The AI Boom Could Create a Financial Crisis

Knowledge at Wharton

Ask about Knowledge at Wharton

Episode Description

ABOUT THE EPISODE The AI boom could transform the economy, but could the way it is being financed create the next major financial crisis?

Joao Gomes, Wharton Professor of Finance and Senior Vice Dean of Research, Centers, and Academic Initiatives, argues that the Federal Reserve needs to look beyond inflation and pay much closer attention to the financial risks building around AI investment.

Gomes explains the parallels with the housing crisis and internet boom, the growing role of private credit, and why the Fed needs to rethink how it monitors financial stability.

ABOUT THE PODCAST This Week in Business features interviews with Wharton faculty about the latest news, fascinating trends, and issues impacting both consumers and the business world. Episodes are recorded at the Wharton School and published twice per week.

#AI #ArtificialIntelligence #FederalReserve #Economy #FinancialMarkets #PrivateCredit #Investing #FinancialCrisis #KevinWarsh ----- The Wharton School of the University of Pennsylvania is committed to sharing its intellectual capital through Knowledge at Wharton, the school’s online business analysis journal.

Launched in 1999, Knowledge at Wharton offers free access to:

- Articles, podcasts, and videos highlighting Wharton faculty research and analysis of current business trends - Interviews with book authors from Wharton School Press - In-depth series of curated content like Wharton Executive Education’s Nano Tools for Leaders - A searchable database of more than 10,000 articles covering all aspects of business - A weekly newsletter that delivers Knowledge at Wharton insights directly to your inbox

Visit the Knowledge at Wharton homepage: https://whr.tn/3qUpSBE Subscribe to the Knowledge at Wharton YouTube channel: https://whr.tn/3DEAgAU

