foundational shifts happen in any industry. Right. When the public spotlight is pointed at something massive like the Fed, the companies that are actually rewiring the financial Internet, Well, we're doing it quietly in the shadows. That is exactly our mission today. We are going to extract the core strategies of three companies, Circle, Coinbase, and Affirm. Three massive players. Yeah. And we are going to look at how they are fundamentally changing the infrastructure, the pricing, and the actual access points of modern finance. Okay. Let's unpack this because this first one is a story of massive, sweeping infrastructure being deployed while nobody is looking.

1:50Right. Happening right under the radar. Let's talk about Circle. A company that for a very long time has been synonymous with just one thing, right? They were founded back in 2013. And for 13 years, they have essentially been known in the industry as the USDC company. Yeah. They're the ones who made stable coins, you know, digital dollars, a trillion dollar adjacent business. Exactly. But this week, Circle, which trades into the ticker symbol CRCL, went live with something completely new called ARK. And this is huge. It really is. Our sources show this is a brand new layer one blockchain built specifically from the ground up for stable coin powered finance.

2:32And for the listener who might just dabble in tech news, we probably need to clarify what a layer one actually is. Yeah, because they didn't just build an app. Far from it. Right. Building a layer one blockchain is like building the underlying operating system. It is not an application sitting on top of Windows or iOS. It literally is the Windows or iOS of this specific network. So it's the foundation. Exactly. It's the foundational base layer that all the other applications, the smart contracts and tools have to plug into. And, you know, they didn't just quietly test this new operating system in a sandbox.

3:02No, they launched with overwhelming force. More than 100 applications and institutional partners were live on day one. Which is a staggering day one metric. Right. And get this detail from the filing because this blew my mind. Circle minted the full 10 billion token supply of the ARC network token right out of the gate. Yeah. That makes them the first publicly traded company to ever mint a network token for a new layer one blockchain. What's fascinating here is a validator list for this new network. Oh, the validator list is crazy. It really is. When you launch a decentralized or, you know, even a semi-decentralized blockchain, the validators are the entities that process the transactions and secure the network.

3:43The people checking the math. Exactly. You can think of them as the decentralized accountants making sure nobody is double spending money. And Circle's founding validator list for ARK reads like a Davos guest list. It really does. You've got BlackRock, Visa, MasterCard. You have the DTCC, the Depository Trust and Clearing Corporation. And you've got Standard Chartered and HSBC. I mean, that is a wildly traditional, heavy hitting lineup for a blockchain network. It bridges the gap between traditional legacy finance and the crypto ecosystem in a way we just haven't seen at this scale before. I mean, when we say the DTCC is validating, we are talking about the primary entity that settles almost all of Wall Street's traditional trades.

4:26And now they are acting as a node on a blockchain. That's a massive shift. It's huge. These aren't just random crypto native startups validating the network. These are the foundational pillars of global fiat finance plugging directly into the crypto mainframe. And the mechanics of how this specific network operates are what really caught my eye on the sources. Oh, the gas fees? Yes. The CEO of Circle, Jeremy Allaire, called ARK the single most significant launch in the company's history since USDC itself. Which is a huge statement. But he used this analogy that I thought was just perfect to explain the why behind this launch.

5:03On most blockchains, if you want to make a transaction, you have to pay a transaction fee, often called a gas fee. Right. And you have to pay it in that specific network's volatile native token. Which is incredibly cumbersome for a regular business. Like it makes no sense. It's a total accounting nightmare. Exactly. But on ARK, you pay the gas fees directly in USDC. Allaire compared the old way of doing things to Amazon forcing shoppers to buy Amazon Web Services stock just to be able to check out their shopping cart. And the friction of that old model has always been a massive barrier to mainstream corporate adoption.

5:42It's just too complicated for them. Way too complicated. If a major corporation wants to use a blockchain to move money globally, they absolutely do not want to hold a highly volatile cryptocurrency like Ethereum or Solana on their corporate balance sheet just to pay basic transaction fees. Because the price swings wildly. Right. Doing so introduces severe currency risk. It creates absolute accounting nightmares, as we said. And the tax complexities of tracking capital gains on every single microtransaction are just paralyzing. Yeah, that makes sense. So by pricing the transaction fees in USDC, which is a stablecoin perpetually pegged to the U.S.

