Jason Furman: AI Tail Risks, Tech Valuations, Private Credit, and the Next Macro Shock

[upbeat music] We sit at the intersection of markets, policy, and the real economy, and the backdrop today, the Strait of Hormuz and the war with Iran. AI valuation stretched and private credit at scale without a full cycle test, and whether any of it's been priced into our markets. Joining me today is Jason Furman, Harvard economist and former chair of the Council of Economic Advisors under President Obama.

Jason Furman, welcome to the pod. Great to be with you. Before we get into frameworks, are you comfortable with where the world is today, or is there something keeping you up at night? You know, the short-run macroeconomic situation, there's always a list of things to be worried about. There is even more than usual with what's going on in Iran. But whatever it is, we'll get through it. You know, where we are in the medium and long run, I'm super excited and techno optimist, but there is a bad case out there and, and, you know, it's that tail risk that I'm most worried about.

Okay. So in the long run, we're gonna be okay. But what could happen in the next 12 months that would change the calculus? Oil prices could go up to 150 a barrel. Honestly, that's not great, but that's fine. It would probably take half a point off of U.S. economic growth. I don't think it would be enough for a recession. Our economy is much less oil-intensive than it used to be. We're net oil exporters. If oil could go up to $200 a barrel, I'd be getting more nervous if that happened.

But, you know, adjusted for inflation, we've seen things like that happen before when we were much less prepared for them than we are now. So for me, the oil stuff is in the known known category. That doesn't mean it's good, but we know how to deal with it. You know, what's going on on the technology side is incredibly exciting, but there might be some unknown unknowns in there too. That's what I'm, if anything, staying up at night worrying about.

So with respect to the Strait of Hormuz, do you think markets have fully priced in the risk at this point, or have they underestimated the risk of escalation?

And then as a follow-up to that, when this conflict is finally over, if we turn the Strait of Hormuz over to Iran, is that a major global risk? Oh, absolutely. I mean, if this war ends with Iran controlling the Strait of Hormuz and the war is over, that is probably worse than what would have happened absent any conflict at all. That's a problem that in some sense, you know, again, it's not like catastrophic recession, but it's a problem that would be with us for a long time.

Uh, it also would guarantee instability in the region. I don't-- That would not be the last word. It's not like, oh, it's now a stable situation. Iran's in charge of the Strait of Hormuz. Everyone returns to their normal business for the rest of time. That would just be a, you know, make what is going on geopolitically in that region a continued open issue. You know, the markets think that there's a certain amount of taco going on, a certain amount of if prices stay high for a while, supply will start to adjust in other margins, which it hasn't so far.

That on the demand side, people will figure more things out. So I think the market is probably right about that. They have a meaningful chance that the price is above $150 a barrel, and think it's possible above $200 a barrel, and we'll see what the president says. That makes sense. And so when we look at markets today,

is there a bubble or, or is there an everything bubble? I mean, where do you think we sit? If we're in a bubble, it's the tech sector. That's where all the growth in the stock market has come from. That is where you see valuations that are completely at odds with what the current earnings are. And now the path for the future, if you're integrating over all the possibilities, one of them probably includes infinity. And when you average infinity into anything, well, you get a pretty high number.

But you also have, you know, OpenAI, you know, making a loss on every sale. Anthropic seems to basically be as well, probably a smaller one. So yeah, their revenue's growing like crazy, but if that revenue is coming because you're pricing something below your marginal cost, 'cause the cost of inference is quite high and they can't come close to keeping up with it at current prices. I'm not sure what sort of model that is.

I'm more bullish on the productivity gains from AI than how they translate, you know, into profits. Because, you know, the other thing, you look at models like Llama, Mistral, DeepSeek, and the like, you know, they're a couple months behind the foundation models. You know, some people, you need to use the very, very, very latest. For an awful lot of uses, being three months behind a phenomenal model will still be, you know, very, very good.

