Subscribe
Share
Share
Embed
Southeast Asia lacks honest signal.
Straits Signal is a long-form conversation platform built around the systems shaping the region: mobility, capital, infrastructure, and the energy transition. Each episode is a single extended dialogue with someone navigating a structural shift in real time: the operator absorbing the friction, the investor pricing the risk, the builder sequencing the impossible.
The format is deliberate. One guest. One thread. No panel, no pivot. What emerges is how Southeast Asia actually works, the version with the capital gap, the regulatory constraint, the human resistance no one writes about.
Listeners leave each episode understanding why something is hard in a way they didn't before.
Hosted by Kim Yeoh. Recorded in Singapore. Built for everyone who wants to understand what's actually being built next.
PANTARAI: THE DECANTED PORTFOLIO
Final Podcast Script - Part 2 Draft 1 (Strategic + Warm)
Runtime: 50-55 minutes | 55% Strategy / 45% Narrative | Warm, Curious, Rigorous
SEGMENT 0: THE OPENING (2-2.5 min)
[TONE: INTELLECTUAL WARMTH + INVITATION TO THINK TOGETHER]
KIM: "Welcome to a conversation about something that's becoming urgent in a way most people don't realize: How do you invest in a world that's fundamentally changing faster than your portfolio strategy can adapt?
See the pun I did there – You've probably heard the advice. Sixty-forty. Hold for ten years. Don't time the market. It's comfortable. It's the industry standard. And increasingly, it's inadequate.
Today I'm speaking with Nicolo Carpaneda and Bastien Seignolles—the founders of Pantar.AI. They met 15 years ago at IE Business School in Madrid, which also happens to be surprise surprise my alma mater as well, & you climbed the institutional ladder at Citigroup, M&G, and blue-chip operations across two continents. And then they made a decision that most people only dream about; they walked away to build something.
But here's what's interesting to me—they didn't just build another fintech. They built an AI system called Cartesio that does something philosophically different. It adapts your portfolio every single day. No static allocations. No 'just hold.' Just intelligent responsiveness, radical transparency, and the courage to bet €1.5 million of their own money in it.
We're going to talk about partnership, about risk, about the future of wealth management, and about why timing in markets is indeed a lot like timing in wine. But more importantly, we're going to dig into the hard questions: How do you actually scale this? What could break it? And is this a product company or something bigger?
Nicolo & Bastien—welcome. I'm genuinely curious about what you've built."
SEGMENT 1: THE PARTNERSHIP ORIGIN (7-8 min)
[TONE: WARM BUT INTELLECTUALLY RIGOROUS - BUILD THE FOUNDATION]
KIM: "Let me start with something that fascinates me about your story. You guys met in 2008-2009, right when the financial world was imploding. You were at IE Business School. And instead of networking your way into recovery mode like everyone else, you spent a year cracking cases together, day and night.
That's pretty cool for a MBA friendship. That's something else.
Walk me through that year. What case were you solving? And more importantly—was there a specific moment where you looked at each other and thought, 'We're going to build something one day'?"
BASTIEN: “Mmmmmm, that was a couple of years ago from now i think. Nico? ”
NICO: “ Few years ago, time flies!”
BASTIEN: “Ummmmm, thank you very much, I think ummmmmm ummmmmm, the way the MBA was organised, i mean we were like an intake of 400 young professionals & ummmm, these 400 guys were divided into different sections. So Nico & I had the chance to be in the famous in my own words “Enthusection”. & we were like 60 guys with 30 different nationalities & the way the work was organised during (not sure word) was that we work in thesis for basically each different topics & groups. So we were divided in groups of 6 people & these groups were changing every 2 months basically.
So first of all like we have a chance to get to know quite well of people in our class in class, like you know sections so like small groups of people & the way basically it's like listening to your peers & your colleagues & I think that they'll value or I think the biggest value in MBA would be your peers & basically who you're studying with, & yes of course with academics basically the rankings the different things. The matter is that the peers & the group of people who are here are super important & obviously the way I meet people by sections – I have a chance to hear him talk, challenging the teacher on answers, providing feedback. & I really liked the way he was working & he was thinking & also we have a chance to work on different thesis & to appreciate the quality of work. & also I think the human dimension because I think you are spending quite long hours together so it's quite intense, it's almost impossible like a little bit goes by per month like 14 to 15 months; that's like long hours long day long night as well! Not only is it in class but also in groups but still spending time together is the right opportunity to know each other to measure & assess in the consideration of value assess I think the value I mean the you know the other can bring and I think that to what I recall and maybe Nico you you will tell us what what you remember as well because after 15 years you know we maybe might have some gaps uh but um I think that like in the end of the programs even if we uh basically myself I I I went back to France uh and Nico, I don't remember exactly when you moved to the UK. If you move straight, but you will tell us afterwards to to London. Uh but uh we said okay, I mean we we like to work each others. We think we could uh you know, we are different. We we have different point of views. We are working hard. Um so there's like maybe it's not something could going on. And at that time you were right to remember also that it was really 2009 because we graduated like almost in in Christmas 2009.”
