0:20

Hello and welcome, everybody. This is Joe from StartupRate.io coming to you

0:24

today with another subject matter expert.

0:27

When Paolo says his first words, most people will recognize him. Ah, Paolo again.

0:34

You most likely don't need an introduction, but I still do it because we're growing like crazy.

0:42

I welcome you back to my podcast. You are the most frequent guest on StartupRate.io.

0:48

You are a thought leader in the world of fintech and banking.

0:52

You're a trusted expert in the intersection of technology, banking, and capital markets.

0:56

With years of experience at the forefront of banking innovation,

1:00

Paolo's insight are shaping the future of financial services.

1:03

Today, we'll be discussing his latest research at the IBM 2025 Banking and Capital

1:08

Markets Outlook and how it will influence the next generation of financial strategies.

1:14

Paolo, welcome again to my show.

1:16

So with such an introduction, I have to start with the right foot.

1:22

Buongiorno, buon pomeriggio, buonasera to everyone,

1:25

Wherever you are in the world.

1:27

As we are at the start of a new era in banking and capital markets,

1:31

one question dominates. How will emerging technologies like AI, blockchain and digital currencies shape

1:38

the financial world in 2025 and beyond?

1:40

Today, we are diving deep into the IBM 2025 Banking Capital Markets Outlook

1:46

with one of leading voices in fintech and banking transformation.

1:50

Stay tuned to discover what the future holds for financial institutions,

1:54

investors and entrepreneurs alike. And we also may tell because most people are listening to this,

2:01

your name tag also say the bank is bookshelf. So you're also a podcaster there.

2:06

Yes, I am. I'm following your inspiring episodes.

2:13

So the Bankers

2:13

Bookshelf is the opportunity to have on the show the most interesting book authors

2:19

and researchers that are helping us to understand the present of banking and its fintech future.

2:25

So happy for you to also attend our episodes.

2:29

But we can share some key insights here with you, Jorn. So this is a wonderful podcast.

2:34

I'm always happy to get back. You know, this is amazing. So let's make it happen.

2:38

Happy to come over for a recording to you. Paolo, in your recent research with

2:43

IBM, you highlighted the growing importance of AI in banking.

2:47

How do you envision artificial intelligence reshaping the landscape of banking

2:52

and capital markets in 2025?

2:54

Well, there's nothing new,

2:56

Right? So just becoming important.

2:59

Let's put it into perspective.

3:02

So maybe this helps to understand where the industry is effectively leaning to.

3:09

If we go back 15 years ago,

3:11

Say at the start of the global financial crisis, which is always an important

3:16

moment in time in the history of banking, back then banks realized that they could not operate the way they were operating

3:23

before because their capability to generate revenues was much more reduced,

3:27

I guess, before there were a few excesses.

3:31

And so there

3:32

Was also the

3:32

Time when digitalization started. The iPhone was invented in 2007,

3:37

more or less. So we saw a growing trend of people be more inclined to use the

3:42

website or even start using a phone for doing some basic banking transactions.

3:47

So for the banks was, you know, Brandon to say, hey, let's close branches.

3:50

We rationalize because we can push some of those activities on a website and then on an app.

3:55

But in essence, that didn't change. The banking was working.

4:00

was primarily a lift and shift of a branch on a digital architecture,

4:06

Which in many cases meant more investment in cloud computing and APIs to power

4:12

up banking in a different way, not in the premises of a branch,

4:16

but mediated by a technology piece.

4:19

So that I would call digitalization of right. Of course, in the very beginning,

4:23

Banks were like, can I use cloud or not? Regulators learned that maybe that's

4:26

important. So they started doing it. Banks realized that just having an app on a phone or a nice website doesn't

4:35

mean that clients are comfortable in navigating through.

4:38

So they started thinking we need to personalize that.

4:41

So we move from an age of digitalization of write into an age of personalization.

4:48

And how do you personalize use AI?

4:49

So they started

4:50

Using machine learning in particular to, in a sense, manage the SEOs,

4:56

to create the content more relevant for the client.

4:59

to streamline the processes, make onboarding a bit easier, let's clumsy.