And so just enormous constraints on pricing power in the sector. So

lot of reasons to be both simultaneously bullish about the technology, but very nervous about how it's gonna be turned into profit. So help us understand this AI build-out. We know that data centers need to be built, we know that energy is a constraint, and we know that these models have to be proven out to lead to profitability. So when we look at 1999 to 2001s and the dot com crisis, that came back roaring quickly with the internet growth.

But now, what about the 2006 to 2007 situation where all this leverage threatened the entire financial system? Does this look more like 2001 to you, or does it look more like 2006? Yeah, I think if a bubble bursts, it will be much more like the dot-com burst than the housing crash. And the most important reason is that people knew that the stocks they were buying in 1998, 1999, 2000 were risky. You know, they were sad to see them go down.

They were disappointed to see them go down. But when you buy a stock, you know it can go up, you know it can go down.

Housing crisis, people had what they thought was money. They thought it was interchangeable for treasuries, whether that was, you know, a CDO they had, whether that was a money market fund, you know, whether it was something they were putting up for collateral, whatever it was. All of a sudden you learned that stuff you thought was totally safe money was not, and when that happens, you can get a real rush out the door, a real jolt to the system, and that's where the government needed a massive intervention, and even that massive intervention wasn't enough to prevent a pretty bad thing from happening.

Now, I look at the interest rates that people are charging on their data center debt right... For the data center debt right now, they're pretty high interest rates. That, to me, says these people know this is not money. This debt isn't being bought by money market funds. This debt isn't being used as collateral with a tiny haircut for overnight borrowing. It just, it- no one mistakes it for money the way they did in the housing crisis.

So that brings me to my biggest concern, which is non-bank private credit. There's been a lot in the news lately about private credit, and there's more concerns and redemptions are up. And when I asked Larry Summers about non-bank debt two years ago, he goes, "Good question."

Okay, two years has gone by. How concerned are you about private credit? First of all, I think private credit is where there's an unknown unknown. We haven't had a sector of this size going through a downturn. Banks had gone through a lot of downturns. They sort of knew how they worked. This sector, it could be a new experience for. It is less regulated than the banks are.

So I am nervous about what we don't know about it.

The things I would say on the defense side, and this is largely what I think, is it's still,

um, a couple trillion dollars. You know, it's still a small fraction of the overall lending. A couple trillion dollars sounds like a lot, but it's a small fraction of the overall lending in the economy. Um, a lot of the duration is pretty well matched on the asset and liability side, and the liabilities can't run in the same way that bank deposits do. And they've done a decent job at taking the riskier stuff and putting them in funds that people know is risky.

Again, this point of whether it's money or not. So do I think a lot of people are gonna lose money in private credit? I would not be surprised if that is how all this shakes out. You know, would one of the less effective companies go out of business? Maybe. Will that be a Lehman moment? It's possible. I'd be terrified about it if I was a regulator, but I'd be very, very surprised. This could be something where it's like a hedge fund and, you know, it loses money, the investors lose money.

You know, too bad, so sad. The rest of the world marches on. But what do you think? And what did Larry say? Did I get the right answer? Larry's answer basically was, "Good question. I, I don't know."

But

look, you're close to the markets, and when we look at what's going on today and what Jamie Dimon said, where he said, "Where there's one cockroach, there's usually more,"

when you look at the news and whether or not it's Blue Owl or Ares or BlackRock talking about redemptions, I mean, redemption requests are up. People are capping redemptions.

What I'm trying to understand is, does this lead to a contagion? I mean, is the market underestimating the risk because the public companies are kind of hedging right now, and if BlackRock and Ares are having trouble, I mean, what about the little non-bank, the, the private credit investors who aren't as sophisticated? That- that's... Looking under the hood, that's what I'm trying to get at. The question is, I'm happy to agree with everything you just said.

The question is, when a run was starting at Silicon Valley Bank, the government had to hold its nose and pay the depositors. Did one of them have a, was it a billion dollars? I can't remember. I don't think the government wanted to give them their cash. I really don't think they wanted to. But they said, "If we don't give the money to the [laughs] Silicon Valley depositors, there's 50 other banks that will- everyone will pull their deposits out.