KIM (acknowledges): “Okay.”
BASTIEN (Doubled down): “And the the job market and like the economic uh conditions environment was just I mean the worst over I mean for the last decades if not more. Um and so it was also a difficult time to find a job. difficult time I think to also to to launch the companies and of course we need at least on my side I needed to to to to start to pay back my investment because I founded myself the the MBA uh and so the priority was really to to generate income for me and my family and at some point I I made the decisions to not really you know look at this thinking about what kind of business I could do but more like you know finding an income and this is basically what I what I recall and What about you, Nico?”
KIM (agrees then focuses attention to Nico): “Okay.”
NICO: “Yeah, thank you. Thank you, Kim. It's a pleasure to be here. Bastian, you say a number of wise things. I will only add that first the MBA years were such a fun period and a very challenging period as well and Bastian happened to become a good friend. He was one of the top students in class especially in terms of curiosity and being well organized. And so just to say something different, fast forward after a number of years, I realize now that finding a co-founder, it is so important and so difficult as you go through something that is a dream as you mentioned Kim is wonderful to think about. Look, we can move out of something managed by others and do something that we care about, but it is incredibly emotional. It is incredibly difficult. And so if you can partner with someone for me Bastian with whom I spent a couple of years working with sleepless nights challenging thinking of different friends, different cultures if you want disagreeing in a number of ways an going through a lot. It was a very good test to go through the challenges that we face every day. So it so much fun to met in the past but it was also a very good dream to go through the daily challenges that we have. So it was a very important pivotal moment for us and we didn't know we didn't know at that time.”
KIM (agrees then focuses attention to Nico): “Okay.”
[PAUSE FOR RESPONSE - LET THEM TELL IT]
KIM (FOLLOWING UP BY COMPLIMENTING): "I think it's always easier looking backwards you know when kind of see how the puzzles all fit together because I I do share the same sentiment as well. I'm still very close to my IE friends as well. So I do share I do share what you guys have in in some ways not fully but
in some ways.
So, you know what strikes me is that the 15-year gap, right? Um, you guys met at that point where you spent a year really thinking about how markets work, how institutions fail, how fear spreads, because that's when Leman Brothers was down, the big institutions were all down, right? And then you both disappear into corporate life because you invested you needed your ROI back which is perfectly something that to heavily consider if you're considering an MBA but um and you became in some ways the system that you were studying right so between this 15 years what were you each thinking about as you journeyed through it because were you guys were constantly in touch I believe to have come to this point to this day right did you guys pitch to one another and say this Is it or do you just both feel that the market conditions shifted at the same time and knew the stars were aligned and this was it?
BASTIEN: “But I I think that um to answer your questions um I mean we have a chance to have a very united uh um you know intake.”
KIM (acknowledges): “Okay.”
BASTIEN (Carries on): “But I I think that um to answer your questions um I mean we have a chance to have a very united uh um you know intake. Yeah. Uh and um and we have a chance to keep seeing each other. I mean when I say uh keep seeing each other. Let me give you an example. Uh next month I'm going to go in Japan uh for for holiday with my family. And I'm going to have a lunch with some other peers Japanese friends and South Korean friends in Tokyo. So it's true that we we we continue to to have you know regular uh you know gathering and thanks also to you know the global alumni forums which you know occurs every um every June um I had the chance to to to come back there several times. Nico also was coming and now he lives in Madrid so it's even more easier and also the the life made me to deliver some projects in Madrid and meaning that I had also the chance and also to go in London so to see to see Nico so we are we are we were and we are still a very close community and that's fantastic because it's it's a worldwide you know footprint and Niko and I continue to exchange through this community and also you know following each other on LinkedIn uh you know sending a WhatsApp time time to time just to have a catch up and and and see how things were were going. It was really like a um and this is I think a big plus of this of this community of this IE community is that we we continue to to follow up and people were talking of course about business but as Nico mentioned about very you know friendship and you know and family and uh and we had just sense to to basically I would say it was completely natural not like you know uh no I need to continue to to to keep contact or okay maybe some for business or whatsoever.
This was a real friendship and this is what really built you know basically this relationship and this is also that uh you know built this business, this opportunity and on that is on my side. I will finish by this. Um I like I I live in several and work in several countries uh after uh
after the MBA and uh my I was in South Africa before being in Singapore and moving to South Africa uh gave me the opportunity to look back in the mirror and give to give me some time to, to, to, to think about what could be my next move uh after my career in in management consulting. And it just happened that at the same time Nico you know had moved forward in his project and his idea because Niko is is the founder and he's really the one who uh you know developed program or engine in the AI system and it's really like you know the stars were aligned as you mentioned like all the flags were green right, uh I could be available Nico was some needs and as as we knew each other and as basically the first door was open, the trust was already there from the beginning. It's really helped to basically you know be in motion and to make take off this project.”