5:04

So that's a period

5:05

Where AI starts entering the banking scene in a variety of fields,

5:11

but in particular there when you have to do banking activity on mobile,

5:15

when you do a bit more automation, you need to streamline the processes. But still,

5:20

That was not

5:20

Enough to shift a lot of important operations from a face-to-face branch onto the digital app.

5:29

Now, if generative AI is

5:32

Famous for something, it's about communication, language. So I think that

5:35

Now we're entering the stage where we had years where we had digitalization

5:40

to core as a key theme, years when we started talking about personalization,

5:44

And we could use AI, machine learning in that respect.

5:46

But now it's about communication in many cases.

5:49

So how can you transform the way you effectively talk to clients,

5:53

As well, I would say, the way you talk inside the organization?

5:56

And that is where AI has now more capabilities to transform the way banks operate

6:02

because up to the past was

6:05

Sort of a linear

6:06

Approach in terms of plugging in an algorithm here and there where you could grab some value.

6:10

Now you can rethink the way effectively everything works inside the bank and

6:15

in front of the customers.

6:17

Actually, as you said, that it kind of reminded me we first talked about cloud. Yes or no.

6:23

Crypto. Yes or no. Now we're talking AI.

6:26

Yes or no. And it was always a slow approval by the regulators.

6:31

And then the banks went in more or less over time.

6:35

And there's quite a misunderstanding there that come to the surface with time.

6:42

We could have seen that at the very beginning, the strong of push of narratives

6:45

around. but we see that. And if you think about cloud,

6:50

Cloud is not what many expected to be.

6:53

The economic model of cloud that doesn't hold through in many cases.

6:58

Now private cloud can do as much as public cloud. And so I don't know what the

7:02

upper scalars are thinking. Actually, I see the upper scalars trying to move on, if you like,

7:06

compared to the propositions before.

7:08

Lots of institutions realized that with growing volumes, you know,

7:12

the thing was not operating as expected. So it's an hybrid cloud mix.

7:17

The crypto world as well,

7:18

Of course, there was a lot of fantasy and excitement. And there is still for the pure crypto world.

7:24

But now we're leaning towards a different aspect, which is around the regulated

7:28

space with the CBDCs and the stable coins.

7:31

So it's more

7:33

Like going there where it is traditional, though with technology instead of

7:38

taking a direction where it is completely revolutionary,

7:42

Right?

7:42

But out of the mainstream. Some of these elements have been,

7:45

if you like, digested inside the system as well.

7:49

That's true. But it's very different from the early stages, proclamations of,

7:53

if you like, total decentralization and transformation.

7:58

And with AI, maybe it's the same,

8:00

Right? So we're going through phases to understand exactly what it does,

8:04

what it can do, what it can't do. It's so easy to get trapped into a hype that doesn't help anyone but a few.

8:13

AI, however, is different from the previous two because

8:17

AI started in the 1950s, so it's a long track record in terms of usage and understanding.

8:24

Even before Generative AI, AI was already part of banking.

8:28

It was just talked maybe with less emphasis from the general public compared to what it is today.

8:36

And what is happening is now

8:38

The technology is becoming more

8:39

Accessible and more flexible, which typically is

8:43

Important when technology starts unfolding real value for transformation because

8:49

it cannot be just rigid, so it needs to be adaptable in a sense.

8:53

Now, data adaptability is not risk-free. It requires a professionality.

8:57

It requires a good risk management approach. It requires platforms and architectures to execute it.

9:03

And that's where

9:04

We will see the differentiating elements. It's not about who has the best AI,

9:09

But it's about who has the best way of using AI, which is a different type of investment in IT.

9:17

AI is becoming more open source. It's multimodal. It's multimodal.

9:22

So that's the story. But for a long while, we thought about the best AI is the best institution.

9:29

It's the best way to use it. It's not the best.

9:31

I actually totally agree with you.

9:34

when you talked about like the new technologies going into the old ways.

9:38

What came to mind is the tokenization of securities that is currently ongoing.

9:43

And I also thought about application of AI.

9:47

Basically, it's a nightmare for a bank because you have to keep track on all

9:51

the AI agents you have there. They're working there because not one agent from one department,

9:56

an agent from another department do something like crazy, You're shifting money

10:00

back and forth in the billions on seconds basis.

10:04

And then you have a big vest. Nobody wants that.

10:07

Before we get too far for our audience, for our listeners, how are you or your

10:14

company preparing for a ice roll in the future?