And once those go down, there's another f- 500 other banks, and then the whole system will be gone." The question here is, if Blue Owl can't meet its redemption, so they freeze the redemptions or they dramatically discount them, will the government feel, "You know what? We're worried this is gonna happen at three other funds, so we're gonna give money to the Blue Owl people"? I don't think the government is going to do that.

I don't think the government should or needs to do that. Could we end up with a run? There's some things that stop that run, you know, the non-runnable insurance money, for example. Yeah. Maybe we could have the first private credit run in history, but if we do, there's a decent chance the government can just stand by and some investors will suffer, but the, you know, the economy will, will be fine. So I agree with you that this is different.

And, and I don't know exactly how this plays out, but what I've been trying to understand and explain is that I, as a restructuring person, see a lot of debt that's out there in the world sitting on people's balance sheets.

And in the old days- There were hard assets that would back up that debt. So if the company struggled and the lender needed to be repaid, there were hard assets to sell.

And when I was a young lawyer, for instance, we were told lenders don't lend money to companies where the assets go home at 5:00. We've now got an economy where over a trillion dollars have been lent out by these non-bank institutions to companies that haven't had to have operating income.

They haven't had to prove that they could amortize this debt and pay it off over time.

And in many cases, because they were venture capital backed, they actually have equity that they raised or even other debt they took on that they are using to pay the interest on these non-bank loans. And, and lately we've seen this with the PIT, you know, the pay in kind provisions where the companies aren't able to pay the debt anymore, and they're just issuing more and more debt.

So what I'm trying to figure out is if we don't reload venture capital with enough money that it can continue to use equity to pay interest on debt, if we don't have the capacity to monetize another $5 trillion and just throw it out into the economy without creating a lot of inflation, as this debt works its way through the system and these companies are unable to pay it off, is there gonna be a reverse multiplier effect in terms of its effect on the overall economy?

And I'm not sure that the market has really priced this in. Yeah. And look, I mean, you could look at all the housing vehicles people invented before the financial crisis and say, "What are the chances that, you know, 100 houses in this pool of 1,000 are all gonna go bad at the same time?" Hmm. And there's a lot of correlated risks out there and, and software is the main one people have been focused on so far. But, you know, what gets disrupted in a positive way, what gets disrupted in a negative way by this technology, I, I just don't think people can confidently answer that, but there are certainly scenarios where it's a lot of things that are, that are highly correlated in ways that maybe no one would've counted on five years ago or even one year ago.

So that makes sense. And going back to AI for a second, it sounds like you're generally in the long term very bullish.

But if AI works better than people think in the short run, won't that be disruptive? And then what about if it takes longer than people expect, won't that be disruptive? I guess what I'm getting at is isn't there gonna be a bump to this economy either way? I mean, first of all, you're raising a question that I've been pondering for the last month or two, and I'd never, ever, ever thought about this before in my whole career, which is if I could pick the pace of productivity growth, what would I pick?

If you'd asked me that a year ago, 20 years ago, the higher, the better. Like, why are you even asking me? You know, million percent, great. Let's do it. Yeah. At this point, you know, the fastest we've ever had for a decade in American history is three and a half percent. You know, I'd take my chances with four, four and a half percent. If you told me we're gonna get 10% productivity growth a year, boy would I be worried that that's coming with some real economic disruptions to people's jobs, that they can't switch and learn new things fast enough.

Maybe dramatic changes in the ratio of income going to capital and labor. Possibly things that go beyond what we can quantify and get to the meaning and purpose of our lives and the things we believe in and, and make us human. And so, you know, I would love the Goldilocks. I'd love faster productivity growth than we've had for the last 50 years, but

I don't want it to be too, too fast. Okay. But there are real constraints now, right? I mean, if you look at these data centers, they require an enormous amount of power, and even with the new projects being green lit,

it doesn't appear that the grid's growing fast enough to meet the demand. So isn't energy likely to be a material bottleneck?