NICO (Agrees & shares his experience): Yes. On the on the 15 years gap I was going back with my memory we were at I in 2008 2009 and the iPhone was not even there.
Okay. So you should know that I I went into IE to learn how to start a company. It was all right popular. Amazon had been around for a while but it was not these stream of the past 10 years where it was relatively easy to do a successful app. It was before these years and I was inspired by a family full of entrepreneurs on my side. Some who did very well, some who failed massively, but the passion that they were bringing back home, it was incredible. And it was very different from my first years of being an employee into big and nice companies.
Anyway, so I decided to do an MBA to start learning about startups, but it was really not as easy to start. And so we met with
Bastian with the story that Bastian shared and I and some of us really wanted to start something but many barriers of entry were present in code, in manufacturing, in telecoms. And so we informally decided to all go back to work, make back some money and also build some experience in whatever the field. And I think there are two ways to do startups. The more American one is that you drop from school, you have an idea, you meet a co-founder at university and you try to make it big, but it is an environment where money is or was widely available and where risk is rewarded.
Europe and Asia slightly different in Asia. Of course, it depends on the country. And so we had to wait and we decided to think about starting a company with more experience. And so we both built that experience digital transformation broadly for Basani. We'll go through that financial markets for me. And I think we decided to remit with an ongoing friendship and a lot of touching points at some point when we thought it was the right time because barriers to entry were lower and we had enough contacts especially and also
independence, from a financial point of view to start this.”
Kim: “So it's patient capital in some ways true friendship.”
NICO (Agrees): “I think so, it's a good summary.”
Bastien (Agrees too): Yeah. Absolutely.
[LISTEN DEEPLY - THIS REVEALS THEIR CONVICTION]
KIM (PRESSING GENTLY): “And um speaking about entrepreneurship like Nicole you were saying you were exposed to a family of entrepreneurs as well and I think Bastian as well in one of our conversations you mentioned that um you wanted to you had a family history there too. Your parents, your grandparent were all were entrepreneurs right?
So in some ways this is not really that accidental although it is accidental if you think about it. Um, and you mentioned as well that you wanted to take control of your own destiny at this point in your life. But here's what I want to know. Like both of you, Nicole and Bastion. What was the reaction from your partners, your spouses, both of you married to executives running APAC regions or managing teams, building portfolios?
What was that conversation actually like? Because you guys are living your jobs, right? I mean, you're going from two big corporations to a two-person startup. That's not a small ask for people you're, you know, building a life with.”
NICO: “I will kick off if you wish. There was a quote, I think the other day coming from Marc Andreessen. of a16z that said something like everyone from the outside believe that doing startups is fun. But it is such a big emotional baggage that when you are inside is not fun at all.
And while I think it is still fun, I can perceive the emotional baggage that in fact is part of the journey. And what you mention about the spouses is absolutely key because one key characteristic I understand now needs to be there for every founder which is you need to be resilient, through the challenges that you will find every day with money organizing the job people not buying people not listening to you friends not caring potential prospects saying that yes they care but then when you start they don't buy and so The resilience touches two or three points. First, your emotional well-being in remaining balanced and centered and trying to carry on despite everything that does not work.
Second, in your professional day-to-day, you also have a life and in this life you also want to remain centered as much as you can and the spouse role is fundamental. So I am married with an IE former, IE student Rosalyn Ramos from Puerto Rico and she has been fantastic as my friends know because she said from the beginning I believe in you. I will support you. I know you are passionate about this idea and starting a company to help people solve a specific need. So I'm with you. I think you're ready. do and try to uh put it put it together and I will always be on your side and this message continues and so for my resiliency which I think is a core characteristic that I'm building uh the support of my spouse she's a senior executive in LinkedIn is fundamental emotionally financially from the organizational point of view of my family and so absolutely critical without that it would be a mess.”
Kim: “Okay, agree.”
