Transcription
I have slightly reinforced S&P Global. If there were a description, it would come more from a more structural element. Every 5 years, we have the impression that it's going to be disrupted. Me, I believe that what I see, we can imagine that it's like at the bank, you have a cap, a nominal, an interest rate, a duration, there you go, finished. No, there are an enormous number of clauses and it's constantly evolving. I feel like I'm a bit of a fool having my MSI on one side. I tell myself, well, they are indispensable. I don't know where it's going to gain, where it's going to lose, but globally, I should come out on top. You have S&P Global and MSI. Don't you think it's a bit of a duplication? Hello Xavier, hello Lauris. Today, we're meeting for a new episode of Bourse Écomode, and we're going to talk about the fall in financial information stocks in a context of fear related to AI. More specifically, we're going to talk about stocks like S&P Global, so Standards and Poors, Moody's, and MSCI. But before giving you the floor today, for the first time on this podcast, we welcome a guest, Paco de Bonire. Now, for those who don't know him, Paco, you are the host of the podcast Investisseur 4.0. It's one of the most important podcasts in the French-speaking world in the world of finance. And before that, you worked for over 10 years at Moody's. And that's precisely why we're inviting you to participate in this episode today. But I won't say more, and I'll let you introduce yourself quickly before we get to the heart of the matter. Yes. Well, thank you. Thank you all. I'm very honored to be here today. I was a credit analyst for 15 years, 5 years in credit insurance, and then at Moody's in ratings for 10 years internationally. I worked in Brazil, London, and then Paris. And yes, I did indeed launch a podcast on personal finance in parallel. And I'm currently a real estate operator, so I've now refocused more on the moat. But I still have this vivid memory of that experience at Moody's. So, just, we asked you the question just before starting the podcast. So, you worked more in the rating part of Moody's, not the analytics part. Exactly. Yes. I was part of the teams that ultimately delivered ratings to clients who wanted a rating when they issued debt, particularly on the market. Okay. And well, I imagine when you were at Moody's, you had shares? No, I didn't have any. I was No, no, I didn't have any. In fact, they had cut the program. They were very generous with the old employees but not with the new ones. And I didn't have a stock portfolio at the time because it was, well, it was complicated to manage with the reporting that had to be systematically provided to them, etc. It wasn't simple. Even the company's shares were complicated to hold, or just the clients' shares? No. Yes. The company's shares were possible, but I didn't want to position myself on the company at the time. Yes. Okay. Well, great. That's an excellent transition to the topic. So, Xavier and Lauris, and you too, Paco, but I'll give the floor to Xavier now. What do you think globally about financial information stocks, especially in the current context? I've always liked these stocks a lot, but they've been very highly valued for years and years. So they've been on my watchlist for years, for a very long time. And then these stocks are repricing, they're becoming less and less expensive. We'll talk about it every year in terms of earnings multiples. So I got in relatively recently, let's say, when the P/E ratios started to go below 30, which is a bit of a psychological barrier for me. So, I have S&P Global and then MSCI. Why not Moody's? I saw it as less diversified than S&P Global and then MSCI, there's this index calculator aspect and financial data that's quite, quite, quite incredible. But for me, it's part of these moats, it's what we love at Bourse et Comod in terms of competitive advantage. For me, it's part of the top, it's part of the things that are inevitable in the financial world. So, there you go, I have two out of three, let's say. And on the moat, what's the big word that appeals to you at Moody's, not talking about indices, but mainly about ratings? On the rating side, it's that they are obliged to be rated, if only to sell their bond issues, whether it's states or companies. And then, it's almost a duopoly, even though Fitch is in third place, Fitch is still very, very far behind. So there's an almost duopoly between Moody's and Standard & Poor's. So, yes, it reminds me of these duopolies that we like. So there's the toll booth aspect. There's the toll booth aspect. It's large amounts, it's serious clients. Then, I know Abdallah, you know the sector well, better than me, and because it was by reading your analysis on Bourse Comode that I learned a lot. I thought I knew the Bourseo club well, on the Bourseo club, so. On the club, sorry, on the Bekocho club. Yes. So, yes, for me, it's the toll booth aspect. You have no choice but to go there. Of course, the pricing, well, for now at least, they set it a bit as they please on these large amounts, it goes by a bit, a bit unnoticed. So, yes, for me, it ticks a lot of boxes. Then there are questions about growth, there are questions about AI, etc., we'll talk about it, but on the principle of the moat, I really find it incredible. And you, Lauris, do you think the same as Xavier? Yes, I won't be original. I have the same stocks, so with pretty much the same convictions. For a while, indeed, these stocks were trading at 45-50 times P/E. So, I wasn't in them, and for the last year or two, it's been improving, and even more so recently. So, I've slightly reinforced S&P Global, and a bit like Xavier. Why can't I go for Moody's? Because it's more diversified, a bit of ETFs, a bit of financial information, a bit of rating. But I'd be very curious to know Paco's opinion on this. Paco, do you already hold these stocks or not at all? And what do you think of the comparison between Moody's and S&P Global? What do you like about one or the other, or not? Yes, I think that indeed, in terms of business model, S&P is more diversified, of course, because of the index aspect that's added. Then, the business model for ratings is really very similar, frankly, there's no big difference. We even see, well, in principle, when we analyze, we always compare ourselves with competitors. So I think both are pretty much on par. But yes, more diversification, I'd say on the S&P side. Can Paco explain a bit how it works? That is to say, how a client, a company, a government contacts Moody's. What are the approximate amounts? How many teams are involved? In short, it's a bit unclear for most people. Can you explain all that? Well, it's a bit unclear for the analytical teams too, I mean, and it's by design, actually, because the idea was precisely after the financial crisis and that caused in 2007, there was a wall of China that was established in these institutions to completely separate all commercial activity from analytical