6:17dollar circle, has completely removed that friction. They've made blockchain transactions look, feel and function like standard software as a service costs. That makes perfect logical sense. But I mean, I have to push back on the market reaction here. Let's hear it. Because if this is truly the biggest launch since USDC, and if they have completely solved the gas fee friction problem that has held corporate crypto back for a decade. Yeah. And if they literally have Wall Street royalty like BlackRock and the DTCC actively validating the network, why did Circle's stock drop 5 % to 7 % on the news?

6:54It seems completely disconnected from reality. Well, it is the classic phenomenon of score the touchdown, still lose the game. Okay. What do you mean by that? Well, Mall Street is notoriously short-sighted, especially in a week where a Fed rate hike is actively dominating the trading desks. So they just can't see the forest for the trees. Pretty much. Analysts are almost exclusively looking for immediate revenue bumps. They are analyzing the next quarter's earnings projections. But building the fundamental rails for the next financial Internet is a deeply long-term play. They're building infrastructure.

7:25Exactly. Circle is out there laying down the train tracks. The market is looking at the heavy capital expenditure of building those tracks and choosing to sell off the stock. They are completely failing to price in the fact that very soon everyone else's trains are going to be paying to run on Circle's rails. Wow. It is a pure infrastructure play, and the public markets routinely misprice foundational infrastructure in its earliest days. That's a great point. Circle and Coinbase are basically rebuilding the plumbing for moving money. But, you know, none of that matters if people don't have a reason to use it.

7:58Right. And that actually brings us to Coinbase, who is weaponizing those exact same stablecoin balances to aggressively defend its turf. The logical connection here is vital. Circle is building the fundamental rails using USDC and the utility of USDC is expanding rapidly. Coinbase recognizes that utility and is leaning into it to solve a massive competitive problem. Let's look at the sources for Coinbase. ticker GoalOnOn, founded in 2012. This week, they slashed their trading fees for active users on their Coinbase Advanced platform. Just completely gutted them. Right. They lowered the qualifying volume threshold from$25 ,000 all the way down to$10 ,000.

8:42And incredibly, holding USDC balances in your account now instantly unlocks VIP pricing. The scale of the fee compression we are looking at here is genuinely staggering for an exchange of that size. It really is. The source notes that at the top VIP 8 tier, spot fees are now as low as zero basis points for makers and just two basis points for takers. That's basically free. Yeah. And for the listener who isn't a day trader, let's translate that jargon real quick. Yeah. A basis point is just one hundredth of a single percent. Super tiny. Very tiny. And a maker is someone who places an order that doesn't fill immediately, adding liquidity to the exchange's order book.

9:19A taker is someone who buys or sells immediately, taking that liquidity away. So at zero basis points for makers, Coinbase is practically giving the trades away for free to their highest volume users. To understand the motivation behind giving away the house like that, you really have to look at how Coinbase has historically generated its revenue. Which is mostly retail, right? Exactly. Since 2012, they have leaned incredibly heavily on retail trading fees. If you were just a casual user buying a little bit of Bitcoin on their standard mobile app, you were paying a significant premium for the convenience.

9:56I love the analogy the source material used for this dynamic. It compared Coinbase's fee structure to Blockbuster Video. Oh, that's such a good comparison. It's perfect. For years, Blockbuster made a massive outsized chunk of their total revenue simply by charging late fees. Which everyone hated. Everyone hated it. It was a terrible user experience, but it was incredibly profitable while it lasted. Cutting these retail trading fees now is like Blockbuster, suddenly realizing that late fees are fundamentally bad for long-term customer retention. Yeah. Except in this case, Coinbase is making the move before they get Netflixed out of existence.

10:33And that distinction changes the entire narrative. Mizuho put out an analysis on this fee cut, and they did not mince words. What did they say? They called this the first step to right-sizing the cost to trade crypto. And they pointed squarely at Robinhood as the primary catalyst. Oh, Robinhood. That makes sense. Yeah, Robinhood has been applying massive, sustained competitive pressure by gamifying and offering zero-commission crypto trading. Coinbase realized that if they didn't adjust their retail take rates, they were going to slowly bleed out their most valuable, high-volume users. So what does this all mean?