Bastien (Partially agrees but elaborates): “No, I I can only second Nico on this. I mean um I mean on my side um and it's it's true that like you obviously it's super important because uh I mean you have like a uh I would say a question about uh you know life balance which is super important. Uh you have a question on of course of money because uh you are not going to um to get as much money as you you had before for sure at least for for a couple of years. So you need to be ready to to know that uh I mean the income of a of a family might be reduced at least at the beginning for for a couple of months years. Um and of course there's a question of stress management and uh maybe call you more resilience but that's also the same idea is like the way how you behave because uh some days you are desperate and some days you are like in a fire and you are super you know optimistic. So um I think that's an emotional roller coaster. It's very difficult same thing for me without of course the support of my wife I would have been able to to do it. I have three kids. Uh so obviously we need to have some you know some oil in the engine to make it work. Um but in the meantime um on my side um as I said I I travel I mean I I lived you know overseas abroad for many years and this um situation of living abroad also gave me some situation where I could not work or there were like a lot of complexity barriers also to to find uh local works for like work permit reasons for like compliances I mean many different reasons which in the end you know happened to give me the you know the opportunity also of thinking differently and also looking at at new options and on that side my wife support me sorted me a lot saying okay met something where the passion is because she really loves to see me passionated uh and uh and she much more valued uh know to see me passionate than rich so um so that's I think that's also an ent like an important entry point.”
[LISTEN FOR THE EMOTIONAL TRUTH]
Kim: “Yeah and I kind of disagree agree more because when passion lives in you the way your outlook you have on life is very different it's it brings a different dimension to the marriage because I myself I'm you know I do feel you on that way and my my partner has said that as well um so anyway before we go forward um just last one though because the reason why I'm asking that about this is um not only your home life with a partner has to be intact your your partnership within the business has to be as well in order to move forward right it's the foundation of everything. So, and yours isn't based on a hot idea or a coincidence. It's based on 15 years of trust, complimentary skills, and intellectual alignment.
So, Nico, um, one thing that impressed me a lot that when Bastien shared about you guys was that you taught yourself to code during the pandemic, um, and you could have hired a CTO and you had a network, the capital, the credibility.
Why do you decide to build a thing yourself?”
[PAUSE - LET HIM EXPLAIN THE INFLECTION POINT]
NICO: “Nico: Good question and I I'm is is it it has been such a wild ride. I have to say I left London after 10 years of working there in mid of 2018 and I was still working full-time and 2019 it was my first year in Madrid and then I organized an agreement with my company to organize you know my, my former employer my time to go part-time.
I really wanted to take a project to study the viability of this idea about autonomous investing and I've been asking them you know if they could just give me some time so they said yes and I started working just a number of days per week having some free time in that free time in the early days I have been looking for a CTO.
So I really tried to look for expert engineers that understood a little bit of financial markets where I could complement more insights and for them to be really good at doing code and we run some experiments where I put some initial money but they failed and they probably failed because of my fault or there was no fault on any side solely not on the engineer side. I partner with some very good people in Madrid. But it was such a difficult conversation to sit down together on a weekly basis saying look I am picturing A, B or C. Can we experiment with code, okay give me two or three days and then it was coming back something which was a good interpretation of my idea but I was changing my mind and the idea was not working and we wanted to put hands immediately on the the byproduct to see how could change it on the go. But of course, programmers are not infinitely available for you unless they are not part of your team. And I couldn't find anyone. I didn't have the money to hire someone full-time to experiment.
Then then when suddenly COVID hit and we could not meet or talk for a number of months, I said no, I this is burning inside of me. I need to do something about it. And I said it cannot be that I don't understand anything about coding ; I will buy some courses online. And so I spent some very little money with UDMI, Coursera or other YouTube videos for free to try many different things and Python and Django front end systems, Gatsby front end, uh, NodeJS versus Python libraries, the initial machine learning and at the beginning I did not understand anything.
Fast forward of a number of years, I can code. I love coding. I would say that I'm not absolutely a senior expert master engineer, but I do front back end. I can now code with some AI help. We have an AI system ourselves. And so there's space for everyone if you really want to do it to become fluent in coding. But the point there is that I really tried to get some external help.
But time is of essence. And while we don't have a story of having an idea going to market in three months and making it big which probably these situations are rarer or more rare than expected. They are more narratives and stories than what really happens. But really we have been using unproductively 8 to 9 months trying to work with external experts to build this and we got to nowhere. lack of communication, lack of skills on both sides, my lack of good expression of what I wanted. And so I decided to do it myself; It was an
investment of one and a half years of coding every day.
I remember that up to one year and a half, I couldn't really do anything on my side. And so I was really depressed. I thought I was wasting my time. But in fact, it was absolutely needed. And again while we finished the MBA with Bastian with no iPhone, there was no ChatGPT, no Gemini Pro, no Claude in 2020 as now we'll talk about it if you wish, but I think it is significantly easier to start. Uh but I I think I had to go through that pain to uh push myself to learn coding and we are the two of us Bastian and I freer with experimentation now as we can implement in the code what we want every day without depending on anyone. Of course we will need to hire going forward but it was a good choice.”
KIM (sings his praises): “That's what resilience gets you. the fact that it gets you this far for you to have to gone through those two years of pain of learning to code on your own. I can't imagine how hard it must have been. But kudos to you guys who have made this alive.”
Bastien (his turn now): “Yes. But it's true. But I I remember now that Nico mentioned it like uh I I if I'm right, I opened my Facebook account during the MBA because like people were -- we were using Facebook to set up meetings.”
