Prof G Markets - $1T Moved on Iran “Talks” — Did They Even Happen?

Episode Date: March 24, 2026

Ed Elson speaks with Justin Wolfers about the potential for deescalation in Iran and whether markets were right to rally on Trump’s word of negotiations. Then he discusses a slate of news out of OpenAI with Alex Heath following the company’s decision to refocus on its core business. Finally, Ed offers advice to investors amid the volatility of the war.  Justin Wolfers is a Professor of Economics and Public Policy at the University of Michigan. Alex Heath is the author of the Sources newsletter and co-host of the Access podcast. Check out our latest Prof G Markets newsletter Follow Prof G Markets on Instagram Follow Ed on Instagram, X and Substack Follow Scott on Instagram Send us your questions or comments by emailing [e-mail] Learn more about your ad choices. Visit podcastchoices.com/adchoices

Megan Rapino here, this week on a Touchmore, our Captain America Hillary Knight is joining us to talk about her storybook career. And the March Madness Bracket is out. And we have thoughts and predictions to share. Plus, we're also taking a look at the NWSL's Blockbuster opening weekend. Check out the latest episode of a Touchmore, wherever you get your podcast and on YouTube. Mormons are having a moment. But some in the church wonder if that's actually a good thing. I'm not so concerned about Mormonism kind of being radicalized.

I'm actually more concerned about it becoming so obsessed with assimilation that it kind of loses sight of what it actually is. What happens when faith goes mainstream? That's this week on Explain It to Me. Find new episodes, Sundays, wherever you get your podcasts. Chances are your favorite websites used to depend on Google for traffic. and money. But that's not really working anymore. Now publishers are scrambling for new lifelines. Neil Vogel, who runs People, Inc., says his company figured it out a couple years ago. You would think, given what everyone said about us, that we would be the guys that would be doing the

worst now. We're kind of the guys doing the best now. I'm Peter Kafka, the host of channels, the show about tech and media and what happens when they collide. You can hear my conversation with Neil Vogel now, wherever you listen to your favorite podcasts. Today's number, 49 billion. That's how many dollars VC firms invested in defense tech companies last year, the largest number ever. New products include subscription-based missiles and nuclear solutions that scale. Welcome to Profi Market's math. I'm not. I'm not. Welcome to Profi Markets. I'm not. Ed Elson, it is March 24th. Let's check in on yesterday's market vitals.

The major indices rallied as President Trump signaled a move to de-escalate in Iran. More on that in a moment. Oil fell on that news. Meanwhile, a yield on 10-year treasuries dropped, and the dollar declined. Okay, what's happening? The war in Iran may be entering a turning point. On Saturday, Trump gave Iran a 48-hour ultimatum, reopened the Strait of Hormuz or the U.S. would destroy Iranian power plants. Iran responded by threatening to take out U.S. energy and desalination infrastructure across the Gulf just two days later. Trump said the two countries were engaging in, quote, very strong talks, and that he would pause all strikes for five days. However, Iran's parliament speaker quickly dismissed the news, saying the talks never happened. Nonetheless,

the S&P rose more than 1% and Brent crude oil fell more than 10% on the news. So here to help us make sense of what is actually happening here, we're joined by Justin Wolfers, professor of economics and public policy at the University of Michigan. So Justin, first Trump tells us that we're having these constructive talks with Iran.

It starts to look like maybe the war is coming to an end. Then Iran tells us, no, that didn't happen. Markets kind of digested that. They sort of fell down a little bit, but ultimately they're still up. What do you make of what we've learned here in the past 24 hours? It's the most extraordinary thing that the most critical issue facing American financial markets is whether to believe the American present or the Iranian leaders about what the

Americans are doing? Should we believe the American president talking to the American people about the actions of America? And I think it remains very much up in the air. And just to be really clear on the stakes here, Ed, as you said, markets rose very strongly on news that these talks were going ahead, probably about a third of it unwound as it appeared that the Iranians were not so confident that anyone had ever called them. So that right there tells you that markets believe that our president is slightly more credible than the Iranian leadership, but not substantially so.