We don't know what the constraint is, whether it's compute, whether it's data, whether it's algorithms, but there's a bigger constraint even if none of those three were a problem, which is just our ability to adapt and use these. I'm reminded by McKinsey used to do these studies. They basically said, "If you do everything perfect in your house, you can cut your electricity bill in half. You know, insulate this, close that window, monitor this setting or whatever." And then an economist would come along and do a study of what people actually did in their houses, and they would get a fraction of the savings because they didn't put in the right insulation, they left the window open, they forgot to turn the thermostat down.

And we're just such imperfect users of all of this. I mean, I feel it myself. I'm spending a lot of time trying to learn the very latest Claude code, Codex, you know, switching back and forth, et cetera. I'm using a small fraction of the capabilities they already have that are directly work relevant for my own personal workflow. I almost like want them to stop innovating so I can just spend like three years just sort of figuring out what is already there.

But every time I start to get to something, it's like they add a new thing and it's like, oh wait, now I need to learn about that one too. And I'm just one person. You take an organization and it's actually much more complicated and much slower. So it's the adoption and deployment that to me, I think is the biggest constraint on all of this, much more than power, data or algorithms. But I agree with those too. I mean, those are on the list too.

So as you're learning your way through this, I mean, do you use AI in your research? I, I mean, has it affected how you do things? Well, yeah. I mean, I use it for, you know, background literature review. I sometimes bounce ideas Back and forth. Use it for data analysis, starting to use it to prepare some charts. Still need to do a lot of checking, all of it. I feel much better writing all my ideas and arguments out, and then I have three points.

Oh, wait, you know, maybe I should've had a fourth, or here's a weakness I can shore up in point two that I didn't think about. That to me is better than going to it and saying, you know, "I have something about to come out in The New York Times on manufacturing." I would not have gone to AI and been like, "I want to write something on manufacturing. What should I think?" Um, you know, I know what I think. I know what my main arguments are.

I've spent a lot of time on the data and the economics over the years. I want to be able to figure out how to share that with readers. Absolutely. So I understand that when it comes to the consumer's use of AI, we may not be seeing a meaningful impact in terms of profitability for the AI companies. But when we circle back to The Magnificent Seven and their use of AI and how it's affecting productivity, is it possible we're already starting to see a meaningful impact on our economy?

There's a, a few ways to look at this. First of all, there's the aggregate productivity data. Productivity growth over this business cycle has grown at, uh, two point one percent annual rate. That's about four-tenths faster

than it did in the previous cycle. We're about two percent above where the Congressional Budget Office, cumulative two percent for the whole cycle, above where the Congressional Budget Office thought we would be. So productivity growth is looking, you know, is looking pretty good. Not fantastic, not stellar. If this is all it is, no one will remember it, but, you know, but pretty good. Um, if you look at the micro studies of, oh, if you use it in a call center, if you use it in software, if you use it in a doctor's office, how much productivity you can get, you see many of them that are ten, fifteen, twenty, thirty percent.

You're not seeing anything like that in the aggregate numbers and, and the way to crosswalk them, of course, is that slow adoption and diffusion. Which is why I think even if there's no technological progress and we just freeze it where it is now, we also have a decade to figure out how to use it in our lives and in our businesses, we would still probably get, um, a non-trivial amount of extra productivity growth. So, you know, we're really just starting to see glimmers of it, um, in the data.

And by the way, same thing on j- the job displacement side. There's been a lot of, a lot of ink spilled on, you know, it's gonna take all the jobs. And maybe for some younger workers in software, but even there it's not totally obvious. But outside of that, so far a, a quite small impact on jobs

so far. Let's pivot to national debt. I mean, given how quickly the national debt is growing, and I think it's growing faster than the two point two growth rate you referenced, how concerned are you about the current interest rate environment, our level of debt financing, and kind of short-term maturities versus long-term maturities? When you look at this, how broadly do you see these fiscal deficits affecting the overall economy over the next few years?

Deficit, sort of crazy that it's six percent of GDP.

When has it been higher than that? World War II, the financial crisis, and COVID. That's it.