NICO (added & chuckled): “It was the new big thing!!!”
Bastien (Carries on): “It was the new big thing was Facebook and uh uh basically Blackberry was like the top notch you know cell phone at that time. And uh I basically I had invested all my money in the in the MBA. I had just like a future phone. So I was like sending text message and I had like a terrible phone. This is what I remember.
So, and like even like almost I mean we we had internet at home but we were just, we like for someone who I mean who was born in the in the 2000s like it looks like stone age. It's like-”-
[GOSSIPS STARTS]
Kim (Added): “Yeah, do we see that transition to what it is now?”
Bastien (Carries on): “Yeah. Exactly. Exactly. Just like we we, we had like yeah we we did our and I mean the first link we had I mean as an intake and in the school was really uh Facebook.
It's completely changed basically the way we were connecting in interacting because the school had a system but it was terrible like a email system was not even working anywhere and like the they had like a student system for trying to organize social life and whatsoever but nobody was using it because it was so poor. uh and in the end Facebook like changed our life. So technology was already there uh for for for social first social network and then of course you know the story I mean but it was just 15 years ago just crazy to think that it's it was basically yesterday!”
Kim: And compared to Facebook friendships I'm sure this would stand the test of time.
Bastien: Yeah for sure!
Kim: it's different because Facebook there's so much of screens here is like this real time.
Bastien: But I think we I mean we are no longer using it anyway. I mean like it completely all the social life from from the for intake was in Facebook at the beginning for a couple of years but but now it's completely stopped. I mean there's no no more use of Facebook.
Kim: I mean Instagram is new Facebook now…
Bastien: Uh but we use WhatsApp I mean we have a big WhatsApp group of 200 names or 200 people and so that's what we use. Yeah mostly. Yeah.
[GOSSIPS ENDS & BACK TO TOPIC]
Kim: "Okay. So I mean the reason so essentially what you're trying to say is execution matters more than anything else. It's always about more fast break things right...So you guys come from institutional regal, compliance, risk management, audit trails. The question I have is this:
"Does that now, now that you've gone through that regal for 15 years, you chose your own path; Do you think that the institutional background makes you more cautious or are you actually faster because you already know you've already got your foundations right?
How does that DNA shape how you build?"
Bastien (Responds first): Hmmm, that's a good question I would say because I mean I think we we we have been talking a lot about it about this like with Nico because like coming from I mean uh I work for several different companies.
But let's say the my my culture at work comes primarily from from Capgemini which is like an IT consulting group, like Accenture and which and there were I think 350,000 employees in my company. So which is like a massive uh you know multinationals, where you have a lot of procedures you have you know everything is organized and things like this. And and and I think that that impact of course uh you know we we I think at least on my side and the same thing in terms of work.
I mean we we learned how to work meaning how to organize how to to look at the problem show to uh sort the data organize the idea uh structure the answer and then deliver but sometimes It's too much sometimes and we we we are like thinking too much and whereas we should I mean the go into action and like do the test and learn model which is essential of course for a startup where you don't have you know a lot you don't have time as you said you don't have money and you don't necessarily have a resources to I mean to to do the work uh at 100% as you think it should be done.
So of course it's it's it's big part of adjustments. uh I hope and I think we are we are getting better you know being more agile being more trying things test things and going into straight okay we have an idea let's try it uh but I think that's one of the challenges we are really facing is this one as we are both coming from uh from from, from, from big structures uh and but of course we learn also many things so I think also and as Nico mentions we we have a lot of experience so
Because of that I think also we are we are uh saving some times. We are also maybe better you know uh understanding on uh you know what is uh important what is not. I mean should we uh invest some time on this or not and uh maybe also in risk management I think where basically uh you know of course you ask yourself is there risk to do it or not? Can I make a try? Can I I mean dangerous or not. we need to find the right you know tipping point and the right triggering point.
I think maybe the last point for me would be also the um the networking dimensions uh because like uh of course when you have limited resources uh like in in a start-up uh you need of course to to develop a lot your connections uh you need to to to make as much noise as you can with I mean the money you have or or basically the resources you have. Um so I I think really um something which um I see I I can appreciate is really the networking part because I I think that's uh um we both know how to leverage and use use the network also to to try to to pitch our product to expand ourselves to to expand the brand uh and to to make people test our product.
But it's true that um I mean I think in like in every situation pros and cons I would say because like um uh if you know to bootstrap and if you have been used to bootstrap for for years uh for sure you're going to be I mean at least when you at you know early stage it's it's it's a it's a big advantage.