Look, 1% on the S&P 500 is $600 billion. Maybe it rose about one and a half percent, actually, if I remember. So that's about a trillion dollars. There's about 100 million American households. So that says whether the president was telling the truth or not makes a difference in the wealth of the average American household of plus or minus $10,000. It's just a way of saying the stakes here are incredibly high and we are sailing through fog at midnight with a blindfold on while listening to the sweet, sweet tunes of our president telling us what may or may not be able. What do you think about how investors are going to make decisions based on what they believe or don't believe going forward? Because, I mean, it's clear that literally the entire world, the entire economy, all of global markets depend on these questions here, on what happens in Iran, what happens in the straight of her moves. And as we're learning more and more, it's also going to, I mean, our entire economy is going to depend on this question. If this, if this, if this area remains closed, then we will see unbelievable amounts of inflation.

I mean, if we see oil consistently remaining at, I mean, let's call that $150, maybe $200 a barrel. I mean, the implications are tremendous here, and it all comes back to literally like, what is this guy going to do? And do we believe him? And it reminds me of Greenland. It reminds me of the tariffs. and then it just brings up this question like, are we in the same position again, where we're basically just hanging on to his every word, making decisions about his every word,

which often just don't pan out to be even true. Yeah. So for sure over the next few weeks, lots of people are going to say financial markets are crazy, their hormone or why are they moving up or down like crazy. Right. But if you're actually in a situation where there's two choices we could make, aggression, non-aggression, and they have huge effects on the economy, and we don't know the probability, and we really, really don't know the probability if we end up here or here,

then small statements rationally lead to large re-evaluations of the value of stocks, of your future forecast for the economy, of bond rates, the whole nine yards. So things are going to look crazy. It will, but I think here this is not a statement about financial markets necessarily being temperamental. It's more a statement about the president's inability to convince anyone that what he's saying he means. And this is actually, I think there's one profound sense. There's many senses in which these are uncharted waters. But one very profound sense is I don't think we've ever been in a situation like this where

the word of the president about what his intentions are is so uninformative about the future. Right. George W. Bush, when he said, let's go ahead, you kind of knew that's what he meant. And when he said, let's pull back, you kind of knew what he meant. And here, I think we're all just guessing. Do you think that this kind of goes back to Taco where, I mean, the whole premise of Taco was Trump threatened something crazy, and then the markets throw up, and then he gets worried because he values the opinion of the markets,

or he values the dollars that are related to those investment decisions, and then he tacos, he chickens out. And then it became a question of maybe he's immune to taco at this point, or maybe the markets aren't reacting anymore to sort of front-run the taco, which means that we no longer have this regulating effect where the markets basically slap them on the wrist and tell them to do the right thing. like, what do you think the relationship between the markets and Trump's decision-making actually is at this point? And is that relationship still strong as it seems to be back during the tariffs

of last year? Yeah. So I think that's a, it's been a question very much on a lot of people's mind. So if your simple model was the president will do what he wants until he hears markets don't like it, when he hears markets don't like it, then he'll undo it. And if markets can think one step further, then they'll see the president does what he wants. It's not very good. They think he'll undo it. Given that they think he'll undo it, they don't need to move. Right. It's the equilibrium of this game. It could be that markets don't move very much and the president becomes hypersensitive to markets. Who knows? This feedback loop is profoundly broken and it would be a lot easier if it was genuinely mechanistic, but it's not.

Maybe, you know, all of this comes back to what, how do you rewrite the rules of the game when you have an unpredictable president? And so the idea of TARCO that people found very reassuring is, in fact, the president's very predictable. Right. I don't know that that's true, actually. So yes, he tacos sometimes, except when he doesn't. And so you can think about, you know, and I do think you're right to bring up all the

past analogies, I think most clearly through the trade war. But if you remember, Canada went from our number one enemy to our number one friend, and it did like three or four round trips on that journey. Actually, right now, the rhetoric is profoundly anti-Canadian, and the reality is not very anti-Canadian. But it's very hard to keep track of, and it's not even clear that the president has. So, mate, if I look a little bit lost, it's because this is very confusing. So, I just, I just, Look, here's the safest thing to say. Markets are reacting as if developments in Iran are tremendously important for the development of the global economy.