activity, so that there wouldn't be any conflicts of interest, like, well, I'll pay you a bit more and you give me a better rating, for example. So, that's what the regulator saw a lot of, and what the rating agencies have really worked hard to display. So, indeed, in my work, I had no idea how much they paid, unless there are always emails that we sometimes receive that we shouldn't, and so we have a little idea of the amounts, but very generally, the system is quite simple: there's a sales team that works with the client on explaining why the rating is important, why they need these ratings, and I agree with what Xavier says, it's essential for large issuers, for example, to have at least Moody's, sorry, S&P rating. Well, two out of three, but generally we like to have the first two, and then once the negotiation has started, there's a message to the analytical team, okay, we need to do this rating, and in the negotiation, there's usually the timing, so the deadline to create a rating, roughly, is 3 weeks to a month, depending on the size of the issuer, that's about it, and then the analytical team looks into it, there's always a discussion with management, and that's actually an important asset, I think, when looking at the business models of these companies, it's, I think, compared to the moat, it's access to management that's really important, and so we can have a retrospective view with the history, the data, etc., of the company's analytical data, its performance, etc., the strategy, but also the things that will be put in place, and that's something that only management and the relationship with management will allow us to do. And so, yes, then there are many, we might come back to it in detail later, but on the, let's say, the analytical tools themselves to issue a rating. In any case, just to have an idea, it's not like we imagine, just a machine, you put in two or three numbers, pop, and you get a rating automatically. It's not a fixed grid like that, it's much more complex, much more nuanced. I think that's also what saves this business model a bit from the danger, the threat of AI, at least in this regard, for then issuing a rating to management. Precisely, I wanted to talk to you about that. One of the big narratives about AI right now is to say, well, tomorrow ChatGPT, Claude, they'll be able to do ratings themselves. It will cost much less than Moody's and MSCI. You're saying it's much more complex. What is this complexity? It's access to sometimes non-quantitative data, actually, just discussions with management, non-public data, in short, what makes things complex? Yes. So, to go back to management, well, there are things they tell us off the record that we can't necessarily put in a report, but it's a kind of commitment that's created between the reputation of the rating agency, which will put its reputation on the line, and that's its biggest asset, its reputation, to say, we think that, for example, the leverage, so the debt-to-EBITDA ratio of this company should fall below four times, for example, even though it's temporarily a bit higher, but because we know this, because we discussed it with management, who told us, well, we're going to sell assets, or we're going to grow here. There are things that are a bit sensitive that we can't necessarily put in, because the business of rating agencies is not forecasting, but the rating includes a prospective element, and it's not just about the past, but it's really about what we think the credit quality of the company will be in 3-5 years, or the default risk of the company in 3-5 years. So you have to incorporate a forward-looking view, but you can't say it's going to happen. Well, you can include it, but you won't be forecasting, that's not your business model. So, for example, that's data that, well, that cannot really be created by AI, unless the management, well, that's very human, it's really about relationships, trust, strategy, and we think that, you see, that's just an example, for instance, that won't be replicable by AI, in my opinion. Beyond that, I had a question I wanted to ask you before Lauris asked you in detail how it works. You've answered it a bit, but today we're talking about a duopoly between S&P and Moody's. In fact, I have the impression that an issuer, especially if it's large, needs both ratings systematically. So, in fact, there's almost no competition. We take both to reassure the market. Yes. Well, that happens, yes, there are large companies that are rated by only one of the two, and that's enough. It really depends on the market's appreciation of whether it will believe one of the two. Is that enough? But in any case, one of the two will be there. That's mandatory. And that's also part of the relationship between creditors and the company. That is to say, if there are creditors who need it, they can also put in the financial documents for the debt. Well, there must be at least two ratings. If the rating drops, you see, in the covenants too, you see, if one of the two rating agencies lowers its rating to such a level, or is it one of the two agencies or both agencies, you see? That's in the negotiation when the company issues debt between itself and the market, that is to say, the creditors, it's not just a transactional relationship between the company and Moody's, for example, but it's really with the market. What does the market want? And generally, they want at least one of the two, if not both. Okay. Now, I'll also give the floor to you, Xavier. What do you think of this threat from AI, especially on the rating activity? I think that, well, I've already had the opportunity to share my point of view on AI, at least regarding software. For me, AI will eliminate jobs, it will eliminate employment, but it won't necessarily eliminate companies. In any case, in this rating system, I don't see how one can imagine for a second that someone will accept a rating coming from Claude or from ChatGPT, or even from a new player, a fourth player who arrives and says, "Well, I'm cutting prices because that's the only interest, ultimately. I'm cutting prices because I'm using AI." But who will accept the credibility of this new player? I don't understand this AI narrative at all, and moreover, I have trouble understanding that investors have really pulled out based on this. Then, there's a growth problem with these companies. There's a problem with private markets, whether it's on the debt side or the equity side. Private markets are becoming increasingly important. Companies, as long as they can remain unlisted, remain unlisted. And the same goes for debt, there's a lot of debt. So to what extent are they able to But I don't think that's necessarily a problem for them, they are also very active in the private debt market. But this aspect, for me, it's more of an aid. It will help them save costs. It will help them reduce costs because I'm sure that in its use, there's a way to use AI. But Paco explained it well to us. There's a lot of informal