So we had I think also to learn or to to be more risk and try to think about, I mean our natural you know behavior which is okay let's do a file let's think let's make research and you know and and then we'll see…”
Kim: “Okay, thank you”
Nico (Chimed in): Yes! on my side I would say a couple of things institutional focus absolutely fundamental as we are in a very specific niche I will talk about it in 30 seconds but about what you mentioned at the beginning risk management audits processes inside big organizations with the benefit of the hindsight absolutely unnecessary in a startup, because it is much more important to try things.
In big companies, you have a theory, you influence others in committees and you very slowly try things through some uh branches of your company. Feedback is very slow. it comes back but there's a consolidated way of doing things and things tend to work anyway. Uh maybe in a way that is not fully understood by management.
When you are new, you don't have a brand, you don't have clients, you don't have anything and you just need to go out and test it before thinking if it works or not.
Okay. So there's a part that comes with experience which is uh putting a break on us and on start-ups but of course we have evolved and we have been learning about it but it's something that yes we observed not useful; What is very useful is coming with some experience because you will have some knowledge about certain fields, knowledge of certain people and you will eventually understand the nuances of your sector, product or industry.
Okay, this is in general specifically we are doing autonomous investing software driven of an investment strategy. So now we are a product company becoming something else. We'll discuss later about the point you were making earlier. But in the first product we are pitching to a professional and personal contacts a solution that should help them managing their own money or savings. And so we go and touch something very delicate in a sector that has been full of frauds, full of wrong projects. And so without an
institutional profile, without the knowledge of the marketplace meaning financial markets that we bring and without those 15 or 20 years of experience in the industry, the trust just to listen to us from friends, high net worth individuals, family offices, private bankers, discretionary managers wouldn't exist and so we would not have a business.
Okay. So in our specific case the institutional source of knowledge absolutely fundamental. Other elements of seniority and expertise some useful some are not.”
Kim: “No that was good. It's it's it it kind of cements found cements what you said earlier. So it it all makes sense. You need to have that especially like you said fintech is so much compliance focused. You need to have some form of trust that you guys already build up through your careers to get that trust from people because you're dealing with people's money. It's not easy to get with that trust. It's something like you said very delicate that you thread carefully.
So, okay, let's get into what you're actually solving. That's a fun part, right?
So, every investor wants one thing. Own the right assets at the right time. Generate reliable returns. Sleep at night. That's it. Simple. The industry changed 15 years ago when you were then in grad school whereby the market accommodated and then the 60-40 became gospel. Everyone accepted it except COVID happened and then 2021 and then 2022 and then now we have 2026...I guess.
Um so the nightmare scenario that nobody talks about is a positive stock bond correlation. the entire premise of diversification just stopped working. So, Nico, you have worked inside global markets long enough to see this play out. I want to hear not the theory, but the moment. When do you realize that 60-40 wasn't just outdated, it was a comfortable lie that we all decided to agree to tell ourselves?
Because the industry isn't lazy. It's not dumb. There's a reason they say don't time the market. Market timing is always hard and for 50 years no one had that technology to actually do it systematically at scale for retail investors especially now that you look at it the war you know no one expected the sudden war and then now it's a shame that you can't be here live that has kind of impacted this whole podcast session.
So give me one specific market moment if you can let me ask you this differently a day a week a month in your testing where traditional advice you have lost money and cardigio what protected it. Numbers, context, what actually happened?"
Nico: Oh, fantastic. I know. Incredibly well said. Quick answer to this is very complex. We have 78 years of experience in modern financial markets. Products have evolved. But it is incredibly difficult to read financial conditions in real time unless you have events like the war in Iran now and you see energy prices going down and everything else going uh sorry, energy prices up and everything else going down. One moment in time for you Spring of 2022.
Crazy inflation going higher everywhere after a build up in the supply shock postcoid.
Inflation is good for stocks up to when it disrupts consumption. Historically accepted inflation in developed economies is up to 5% year on year rising prices.
In December 2021, the US were experiencing over 7% of inflation for the first time or above 6 and a half. So when inflation is so high, it is
expected to destroy demand. So companies would sell less. So earnings would go down. So the stock market started plunging, a 60-40 portfolio where 60% is stocks and 40% is bonds, you would expect the bond side to protect you. When there's fear, you take money away from stocks, you put it in fixed income coupons, but fixed income suffers with inflation because inflation going higher, it erodes your yield benefit of fixed income and so also bonds started tanking. So you had the two parts of the 60-40 going down, no diversification benefit and no one was buying gold now popular, not many were buying a commodities and there are some nuances like floating rate notes which are very short duration fixed income instruments with a coupon that resets every three months and so not exposed to any inflation problem.
Cartesio in that time in back testing but we added these tools on purpose went completely out of stocks reduced the exposure to long duration government bonds with gold at the beginning of 2022, more commodity instruments in the second half and floating rate note cash proxies it came out with a positive return when the average portfolio would have done between minus 10 and minus 20 and is very
early in the Iran war.