Yes. When you see the move one and a half percentage points based on a truth social post, and that takes into account that he might taco, that one and a half percent is a profound underestimate of the true effect of going to war because they don't think that he pulled back from war. they think there's some possibility he did and some that he didn't, in some sense, what's already priced in. The economic implications here are numerous, and it's kind of hard to put any numbers on it. I'm just wondering, when you teach your students over at U Michigan, I assume that what's happening in the news is making its way into your classes and into your lectures, what kinds of takeaways are you,

trying to convey to your students right now in the middle of this moment. Yeah, Ed, I'm not teaching this semester. Sorry about that. Okay, fair enough. I am talking to a lot of people. So let me give you a couple of very quick answers. The first is the orders of magnitude involved with war are very, very large. Second is what I think is being the most important macroeconomic skill, which is being able to keep

track of orders of magnitude. to give an example, I was interviewed a bunch of times a week ago just after it came out that the Pentagon had said that the first week of the war cost $11 billion. And commentators like, oh my God, $11 billion, this is outrageous, it's terrible, it's too much, and I'm like, you know, actually 11 billion is not that much. 11 billion is, well, if there's 100,000, 100 million households, you can do this for me, Ed, come on. Don't put me in this position.

It's a hundred bucks to household. It's a hundred bucks per household. Not much. But then you see markets today rose one of the half percentage points. And as we talked about, that's a trillion, that's, you know, a trillion dollars. Right. And so the stakes aren't tens of billions. The stakes are hundreds of billions and plausibly trillions.

And so therefore the stakes for the average, and no one is average, the average, the average American household thousands and plausibly tens of thousands of dollars. So two things that you learn out of that. One, big deal. Two, keeping track of orders of magnitude is actually the most important skill here. I can probably guess how many zeros are involved. I can't, anyone who thinks they know what number is in front of the zeros is kidding themselves about how precise they can be. All right. Justin Wilf is Professor of Economics and Public Policy at the University of Michigan. Always appreciate it. Thank you so much.

Great pleasure. After the break, OpenAI locks in. And for even more markets, insights, you can subscribe to my weekly newsletter, simply put, at simply put.profgemedia.com. Support for the show comes from ShipStation. When your company is growing, nothing stops that growth quite like a bad customer experience because they couldn't get your product. Luckily, Shipstation's intelligence-driven platform

form brings order management, rate shopping, inventory and returns, warehouse systems, and comprehensive analytics all in one place. They say they can even save you up to 15 hours of time per week on order fulfillment. Shipstation automatically chooses the right carrier, finds the best rate, prints labels in bulk, and even sends tracking updates to your customers. Shipstation even compares rates across major global carriers like USPS, UPS, UPS, and FedEx, including any discounts you already have to select the best shipping option for every order with savings of up to 90% off. You can try ShipStation free for 60 days with full access to all features. No credit card needed. Go to Shipstation.com and use code markets for 60 days for free.

60 days gives you plenty of time to see exactly how much time and money you're saving on every shipment. That's shipstation.com code markets. Shipstation.com code market. Shipstation.com, code When you think of the most influential American politician of the 21st century, you probably think of a man. But there's one woman who might fit the bill. She never thought she'd be in politics, but went on to break what she calls the marble ceiling. Then they'd say, well, why don't you all just make a list of things that the women want and will do those? What? This is in this century.

Really? So the marble ceiling, it's not a glass ceiling. It's a marble ceiling. And they all had it lined up, but you go next and I'll go next and I'll go next. And we're like, well, you know what? We've been waiting over 200 years. We're breaking in line.

