Zubin: Alright Aman, I'm telling you, you can't just swan back onto this podcast after three months of globetrotting and pretend the scorecard doesn't exist. The listeners have receipts, I have receipts, and you know what? Ben and David from Acquired probably have receipts as well.
Aman: You know, Zubin, I've been waiting for this episode. I've been waiting for this episode since the day we recorded it, the season one finale back in December, because here's the thing about making 10 predictions on a podcast. Most people hope you forget. We're the idiots who wrote them down, marked them up and invited the internet to grade us.
Zubin: Right. Welcome back to A to Z Fintech. This is the special Q1 Hot Take. No cold opens, no cinematic scene-setting, no history sweep, none of the usual theatre.
Aman: Yeah, just a scorecard, about 15 minutes of accountability. Think of this as the quarter-pole of the 2026 season. In F1 terms, we're three races and two suspensions into a 22 race calendar. The tyres haven't even warmed up. The reset regulations are still being tested. The pit wall has the data and it's time to look at the standings, Zubin.
Zubin: Yeah, and the headline before we get into it is 4 out of 10 fully playing out, another 3 directionally correct, one spectacularly wrong and one where the outcome is, shall we just say, pending.
Aman: Which, between two guys with 50 years of combined scar tissue across banking, big tech and payments, I'll quietly submit is not a bad hit rate for 90 days of elapsed time. But look, let's not grade on the curve. Let's go through it.
Zubin: Yeah, let's start with the wins. The P1, P2 and maybe the P3 of Q1. Because if we're gonna eat humble pie later, we'd earn a lap of honour first.
Aman: The podium finishes. Start with the biggest one. Prediction number three, Visa and Mastercard settling on-chain.
Zubin: Dead right, and not in some vague directional way. Like in March, Mastercard walked into the market and acquired BVNK, a stablecoin infrastructure firm, for $1.8 billion, with a B, which for those keeping score at home is the exact thesis we laid out in December. They would stop fighting stablecoins and opt into them.
Aman: And we covered that, if you missed it, in our last episode. Now, Visa didn't just watch. Visa Direct is now running at $3.5 billion in annualised stablecoin settlement volume. That's not a pilot. That's not a press release. That's a product line.
Zubin: Yeah, and here's the non-obvious bit that I don't think the market has fully priced in yet, right? The card schemes aren't becoming stablecoin companies, they're becoming the regulated settlement layer for stablecoins, right? The new interchange, the same toll roads but different asphalt.
Aman: Yeah, well, win number two. Compute nationalism. Was prediction number eight.
Zubin: Right, this one wasn't even close. Like Gartner reported sovereign cloud spending will hit about, what, 80 billion this year, up 35%. The Middle East and Asia are holding H100s and B200s like central banks used to hold gold. But when you stop and think about it, that's exactly what they are. GPU clusters are strategic reserves now.
Aman: Yeah, and the tell, Zubin, was never going to be a single headline. It was a pattern. When France starts talking about AI factories in the same tone they used to talk about nuclear plants, you know the metaphor has shifted. Compute is the new oil and sovereignty is the new OPEC.
Zubin: Right. And let's go to podium spot three, the OpenAI-Microsoft divorce, which was our prediction number four.
Aman: Messy, public, in progress, exactly as called. The late March news broke that they restructured the partnership. Microsoft secured $250 billion in contracted Azure purchases, which sounds like a win until you read the fine print, slipped in, which effectively granted OpenAI compute autonomy. Microsoft stock slid in Q1 because investors finally understood what we said in December. They are funding a company that is actively swimming across the moat.
Zubin: Yeah, the word divorce is doing a lot of work here. It's not a clean split. It's this conscious uncoupling with, what, let's just say, a $250 billion prenup.
Aman: And the fourth podium, because I'm counting four wins here, not three, is the one I, Zubin, am quietly satisfied about. It's the Ive-Altman hardware baby, or the lack thereof. Prediction number five.
Zubin: Yeah, this is the paperweight prediction, right? You called it a paperweight in December. Yeah. In March, last month, the first prototype leaked, right? The reviews from people who've actually held the thing are calling it, and I quote in spirit, beautiful, minimalistic, screenless ambient device that feels like a luxury object rather than a smartphone killer.
Aman: Yeah, so minimalistic that it didn't even appear, so far. It's the politest possible way of saying pretty paperweight. And look, I want to be careful here. I'm a lifelong fan of Jony Ive. I was a fanboy before the iPhone was launched, but the market is sceptical. The market is sceptical for the exact reason we flagged in December: melding hardware and software into a magical object is infinitely harder than Steve Jobs made it look. And Sam Altman, for all of his talent, is no Steve Jobs.
Zubin: Alright, so 4 for 4 on the podium. Let's get into the mixed results before we get too cocky.
Aman: Okay. The in-the-points bucket. Not P1, not on the podium, but we're on the board. Let's go.
Zubin: Yeah, prediction one, right? Like, the CUDA Killer. Google and Huawei.