We have seen markets doing very poorly for two days. But if the energy prices continue to rise, gold is not doing well, stocks are
down, bonds are down, it will be a very similar environment. In fact, the Cartesio is reshuffling in these days going towards floating rate notes and very short duration government bonds.
Private bankers don't have this flexibility managing their portfolios by design. They need to ask for the approval of the end customer to move the portfolios. It's also very difficult to know when to move. Uh European stocks go down. US stocks don't go so much. Then Asia is closed for one or two days and then it opens and the stock indices tank. Where do you go and when is incredibly complex now we have the technology.
Bastien (Interjects): "Something I want to highlight is on the why also I joined this project and why also I I believe in in Cartesio and what we're doing. So of course I I I don't have the technical and background of Nico and his knowledge in markets but I had the chance of course to to be I mean one of the customer of of this you know private bankers and private bank, and um at the time basically we
Nico and I started to talk about this uh I was very disappointed by the basically my banker and so it I I was selling directly on my portfolio of of this high inflation period, the impact on of my on my stocks on my bonds and after 10 year of investing uh following uh the advice of my of my personal advisor I had like almost zero performance after 10 years uh which mean I lost you know like 20% because you know if you have zero performance after 10 years media was like five plus five I mean I mean almost zero it mean that with inflation you lost I mean a lot of money and that's why also I I mean the uh the vision of Nico and and of this project really you know resonates with with me when when we talk because I was very disappointed by the work done uh by by I would say by my advisor and because he he had this typical uh behavior that you you mention uh which is basically you you you stay with a static portfolio. I make the I make the pitch, I make the sale uh to the client and then I try to sleep on it for for 10 years, because as the market is always going up in the end.
It will make money but it's not necessarily the case and this is uh also why I strongly believe uh in what we do and on the uh you know the impact that you know this autonomous uh you know investment system will will give to to our clients."
Kim: “Interesting. So back by again your experience that didn't help and I guess this would definitely speak to retail investors who actually go through what you did as well. Right. Essentially correct.
So um Nicole you built again 253,000 lines of code into Cartesio. This is crazy. I mean the way you did it was wow mindboggling.
But you didn't build a neuronet network. You didn't build a deep learning model. You didn't follow the hype. Basically, you're a hipster in your own way. You chose a rule-based expert system. And in 2024, when everyone's talking about AI agents and black boxes and the machine knows best, you chose to be explainable.
You guys both chose transparency over optimization as far as I understand what kind of how it works. So that's pretty contrary to what the market is already moving towards.
So why not just let the neural network do its thing because here's my theory. You made this choice because you guys came from institutions and institutions demand auditability, right? So they want to know why your decision was made, not just that it work. And you thought correctly that retail investors deserve the same respect. But I'm just going to push a bit a little. What's the thing that Cartesio can't do that a blackbox neuronet could do? What's the trade-off that you're making?”
Nico: “Thank you very much. This is such a peculiar interesting question but also a challenge for us to explain and I'm glad we will have the chance here. Hopefully I can be clear. Uh
I will start from the premise that we are all now using AI which is a sub sector of AI which are large language models through ChatGPT, Claude, Gemini Pro and similar where we all go to an interface that replicates human language and replies to us with some wise advice or answers.
The technology was born with a paper from Google in 2017 called attention is what you need or all you need. There was a movement going from neural networks into transformers. Transformers is the technology behind the LLMs. What I perceive without being inside the engineering team of Chad GPT is that LLMs are an incredible statistical machines. I absolutely don't believe we are close to AGI but we found a way as humans to replicate the human knowledge almost in a perfect way.
Okay, if we put hallucinations on a side but here's the point. Statistics can help us humans to identify 20 million pictures and recognize what is a window from a door because we as humans we would all be able to recognize a window or a door.
Whatever the shape and form. What we as humans are not capable to do by definition if to say if Bastian or Nico are nice or fun people based on this podcast because it's subjective.
And so how can we define if gold is cheap or expensive? Based on what normally is demand and supply gold in 2025 has been incredibly successful because of demand. supply was kept and there was if you wish a loosening of trust in US policy making around the world and
so many people left the US dollar to go into gold.
Can someone explain to me why gold has lost 4% in the two days of the Iran war? Because it doesn't make sense at all. It shouldn't is a hedge. It is both as a hedge and it went down. So statistics telling me that gold normally goes up in the early days of a war don't work in this case. Why? Because we have seen a very peculiar phenomenon of retail money jumping into gold one year after very strong performance. Many people piled up into golds above $5,000 per ounce in the early part of 2026. And in the panic selling of the early days
of the war, they also panic sold gold.
Statistics don't help here. So I really tried to use all the algorithms of Google TensorFlow and other software to test machine learning on markets and they cannot provide stable answers to all market scenarios in terms of mixing your portfolio.