Nancy Pelosi, the longtime Democratic Speaker of the House, live on stage at South by Southwest in Austin. Today explained, every weekday, and now on Saturdays, too. This week on version history, our chat show about the best and worst and most important products in the history of technology, we're talking about a gadget that was meant to be used on phone lines

and was eventually used by the military and then finally changed the music business forever. That's right, of course, I mean the vocoder. The thing that let us all play our voices like an instrument and change the way that we think about our voices. We have a really fun guest, we have a really fun story to tell. All of that is on version history, on YouTube,

and wherever you get podcasts. We're back with ProfG Markets. Open AI is changing course as the AI race intensifies. In an internal memo, the company's chief of applications said that OpenAI needs to ditch its side quests in order to, quote, nail productivity on the business front. That new strategy includes building a super app that combines ChatGBT, BT, Codex, and the Atlas browser. The company is also planning to double its headcount by the end of the year, as it works to keep up

with Anthropic. Here to discuss Open AI's new strategy. We're speaking with Alex Heath, author of the sources newsletter and co-host of the Access podcast. So Alex, this new news from Open AI, they are trying to ditch the side quest. This is sort of the initiative that's being spearheaded by Fiji Simo, who I know you have interviewed and who you've been talking with. Tell us a little bit about what you know about what they're trying to do at this point. How has the Open AI business strategy actually changed here? Yeah, I mean, I've been seeing people calling this a pivot, and I guess you could say it is. But if you look at what's really happened at over the last couple years in AI, Open AI went square for consumer, being the biggest consumer AI platform.

And they did that, right? They have 900 million users, and they're growing there still. Meanwhile, what Anthropic did is during that same period, it focused almost entirely on enterprise. And now Claude has, you know, some traction, especially off of, you know, people wanting to support the company due to what happened with the Pentagon. But Anthropic was B2B, Open AI was consumer. Now that OpenAI has one consumer, guess what? The vast majority of its users actually cost it money to serve. This was something I was talking about with Fiji when she was on Access a couple months ago.

And that's a problem. If you were a venture-backed, unprofitable AI lab who, needs to IPO, right? And where is the money in AI? It's in enterprise applications. It's in putting agents in large companies, doing forward-deployed engineering, building all these kind of bespoke products for enterprise usage. That's where, I mean, in Silicon Valley, people are talking about token maxing, right? Companies literally incentivizing their employees to burn as many tokens, you know, which are just the atomic units of AI as possible. And there's literal leaderboards of, you who is doing this and spending the most money inside the company.

That doesn't exist in consumer land. So Open AI needs to make money. It needs to become profitable. And short of the ads thing that they're doing, which we've talked about a lot, this is the way to do it. It's to go after Enterprise. And they finally have a model and a product in Codex that is growing on that front, on the coding front, and finally can challenge what Claude has been able to do,

specifically on coding, definitely doesn't have the mind share, but has exploding usage. And so, yeah, the super app idea makes a ton of sense, and as does, you know, growing the enterprise push, if you look at it that way. And I think, I think this is just the beginning. I mean, I think the super app idea is like, this is the future of all of these assistants. I think Claude code is going to be the interface to access what we think of as Claude, the chatbot. and I think the same will happen for Chad GPT, Codex will eat it. Do you think that they see their push for consumer versus enterprise as a mistake at this point? I mean, that's really how it seems on my end.

I look at how anthropic, how Claude is taking over the enterprise market. But I always thought that this would happen. And one of the predictions that I made a while ago, about two years ago, I thought that ChatGPT as a consumer product was going to kind of fall by the wayside. And my prediction was opening I was going to sort of shift their focus. They had this awesome hit product, but they're going to go after the real big fish that everyone wants, which is enterprise customers. So chat GPT will kind of fade out of view and we'll see more emphasis on enterprise. That's not what happened at all. But if they had gone with what I thought

they were going to do, then maybe things would look better for them. But do you think that they view that as a mistake, that they went after the consumer and they realized this actually is isn't that great of a business compared to Enterprise, and now they're switching over? I think they view not focusing on coding and coding model progress earlier as a mistake, as a huge strategic mistake that they're now catching up on, and they gave Anthropic the opening with Claude Code that it has.