10:17

Feel free to share your thoughts on LinkedIn or Twitter or Let's Connect.

10:21

No, I think

10:22

That you hit the bottom of a propeller into the future.

10:26

You talked a lot about agents. as agent TKI is coming as well,

10:31

But not yet there,

10:32

You know, to have that complex system that operates through agents.

10:36

But that's another interesting evolution of technology, which is getting more

10:42

and more prone to make decisions which are unstructured, that are not part of

10:47

a defined rule set, and take responsibility for that,

10:50

which could only be dreamt about years ago.

10:53

So now it's becoming effectively a reality where people want to engage,

10:57

you know, down the road to that type of configuration is another story,

11:03

but effectively, this is happening as well.

11:07

When I was thinking about how. a bank could use ai of course what came to

11:13

mind is like a lot of menial tasks done in

11:16

the background um but then a lot of

11:18

banks would need new systems where agents could

11:22

access plus you need to kind of a controller of all the ai agents make sure

11:27

they all work properly you need to document them you need to monitor them so

11:32

i'm not too sure if like there's a huge win in efficiency there.

11:38

What are your thoughts on that?

11:40

Okay.

11:41

It's another key point here. The way you measure it, but also what you want

11:47

to do with that, because if you don't unleash new value, you know,

11:51

it's a zero-sum game, in essence, right?

11:53

Or you reach a floor, so you cannot expand further.

11:56

So the point is, how can you do it not just to simplify what you're doing,

12:00

but then to create something new? Because that's where you justify most of the investments.

12:05

If you look at the cost structure of banks, so the total operating expenses,

12:11

I give you a number, which is about banks worldwide, and there could be differences

12:14

bank by bank or across the regions. From 2007 until 2024,

12:21

The cost, the percentage of cost that banks had

12:24

To spend for people, which is a good cost, salaries and benefits, went from 50% to 54.6%.

12:32

The percentage of expenditures in technology that can come went from 6% down to 6.6%.

12:40

So the amount of costs allocated to the workforce increased over time,

12:46

notwithstanding all of the investment technology.

12:48

There are a few reasons for that,

12:49

But that tells you that essentially banking is a people business anyway.

12:53

And if you want to

12:54

Get benefits out of automation with AI, you always need to ask yourself,

13:00

How do I also augment the employees to do something more, something different?

13:03

Because if you don't

13:04

Put both elements into the right perspective, you may not justify the economic

13:08

model anyway. For the good and for the bad, right?

13:11

So that is, I believe,

13:12

The most complex things that bankers have to do because they not only have to

13:16

innovate what they're already doing in automatizing it because they have to

13:20

change it in automatizing it that way.

13:21

But they have to imagine how to

13:22

Do things differently, okay, or new things. And, you know, you really have to think out of the box.

13:29

But, hey, we have hundreds of banks worldwide,

13:31

You know, thousands of people working on that. I'm sure that here and there

13:35

we see good ideas popping up.

13:38

We've spoken quite frequently about neobanks and fintech startups disrupting

13:44

traditional banking models. What trends in 2025 do you see that will allow these fintech players to continue

13:51

to challenge incumbent banks? And where do you think established players need to innovate?

13:56

That's me a lot.

13:57

Because end of last year, in September, October, I published a new research

14:04

dedicated to banking for small and medium enterprises, commercial banking,

14:08

putting on the side the sole entrepreneurs, which are more like retail,

14:12

so small and medium enterprises. Say companies with more than 50 employees and below 200 million of revenues,

14:19

maybe 100 million in places like Brazil. Why does this excite me a lot? because this is where the competition between

14:28

bank and new bank is intensifying in the next five years.

14:32

And there's a reason for that.

14:33

The last 17 years were about retail markets, which are overly commoditized.

14:39

And there, to move the needle with technology, you really need to get into larger

14:44

volume. That's why Revlon needs to have more and more clients.

14:46

New bank got 100 million customers in Brazil, and it's still growing.

14:51

So you need to get to a very large scale.

14:54

to grab all the benefit in retail

14:56

markets with this type of transformation technology. So it's not for all.

15:00

Corporate banking is not for every player. It's really a relationship on a few.

15:05

And capital markets, again, is for a few banks that have the shoulders to basically

15:11

do trading and become more prominent in that highly volatile space.

