AI and Financial Index Innovation

Stefan Wagner: 00:43 

Good morning. I'm here with Christian Kronseder. He's the CEO of AllIndex, and we're going to talk a lot about indexing, as you can tell from the name. But first of all, thank you very much, Christian, to be on the call. Now, I always like to ask people in the first stop, you know, what actually inspired you, the founding of AllIndex and how, you know, you started with an idea and how did actually the vision over time had to adapt to the market or your application in a sense changed?

‍

Christian Kronseder: 01:12 

Well, hi, Stefan. Thanks for having me. So, yeah, what inspired the founding of all Linux.com? So, like you, I have a background in structural products. And all the guys that are involved in AllIndex, they all have a structural products background, because we felt that there is not enough technology to help structural products desks to create baskets and instances quick enough. There's this lag, and I'm sure you have experienced this as well. So you go, you have this great That's the one idea. You go to an index provider, like it takes six weeks to get an answer, although you could issue the product within the day. So that was some sort of the initial core where we said we need technology around it.

‍

Stefan Wagner: 01:55 

And that's where all the innovation right now in structured products is, is in the underlying, not in the payout. 

‍

Christian Kronseder: 02:01 

Exactly, exactly. So, we were a bit too early, because we did this like four or five years ago, where it was like, oh, I have my spreadsheet, but our spreadsheet has a few issues.

‍

Stefan Wagner: 02:10 

I always say, show me a spreadsheet without a mistake in it.

‍

Christian Kronseder: 02:15 

There is none. So, that's it. Now, what has changed in the meantime? So, one thing is we learned an awful lot about the buy side in the meantime. So, asset management, passive investing, we know what passive investing is, of course, and fund management. And we suddenly realized that this problem is not only a problem for the investment banks or sales side to create a product quickly, but there's some sort of general problem of how do I bring an idea into a portfolio. 

So, we are basically a laboratory of creating baskets, indices, portfolios, whatever you like, we don't care, where you can have, oh, here's my idea, and it's what I would want to try, or here's a model, which I would like to change, or here's a set of ETFs, which I'd like to build into a certain set of exposures for my clients. So, this is how it changed from like our single stock basket, very concentrated to actually you can try a lot of things. And I think the best way to describe is like from Very focused on baskets now to laboratory for R&D. And then we hand it over to whatever you like, PMS, OMS, or we do it for you.

‍

Stefan Wagner: 03:29 

Fantastic. Now, you told good about the company, but what actually drives you and can you tell our listeners a little bit something about you that maybe most people don't know?

‍

Christian Kronseder: 03:41 

Well, I think what really drives me is, and I really get a kick out of this, is solving hard problems. Honestly, this is like, I have a solution. I really feel very good about it. This is, I don't know, if you ski, the feeling is similar to going through some sort of a deep powder snow type of line. Exactly the same feeling for me.

‍

Stefan Wagner: 04:03 

I do ski, yes. I know the feeling.

‍

Christian Kronseder: 04:07 

So that's what it's like. Our problems are like skiing.

‍

Stefan Wagner: 04:09 

Okay. So, you mentioned, you know, you see all indexes, what you offer sort of as a laboratory in which people can conduct R&D. So, is all index then actually also still an index provider? Because the name index is obviously in the name.

‍

Christian Kronseder: 04:25 

So, we are not. And we very early in our journey realized that this is actually not a really smart idea. A, because there's tons of index providers out there. the big brands that everybody knows, and the smaller brands that are super happy in their niche, and in wherever they want to compete. We also realized if you really want to compete effectively against the brands, you need an awful lot of money because it's a brand building exercise. It's not so much a technology exercise, but it's a brand building exercise. And nobody's going to give me like $50 million to build a brand.

‍

Stefan Wagner: 04:58 

Yeah, and nobody got fired ever to invest into an S&P index. Exactly, exactly.

‍

Christian Kronseder: 05:04 

So, this is what we did early on in the phase where we realized, actually, it's a technology play we want to do. And everybody can use that. So, you could give this to one of the big brands, to a smaller brand, and help them to grow certain parts of the business. So, if you want to enter like wealth management, good examples in the States, direct indexing, tax-driven indexing. This is one way how you could address the market with this. You can roll this out to hundreds of thousands of users and it will perform the back test, the changes in the model and everything else. So, we're not, we can, we have the ability, we have the operations team, we have the technology, but that's not our core business. Our core is like R&D.