I'm not so sure about the consumer versus enterprise thing. I mean, look, when you get to a billion-ish users, you can kind of just figure it out. That's like the hardest thing to do. Right. And, you know, what is harder? You know, Anthropics Enterprise deployments or getting hundreds of millions of free users to be very retentive in your product. It's hard to say.

I mean, I think, you know, it's incredible long-term optionality. If they can figure out the ads piece, which they will, they just hired a senior meta ad exec this week. Fiji ran, you know, monetization at Facebook. They'll figure that piece out. That's the thing Anthropic doesn't have. And that's probably where all the margin is, really. If you start doing really good ads on a surface that high intense, that data rich for hundreds of millions of people, you've got the next massive scaled consumer internet business, which Anthropic just doesn't have by the function of Claude not being huge as a consumer product. Yeah.

So I'm not sure. I don't know if they really feel like, oh, my gosh, we forfeited, you know, enterprise and now we got to scramble. I do think, though, the coding piece. I think these companies judge themselves based on the model progress. And I think Open AI got distracted with, I mean, Fiji said it, all the side quest, but with, you know, being a massive consumer company and hiring all of these consumer, you know, previous era tech employees to come in and do all these different things and SORA and all this stuff. And meanwhile, Anthropic was singularly focused on coding and the enterprise market. And now that's paying off. So they're playing catch up there for sure.

It does seem like focus would also just mean. building out this ad business. If they want to win in the consumer world, which they have done, and they want to figure out how to monetize it, now it's time to get into ads. And as you just said, they just hired one of the top advertising executives from Meadow, who's now going to lead their ad sales. Would you say that advertising is still top of mind for Open AI in spite of being made fun of for it after the Super Bowl with those Anthropic Super Bowl ads? People saying we don't want ads in our AI apps, do you think they're still going for the ads? Yes, they are. I mean, they're testing it and the early reporting on the tests so that they've been super low tech, surprisingly, very bespoke and white glove.

And I think they're just trying to feel it out and feel out the limits of advertising on a surface like this. And now they're productizing it and they're building the system to kind of deploy it at scale. They have to do it for the reasons we talked about. Yeah. I think if you bought the theory from Sam Allman that's super intelligence or some version of it is just around the corner and this technology is really going to automate all labor and replace, you know, scientific drug discovery and all the things that they talk about. Would you be doing a concerted ads push? Maybe not, right? Like, if you've got like AI God right around the corner. So I think it also suggests that this model progress that gets hyped a lot in the industry is maybe not as dramatic as, as it said. But there's no, I think

ChachyPT is going to be around for a long time. I think it's going to be big. It's just ingrained. You know, it's like the Kleenex of AI for a lot of people. That takes a long time to unseat. That's what Google managed to accomplish. Yeah. And sure, like there will be nipping at the heels and at the edges of it, like what Claude is doing with a lot of like tech-related adopters right now, who just anecdotally, I'm also part of this, but here just like switching from chat to Claude.

But in terms of your mass average consumer, chat is still the only AI experience that they have and that they know. and that's just going to be very hard to unsee. Yeah, the other final news, I just want to get your reaction to that I saw today, this is exclusive from Reuters. Apparently Open AI is offering private equity firms investment deals with a guaranteed minimum return of 17 and a half percent.