Aman: Yeah, well, we're kind of halfway right and halfway wrong, but fully interesting. There is no signed Google-Huawei joint venture. That would have been, you know, we've got still about three quarters to go. Geopolitics made that one a bridge too far, as it probably always was. And here's what happened. In late March, Huawei launched the Ascend 950PR, the Atlas 350, claiming that it's 2.87 times faster than Nvidia's H20. Meanwhile, Google is aggressively pushing its OpenXLA stack to decouple software from Nvidia's hardware, and Huawei is building its ecosystem to be compatible with those open standards.
Zubin: So, I mean, what we're saying is the alliance is real. It's just not a press release. It's also sort of this open-standard handshake, which if you think about it is actually a more durable form of partnership in this sort of geopolitical cold war.
Aman: Yeah, the non-obvious insight here we missed in December. In a world where American export controls make a formal JV impossible, the new alliances form through shared standards, not shared equity. That's a lesson I'll take.
Zubin: Prediction 7. Agents get a license.
Aman: Well, also directionally correct, Zubin. Wrong, you know, in the specifics. We predicted regulators would create a dedicated license for AI agents to hold money and sign contracts. What actually is happening is more interesting. In March, the UK's Competition and Markets Authority published guidance stating that businesses are 100% legally liable for their AI agents' actions.
Zubin: You know, different from a license, I'll give you that, like arguably more consequential, because if you're a company and you sort of deploy this agent that goes off the rails and, what, books 10,000 flights to Kazakhstan on a corporate card, you can't hide behind "the AI did that," right? Like, that's the regulatory thunderclap, right? The legal framework for agentic commerce is being built faster than we expected, but through liability, not licensing.
Aman: [Al-Mahdi here we come.]
Zubin: Yeah.
Aman: Prediction number six. Quantum breaks encryption. Q-Day scares.
Zubin: Right. Yeah, no, we've not had a Q-Day event yet, but you know, I think the panic upgrade cycles we predicted, I would say completely real, right? Like, we saw this in February with Google's leadership issuing this massive call to action, warning that, you know, the store now, decrypt later sort of attacks are accelerating. We've also seen banks spending record amounts on post-quantum cryptography this quarter itself. Now research published in Q1 dropped this physical qubit requirement to crack, you know, what, 2048-bit RSA, from what, 20 million down to under 1 million. That's a factor of 20 reduction in security margin.
Aman: The actual break is closer than the industry wants to admit. The scare is here. The event isn't. I'll take half credit and a nervous glance at the calendar.
Zubin: Yeah, now this may have a longer horizon than even 2026. Now, tucked in a quick one on our prediction nine, AI enters real F1, right? Mercedes announced a massive expansion of their Microsoft AI partnership for the 2026 season. They're running 1.1 million data points per second, which is staggering for live decision-making. Red Bull and Oracle revealed their own AI strategy agent. Whether it wins the championship is a Q4 question, but the AI strategist is officially in the garage, right? Directionally, correct.
Aman: Alright, the part of this episode where we eat pie. Alright.
Zubin: Hand me a fork. I'm on prediction 2. The mega-merger. JPM buys Nubank.
Aman: Dead wrong. No movement. Jamie Dimon's Q1 commentary was all about sovereign AI infrastructure. Nubank remains independent, growing, listed in the US, that's material. And frankly, looking more like the acquirer than the acquired. David Vélez isn't waiting to be bought. He's preparing to invade the United States, which we covered in the Purple Revolution episode back in January. Yeah, yeah, yeah.
Zubin: So, directionally the arrow was wrong. We had JPM swallowing Nubank. The reality is Nubank might just be the aggressor, right? The hunter becoming the hunted becoming the hunter again. Which, by the way, interestingly, is, I would argue, Revolut's playbook too.
Aman: But I want to say something that the research on this turned up. Our corollary to that prediction, the one that we almost threw away as an afterthought, was dead right. We said the JPM move would trigger reverse mergers, where large neobanks would buy small traditional banks just for the license and the deposits. That's exactly what's happening, Zubin. Several European neobanks moved on distressed regional legacy banks in Q1 for precisely that reason. I'm being defensive.
Zubin: Yeah. And so you're saying the headline was wrong. The footnote was right, which is, you know, a very podcast way of hedging, Aman, and I'll call you out on it.
Aman: Fair, fair, you know, harsh, but fair. Mark it down as a loss, with a silver lining. Prediction number 10, Zubin, the hostile takeover of A to Z Fintech by Ben and David from Acquired.
Zubin: Let me just make one last check. Ooh, no call from Ben or David, which, I'm going to be honest, slightly hurts my feelings. I had sort of a valuation range ready.
Aman: Yeah. And by the way, it's not like there haven't been any acquisitions, Ben and David. Like, come on. There's been quite a few, OpenAI, this week in Fintech.
Zubin: Yeah, this week in Fintech with Claude, that was a good one.
Aman: So, so we remain stubbornly, defiantly, irritatingly independent. Power to the people.
Zubin: Alright Aman, final lap. 4 for 10 fully, 3-ish directionally correct, 1 spectacularly missed. That's a scorecard. But I wanna push you on the pattern, right? Because anyone can call a scorecard. What's the non-obvious insight underneath?