information, and I remember when France's rating was lowered, and I read, it's very rare, but when I read the analysis report in detail, they explained things like, for example, the ability to raise taxes. So there were things like, can the French population accept further tax increases? That's a judgment call. You have to follow politics, you have to look at election programs, you have to look at candidates' platforms, etc. Of course, you can get help from AI, but I completely understand Paco when he says that without forecasting, there's still a part that's a bit qualitative. AI will be a big help. I'm sure for these companies, it's a big help, but it doesn't scare me too much. I don't know what you think. For me, in the area of concern, there's also the aspect, like with software, could it have an impact on the growth of these companies? I agree that for complex cases of states, large companies, it seems difficult to eliminate the human element, but for simpler cases, smaller companies, could an investor not just click on Claude's financial analysis in 2-3 years, say 5 years, it will be much more developed. Isn't that enough in simpler cases, and I don't know what percentage that represents of these companies' revenue. There's something that's important and that really creates the barrier to entry for me, and that's all the historical data because how are these companies judged, their ability to have been good or in line with the default rate? And that's historical data that's not from 5 years ago. It's historical data that, for the market, is important to have, they're looking at 30, 40 years, and that's something that can't be disrupted. The other thing, I think if there's a disruption, it will come more from a, let's say, a more structural and political element, that is to say, we no longer believe in their reputation, you see, because that's where the risk really is, it's if we no longer believe that these companies are capable of producing independent and judged ratings, and so on. But as for, you see, we see in the private debt market right now, there are many new entrants taking a lot of market share because there's a big boom in it. But these are segments where the big ones, Moody's and S&P, haven't gone because they weren't sure, in particular, that they had enough data. You see, these are ratings that are given based on third-party information but without access to management. And they weren't comfortable as agencies to go into that because, in fact, the problem is that if you go into something like that, and they suffered it in 2000, in 2007 when they rated certain products with the subprime mortgages where they crashed, it's very hard afterwards to regain trust, their reputation. And so, when I think they are sitting on a, well, on a monopoly position where they say, well, we know our business model, we have this market confidence, we'll continue. Very good point on growth, it's indeed where I would rather have a negative point, is whether we will continue to have debt that will be rated continuously like this for large companies? Do people not want to remain private, not be listed? But no, no, I think, well, I think historical data and track record are important, which ultimately means that even if a new entrant has 5-10 years, it's worth nothing compared to companies that have 30-40 years. And what about after the subprime crisis? Because there was a big, well, a big degradation of their image. Yes, it's true. And indeed, it's surprising that they've recovered so well. I wanted to talk about that precisely. I wanted to talk about it because when it's, they went through the ultimate stress test. I mean, when I was really in the industry at the time it happened, I was convinced that the rating agencies were dead, finished. We made them the main accused of the subprime crisis. And indeed, a small point regarding AI, why were they blamed so much? Why did we say they were wrong? It's also because they used a lot of quantitative models, purely quantitative, to say, well, when you had a CDO, we took the safest tranche and then we rated it AAA, and so on. And that's where we realize that they made big mistakes due to over-reliance on technology. In short, to summarize, AI would probably have fallen into the same trap, and a human with a bit of perspective would have been more able to see the problem and the limitations of the models. But for me, the fact that they survived 2008, I tell myself that before they have a reputational problem that affects them more than that, because even if it's a big bribe that caused a rating change, let's say it's an individual, it's specific. And here, it's the system that was called into question. So, I find it crazy that they survived. And indeed, just on that point, it happens, it happens to agencies to mess up. I've already seen ratings that were given. The market did not at all appreciate, how to say, the credit quality at the level of the rating that was given. Competitors were completely one or two notches off, and then it was completely questioned. We go back to credit committees, what did we miss? And then we evaluate. So there's really this aspect of adaptation, it's something continuous in ratings, it's not something fixed at a given moment, the rating must always have that prospective credit quality of 3-5 years and always be current, you see, always ongoing. So there's this ability to be flexible according to and to adapt. It's not that the rating is given, it's inscribed in stone, you see. That's what's important. And so, consequently, are you currently invested in Moody's, in the sector, or does it not interest you at all, or do you not do stock picking at all? No, yes, not at all. For now, I'm putting all my eggs in real estate, but I think eventually I'll reallocate to the stock market after RLV. For now, no, I'm not invested in these companies. And on the regulatory side, increasingly, we see, well, Europe wants to have its own rating agency, today Scope Ratings, but it's very small, I think it's German, I'm not mistaken. There are also the Chinese who want to do this. What do you think about each bloc having its own rating agency to avoid American domination? Abdallah, what do you think about the emergence of new players, perhaps linked to AI, but especially to regulation where we regain sovereignty and we no longer rely solely on American companies? In fact, I think that even if it's a risk, it will materialize in the long term. And then, for me, the main risk is where the capital comes from. If the capital is American, they will ask for ratings from Moody's and S&P. If, on the other hand, the capital is European, we can possibly insist on having a new European rating agency, or Chinese if we are in China. So, for me, the whole issue is more about where the capital comes from. And if the capital is national or European, well, then we can perhaps start to impose these kinds of things. The problem is that today, well, my understanding is that the capital is predominantly American, at