Which is why I also struggle to see right now how LLMs would be capable to have a good opinion every day on financial markets. High level yes but to act every day no. So with this premise we tried to put into Cartesio a web of human knowledge which looks at everything financial markets, macroeconomic variables like energy prices, geopolitical noise which can be captured through daily measures of
newspaper articles.
But also where we are in terms of economic growth where we are in terms of inflation where we are in terms of Vicks and economic sentiment to try to read context. text. Okay, that could probably be simulated with LLMs and statistics. But then we had to add fear on the market. We had to add the momentum trends of different prices. And when you put everything together, we are talking about hundreds of different variables. You need an opinionated reading of the context.
So we are not asking Cartesio to do the training as a neural network to let Cartesio understand and then give real time data to provide an answer. We expert humans have put all the human knowledge inside the system as a training plus we let Cartesio see the daily flows the
interpretation of the human historical knowledge of all the parameters judging when correlation is causation and when is not plus the reading of real time events.
We let Cartesio process everything of this in real time to give us the answer on where to invest today. But a good chunk of the training is done by humans who can recognize causation and correlation. Machines can't.
Okay. So this is where we tried to put together this quirky opinionated system because for example you could not see yesterday everything tanking energy prices is up but gold is down bonds are down and inflation is not showing up yet.
There's a fear of inflation emerging from probably the straight of our moods being blocked for longer days. There's no data giving us that fear, if not the fear on the stock market. So you need to complement many different feeds to let the system really understand where we are and I think for a number of years we will have an edge.”
Kim: “Again building on patience and your resilience to build that because why I'm trying to get that is that it's beyond just the three main signals. It's beyond global macro. It's beyond momentum. It's beyond sentiment as well. There's so many permutations like you said that come into play when it comes to markets.
So because I noticed something in your materials, you guys talked about modeled human expertise. You say the system is capable of checking small correlations and interactions between variables on a daily basis like you mentioned earlier, right?
So that's the algorithm, but you're also programmed to focus on the sudden events that have no immediate meaning until their temperature, temperature changes...that's a human part, is that I mean I know you can't say about--you can't talk about your secret source but given that you've taught yourself to code during the pandemic what's one role in cardio that you initially resisted adding and now that you basically what was that one thing that you initially resisted adding and now you followed your gut and now it's like you know it's actually it all made sense come was there any moment such as this any inflection point that but now it's like oh yeah it actually does make sense!?”
Nico: "Yes, we are trying we are trying you know to work in partnership with AI and if we all want to be more honest AI is not coming out of nothing. It has been coded by humans. You delegate to a system the training to see things but the training has been implemented by humans and LLM models become better every day because humans learn and improve the algorithm. So humans are behind these gigantic statistical machines.
It is just more fascinating to say there is magic and there will be the end of humans soon because these magic box has come but humans are coding that and so we are just doing the same instead of closing this as as a black box.
We are trying to convince on top of the trust that we have as institutional senior if you want executives from our past. We want to say look this is the way we interpret variables and this is the way the machine is connecting the dots. In fact we are or we just published Cartesio can be accessed through our website right now is the thing of the past week.
Okay. So that transparency is important. We got to that point through some errors. One example that you ask me, we add an algorithm to pull out of market shocks instantly or after one day like yesterday and that mechanism depends not on macro data will arrive later not only on market valuations which tend to be wobbly daily but on sentiment when sentiment indicators tank altogether there's a number RSI indices the VIX you can look at cyclical versus defensive equity sector performance.
You can look at different credit spreads in corporate bonds in different regions between higher quality and lower quality companies. You can put all of this together. When they snap over one day, you know that there's a problem. And the algorithm was coded to protect first.
The idea of riding uh market conditions generating reliable performance is first not to lose when things tank and then accelerate as soon
as you can. The algorithm last year in our Bastian & Niko and teams money experiment was coded not to reinvest in stocks with high VIX. Okay. When the US president declared tariffs in liberation day last year in March, we have seen the S&P going down up to minus 20%. Cartesio did minus three in those days. Okay, fantastic. But then there was a tweet of the US president saying, "Hey, feel free to buy what you want." How do you code this in an LLM?" We also didn't code a president tweet into Cartesio.
And so markets started recovering big time despite the VIX in the US being significantly higher than the 2022 gate points of normal conditions. We missed that in our experiment. We missed the April rally because it was unsafe to trade in normal 70 years longer conditions. But markets traded anyway. And so there we tried to become more pragmatic saying yes we will need a number of rules coming from human experiences.
But if Cartesio will detect that the rules are not necessarily working or congruent we will patch the algorithm with a number of extra solutions.
In this case, when VIX is high but fading, we go back with partial isk on into the market. Okay? Which is also how LLMs have improved from the first versions to the current versions because humans have detected problems and they put patches in their algorithms. It's just less transparent the process."