As the adoption of generative AI in business continues to grow, the banking sector is the latest to embrace the technology, aiming to boost productivity and streamline operations. Wells Fargo Managing Director and Head of US Large-Cap Bank Research Mike Mayo joins Yahoo Finance Live to discuss his top AI picks in the banking industry, emphasizing that "if a bank does not have a good AI strategy, I would say they don't have a strategy."

Mayo names JPMorgan (JPM) as his "best-in-class" pick, stating that the bank has the potential to become "the Nvidia of banking." He highlights that "more than any other bank" JPMorgan possesses the necessary resources, spending power, data, processes, and personnel to successfully implement and leverage AI across its business. Mayo notes JPMorgan's massive amount of data "is the fuel for AI," which the bank can utilize to gain valuable insight into its operations. Like TikTok's recommended content algorithm, JPMorgan can take advantage of "predictive behavior" models for its consumers.

On the other hand, Mayo identifies Citigroup (C) as his "worst-in-class" pick, stating that "they've failed the last twelve restructurings." He notes that the bank has fallen short on returns and efficiency. However, Mayo says that their consolidation into "five lines of business" could make it easier to tackle problems in each segment. "If they simply meet their targets, they will probably have some of the best year-over-year performance of any large bank."

For more expert insight and the latest market action, click here to watch this full episode of Yahoo Finance Live.

Editor's note: This article was written by Angel Smith

Video Transcript

JULIE HYMAN: Well, AI has been tipped to be a game changer for most industries. But banking may not be the first that comes to mind. Though financial and bank services comprised a quarter of AI spending in 2023 amounting to around $154 billion. That's according to Wells Fargo. Who says we're just about to see that investment start to pay off.

Mike Mayo is Managing Director and head of US large cap Bank Research at Wells Fargo. He's the one who gave us that number. And he's here with us in studio. Mike, it's great to see you.

MIKE MAYO: Nice to be in this new modern studio.

JULIE HYMAN: Thank you so much. I appreciate it. So you put out this large piece of research on this.

And I have to admit, I'm one of those who I did not think of banking first off when you think about AI spend. But I'm curious concretely, what kind of changes are customers going to see? And what kind of efficiencies are these banks going to get when it comes to AI?

MIKE MAYO: Well, shameless plug for my hobby of weightlifting. I have a barbell strategy. We have the best in class.

JPMorgan as my number two pick. And the worst in class Citigroup has actually my number one pick. Yesterday, we hosted a New York City conference. And we had the head of JP Morgan's head of AI basically.

And I think JP Morgan could wind up becoming the NVIDIA of banking because they are a Goliath. Goliath is winning. They have more data than anybody else.

They've been at this for a decade. They've extracted AI from the rest of technology. They're going through all sorts of use cases.

Only one idea out of 3 makes it from lab to production. And they should see-- they targeted $1.5 billion of benefits last year, which would be a doubling from every year since 2020.

And so they have the resources, the spending, the data, the processes, and the people in place. And I think you guys reported that Jamie Dimon said this will have unbelievable potential. Maybe they'll split it off as a separate business line one day. But JP Morgan has optionality.

And they start from a position of strength more than any other bank. So that's your best in class example. My worst in class example, the other side of the barbell is Citigroup. They've had worst in class returns, efficiency.

They've fallen short of all sorts of targets. And for them, it's more basic. At my conference yesterday, they invoked Conway's law which says that your technology architecture is a function of the architecture of your firm as a whole.

So Citi has sold off businesses. Now, they've simplified the firm. They'll be done with their org simplification two weeks from this Friday.

And then that allows for more simple technology to which they can add on AI. So different examples of how two firms are approaching it. But for the industry as a whole, it should be able to take banking efficiency toward record levels.

It's the most obvious use cases are coding. This ancient COBOL code, even I learned back when I was a computer science major. You can transition from COBOL code to C++ or Python in a very precise manner. So AI is very good at doing that. And then all of us are going to have AI co-pilots to make us more productive.

JULIE HYMAN: OK, so just a quick follow for you. And it's a selfish follow. I have to admit.

I'm a Citi customer, a long time Citi customer. And I had occasion to call the bank yesterday with a question. And I got an automated system.