least what is invested in listed markets, bonds. So, for me, it's not too much of a threat, at least in the medium term. It doesn't scare me too much. As often with these magnificent moats, we always have the impression every 5 years that it's going to be disrupted, and I believe that what I see. So, Abdallah, you've summarized everything by saying where the money comes from, etc. I mean, a reputation isn't built like that. Europe doesn't know how to create a good independent private company, a good independent multinational. Europe doesn't know how to do it. So it will be a half-state entity. In any case, it will be pushed by the states. It won't be, and it will be a big problem. Obviously, no one will trust this organization that will have been pushed, subsidized, etc., by the states. So, I don't believe in it. I don't believe in it at all. But it's funny, it really reminds me of companies like Microsoft, Visa, Mastercard, etc. We say, "But why isn't it disrupted?" It's quite simple. They don't necessarily have the best product, etc., but in the end, you realize they are unshakeable. And for me, you can create a European, Chinese player, etc., as long as the center of the financial planet is in the United States, and even if it shifts a bit more to China or Europe, there will be at least a double or triple rating. For me, if there's a European agency, it will at best be an addition. At best, and even then, I don't believe in it. And so, I wanted us to now move on to, we're talking about AI threats, globally we're not finding any on the rating side, but on the analytics side. First of all, do you know the analytics side at all, or have you never been in contact with them, Paco? Yes, it's the same for me. You're really based on sales, you see. We were really separated from that team completely, so I know what they do, but it's, yes, we were really not in the same team. Xavier, do you think there's a risk related to AI? Well, I'll answer with my limited knowledge, but then Paco can perhaps enlighten us. I always think that, well, what I hear about AI and what I see working is that you always need data, and you need someone who can, who can put it in order and give you a certain security to tell you, well, it's AI, but don't worry, well, AI has done its job, but here, I'm giving you more reliable data. I've sorted the data, etc., and it's not because, you have to see the sorting of ChatGPT, etc. The sorting of data is almost nil. So, I think AI can only boost these companies. Those who use AI will need clean data that will come from these companies, from these companies. But then, it's not necessarily super clear what they do. So, if you can enlighten us a bit, Paco, it would be a pleasure. I'd really like to, frankly, I'm not either, well, I know it's analytical data for companies, for funds, in particular. It's very, how to say, it's distributed intravenously into the systems for funds, etc., but it's quite, yes, indeed, as the name indicates, it's analytics. They are really more focused on pure data, and there, I think there's more risk of disruption. That said, in fact, I'm, compared to what you're saying, Xavier, is that, at least on the rating side, it must be the same on the analytics side. In the end, there must always be a counterparty who commits to the paper, to the result they give, the data they give. So, who is really a counterparty, a company, ultimately, it's not data. For example, when we talk about disruption, we imagine it's a third-party entity like Claude, it will arrive, it will do its Claude rating agency. No, in reality, it's really a tool, and behind it, there must be someone who uses the data and gives a result to a company, analyses or otherwise, and commits personally. So, in the end, it will always remain like that. There's no, well, I don't know what disruption by AI means. If that's what we mean, then my opinion is that it will have to be a tool, a counterparty that will commit. I believe we don't realize how much having good data is a limitation of AI. I came across a graph, I don't have the sources and I don't have it in front of me, but it showed the percentage by site of data used by LLMs. So, first, if I remember correctly, it was Reddit. I think very quickly after, there was Wikipedia that was on the podium. Then, there was YouTube, etc. Imagine, and I say Reddit, it's very good. I really like this network, it remains a social network. Imagine the difference in data quality between financial data professionals, sell-side analysts, S&P, etc. And then what we can find on, even Wikipedia, I mean, Wikipedia, there are many things that are false, that are not up-to-date, that are approximate, that don't cite their sources, etc. So, I tell myself, but AI can only make those who have real quality data explode. And we're not talking about market cap, we're talking about very, very fine, very precise data. Moreover, I believe that S&P Global, you'll correct me if I'm wrong, Abdallah, but I believe it's 95%, something like that, of the information that is true, that is really owned, and it's really, as you said, Xavier, a quality of information that is truly superior to everything else. And what worries me a bit more on this part is, well, when I was in a fund for MSCI, for example, for ESG ratings and others, well, we take, let's say we take a license, but the less active management there is, the less potentially there are licenses, and it's true that in the long term, I don't have strong conviction in active management, at least in funds, and I tell myself that eventually, there might be fewer analysts on the S&P side to exploit this data. So that's more of a benefit on the margin, but I also think that in funds and in banks, there will also be fewer financial analysts or portfolio managers. So I don't really know how to weigh the balance, but I still think that it's perhaps a part where there will be less growth, and much more so than rating. In fact, I am relatively confident about rating, but on the analytics side, it's even financial data, I tell myself that there will be fewer licenses in the long term. And it's a bit the same issue.