Every other time in our history it's been lower, and guess what? We're not in the middle of a world war. We're not in the middle of a pandemic. We're not in the middle of a financial crisis. How does that compare to other rich countries right now? It's the highest. No one else is as high as us. So we are

in a bad place. We're also the United States of America, and so our error tolerance is just much higher than other countries. Any other country that ran itself this way would be in a world of hurt. We're in a little bit of hurt because of this. I believe interest rates are higher. Mortgage rates are higher. It's a constraint on business investment and the like. Each year is a little bit worse than the previous year in that regard, but it's closer to termites in the woodwork that are eating away slowly rather than something dramatic.

And that, of course, is why we don't deal with it, because, you know, any given president, if they want to punt, you know, it's just not that bad for the next year or two or three. It just isn't. It's, it's, you know, we'll be fine if we don't deal with it right away. But one day we won't be fine, and we don't know when that day is. No economist, no markets person, no one can tells you when it is. And so, you know, to me, this is not the number one problem that keeps me up at night, but it is in some ways the most obvious problem that we have and is the most easily solvable problem we have economically.

Politically, of course, there's a whole, whole different blowfish. What's the number one problem that keeps you up at night? I think it is the s- tail risk associated with AI. And when I say tail risk, by the way, tail risk might be

understating the probability. I mean, it might be a ten, fifteen percent chance that it is so fast that it is really disruptive and we have a hard time adapting our economy and our lives to all of it. That is, you know, not where I think most of the probability mass is. Historically, you know, ninety-five percent of people that want to work at any point in time can work at any point in time, which is to say the unemployment rate historically is about five percent, you know, give or take a few percentage points, grand sweep of history.

Um, there's a lot of good reasons for that. Technologies create new types of jobs. They create demand for old types of jobs. They replace parts of jobs. They make people more attractive to hire by raising productivity. So I can give you all the reasons why this time will not be different, that we will continue to have everyone employed, and we'll all be happy and satisfied while we're employed. I think that is very, very-- I think that's very likely to happen.

But I can't guarantee you, um, that's gonna happen. Can't guarantee you that Dario Amodei is wrong. Well, I think on his timescale he's almost certainly wrong, [laughs] but on some sort of timescale. You know, I- so I am, I am worried, and I'm also worried 'cause I don't know what to do about that. I don't think we can regulate the technology and direct the technology. I think the technology's gonna be what it's gonna be.

Maybe we can come up with government programs to help connect people to jobs and enable them to work, but if the change is so massive and so fast, that'll get harder. I don't really love giving up on work and just shifting over to UBI, but if we're prosperous enough, one day we could do that. And,

um, so it's, it's the combination of that tail risk and the uncertainty about how to deal with it. In some ways, what you're describing is a kind of Star Trek future, you know, where AI and robotics provide food and housing and our basic needs, and so it does become a question of, you know, do we all become explorers? In the long term, I'm optimistic about human creativity, and technology does tend to create new industries that we can't yet imagine.

Are we gonna be colonizing Mars? Are we gonna be ex- scou- you know, expanding the galaxy?

Uh, that's great, but, but in the short run, I do worry about disruption. I mean, we have gone a long time without a recession, and we've got an enormous amount of money globally that's been printed, and to me, the world is facing real fiscal and geopolitical pressures. Britain lost its reserve currency status. And when I look at America today, I think we are pretty stretched, and I think there could be some real turbulence going forward.

Yeah. And look, it, this has always worked out for us in the past. It's not a guarantee that it always works out, but it's... Where you wanna start is with history, especially when you've been through something over and over again. But some of these changes have been pretty dislocating and pretty disruptive as people switch from agriculture to manufacturing, as people switch from manufacturing to services. Moreover, we often had a huge response to that.

As people shifted from agriculture to manufacturing, we made high school universal. As they switched from manufacturing to services, we dramatically increased the number of people going to college.

What, at that scale, are we doing today about the next transformation we're going through? I, I can't name it. When you talk about universal basic income or UBI, I mean, that is at least one possible response. If AI and robotics create real abundance where we can produce all of these things we need, then maybe human labor becomes less central to the world, and therefore, yeah, I can imagine some situation where we have direct income support becoming much more plausible.