Aman: Here's what I see. When I look at what we got, where we got it right and where we got it wrong, there's a signal in there. We got it right where we read the structural logic of the system. Stablecoins displacing card rails was always going to be a capture, not a war. The OpenAI-Microsoft tension was always going to be a divorce because their incentives had diverged. Compute nationalism was inevitable the moment GPUs became strategic assets. These were calls based on 50 years of watching how big institutions actually behave under pressure, not reading the headlines.
Zubin: You make a valid point. And where did we get it wrong?
Aman: I think we got it wrong when we substituted the narrative for the structure. The JPM-Nubank merger was a beautiful story. We'll grant you that. Jamie Dimon finally gets his global retail bank. David Vélez cashes out at a premium. The symmetry was irresistible, but symmetry is a storytelling instinct, not an investing instinct. And the actual logic of the system was the opposite. Nubank has lower cost-to-serve, a better tech stack and a founder-CEO with 20 years of runway. Why would he sell?
Zubin: You know what, Aman, that's the single best self-diagnosis I've heard on this podcast. "We got it wrong where we mistook narrative for structure." Write that down.
Aman: And that, for our listeners, is the whole game. Looking around corners isn't about being psychic. It's about knowing which parts of the system are load-bearing and which parts are just decoration. Zubin and I have spent 50 combined years, you can tell it by our hairlines, inside these boardrooms, across five continents, watching people confuse the two. That's the moat of this podcast. We don't always get it right, but we get it right enough, and often enough, that when we tell you something, it's worth writing down.
Zubin: You know what, I think the most natural next question is, what are the quick call-outs for Q2, Q3, through Q4? Because I know everyone's going to be asking. I mean, what are you watching?
Aman: Well, three things. One, the Mastercard-BVNK integration. Talked about it last episode. Watch how fast corporate treasurers move from pilots to production on stablecoin settlement. That's the real tell on whether it's a 2026 or a 2028 story.
Zubin: And Aman, you're looking back at the Stripe-BVN, sorry, Stripe-Bridge speed of integration. I think this is a 2026, early 2027 story. What about two? What's the next trend you're watching?
Aman: Two is the agentic liability question. The UK moved first. Singapore, the EU, and the US will follow in, I suspect, Q2, Q3. How they frame it will determine whether agentic commerce scales into enterprise or stays stuck in demos for the next 18 months.
Zubin: We need another pod on agentic commerce. We do. So what's the third one?
Aman: Third is, watch Nubank in the US. With a national charter, they could launch with meaningful product-market fit in a single or multiple states by Q3. Every American regional bank CEO is going to spend their Thanksgiving writing a strategy. If they don't, the Purple Revolution pauses at the border.
Zubin: Yeah. So, you know, honestly, those are the three areas I'm watching too. And for me, maybe I'll add one more, right? Like, and we covered this a couple of weeks ago, which is the PayPal endgame. We did an entire episode on this. I'm not going to get into that, but Enrique Lores is the accountant. The accountant's job is to find the highest bidder. By Q4, I think we'll get an answer on which of the three scenarios we laid out actually plays out.
Aman: You know, Zubin, it occurred to me. I want to tell you something. If you ever made, if we ever made Zubin's dream sort of, like, a Game of Thrones kind of story about payments, the hero of that story would be Dee Hock and the villain of the story would be poor Enrique Lores.
Zubin: Man, yeah, he's got a hard job ahead. Alright, let's bring us home, right? I almost feel like there's this mic drop coming and we've got maybe a minute or so that we wanna cover this on.
Aman: So here's the thing about predictions. Most of them are performance art, right? Somebody in a suit tells you where or what's going to happen. And by the time they're wrong, everybody has moved on to the next shiny forecast. We don't do that really at A to Z. We write it down, we come back 90 days later, and we check the ledger in public. Because the only thing that separates a non-obvious insight from a lucky guess is the discipline of being graded. Four for ten fully playing out in Q1, with another three directionally correct. It's not a victory lap. It's a reminder that between Zubin and me, we've spent 50 years inside these companies. That experience isn't decoration. It's the reason we could see Visa and Mastercard capitulating to stablecoins before the press releases hit. It's the reason we flagged the OpenAI divorce before the divorce lawyers showed up. And it's also the reason we got JPM-Nubank wrong, because 50 years teaches you that sometimes the story you want is not the story the system is writing.
Zubin: If you're the kind of person who wants to look around corners instead of reading the rear-view mirror, you are in the right place. Stick with us. Tell a friend. Tell the CFO in your life who still is trying to figure out what agentic commerce means. Subscribe, share and send this to that one person in your network who you think deserves a seat at the grown-ups' table on this stuff.
Aman: And we'll see you at the half-season review in July. By then, a couple of these directionally correct calls may have graduated to dead rights. And a couple of the dead rights may have gone backwards and wobbled. That's the beauty of making real predictions, in real time, in public. The scoreboard never lies.
Zubin: On that annoyingly profound note, I'm going to go update my spreadsheet and maybe quietly brag to my kids that Daddy called the Mastercard-BVNK deal three months before it happened.
Aman: They're going to be totally unimpressed, Zubin.
Zubin: Utterly, utterly. Stay purposeful and stay on the grid.
Aman: Stay curious.