On Adobe and others, right. It's a bit it's always is it will we need it? Yes, I think it will always be essential but maybe less so. So that's what with pricing power too that worries me. Me, perhaps. We're perhaps biased by our very equity-focused side, but there's one thing to consider, which is that fixed income and passive don't work. Uh, incidentally, it turns out that for once, fixed income managers tend to beat the market, and I'm not just talking about ratings, but I'm talking globally. On commodities, it's the same, you can't, I mean, you're a manager for commodities and so on. You're not just going to buy a commodity index. So I agree with you on the active equity management part, but there are still, you see, debt problems because we haven't talked about it, but I mean states are massively indebted, companies are doing better. We've forgotten about credit risk a bit, but I think it's going to come back. It's going to come back at some point. We'll get scared, and then, and then everyone will need data, and they'll have to put it into their models. On the fixed income side. What worries me is that, at least from what I've seen, competition is actually stronger, I feel. So we have what we just mentioned, Moody's, S&P, etc. There's London Stock Exchange with Refinitiv. There's FactSet, and there's Tradeweb. I mean, I feel like there's a bit more competition. Uh, I'm not a manager, I've never been a fixed income manager, so I don't know the subject at all. So, feel free to comment. But that's more it for me. I feel like they really had almost a duopoly, and that little by little, even Bloomberg, right, I feel like it's going to get a bit eaten up. Uh, so there you go. Very curious to hear your opinion, Paco. Yes, I think fundamentally on debt, you have to understand that ratings are just a negotiation tool between a company that needs to borrow and creditors. The market, and in fact, we see that it's not something fixed, you see. We can imagine that it's like at the bank, well, you have a nominal cap, an interest rate, a duration, there you go, finished. No, there are a lot of clauses, and it's constantly evolving. We go from, we talk about financial sophistication, you see, but there are times when it's easy to issue debt without covenants, without anything, because the market needs to have fixed-rate products, and and and you see, rates are very low, and people need to take risks. And so, companies issuing debt are in a favorable situation. They will get very good terms. Not just on the rate, but on the conditions that creditors have. That's the problem with private debt right now, is that all the debt that has been issued, you see, I was notably an analyst for the Altis group with Patrick Drahi when he started building his debt empire, and they were able to access the market, to take on debt with completely open terms, where people actually have no control, the creditors. They can just sit on their papers, say, well, I hope I'll be repaid and that the interest will arrive. And then they have no dominant position. And right now, there are, well, constantly, there are also disputes between those who hold debt. First lien, you see, in a capital structure, in debt, you have debts that are more or less senior, some that are more mezzanine, others junior, so they don't have the same position, not the same chance of being repaid in the end. And in fact, there's constantly negotiation, a battle between creditors, between companies, when they have to refinance debt, who will get paid first. So you have, you have, well, there are a lot of things that are important to know in debt documents, in fact, once it's in the documentation, when the company makes the debt contract and says here's how it's going to happen if we manage to do this and creditors have this power, etc., and well, it's just to say that it's very complex, and I think it still deserves dedicated analysis teams, people who interpret for the market, who explain what the real risk is, not just at the time of issuance, but afterwards, when things go wrong, for example, and I think it's difficult for AI to disrupt that. I think, on the contrary, it will be more of an interest to have LLMs, etc., in analytical decision-making structures for ratings, because one of the things we do in credit committees, in particular, is always challenge ourselves. In fact, we always have to have a minimum quorum in committees, you know, 4 or 5 analysts. You need to have people who are more or less senior, you know, who have a lot of experience. So, all of this is done to ensure that it's not just one position. There's the analyst who will give their position, well, their recommendation, but it's just their recommendation. Then, they will be constantly challenged by their peers, and that's perhaps something that will be truly disruptive, because LLMs, with their ability to provide different reasoning, counterarguments, etc., can lead to this process perhaps being faster to ultimately determine a position, but in the end, it will always be a credit opinion that the organization, so Moody's or Fitch, that the company will ultimately grant, will give to the market. And so, I think it will be more of a productivity tool with less personnel in these companies, businesses that are already huge and could potentially grow even more. So I see it more as an upside, in my opinion, on the rating aspect alone, as I'm talking about. But do you think there's really a lot of productivity to be gained with AI from these companies? Yes, there isn't, well, there isn't a huge amount either, to be honest. But but yes, if you just reduce personnel, mechanically, that's what it's about, right, in fact. But well. So, you think there's potential? There's potential. Yes, yes, a little bit. I say that because right now, we deliver a rating in a month, you know, the new entrants in private debt, those, I don't remember the name, but they arrive, they are more aggressive, they try to enter. They give ratings in 10 days, 15 days, sometimes it's appreciated by the market. If you imagine that you deliver a rating in twice the time, maybe you can say, "Well, that could work." But I think, once again, there's such caution from these companies not to make mistakes that I don't know if they would take this route of being more productive, being faster with respect to you issuers, because that's not really what the markets desire. So I think there isn't a huge gap, to be honest, thinking about it a bit more. Yes. And to go back to the analytics part, Xavier, in fact, there's a real growth issue, they even talk about it, MSCI and all, they talk about it for ESG ratings, for financial information. Well, already what happened is that there's been a market consolidation in general, there's been a consolidation. They talk about it in their transcripts. And moreover, in 2024, when there was the absorption of Credit Suisse by UBS, there was a big drop in retention rates because, in fact, quite a few licenses were merged. So honestly, I'm a bit worried about their business related to the number of licenses. The question is, with AI, will they be able to capture a higher margin, a higher, well, a larger share of the value they manage to create with their data? That's a question I ask myself. And then, there's something I like about Moody's that I think isn't present, at least at S&P, or at least less developed, which is the compliance part. They have a whole KYC activity, to support banks, insurers, financial institutions in general with their compliance, fight against money laundering, against financing of terrorism. And today, it's almost become an obligation given the sanctions. We saw it with BNP Paribas, which has been regularly sanctioned with record fines, and it's an activity that's growing enormously at Moody's, and I think AI, for once, is not something that will be disrupted by AI because the stakes