11:10Are they just panicking? Like, are they just giving away their primary revenue stream because Robins would back them into a corner? It is definitely not charity, and looking closely at the structure, it isn't a panic move either. It's chess. Chess? How so? This is a highly calculated defensive fee war. By offering these radically cheaper advanced tiers, Coinbase is building a structural moat around its high-volume traders. They are giving the whales a reason to stay. Keeping them from bolting to Robinhood or to offshore exchanges. And keeping those high-volume traders is critical because they are the ones who actually provide the deep liquidity to the platform, which makes the platform function smoothly for everyone else.

11:49Precisely. But the truly strategic part of this move is the USDC integration. Right. Tying it back to Circle. Yes. By allowing users to qualify for these elite VIP tiers simply by holding USDC balances, Coinbase is heavily incentivizing users to keep their capital parked on the platform in a stable asset. And behind the scenes, this quietly but massively deepens Circle's stable corn distribution. Because for the listener who might not know the history here, this isn't random coincidence. Coinbase and Circle actually partnered years ago to create the Center Consortium, which originally launched USDC in the first place.

12:26That shared history is the absolute key to this whole move. It is a perfect synergy of their two overarching strategies. They're working together without merging. Exactly. Circle builds the new layer one infrastructure to make USDC the default currency of the financial internet. And Coinbase uses its massive retail footprint to distribute that currency, locking users in with zero fee trading. Wow. It is a defensive move against Robinhood that simultaneously expands their structural dominance over the entire stablecoin market. The coordination is just brilliant. But, you know, Circle and Coinbase are essentially rebuilding the plumbing for moving money.

13:03None of that matters if everyday people don't have the capita or the credit to participate in the financial system in the first place. That's a very fair point. And that brings us to a firm who is quietly rebuilding the infrastructure of trust and credit. This is a crucial pivot. We are shifting from the mechanics of moving money to the intelligence of deciding who gets to borrow it. Let's look at a firm, ticker AFRM, also founded in 2012. They're best known for their buy now, pay later or pay in for checkout models, which are everywhere now. Everywhere. Yeah. But according to our sources, they just rolled out something profoundly futuristic.

13:38They launched a transformer-based machine learning model specifically designed for real-time underwriting. That's fascinating. Yeah, and the tech nerds listening are going to appreciate this context. A transformer model is the exact same architectural family that powers large language models like the AI chatbots everyone is using right now to summarize articles or write code. It is the specific technology breakthrough that revolutionized natural language processing over the last few years. Yes, but instead of pointing that transformer at a massive database of internet text to write you a poem, a firm is pointing it at your personal credit history.

14:16Wow. It is underwriting your paying for checkout at a clothing store instead of, like, generating your Spotify wrapped. If we connect this to the bigger picture, the underlying mechanism of how a transformer works is what makes this application so revolutionary. Okay, walk us through that. So the way a transformer model interacts with human language is by deeply understanding context and sequence. It doesn't just read isolated words in a vacuum. It mathematically understands how a word at the beginning of a paragraph fundamentally changes the meaning of a word at the very end. So look at the whole timeline of the thought rather than just flipping through flashcards.

14:52That's a great way to visualize it. Now apply that exact same sequential logic to a credit history. A traditional credit score, like a standard FICO score, is often just a static, blunt snapshot in time. It basically reduces your entire financial life to a three-digit number. Which leaves out a lot of context. Exactly. But a credit history is not static. It is actually a complex sequence of behavioral events over time. It's a narrative. For example, a missed payment that is followed immediately by a full payoff means something very different than a missed payment that is followed by the user opening three new lines of credit.

15:28It is literally reading the context and the plot of your financial life. Yes. The transformer doesn't need a human loan officer to hand design specific risk patterns to look for. It naturally finds the hidden behavioral signals in the sequence and timing of your financial history the exact same way an AI finds contextual patterns in human language. That is incredible. And the only reason a firm can pull this off is because they have spent 14 years underwriting every single purchase individually. Right, because unlike a traditional credit card, a firm doesn't do revolving debt and they famously don't charge late fees.