15:17

But small and medium enterprises are everywhere in every country,

15:20

in Germany, in Italy, in Brazil, in Portugal, in China, in Australia,

15:25

And they were not served that well by banks,

15:29

Or banks could not invest in transforming their services. So the fintech

15:34

In the past years for a few reasons, because all of these clients,

15:39

The SMEs are very different. You have a restaurant, you have a dentist,

15:43

you have a lawyer, you have a travel agent.

15:45

So for the bank or

15:46

The fintech to talk to these communities is complicated, because you have pockets

15:51

of communities at different ecosystems which have different problems that pop

15:56

up at different times in the economic cycle,

15:58

Which have different ways of reporting their

16:00

Numbers and their data. It's not structured. So the cost to serve in

16:04

Banking for the

16:05

SMEs is the highest across the group of clients.

16:10

Now, given the latest two,

16:12

Three years advances in the AI capabilities, I think that now we're at the point where

16:19

That problem can

16:21

Be more easily tackled because with the combination of traditional regenerative

16:26

AI, now you can basically have new means to systematize the process of onboarding

16:30

and understanding these ecosystems and therefore lowering the

16:34

Cost to serve, which can be

16:36

Passed to the clients in terms of a lower lending rate, making the bank more

16:40

competitive or the fintech more competitive without consuming the prudentiality,

16:46

okay, because giving away for free has never been a good idea and you see clearly,

16:51

you know, the marcuses of the cases and so on and so forth.

16:54

So now, given

16:55

The fact that AI made a leap in terms of capabilities, especially with the larger language models, and

17:02

The fact that

17:03

That segment of the market is where the cost of service is the highest in banking,

17:07

Any dollar invested

17:09

In the transformation of the business can generate the biggest benefit.

17:14

So they are the highest return investment.

17:16

But two, three years ago,

17:17

I guess, was too early.

17:18

So it was not yet possible.

17:21

So the economic model was not working.

17:23

And then last but not least,

17:25

These companies are like the, you know, 3TP in the gold mine that are the first

17:32

to suffer when There's an economic downturn, so of course, they tend to be riskier.

17:36

And now is

17:37

Where the opening of a lot of ecosystems through the API economy allows to ingest

17:43

new data in the risk management model, even just with a traditional machine learning model,

17:49

That might allow to

17:50

Better risk manage the positions with these ecosystems. There are varied,

17:56

and they need to engage a bit more on digital compared to what is happening so far.

18:00

So even in that front,

18:02

I think because of the API economy spreading around banking more than in the

18:06

past five years of open banking and open finance, now things are possible that

18:11

were not possible before.

18:12

Mm-hmm.

18:13

When you've been talking about the capital markets, what I had in mind,

18:17

what used to be called, for example, flow monsters and some research,

18:21

big banks with big organizations, where it doesn't make a difference if you

18:25

have two or three additional billions flowing through a day,

18:28

but they generate additional profits for their talk about economies of scale.

18:32

That kind of leads me to the capital markets transformation.

18:38

The capital market sector has faced various challenges over the last decade,

18:42

according to IBM's outlook. How are emerging technologies expected to impact investment strategies and asset management by 2025?

18:53

So when we talk about asset management,

18:56

We need to tell us also a bit of the truth about capital markets.

19:02

So, this is an

19:04

Industry where few people pretend to know too much about the future and know

19:09

nothing, because fundamental uncertainty is a structural market.

19:13

And a lot of people, all of us, are afraid not to know anything about the future,

19:17

but we know enough to be prudent in essence, right?

19:20

It's just that

19:21

In the necessity of positioning all financial markets, capital markets,

19:26

because in the end of the pension funds are for the attorneys,

19:30

There is a way narratives are created and opinions are generated to talk to

19:33

people. Now, the question becomes,

19:36

Do you have now

19:36

A better capability to predict the future because of technology?

19:40

That's not the point. I think you're not getting into more predictions.

19:43

You can have more capability to compete, algo to algo, which is different than predictive,

19:50

But you still have the

19:51

Same old problems to basically position yourself in terms of understanding market.