are too high, and on the contrary, it will even be a catalyst to continue to gain productivity in this area. So there you go, I think there are different activities in the analytics part, but that's what worries me the most today. And so, if we go back to our basic subject of valuation, overall growth, debt, honestly, I'm not too worried. The problem with debt is that it's always cyclical. So there are financing cycles. Uh, right now, Moody's and S&P Global are talking a lot about the refinancing wall, because quite a bit of debt was issued at very low rates and will come to maturity and will need to be refinanced. So, we're really talking about debt cycles that will continue to grow, by a little less than 10% in the coming years. And then there's also the subject of AI, and AI today, you've heard about it, I imagine, but the hyperscalers who will turn to debt to finance data centers, the purchase of more, and so on, and will therefore probably be rated by Moody's and S&P Global. What do you think, ultimately, about the impact of AI investments on these rating agencies? Rating agencies? Is it a positive thing? Uh, Lauris, what do you think? I don't know, the proportion seems relatively modest to me compared to the overall activities of these companies. What's certain is that we're probably going to start a favorable debt cycle again for these players. So that's a good sign. After that, I look at the long-term trend, I think there will be more and more companies that will require these ratings. So that's a good sign. After, as Paco said, it's true that today, I think there's a bit of a lack of, a lack of companies on the sidelines. So most companies today want to stay a bit on the sidelines. Is this a market that, well, is this a market that can escape these rating agencies? It's not impossible. Even if for very large companies, it remains essential. I don't know what you think, Paco. In fact, from when does a company decide, "I'm going to call Moody's, S&P Global"? From what point? In fact, it's an arbitrage they make based on the type of creditor they want, the rate at which they can get debt. And so the lower, the lower it is, the better. The more accommodating creditors are and the less conditions they ask for, the better. And so, in fact, you have several markets you can go to by issuing bonds, Asian, European, or US markets, to be simple. But basically, basically, if you have a small company and you have the confidence of your banks, and then of your investors, because there are companies that constantly use private debt, they don't need to. So there's also a size factor that means that the bigger you are, the more you need to access a larger pool, and that's where you go to the bond markets, and that's where you need ratings. You can very well, if you have a company with the confidence of your creditors, your banks, you can very well decide, because after all, it's a cost, the cost of the rating, the maintenance of the rating, it's not just at the time of issuance, you have to, it's also work. So I think if, let's say, at equivalent cost and equivalent gain, companies prefer not to have a rating, clearly, not to go to the markets. But it's still important to have several market depths, several different markets. You see it recently with Amazon's issuance, a huge, very successful issuance. Well, it paid less for its interest rate in Europe, even though it's an American company, than in the United States. So for them, there was a real gain, and I think it was 100 basis points. It was really not negligible in terms of It's related to the euro and dollar, isn't it, if I'm not mistaken? Yes, interest rates were one point, 100 basis points. That's what I read about the two. Maybe I'm wrong. You tell me, but it struck me because I thought, wow, the gap is significant. Yes, there was a tranche in dollars and a tranche in euros, but in any case, they have access to the European pool, and so diversification can sometimes create effects where even if the risk is seen and perceived in the same way, you get cheaper debt in certain markets. So this market depth can be interesting, which means that for large groups, but you see Amazon is huge, it's a behemoth, now small or medium-sized companies might be better off staying private and not going through agencies in general. And you, Paco, when you were there, did you deal with a lot of unlisted companies? In fact, I don't realize the size. Are these companies that do 5, 10, 20 billion, or do you sometimes deal with small caps? I mean, I don't know if you have, well, not necessarily examples to give, but Yes, well, it's, well, we, well, I was precisely a bit on the small-cap PE, so it's 500 million in revenue, let's say minimum, listed or unlisted, they were, they were with a rating, they weren't listed on the stock exchange, their shares weren't necessarily listed, but they were companies that accessed the market, well, the debt market, that's what I mean. And so, yes, I wanted to ask you, Paco, when you were at Moody's, did you have an order of magnitude for the prices paid by, well, by the clients of these agencies? Yes, there, you're asking me to reveal secrets. No, very concretely, we were really intermediaries, we weren't supposed to talk to the sales team, it wasn't just out of lack of interest, but we had to be really kept apart. So, I have no knowledge other than occasional slip-ups in emails I received that I shouldn't have received, which then had to be reported to compliance, and which obviously snowballed, but from memory, we had an issuance, it was a small cap, about 20 to 30 million issuance, and I think for the rating, we were around 40,000. I had five basis points in mind. So, for example, on, I don't know, a city that wants to issue a bond or a company of 100 million euros, that would be about 50,000. Well, I don't know if that's right or not. I don't know, especially since I imagine that for high-yield, it might be more expensive than for triple-A companies like Microsoft. And then, from what I understood, we agree, Paco, there's the rating at the time of issuance, and then there's an annual subscription of maybe 20,000, 30,000, 40,000 euros for the rating. Exactly. Yes, that's it. You have the cost, I think, at the time of the first issuance, and then there's really the monitoring of the rating, which is constantly monitored, and so until it's dropped, there's an annual cost to pay. Yes. Yes. They talk about it in the publications of Standard & Poor's and Moody's. They talk about transactional revenue, so at the time of debt issuance, and non-transactional revenue, which is more recurring, and that's what you're talking about, PO. And by the way, there's also another thing that S&P Global talks about a lot right now, which is that sometimes companies can ask for a consultative opinion before requesting a rating, before a debt issuance, for example, before an IPO or a sale, well, that type of operation. And that's also monetized. I mean, they charge for it. Were you able to work