To me, though,

you're gonna have to drag me there kicking and screaming. That's not to say I won't get there, but I'm not going to, based on a speculative theory about something that might happen 20 years from now, say let's redesign not just our economy but our society centering around a very different concept about how to organize both. But if we solve not just the AI problem but the physical instantiations of it in robotics, so you can have robot caretakers for, you know, aging parents, maybe even for children and the like, then all of a sudden maybe they can do everything and we can have UBI.

But I, I'm not sort of looking forward happily to that day, but I'm resigned that it, it might come at some point. This notion of shame where we're taking away people's sense of accomplishment and essentially undermine kind of the work ethic, is that what you're most concerned about? It's two issues. One is if we're in a world where we do need at least half the people to work,

but it's a hard time finding jobs for the other half, that's sort of the worst of all worlds because you give UBI to everyone, then why does anyone go to work? You know, let's say, like, garbage trucks can't be automated. How-- You know, the reason people work in garbage trucks is 'cause you pay them to work in the garbage truck. They don't do it for the, you know, the, the fun of it. And, you know, once you have UBI, all of a sudden how do you pay for all...

You know, how do you have a system where society is partly working, partly non-working? Now, we have that society now, but that's mostly people above a certain age not working and people under a certain age working, so I imagine maybe retirement ages just get earlier. That's one way to organize it. Your UBI kicks in at age 55. Yeah. That's one. There's the meaning and purpose questions. My guess is on some timescale that is solvable.

You know, people love their retirements. They're happier in retirement than they were in their working lives. Maybe we could make our entire life one long retirement. That might be possible one day. But, you know, if you were raised thinking of the world in a certain way and then all of a sudden you're not working, you know, developing those new hobbies, those new social connections. Then the last thing is work for a lot of people is their connection to other people, and the other ways we connect to other people have just gone down and down over time.

I mean, Robert Putnam, it was decades ago that he talked about bowling alone, and the problem has only gotten worse since then. And so, you know, other ways of having connective tissue, again, I'm sure we can develop them. I don't think we're talking about the next, you know, one to five years. But, you know, another thing on my worry list because unlike the debt or climate change for that matter, you know, it's not like I know- ...

exactly how to solve it. Yeah. A- and it's interesting because the printing press was incredibly disruptive, and we're already seeing a kind of baby AI where algorithms and social media are, you know, really contributing to polarization. So what's also striking is that polling shows that China is far more bullish on AI than Americans are. Yeah. No, I mean, the printing press, you know, it's a good thing. I'm glad we have it.

There was also the Thirty Years' War and an awful lot of other wars, some of the worst wars in millennia- Mm ... that probably happened because of the printing press and what it did for the rise of Protestantism, Protestantism versus Catholicism. I mean, I don't wanna dwell on all of this too, too much 'cause I'm a techno-optimist. Yeah. But I also think if you're looking at something where 85% of it is people are working, they're happy, wages are going up faster, we have amazing new products, maybe even the deficit solves itself, um, you don't need to think too much about that part of a distribution.

If it's the 10% where things go wrong,

we should maybe be spending 85% of our time thinking of the things that could go wrong and, and 15% thinking about how great the right things are gonna be. You know, when I first invested in companies in a small way as a law student, I remember buying, I think, Ralston Purina at, like, a price/earnings ratio of eight. And then I remember somebody saying to me once, "You know, there are a few companies like Costco or Ulta that can sustain a 30 times PE ratio, but most companies can't.

They're not gonna be sustainable. They're not gonna grow fast enough to really grow into their valuation." So when we look at Nvidia and other high-tech companies today, can they grow into these multiples? Are these valuations sustainable? Or is this more like the housing crisis where we convinced ourselves that houses could continue to go up and up and up and avoid a reckoning? I'm torn here because everything you just said resonated with me, but unlike you, my investing strategy is just incredibly simple.

I buy and hold index funds, and that's worked pretty well for the last 30 years. In fact, I buy and hold very international and diversified ones, which has worked out a lot better in the last year- Yeah ... than it did in, say, the previous decades as the dollar was strong and, and the US market was strong. So there's a part of me that just believes in the random walk theory, efficient market hypothesis. That's not to say that I don't think there's any investors that are better than everyone else.