19:56

What is changing, however,

19:58

To me is the following as every activity that happens in the upper space of

20:03

capital markets needs to boil down into a conversation that has to be had with

20:07

a client that needs to make a decision about their investing is the capability

20:11

to better inform people about how they can make their decisions in a world that is very uncertain.

20:17

by testing the hypothesis, right, and creating more easy conversations right

20:21

to grasp, that can transform the way we are building the various bets in capital markets.

20:27

So that is where I see the biggest transformation happening.

20:30

For the rest, of course, modeling is always en vogue.

20:36

What matters there is that the model of the bank that has the biggest shoulders,

20:40

it becomes the dominant model because it's paid for pricing,

20:43

and everybody else is a price taker. But I think there will be four banks and business as usual.

20:49

We're not going to crack, if we like, the market.

20:53

We have just one case of somebody that apparently has solved the market.

20:58

We don't know yet how we did that, whether we did that. Renaissance technologies

21:01

will be interesting to learn more, but we don't know much about that.

21:05

But for the rest, I think that the structure itself doesn't change,

21:08

but what can change is the way people can access and advisors can talk to people for a lower cost.

21:16

They might transform the way they ask, basically to tap into opportunities in

21:20

capital markets, therefore changing, if you like, the average tower of the trading floors.

21:27

We talked about the person who sold the market. I'm very sure there are a handful

21:32

of projects out there in the world where somebody is trying to train an AI to

21:36

think like Warren Buffett.

21:38

Well, Warren Buffett did

21:39

Not solve the market in a sense. There was more a relationship,

21:43

but Warren Buffett as well was very risky.

21:45

Now, successful, you see,

21:48

It's a very different process. you can make a lot of money because you're very

21:56

concentrated in your bets, or you can lose a lot of money because you're very

22:00

concentrated in your bets. And then overall, in a long time, if you buy the market, more or less, you are there.

22:06

So Warren Buffett is an example where he was very specific in his bet,

22:10

which could be very risky, but it's a very complex process.

22:15

It doesn't conform with high frequency trading or automated trading in that respect.

22:25

And again, if you look,

22:27

If you try to imagine what's beyond the curve,

22:29

Not knowing what will be on there, you try to imagine something which is not in the data.

22:34

The data may inspire you,

22:36

But you get your intuition to go beyond that, and that's very hard to be replicated,

22:41

whether it can be replicated.

22:44

For our audience, what changes do you think AI and blockchain will bring to the investment space?

22:50

Drop your comments below the episode to let us know what you think.

22:56

Paolo, we'll be back after a short ad break.

23:04

Paolo, guys, everybody, welcome for coming back. We're still talking to Paolo

23:09

Signori, thought leader, podcaster, and awesome person here on StartupRate.io,

23:15

plus the most frequent guest. Now we get into, unfortunately, a little bit the boring but necessary part of

23:22

this, the regulatory and risk management in AI.

23:27

What AI and automation transformation?

23:30

How do you foresee regulatory frameworks evolving to ensure that AI-powered

23:37

systems in banks remain secure, ethical, and compliant?

23:43

I never thought that risk management is boring.

23:46

Maybe a regulation could be.

23:50

I completely forgot you are a former risk manager. I'm so sorry.

23:55

A lot of work to do and stuff. It can also be fun.

23:59

And you know why it's not boring? Because being based on data and data is always corrupted.

24:04

So, you know, you can get entertained every day with something new,

24:09

new problems popping up that you have to resolve.

24:12

But it's a very

24:13

Important question because this question contains the solution to most of the problems.

24:21

Let's start saying this.

24:23

In the 1990s, CEOs of banks used to say that every one of their employees was

24:29

a risk manager because risk management is the core business of a bank,

24:34

managing the bank's risk or the risk of the clients.

24:37

Now I'm saying every risk manager, every banker must be an AI risk manager.

24:42

That means he needs to understand how to manage the risk of AI,

24:46

also the risk with AI, but how to manage these algorithms or this new way of working.

24:50

And why is that important? I was actually talking to the global CIO of one of

24:56

the top 10 banks in the world a few days ago.

24:59

And we asked him the question,

25:02

What is missing to scale AI enterprise-wide in financial institutions?

25:06

and basically said, well, it's the reinterpretation of the role of the craw

25:11

and the risk management area. Because AI being an algorithm, that's about validating the algorithms,

25:19

creating the confidence that the algorithm can be used and deployed,

25:22

Right, in a business process. And the confidence is to be

25:25

Higher and higher, the more the business process is highly relevant.