on that too? Was it a significant activity? I mean, can you tell us more? Yes, yes, yes. So that's more of a lead product for people who want to be rated. You had the possibility of having what's called a point-in-time rating, meaning that this time it's not monetized, it's just you give us all the information at a given point, we give you the credit rating based on that, at a given point with all the types of scenarios and assumptions that are made. It can be in private form, meaning it's not disclosed to the market. We deliver, instead of giving a credit rating as we do for the market with a letter, we give a private letter explaining all the assumptions that were taken into account in these scenarios, because there are also companies that do M&A, they are already rated and they ask, for example, if we acquire with this level of capital structure, this level of debt that we will put in place, how would our rating evolve? And they haven't done it, so obviously, it's confidential, they haven't announced anything to the market, so again, access to management is important in this case, and we give them, well, based on two or three scenarios, if we put X percent debt versus Y percent debt, the senior, mezzanine, junior tranche, you see, so there are many scenarios that can go up to three, we could rate, and we deliver a letter saying, here's our opinion based on that. Then it's not, we don't commit 100% that if they issue a rating with the conditions they gave in the scenario A that they gave us, we will give them exactly the rating we gave in the letter for the scenario, because out of prudence, but it's, you see, 90% chance anyway. It gives comfort to the issuer that if they go to the market based on what they gave us, you see. And sometimes they decide to adjust their capital structure or their scenario. But it was still very sensitive because we are really at the limit of the difference between a rating and advice, and the regulator is very vigilant. You see, the agency must not give advice. That's why everything is given based on hypothetical things, point in time, at a given moment, you see, and all the safeguards are in place for that, but it still provides a more malleable tool, commercial flexibility with the issuer to say, "Well, we're not a black box because that's often one of the criticisms of agencies, it's, well, black box, we don't really know what you do, and you give us a rating like that." So, this tool was very practical. And just to be sure, so the idea is to have several scenarios and then have a bit of the breaking point and take the best possible scenario. But for example, Moody's doesn't give advice by saying if you do this, you'll get a better rating. In fact, it's the company or the city or the country that says, "We have these scenarios, what do you think?" But Moody's doesn't say, "Do this scenario instead, which doesn't exist." So, we agree, there's no advice. No, no. Exactly. And that was the result, in fact, of, well, it led to a lot of discussions afterwards with the issuer, because often the issuer prepares their operation, they have two scenarios, they are allowed up to three scenarios, but you see, they will have two scenarios, and then in the middle of the work, they will say, "Oh no, actually, we've thought about it, we have a third one," or when you give them the results of scenario A, scenario B, they say, "Yes, but on scenario B, if we have just a little bit more, say 10% more for," and that's where we have to be very careful, saying, "No, no, we're getting into advice, and we can't afford to give you the scenarios. You define the scenarios, they are fixed, they are not, we don't move them, we give you the results based on that, but no advice afterwards." So, if we go back to the stock market aspect, investing in these values, we still have, despite everything, from what we see, very strong values with few weaknesses. If we have to play devil's advocate, Xavier, what do you like less about these values? It's a shame, it's a shame. I wanted to build on that, I wanted to build on the price if you allow me, Abdallah, because when we talk about 0.05%, you see, 5 basis points, what I love is that I'm sorry to always compare to this, it's like Visa, Mastercard, credit cards, they cost a lot. I mean, credit card payments are 1.5%, but the share that Visa, Mastercard takes is ridiculous, and it's purely, but it's huge on the amounts. And frankly, when I hear the amounts, either what you're telling me, PO, or I think with the scale effect, we're at 5 basis points from Lauriss, I say it's not expensive. I can't imagine someone paying for a large issuance who would save these 5 basis points to go to an unknown player. I imagine it seems very logical to go through Moody's or S&P Global for 5 basis points. Conversely, you see sectors like Airbnb, which will charge hoteliers 16%. That's huge compared to the hotel's profits. It's almost equivalent, perhaps, to the hotel's profits, or at least not that far off. So, I understand that the pressure on prices, if the model disintegrates a bit, is enormous. I have the impression. And the same for index calculation, because here, we're talking a lot about ratings, but there's a small part of S&P Global that's about index calculation, and MSCI, where there's much more index calculation. It's the same, we're talking about small amounts for management fees of 5 basis points, when Amundi or Vanguard and others do the ETFs, there's a small part that's done at MSCI. So I say, why disrupt these players where you have the stamp, there you go, it's an MSCI, it's enormous. Why look for something cheaper, knowing that ultimately the cost is low? That's for the positive side. After the devil's advocate, I wouldn't look for it on the AI side. I would, I would indeed look at the growth aspect, the fact that private is increasingly present, because we know that being listed has enormous constraints, and whether it's for debt. And as long as you can avoid it, as Paco, you explain to us, and I understand perfectly. As long as you can avoid issuing a bond to the whole world, and you expose yourself completely, or the same for IPOs, well, as long as you can stay private, you do it, and you might think that someone who buys private debt can rely on a small player, or rely on their own calculations. Not that it's not complicated, but there you go, for me, the private debt part would be more frightening, but we haven't talked much about indices. That's also why I'm not on Moody's and I'm more on S&P Global and MSCI. I say, you mentioned active management, passive management, and I completely agree with you, and I feel a bit like I'm hedging by having my MSCI on one side, even though there's also pressure on prices. I say, globally, they are indispensable. I don't know where it will win or lose, but globally, I should come out on top. It's a bit like when you have all the hyperscalers, you say, "I don't know who will be the big winner, but if I have Alphabet, Microsoft, Amazon, etc., they