And man, did I go down a rabbit hole. So much so that I had to hang up and start over again. It just seems like there's so much low-hanging fruit when it comes to things like customer service. Are they taking advantage of that as they should be once they do that simplification process?

MIKE MAYO: The good thing about Citi is they've-- well, the bad thing is they failed the last 12 restructurings. This is restructuring number 13. I think--

JULIE HYMAN: Lucky number 13?

MIKE MAYO: For Citi, everything is upside down. OK. But this one is different for two reasons. Number one, they're selling off businesses equal to 10% to 15% of revenues.

So that simplifies them. And they're going from this matrix multinational mishmash structure to five lines of business. Services, banking, markets, consumer, and wealth. And those five heads report to the CEO Jane Fraser.

So now, when you ask a question like how's customer service, well, how is it in each business line? So services is the number one player in global wholesale payments. So that looks good.

Credit cards, they're a top player. And you see with the Discover acquisition. That's a value.

Banking and markets is a player. And wealth is a player. But They have a lot of work to do in consumer US banking.

That is a whole. The wealth business, that's kind of a whole. That's 20% of the company.

So is the glass 80% full or 20% empty from a strategic standpoint? And you don't have to wait forever, I think. This year, they've given the best guidance for revenues to expenses than any other large bank if they simply meet their targets.

I know that's a big ask for Citigroup. They simply meet their targets. They will probably have some of the best year-over-year performance of any large bank.

So sometimes it feels like in recommending Citi as my number one pick, I have hundreds of pounds on my back, you know. And I--

JULIE HYMAN: Bet you've had some recent experience with that. So it's OK.

MIKE MAYO: So as long as we can lift that weight. And I think they are going to lift that weight. And I think Jane Fraser come two or three years from now. She has a risk of being fired if she doesn't improve the returns. But she also has a chance to become banker of the year.

JOSH LIPTON: How long, though, do you think, Mike, that restructuring takes just to play out? What's your timeline for it?

MIKE MAYO: I think what's underappreciated it's we're talking two weeks from this Friday. And you say, well, no banks ever done this. Procter & Gamble did this. We have an analyst--

JOSH LIPTON: Is that the model you would use for this one?

MIKE MAYO: Absolutely. It's like finding a needle in a haystack. I was wondering our research floor at Wells Fargo. Chris Carey covers consumer staples and Procter & Gamble.

And I described to him what city is doing. Because, well, Procter Gamble sold businesses equal to 10% to 15% of revenues. Procter & Gamble went from their matrix.

They called it a thicket to six lines of business. The similarities are there. So anybody who's ever owned Procter & Gamble for the last five years, take a look at Citigroup because there's a lot of similarities.

JULIE HYMAN: I want to circle back to JP Morgan for a minute and the idea that it could be the NVIDIA of banking in terms of the AI opportunity. You said they have more data than anybody. How do they use that data to optimize profits, to optimize the business? How is that going to play out?

MIKE MAYO: Well, they have 500 petabytes of data.

JULIE HYMAN: I don't even know what that means.

MIKE MAYO: It means a lot of data. And data is the fuel for AI. And you do have to go ahead and use that data.

But JP Morgan is in a position. When they deal with the business, they can use that data potentially to have more insight about the business than the business itself has. They can use that for predictive behavior.

So you might be in the market to buy a car. They should be offering you low price loans, or whatever.

JULIE HYMAN: They should know I want to buy the car before I know I want to buy the car.

MIKE MAYO: Yeah, look, it's cliche, but it's true become the Amazon of banking. OK. Or TikTok, what's going to connect you to the next one, what's that recommender. I had a speaker from NVIDIA at our conference yesterday.

And the idea is to have a recommender that the secret sauce. What's going to anticipate what you need? Not just based on your prior history, but based on everything that should be known about you on social media and all the clicks that you have. And the audio, and the video, and you put it all in there.

And then you go ahead. In our language, not in Pascal, or COBOL, or assembler, or C++, in English, you ask a question and get an answer. So this will be a game-changer for the industry over the next 5 to 10 years.

I'm not changing my earnings models yet the next year. But I think over the next three years, it's in sight. And if a bank does not have a good AI strategy, I would say they don't have a strategy.