25:29

And typically, wherever you touch a client,

25:31

It's one of the most relevant,

25:32

Okay, model.

25:34

And so, because in the past,

25:38

The old risk management infrastructure was based on validation of defined algorithms.

25:43

Think about BASEL 2, BASEL 3, BASEL 26, which

25:47

Is a large

25:47

Set of rules that stick together.

25:50

Then you run them,

25:51

You test them, and then you can validate them. You know what they're doing,

25:54

and you can prove that basically they do what they're expected to do.

25:58

And that's a lengthy process,

25:59

But it's very focused. With AI instead, it's not that you validate one AI.

26:03

You have hundreds of different use cases in utilizations of that AI that needs

26:08

to be in the studio validated, where the data is used in slightly different

26:11

forms, where the reactions and the positioning of the algorithm is differing

26:16

from what you might expect.

26:18

So how do you

26:19

Transform the risk management into function that interacts in the continuum with the process?

26:25

Now, if I now think about the regulators,

26:27

That are looking at all of this, the biggest challenge that they have is that they

26:32

Need to start understanding that

26:34

The risk management function and the validation function needs to get transformed.

26:38

It's not anymore

26:39

A process where you define the rules, things happen, and then you revalidate

26:44

what happened, conformed with those rules, and so the algorithm also performs.

26:48

It has to be more in the continuum,

26:49

Which poses a problem of independence, right, of the various functions.

26:54

But the point then is,

26:56

Therefore, or what are the platforms and architectures that would enable to

27:00

make this renewed interpretation of the risk management function that interacts

27:04

with the way technology now asks you to basically interact with the problem

27:09

or the opportunity that is different from the past.

27:11

So I was very, very inclined to try and understand how many banks are now

27:17

Making those investments in a way that

27:19

They can position themselves more capable of embracing this change anytime it happens,

27:25

Been strategic in their approach to AI and to generative AI.

27:31

And what I found is that Only 8% of the banks

27:35

Worldwide have a more strategic perspective and approach to the deployment of

27:42

the latest advances in AI, while 70% of the banks wants to do it,

27:46

but they're very tactical. That means that they start here or there, where there's more interest or some

27:49

money on the table. And that's not enough, because it

27:52

Will not enable

27:53

Them to build a risk management culture that differentiates. In fact,

27:57

What I saw is

27:58

That it's not about the biggest banks versus the smallest banks. is

28:02

About the banks

28:02

With a certain culture compared to banks without the culture.

28:07

And the culture, big or small,

28:09

Is complex to build,

28:11

Right, and to, if you like, promote in an organization. So that's my answer to your question.

28:18

That's how the regulators should change,

28:20

But that is exactly where you find the solution to the problem,

28:25

not because you have a compliance that makes sense, but because you have a culture based on new interpretation of risk management

28:31

that enable you to tap into what the technology can enable you to do with transparent comfort.

28:38

In your research, you mentioned digital currencies, central bank digital currencies

28:43

will play a big role in reshaping the financial ecosystem.

28:48

Can you give us like a very small wrap-up?

28:51

What are you thinking forward from here?

28:54

Well, it's like hard in Europe, at least for the European audience,

28:57

said that in September, October, they will define the remaining shape of the roadmap.

29:03

OK, so on whether and how, basically how the weather they will go ahead.

29:07

So there are serious discussions in terms of the creation of CBDCs.

29:12

Some of them have been created, not totally successful, right?

29:15

But there is a needed moment of experimentation.

29:20

The only thing that

29:21

I believe is to be considered here is the following. If

29:26

You look at

29:26

What the Brazilian did with the PICS, which is an instant payment framework

29:31

that will evolve at some point into CBDCs, but it's not,

29:36

they got massive adoption in a few years because they just resolved the real

29:42

problem with simple technology in essence.

29:44

That was the first thing to allow anyone to use instant payments to bring their

29:50

security and clarity in the transactions.

29:53

CBDCs are a bit over-engineered as far as I see.

29:57

So every time it is over-engineered,

29:58

That is a harder time to fly. So still have to see where,

30:02

You know, the ambitions map,

30:05

The simplicity needed to operate these things on a very large type of infrastructure.