will capture future added value." And I have that feeling about this set of financial data, indices, ratings, etc. I say it won't come from outside. The money might be more with one than the other, but globally, there's this hedging aspect of having several. But I find it hard to really play devil's advocate, apart from the valuation part, but here we're just talking about the quality of the companies. In terms of company quality, I find them really quite robust. But so you have S&P Global and MSCI. Well, like me, actually. But don't you think it's a bit redundant? Because one could say, well, if you want indices, you go to MSCI, which is heavily exposed to indices, and you add Moody's for better exposure to the rating part, or rather have only S&P Global, so you have a single, more diversified value of the three. What do you think of this argument? I think there won't be a monstrous dispersion in this sector. That's the impression I have compared to other sectors like luxury, where I think on the contrary, we're heading towards enormous dispersions. Here, I say it's diversification that costs me nothing. And you, Lauriss, are you more interested in Moody's? No? Because it's the least diversified. Because you also like this exposure to indices that you don't find at Moody's? Yes, that's it. And honestly, I don't know it at all. Here, it's really MSCI, I used it. S&P Global, well, good, there are indices that are easy to know, the financial data part, with Capital IQ, etc. Well, I know that too. So, frankly, a large part, apart from the rating part, I know what it looks like, I know how people use it, I know the prices, I know the margins, why people like it, why people don't like it, and Moody's, in fact, I don't know it that well, because the rating part is relatively vague, and so when you don't know the rating part of Moody's, for me, you don't know much about the company. So I'm more in favor of this opinion, but like with Visa, Mastercard, like with many sectors, I don't mind having redundancies at worst. Because I think globally, that is, I'll put 5% of my portfolio into these companies, and I'll do 2.5% and 2.5% instead of 5%. So it doesn't bother me that much. Yes, I. And you, Bo, at the end of our discussion, when you start stock picking, will you consider this sector in your portfolio? Yes, but I think in terms of valuation, it might be a good entry point, I would say for this kind of, if we look long term, because, well, the business model, we see it, and it's solid. I think there are disruptions to anticipate. I mean, we're not immune, I think, since we're at the cycle level where there's been a lot of financial sophistication. Again, we see that it's starting to tighten up at the private debt level. You see, there are also, well, products, notably everything related to banks and insurance, so the ratings of these banks and insurance companies, or products where people, it's more sophisticated, there's less knowledge. I think we're not immune to reputational risk for one of the three agencies, and that could disrupt things, as we saw in 2016. I don't think it would bring them down, but it would create tensions. But beyond that, in the longer term, I'm not really worried, and I think, yes, yes, I think these are defensive stocks, where at the bottom of the market, it's good to enter. So, for now, it's still a bit expensive for you, you think? Yes. Okay. Okay. So, well, Xavier, how do you consider the valuation of, well, let's take S&P Global, which has fallen quite a bit after the results. Yes, I bought a little bit before, and I knew it was still a bit expensive. It's part of these sectors. What I like is that these companies, in terms of stock market, have been going back and forth for 5 years, and then indeed, it's falling well. But revenue, profit continues to rise. You see, the valuation, in fact, remains at the same prices, but if you forget the stock prices and look at the balance sheets, income statements, you see it climbing every year, gently, etc., except that the price is stagnant. So, necessarily, at some point, there will be repricing. I'm a bit like Paco. I find that these are good levels, a bit expensive, but I'm wary of these high-quality companies. We always feel like they're expensive. So, since it's a bit less expensive right now, well, I'm quite happy to be on it, but not serene enough to increase my position, to make them very large positions. But they are in my mid-tier, let's say. That is, 4-5%. No, less than that. No, less than that, because I have, yes, rather rather three. Yes, but if you had to keep only one, which would it be? S&P Global, I imagine. Yes. Yes. I have the same values. I have MSCI, I must be at 480, I think, for the purchase price, and S&P at 450, around there. It's been about a year and a half that I've been on it, and I've increased my S&P position. So, in fact, it's a bit like Xavier. I like the cases where there's no big drop, it goes up and down a bit, 20-30% increase or decrease, and in fact, the P/E ratio decreases little by little. So, for me, that's perfect. So I'm ready to wait. I must have, yes, 4% on S&P, no, 4% on MSCI, sorry, and 3% of my portfolio on S&P Global. But for me, it's a core portfolio. It can continue to grow like this. Maybe it won't do 12-15% per year as I'd like for these cases, but that's okay, it's solid, it generates margins, and worst case, if it falls, we have to buy more shares, and I'm very, very happy. So I think it's a first interesting point for those listening to us. But if it were to fall even more, it wouldn't shock me, honestly. We're still at 25-30 times earnings. We're not at 15 or 20. In any case, we will continue to follow them on the Bourseo club. And by the way, Paco, you joined us recently in the Bourseo club. I hope your first days are, well, your first weeks are going well. Have you been able to read some articles? Impeccable. Well, great. No, honestly, how have you been able to read some articles, how have you found the Bourseo Club for your first days, first weeks, sorry? Yes, there's a lot of content. I haven't been able to read it all, but I find it very detailed, and you're not afraid to take a stance. That's what I appreciate. It's opinions, it's not just "it could if maybe." So I like that. Thank you, Paco, for coming on the podcast. You're really great to have your insider expertise. I hope you had a good time. Excellent. No, honestly, very happy, and yes, thank you for giving me the taste for markets and finance, which I had a bit left behind, but it makes me want to get back into it properly. Thank you. We'll be attentive. You'll tell us when you start investing in the stock market again. Well, Xavier, Lauriss, I think we've covered the sector well. So, as you can see, we are generally positive on these values. I hope you learned a lot, especially thanks to Paco's intervention. If you have any questions, you can leave them in the comments, and we'll meet again soon for a new episode. Bye! Hello everyone. Hello! Hello! Hello!