16:06They have to approve or deny you per item right there at the checkout screen. Which means they have accumulated 14 years of pristine, hyper granular item level data to train this artificial intelligence on. Traditional banks just don't have that kind of sequential data set. Because they just look at the monthly statement. Exactly. They just know if you paid your monthly minimum on time, they don't see the micro behaviors. Here's where it gets really interesting, though. A firm's new model isn't just looking to optimize prime borrowers with great credit. They're intentionally using this AI to evaluate thin file shoppers.

16:40The thin file demographic is notoriously hard to underwrite. For the listener, a thin file shopper is someone who lacks a traditional credit history. We are talking about recent college graduates, new immigrants, or just people who actively choose not to use revolving debt, leaving them with literally zero FICO score. Right. Doesn't it sound incredibly risky for a financial institution to point an AI at their data and suddenly start approving microloans for people who have no established credit history? Well, traditional banking intuition would say absolutely yes. That sounds like a fast track to massive defaults.

17:15Right. But the data from the rollout tells a completely counterintuitive story. In their initial deployment, this new AI model safely approved 3.4 % more applications that would have otherwise been auto-declined under their old legacy system. So they increased their overall loan volume simply by saying yes to more people. Yes, but that isn't even the most impressive metric. The critical part is that those exact loans, the ones granted to people with thinner credit histories, actually performed better than the risk expansions they had attempted under their older traditional underwriting models.

17:49Wait, I want to make sure I have this straight. They approved more people with less traditional credit history and their actual toll rate went down. That is exactly what the data shows. Libor McCulloch, a firm's president, summarized it perfectly in the source material. He said that seeing a consumer's credit history more clearly means you can say yes to more people responsibly. That's a huge paradigm shift. It completely shatters the legacy banking assumption that expanding financial access inherently requires taking on more institutional risk. A firm just proved that with highly advanced AI underwriting, you can grow your approval rates and improve your loan quality simultaneously.

18:26So it's not a zero-sum game anymore. It is no longer in either a proposition, no. That is wild. It basically means the algorithm understands your financial habits and your likelihood to repay better than a human underwriter ever could. And significantly better than a static three-digit FICO score ever could, that's for sure. So let's bring this all together for the listener. Think about the contrast we set up at the very beginning. The macro noise versus the builders. Exactly. In a week where the traditional financial world was hyperventilating over a Federal Reserve rate hike, and everyone was doom-scrolling their savings accounts waiting for the dust to settle, These three companies were operating on a completely different timeline.

19:06They were just executing. Circle laid down the brand new operating system for the future of stablecoin finance. Coinbase aggressively cut the tolls to drive high-volume traffic onto those specific tracks. And Affirm got remarkably smarter about who actually gets a license to drive on the financial internet. And synthesizing those three distinct strategies leaves us looking at a truly fascinating horizon. How so? Because right now we are analyzing these three technological advancements in isolation. We have a firm's transformer models proving with real data that they can read sequential credit histories like language to approve loans instantly and accurately.

19:46And we have Circle's new ARK blockchain proving that you can facilitate frictionless globally settled financial transactions without the baggage of traditional crypto gas fees. So if we project this out, what happens when these technologies inevitably merge? Imagine an artificial intelligence that not only underwrites your credit in milliseconds, understanding your unique financial narrative better than any human loan officer ever could, but then the moment it approves you, it instantly issues that microloan directly to your digital wallet. On a layer one blockchain like ARK, settled immediately in USDC, completely bypassing the traditional banking system.

20:26Exactly. No waiting three days for an ACH transfer to clear, no hidden origination fees, and no revolving credit traps. It completely changes the definition of what a bank even is. The infrastructure for that future is being deployed right now. That is a wildly provocative thought to leave on. And a great reminder for you, the listener, that while the macroeconomic noise is always loud, the real signal is almost always found in the technology being built quietly in the background. That's true. Stop doom-scrolling the rate hikes and start watching the infrastructure. Thank you for joining us on this deep dive and keep seeking out those fascinating intersections of technology and finance.