30:10

The Americans now are moving in the directions and no civil rights,

30:14

but a stable car. So it's interesting to see how they are positioned and they

30:17

Are all still positioning, right?

30:20

Positioning is moving, but it's not executing in full.

30:23

But I think the 2025 is the year where we will get clarity on this,

30:30

at least across the Atlantic, the perspective of the European versus the North Americans.

30:36

Also a topic we've been talking about years before, embedded finance is another

30:43

trend highlighted in your report.

30:46

There's a growing integration of banking services into non-financial platforms.

30:51

How will this change how consumers and businesses engage with those financial products?

30:59

Okay, so embedded

31:01

Finance is an important element of transformation. It is not a one-size-fits-all in terms that not

31:08

All of the

31:09

Banking services can be easily or seamlessly embedded into a non-banking conversation.

31:16

And it's not about technology.

31:18

It's about the way effectively people deal with uncertainty around the consumption

31:23

of certain financial instruments.

31:26

And it's not because people are not capable or need to be wired.

31:29

It's just structurally so, logically speaking.

31:33

So, it has been

31:35

An opportunity for banks to say what we have now can be consumed at the time

31:41

and in the moment and in the location of need.

31:44

And that remains a particular relevant for types of payments,

31:48

the need of money like lending and sometimes also insurance,

31:51

some products and some investing.

31:54

And what I

31:55

Think is happening is that. It's not enough to have a platform that allows your banking capabilities to

32:04

be consumed by somebody else to be successful. Because then it needs to be wrapped into a business case that makes a lot of sense.

32:11

So when I interviewed the head of open banking of the major players in the world,

32:15

that's what they basically said. They said we're moving from a transactional perspective into the realization

32:20

that what matters is that use case.

32:23

So it's a different way of selling the technology from a bank to a non-banking

32:26

player. But also we saw banks effectively investing to orchestrate ecosystems.

32:32

And that has been growing in

32:35

The last five years. Not only DBS attempted, they made five of these.

32:39

We have Chinese banks doing so. But also Bradesco with Iagro that takes care

32:44

of 50% of the Brazilian economy.

32:47

Strahe launched it two years ago, successfully making a road into that type

32:52

of business and then thinking about how to expand to different ecosystems.

32:55

So we have State Bank of India with the ONU, which is a marketplace for retailers

33:01

that now is aggregating SMEs, so the merchants, into the platform.

33:07

Both of them created an external entity to operate the platform,

33:12

but that built the platform. Okay, so that's interesting.

33:17

One thing that I want to say here,

33:19

Which I believe is important, I was talking to the head of a very large bank,

33:22

I just don't want to mention it, They launched a

33:25

Few months ago

33:26

A platform for the supply chain of SMEs. And the platform is intended to allow

33:32

the SMEs to trade the one with the other. Of course, there was a round-the-table excitement where we heard about that

33:38

in terms of you're competing on payments, right?

33:40

Because every time people transact, right, that's where you make the money.

33:43

But the response was no. The point is not the payment. The point is my role is to lend

33:48

As much as I can, keeping the risk under control, okay? But to do so, I need to have data. So,

33:53

Can I better understand

33:55

The ecosystem if I have that ecosystem to be concentrated digitally so that

33:58

I can go back to the CEO of the bank and to risk manager and say,

34:01

Hey, now maybe we have a way of understanding everyone in the context of the ecosystem.

34:06

So, we are more, if you like, comfortable in doing that business online.

34:11

So, you see, the value of these non-banking platforms not necessarily is the

34:16

non-banking transaction, but maybe what it enables you in terms of risk management

34:21

to do your more traditional business

34:23

In a more

34:24

Substantiated way. So that I found very interesting. I always say that it is the opportunity

34:31

to basically embed that to eliminate the friction that makes the banking embedded

34:36

to generate new value, that the definition of having when banks have a tech or platform economies.

34:41

So when I say new value,

34:42

I mean new value is not a transaction, it's somewhere else. And that is an example

34:46

that I just like to share.

34:48

And I see

34:49

More and more banks are understanding this. It's a steady, long-term process, but it is happening.

34:55

Hmm.

34:57

Talking about mentioning here. I have one more question for the role of AI and

35:02

wealth management for you. And then I'll just hit a few highlights of the topics you will be talking about.