So basically saying, if you invest in us, we guarantee you we will return 17.5% of your money, which, for those who don't know, that's like a ridiculous return that's that's significantly higher than the S&P average. This just seems ridiculous to me. How can you guarantee 17.5%? It makes me think maybe they're just going to lose money on this thing or maybe some sort of foul play. I don't really know. I just wanted to know if you know anything about this and if that sounds right to you. I don't have reporting to share on it. I do agree with you that it's unusual.

and maybe a red flag. Someone has been picking up the bill constantly. Yeah. And who is, like, going to keep picking up the bill? That comes back to why they need to do enterprise, why they need to do ads. Like, at the end of the day, you need to eventually control your own destiny,

which as a company is producing free cash flow, right? Which is what the meta and Google and apples of the world have, is they have control of their destiny because they have massive money cannons. And Open AI right now is contingent on Sam's ability to convince another large pool of capital that he has somehow not already tapped to invest more money until they become profitable, right? And like maybe the IPO is the next phase of like someone paying the bill. Then it's dumped on retail. And maybe that's, you know, if you're going to IPO it by

Q4, maybe you feel confident giving a 17 whatever percent, you know, guarantee. But certainly strange. You know, I'm old enough to remember when Open AI was a nonprofit and all investors, up until like a year ago were labeled with like disclaimer, this can go to zero and you should treat this as a donation. So definitely a new territory we're in. It was the capped profit company. Now it's, there's a flaw, 17.5%. It's really, really incredible.

Okay, Alex Heath, author of The Sources newsletter and co-host of the Access podcast. Alex, appreciate your time. Thanks, Ed. Well, we started the week on the right foot. Trump said he had, quote, very good and productive conversations with Iran, and that is what created $2 trillion in market value in a matter of hours. The implication was that the war with Iran might be coming to an end, or at least that is what investors thought. But then, as usual, we got some clarifications

that complicated things. Iran denied having any direct talks to end the war, and the nation's foreign ministry said they'd had no negotiations with the U.S. that was when stock started to fall back down again, as investors recommend the possibility that the president might be, once again, talking out of his ass. Is that really what was happening here? Was he lying? I guess we don't know.

But I guess that is also kind of the point. And this was yet another reminder that anything that comes out of the president's mouth cannot really be taken seriously, especially if you are an investor. And we know this because we've seen, seen it over and over again. Two weeks ago, for example, Trump told us that the war with Iran was, quote, very complete. Investors bought that news, and then here we are, two weeks later, and the war

continues to rage on. Before that, we had the debacle with Greenland, where Trump suggested America would simply take Greenland, and he refused to rule out the use of military force. Investors sold on that news. One point two trillion dollars in market value was. erased. But then he got bored of that idea, and then he decided to move on to the next thing. Last year, he suggested firing Jerome Powell. He said his termination, quote, cannot come fast enough. Everyone sold, and then he turned around, and he said he had, quote, no intention of firing him. He did the same thing with, of course, the tariffs. The tariffs were on, and then they were off, and then they were on again and off again. And with each announcement, trillions of dollars worth of

stocks and bonds were traded, each time it happened. And so here we are again, this time betting not on trade policy, but on all-out war. But if there's anything we've learned at this point, is that if you want to actually understand things, if you want to understand the global situation, if you want to get closer to the truth, well, then there is one thing that you should not be doing. And that is you should not be listening to the president. Because what we know now is that despite the power he has, his words genuinely mean nothing. That doesn't mean he's necessarily lying right now, and it also doesn't mean he's telling the truth. What it means is it means nothing.

There are no conclusions you can draw. There are no predictions that you can make based on what he says. And if you try, well, then you will fail, just as millions of traders have failed before you. So we'll end here with a message to Wall Street. If you want to stop losing money, it's quite simple. Stop listening to the president. And for one simple reason, meaning cannot be made out of that which has no meaning at all. Okay, that's it for today.

This episode is produced by Claire Miller and Alison Weiss, edited by Joel Patterson, and engineered by Benjamin Spencer, or video. editor is Brad Williams. Our research team is Dan Shalan, Isabella Kinsel, Chris Nodonohue, and Mia Silverio. Our social producer is Jake McPherson. Thanks for listening to Profg Markets from ProffTee Media. If you liked what you heard, give us a follow. I'm Ed Elson. I will see you tomorrow.