It's just most people that think they're better than everyone else aren't, and I'm happy to put myself in the most people category, not the special category that really can do something. I mentally shorted Tesla. I think I mentally started a short position in Tesla maybe a decade ago.

It's a good thing I didn't do an actual short position in Tesla 'cause, you know, I wouldn't be doing very well right now. So when I think about this, yes, everything you say resonates, and so I'm happy to put a mental short on this, and if it goes down, I'll tell everyone that's what was in my head. But if it goes up, I think I'll be intellectually honest enough to tell them that, too. When I was doing work back in '98 through 2001, I, I really thought it was unsustainable and kind of saw 2001 coming.

And then again with 2007, you know, I realized things were starting to get overheated and I, I saw 2008 and 2009, you know, becoming a real possibility.

But, you know, I didn't know exactly how to time it, and so, you know, I started my restructuring firm. And then, you know, when 2019 came along, I pulled a lot of money out of the market because I really thought that things were overvalued. Well, you know, I, I don't look so smart now seven years later, and I've really tried to figure out, am I wrong, you know, or is it a matter of that these things take time? And one of the things that I think's contributing to the fact that it's gonna take a long time is that a lot of young people today that invested in Google and Microsoft and other companies, you know, are investing because they're working at good companies where they get their paychecks and they're reinvesting it through their 401Ks automatically back into the market.

And millions and millions of Americans are doing that now. So as confidence change, changes and the pendulum swings back, will there be a real day of reckoning? First of all, I'm not, like, a totally doctrinaire about the efficient markets hypothesis. I was just applying it to myself. I think for most people it's the right way to live their lives, but I'm not gonna tell you that Jane Street doesn't know how to make money in the market.

So but I, but I do think maybe they charge everyone a high enough fee that even if they invest in the people who know how to make money in the market, not-- the benefits don't necessarily translate out. That was not a comment on Jane Street in particular. That was just a, a narrow comment about hedge funds. You know, so that, that being sort of my first point. My second point is, yeah, there's, you know, there's a herd mentality, but you could've made that argument around, you know, Liberation Day a year ago, and for a week that looked like a good argument, but two weeks later it didn't.

You could've made that argument during COVID for a month or a couple weeks at least it looked like a good argument until it didn't. And, oh, December 1996, Alan Greenspan says irrational exuberance, and if you put your market, money in the market that day, you would've made money in nominal dollars at least no matter when you sold. Even if you sold at the very bottom of the market, it went up enough before it went down.

And so

I just have for myself just a lot of humility. Now, I don't think the world would be great if everyone goes around and says, "Hey, everything's perfectly priced all the time," because where does the information come from? So when I give an inflation forecast, for example, I spend a lot of time thinking about inflation- I don't just read off the breakevens. I tell you what I think inflation will be and maybe why I think the breakevens are wrong.

And,

you know, would I tell a friend to just listen to me and not look at the breakevens? No. I'd say, "Look at the breakevens, listen to me, put the, the weight on the two that you think is appropriate." But when it comes to the market, I don't know, I guess I put an even lo- lower weight on my view and an even higher weight on what the market pricing says. Do you worry about the Fed put? I mean, I understand your point that the Fed may not step in and buy private credit the way it bought toxic debt in the past.

But what happens when we do have a recession? I mean, do you still believe in normal recessions? And, and given our current debt levels and modest growth, unless AI drives growth to 5 or 10%, isn't there a point where the economy slows and the Fed faces pressure to print more money? Wanna be a little bit more careful to how we define the Fed put. One is if we go into a recession, they cut interest rates. It's not the Fed put.

That's their statutory mandate of maximum employment and stable prices. And of course I think recessions are always a possibility. And if you take a long enough time horizon for your forecast in absolute certainty, and the Fed will cut rates. We're gonna discover in the recession that the economy is less interest rate sensitive than we thought it was, maybe than we thought it was decades ago. Hopefully we already discovered this, that maybe the neutral rate isn't quite as high as we think it is.