35:09

So the people need to go to read your research.

35:13

And of course, we'll link the report down here in the show notes.

35:17

I also do have taken some notes.

35:19

We'll link your research, your LinkedIn profile, a lot of your books,

35:23

your podcast, Banker's bookshelf, plus your SME research that you just mentioned here.

35:30

And now let us get to the last question I'll be asking you, the role of AI in wealth management.

35:36

Looking ahead, how will AI and generative technologies impact wealth management, especially,

35:46

for individual and small to medium enterprises?

35:50

Okay, so a very broad definition here, but let's say that in the world manager

35:54

space, you have the need of creating the products, okay? And there's a lot of

35:59

compliance around that.

36:01

So I do

36:02

Believe that now there are possibilities to better streamline and systematize

36:06

the way a lot of this documentation is prepared and orchestrated.

36:10

It doesn't mean that

36:11

The banker is out of the equation, but the capability the bankers have to perform,

36:16

to systematize that documentation is effectively augmented.

36:20

And that's where I see a lot of basically use cases being deployed by institutions

36:28

that are prominent in the well-management space.

36:31

There will be a

36:31

Time where there will be more effective communication. I think it will still

36:35

take some time for that, at least in terms of letting it get the compliance

36:39

burden. And that is an opportunity right now.

36:43

And of course, as promised, I'll hit some of the other highlights you have in the research report.

36:49

sustainability and ethical innovation in fintech, future investment trends,

36:54

role of digital transformation in banking, global banking insights and regional

36:59

differences and emerging risks in fintech.

37:04

As we're wrapping up this insightful discussion with Paolo, we want to hear

37:08

from you. What are your thoughts on the future of AI and banking?

37:12

How is your business preparing for those technological shifts,

37:17

let us know in the comments below or connect with us on LinkedIn or Twitter

37:21

using the hashtag Banking2025.

37:24

Paolo, your last words, please go ahead.

37:31

I published in 2019, launched in the Off-World Economic Forum and a week before

37:39

in Hangzhou, China, as a preview, the Financial Market Transparency Theory and Principles.

37:47

So transparency builds trust. Not only transparency is a matter among people,

37:55

but definitely is an issue with technology.

37:57

And I'm not surprised that most of the business people are asking AI to be extremely transparent.

38:06

So that is a principle that we all need to look at.

38:11

So if the algorithm is transparent in the incentives, first of all,

38:16

and therefore in the positioning of the conversation with the clients,

38:22

which brings with itself then cascading down different elements that can be on the bias,

38:28

on the race, and whatever. So there's a multiplicity of elements.

38:32

But the question is, how did you

38:34

Embed the principles of transparency in your technology process?

38:39

There is a reverse effect now because by looking at technology and asking technology

38:45

to be as transparent as it should,

38:47

I see the bankers looking

38:49

At themselves and therefore

38:50

Are reviewing the way they position themselves in front of the client.

38:55

Because the moment you ask technology to replace what you do and be very transparent,

38:59

they basically declare the incentive of the system.

39:01

So there is a very interesting,

39:03

If you like, reverse effect, as I called it.

39:07

So I'm looking closely at how this evolves because that is a key turning point

39:13

in the transformation of the business models that is essential,

39:19

especially when you want to use technology.

39:21

And that's the reason why I published the

39:24

Financial Market Transparency Theory as a positive theory that enables to look

39:29

at the markets differently and it's positive theory because it can be transformed into a business model.

39:34

And I discussed those business models in my latest book, Banks and Fintech on

39:38

Platform Economies, looking at contextual banking and conscious banking and

39:42

by the finance and advisory platform. So it all is consequential. So really, by asking it to be transparent,

39:48

we have as a reverse effect, if you like, an opinion about how transparent we've

39:53

been, they've been as bankers.

39:55

And from there,

39:56

A new business model can be generated the way that value is created for everyone

40:01

in the ecosystem, not just for the bankers.

40:04

Amazing closing words, Paolo. As always, it was a pleasure talking to you.

40:09

And maybe next time we talk, you're the host and I'm the guest on your podcast.

40:13

Okay, welcome.

40:14

Great have a good day bye bye ciao paolo,

40:23

that's all folks find more news streams events.