Read the full transcriptHide the full transcript
  1. There seems to be a bit of a collision ready to occur between the growing world of artificial intelligence and how many of those
  2. projects are being financed and how the Federal Reserve does its job and how it should
  3. think about the impact that AI is having on economies around the globe. Joao Gomes is a professor of finance and
  4. as well senior vice dean of research at the Wharton School. He recently wrote an article for Fortune about
  5. this topic and he joins us right now. Joao, great to catch up with you. How are you, sir?
  6. I'm very good. Thank you for having me. Always great to see you. Always great to talk with you.
  7. Now, you say that a better understanding of who is financing the AI boom is really
  8. important for the Federal Reserve from an economic perspective. Please explain that.
  9. Yes, I think that actually should be the question number one in the Fed's mind right
  10. now. It I think one of the reasons I wrote the article is we've had a lot of discussion about the way AI is doing
  11. and it bothers me a little bit the monetary policy in the last 12 months, 18 months, since even before Kevin Morse took over.
  12. It's been so focused on this one thing, which is what is inflation doing and it's
  13. sort of bypassing this entire transformation that is taking place in the US economy that has
  14. enormous long-term potential to really transform our lives for the better and solve a lot
  15. of, frankly, serious economic problems like the national debt, social security and things like that.
  16. It may or may not happen, but I think it has to be taken more seriously. What's happening there and in particular, and this
  17. is really what it ties to the Fed's mandate, which people often ignore to ensure financial
  18. stability. There's no question that these investments are massive. We hear about them.
  19. We have very little idea, very little understanding because so much of that is taking place
  20. in opaque forms through private credit that the Fed doesn't really see the data on, doesn't
  21. really understand the pricing. I think it is something that we need to take much more seriously and think much
  22. more about the consequences of what's happening and how it's happening when we think about what
  23. the Fed's role is. I think the inflation discussion has really got out of hand. It's an obsession about what's happening, what's 3
  24. .4, 3.2, whether PCE does this this month or next month. None of that really matters five years from
  25. now. I think it's time to start that discussion. I think it's time the people at the
  26. Fed really pay more attention to this and start thinking about ways of monitoring them, at
  27. least of making sure the markets are informed. I see it very much as the start of a conversation.
  28. Is it a greater challenge because that monitoring that needs to occur, you're talking about something
  29. that is still very much in development and we don't really know 100% what that final spot, that final path is going to
  30. be. So I think that's as big a challenge as any for the Fed moving forward. Yeah, it's two things.
  31. One is we don't know how transformative it's going to be in terms of productivity, of
  32. long-term standard of living, how much is our income going to go up, how much more productive we're going to be, how many
  33. fewer people are going to have jobs, not have jobs, how much they're going to get
  34. paid. That's one conversation for sure that matters. But a second conversation is to accomplish that
  35. we need to pour a lot of dollars into this and where are we going to find those dollars and what's the implication of
  36. that and how disruptive this is going to be. And my comparison is this sort of happened
  37. during the housing boom of we're building a lot of houses, but there was also piling
  38. up a lot of mortgages and a lot of things that people did not understand, did not have visibility on.
  39. And there are differences. I don't want to draw the parallels too far, but the Fed had no understanding of
  40. that and had to react ex post in crisis mode, trying to do all sorts of interventions for which there was questionable legal authority
  41. and sort of that delayed the process itself. And now we're back in the same position.
  42. The Fed, every three months, there's a sort of a financial stability report is produced.
  43. The Fed looks at it, it ignores it. It has no real relevance, no real content. There's no instructions about what should be done
  44. with it. My best example of that is there was one that was produced just a few weeks
  45. before the SVB crisis and it said everything was fine and then SVB blows up. And so the Fed has to sort of
  46. figure out what to do a little bit hand-waving, HADOC interventions. And I think that mode, that really needs
  47. to change. There's just not enough attention, not enough expertise, no good modeling.
  48. You know, you can tell me inflation is 2.5. Oh, the Fed should do X. Okay. We have that precision there.
  49. On the financial stability side, we have no sense of what we want to do. No sense of, I said, okay, it's not
  50. okay. No, no, no. Forget the precision of, let's just look at this number. Well, and you also noted the productivity boom
  51. back in the 1990s as something that maybe we should reference as well. Yes. That's the second point that I think is
  52. really important that people take back and look at how, and there was a much more
  53. sort of counter-current view of what's happening. I think people are too scared about the
  54. AI boom, economists in particular, and the folks around the monetary policy circles, the fiscal policy
  55. circles. I think they worry too much about the jobs picture and how this is going to
  56. disrupt jobs and sort of the damage it could do to the economy, the need to regulate, to oversight.
  57. And I think you forget that we don't have that many people in the pipeline. You know, our demographics look terrible going forward,
  58. and that's not going to change the next 20 years. I mean, people can start having kids today.
  59. They're still not going to get jobs in the next 20 years. We need help as human beings.
  60. We need a lot of help in lots of places. So I think this is really important for
  61. us to take seriously as a potential transformative for the better development. And just like the internet was, and I
  62. think disrupting that long-term path by being over-aggressive in how you hike rates and
  63. how you increase the cost of capital and how you discourage investment. We also need to think we're in a
  64. fierce competitive battle in an industry where it's pretty much winner takes all.
  65. It's not just open AI, Anthropic, Google, and so on. It's really U.S. versus China.
  66. China will not stop. And there are huge advantages to be the first, to be the platform of reference in
  67. a lot of these industries. So then what does the Fed need to consider? I won't even say do at this point.
  68. Yeah. Consider might be the better term in terms of looking at potential policy moves that needs
  69. to at least have ready to go. So two things. One is the models that the Fed use,
  70. and this comes down to the way the Fed structures his thinking is through models of
  71. how the economy behaves and how it responds to changes in interest rates, roughly speaking, different
  72. types of interest rates, but interest rates. So he has some models of the world.
  73. All the models that Fed uses are essentially agnostic or they take no view on the
  74. long-term impact of policy. They sort of assume that, and this is how they're trained to think that if I
  75. play with interest rates here and there, I may or may not create a recession. I may or may not reduce inflation, but
  76. I will have no impact on the long -term growth path of this economy, a long -term productivity and standards of living that needs
  77. to change. We need to entertain the thought that there are moments where really, yeah, you disrupt the
  78. capital accumulation process, you disrupt investment, you disrupt financing that could really measurably affect the long
  79. -term standard of living here. That's one topic. The second one is people talk about a
  80. dual mandate. There is no dual mandate. There's a triple mandate and financial stability is
  81. the most important part of that. And the Fed needs to rethink that and really think about how do we, that actually
  82. have enormous influence in financial markets, understand what actually is going on.
  83. We failed in the mortgage crisis on understanding the complexity of the financial instruments.
  84. We just have no visibility into that. That shouldn't happen again. We think we regulated that.
  85. We think we understand that, but it was always a game of sort of chasing, you know, chasing around where the risk was going
  86. to be. We sort of shut that down, the risk piled up in private credit and it will
  87. only increase. So we need to understand that and we need to focus on prioritizing the financial system.
  88. When somebody then, I'm sorry, when somebody then brings up the term AI bubble, is there
  89. a concern that should be there that the investment is put in and it doesn't reap
  90. the rewards that everybody believes it's going to? And then the Fed has to be ready
  91. for some sort of crisis. And on top of that, can we take from other crises to better understand if there
  92. is a crisis around AI in the future? Yeah. And I think that the 1990s and sort of the unravel of the internet boom is
  93. a really good example. There was, many people refer to it as a internet bubble in the end of 2000s.
  94. It didn't, it unraveled in 2001. There would have been no recession in 2001. There was a stock market crash, but there
  95. must be no recession. Were it not for 9-11 and all that sort of that catastrophic fallout from there,
  96. there would not have been a recession in 2001. I think we handled the cleanup of that
  97. really well because people sort of understood, they had a lot of visibility of what was
  98. going on. Everything took place through public markets, a lot of IPOs that, you know, sure, there were
  99. mistakes, people lost money, but it just didn't spill over to the rest of the economy.
  100. I think now we have no visibility. We have no sense of whether things are being priced correctly or not, whether the risks
  101. are being allocated the right way. And that smells a lot like 2006, 2005, 2007, that sort of window where, now it
  102. may be we're fine, but we just don't know, right? We just really don't know. And so I think we have, we definitely
  103. have a lot of lessons to learn, but we have to be more, and listen, I'm super optimistic about AI in the long-term.
  104. I think I would tolerate, I certainly would trade off a short-term recession for, you
  105. know, incredible growth in our income standards of living. I will take that risk, but not a
  106. financial catastrophe like we did in 2008. But that being said, this is still something
  107. that has to be on the plate of Kevin Warsh and the other leaders of the Fed right now.
  108. Exactly. And it is not. You'll go through, the entire discussion continues to be about what is inflation going to do
  109. in the next six months. It's just, it's really looking where the problems are not and where the opportunities are not,
  110. in my opinion. It's just so off-center that it is frustrating, I think. João, great to talk with you and get
  111. your insight. Thanks very much. All the best. Thank you, Dan. Thank you for having me.
  112. You got it. João Gomes, who is a professor of finance and a vice dean of research at the
  113. Wharton School.

Brief Summary

João Gomes discusses AI financing and its implications for the Federal Reserve's economic policies.

Key points

  • AI and the Federal Reserve. Joao Gomes argues that understanding AI financing is crucial for the Fed's economic strategy.
  • The AI Bubble Concern. Joao warns about the potential risks of AI investments not yielding expected rewards.

Tension

60/ 100Some tension

Related episodes