And so being back at the zero lower bound would not shock me. Everyone who hates QE,

if they're in a recession, all of a sudden a lot of their sort of holy feelings diminish and a lot of your pragmatism rises to the fore. So I would not be surprised if a recession means zero rates and even QE,

even under Kevin Warsh and Scott Bessent. The Fed put on the market

is, I'm less sure about. I hope we don't have one. That's not part of the statutory mandate. The stock market's gone up, what, like 40% in the last, you know, two and a half years, something like that. If it went down 20%,

you know, I don't think that's something that the Fed should try to do something about or solve that problem. That's how you get bubbles. That's how you have even bigger both market and macro problems down the, down the line. So I hope there's no Fed put on the market. I'm not totally positive one way or the other. So does QE become more inflationary in today's environment? I mean, it worked before without creating runaway inflation, but that was during a period when China was growing rapidly and globalization was putting downward pressure on prices.

Now we seem to be moving in the opposite direction with more fragmentation and de-globalization. So would you worry about another round of quantitative easing, or do you think the economy can absorb it? Yeah. And the main reason why quantitative easing

wasn't followed by inflation in the financial crisis and was followed by inflation in COVID is that the unemployment rate was still really high in the post-financial crisis period, and it came down really quickly in the COVID period. And then the fiscal side was underpowered in the financial crisis and overpowered in the COVID period. You know, I'm not a big monetarist, but if you look at the money supply, um, it didn't go up very much in the financial crisis because when you do QE, banks have more reserves.

That doesn't count as the money supply, as long as it's in their vault. It's only when it gets out their vault into people's bank accounts that it does. That happened in COVID. Why? 'Cause we had QE plus fiscal together. That's what led to the money supply increase. So going forward, you know, if we did massive QE tomorrow at four-ish percent inflation, unemployment, I think we'd get inflation. If we did it because we entered a recession, probably not.

And, you know, it'll be interesting. I mean, uh, Kevin Warsh really, really doesn't wanna do it. He re- genuinely, he's not just pretending he doesn't like it. Scott Bessent doesn't like it, but, you know, there's no atheists in foxholes. If we're sitting here three years from now, what does the story look like from your perspective? Does something break along the way,

or does this economy continue to hold together? I mean, what do you think we're looking at, a much rosier outcome or more turbulence ahead? I mean, markets anticipate the future, so we've seen already enormous movements in markets, and you could see a large movement in the other direction, as we talked about. That wouldn't shock me. Uh, productivity just, it just takes longer. You don't get, like, huge bursts. It doesn't jump up 20% in a year.

You know, it's possible this time it'd be different, but I don't think so. And so I think three years from now, you know, instead of the 2.1% productivity growth we're talk- everybody's talking about, like, "Oh, it's been averaging 2.4. Isn't that amazing?" And, you know, but it'll be creeping up a little bit. So this watching it, you know, it's not you press a button and every single industry is transformed overnight.

It's sort of this business in this industry this day, another business in that industry, plus the beginning of a different industry on the next day. And so, you know, both the excitement and upside of that, but also the types of you wake up and, you know, IBM and a bunch of SaaS companies, like, way down one day 'cause all of a sudden you were worried Claude Code could put them out of business. I don't think we've seen the last of that either.

That's a great way to end, Jason. I really love the way you're coming at this from a market's perspective. I really do come at things mostly from a macro perspective. Ultimately, the two are linked up, and you really need to think and understand both parts of it, um, either as an investor or a policymaker because of, of the way they in- they interact and work together. People really haven't focused on this until the last few months, and so it feels like a very new and important issue.

I really enjoyed our discussion, and again, thank you so, so much.

The Puck: Venture Capital and Beyond is brought to you by CMBG Advisors. If you enjoyed the conversation today and haven't yet subscribed to our show, you can find us on YouTube, iTunes, Apple Podcasts, and SoundCloud. Feel free to leave a review while you're there, and maybe even a five-star rating. You can also follow us on TikTok, Instagram, LinkedIn, and YouTube for updates, highlights, and more content from our team.

Thanks for listening. We'll be back soon with a new episode. [outro music]