‍

Stefan Wagner: 05:51 

I see. And what are sort of, let's say, the top three benefits of what you offer? And have they may be changed over time? You found out suddenly our clients are using for something that was unexpected, which we didn't have in mind when we built the whole thing?

‍

Christian Kronseder: 06:06 

Yes, there's a couple of things that I can give you the anecdotal evidence. But the really three top benefits on this are it's fast and intuitive. So, you don't have to be a quant in order to build industries when you use our tool. And if you want to be a quant, we can help you too. And what we found over the last, let's say, 12 to 18 months is that the combination of investing and AI, AI in a sense of research support, is something that a lot of people value. And we had this really, really interesting experience that we were building. Well, we have a search engine that helps you in building thematic indices really quick. And one of our clients didn't use it for listed stocks, he looked for startups. It's like, hey, I'm looking for quantum computing startups.

‍

Stefan Wagner: 06:58 

So, he went to the private market.

‍

Christian Kronseder: 07:01 

Absolutely. And we were totally surprised because we did not build specifically for outside of that, but we thought it would be. And I showed him, look what I did. I'd created a research report on quantum, like quantum startup cards in Germany. And it contained everything, shareholders, SWOT analysis, whatever was in there. I was like, oh, that's interesting. So, we keep this in mind. So, we're expanding on this at the moment. But coming back to top three, things are, it's intuitive, it's fast, and it's AI supported in the meantime.

‍

Stefan Wagner: 07:30 

I mean, AI seem to be, you know, people are using it now in the index or the picking of underlying in the basket, whatever the difference is between an index and a basket. And it's sort of this active versus passive. The lines are getting very blurred. Is that a challenge? Because how do you position yourself? Maybe everybody wants to put you in a bucket. You do active management, we do passive management, investment.

‍

Christian Kronseder: 07:56 

So, we are closer to the passive market. At least where we are. So, we focus on the wealth segment, so we're not doing any institutional business. This whole discussion about, oh, we're going to have totally hyper-customized portfolios for every client does not really materialize unless there's a real reason. And the only reason I've seen so far in the States is tax. But here in Europe, if I have some sort of a certain preference, I'm not going to build my own index or my own thing. So, this is going to go away. So that's why passive investing for us is right now a bit more interesting. Of course, we can do this, but we are really more focused on the passive part.

‍

Stefan Wagner: 08:40 

You already touched on it, but maybe you could go a little more. Who sort of is your ideal customer and user of it?

‍

Christian Kronseder: 08:47 

So, on the sales side, it's the investment banks and structured products. And on the buy side, it comprises of investment advisors in the broader sense. And the third segment where we really, really like to get more traction is the index providers.

‍

Stefan Wagner: 09:06 

So, you see, okay, that's interesting. So, you don't see the index provider as your competition, you could help them in a sense. I think that's something that's good to clarify here. And can you sort of give maybe four, maybe how these people then use it, the IB and all structured products, people in the investment banks and the investment advisor, and maybe also the index, the traditional call them index companies.

‍

Christian Kronseder: 09:35 

The discussion with the structural products providers is much easier because investment banks are more tax-savvy. So, if they see an advantage in a platform, they always take a close look at it and then they might adopt it because it gives them more business opportunity. So, investment banking, in that sense, at least structural product area where you and I come from… It helps when you speak the same language.

‍

Stefan Wagner: 09:57 

Exactly.

‍

Christian Kronseder: 09:58 

So that's one thing. So, I talk and walk like somebody out of this area, but it's also it's a competitive area. So, innovation, innovation is one of the key things you have to have. We're finding this on the financial advisor side, we see that a lot of them use these customization capabilities or, or research capabilities. in order to justify their existence. And I don't mean it's derogatory at all. But if they face the problem of, like, if I give you iShares or Vanguard or Amundi ETFs, the client will ask, so what is your value at here? You charge me 80 basis points for a service where I can go on the next website and buy it myself. 

So, they have to have this capability to really innovate in the way they service the clients. This can be, amongst other things, this can be in the way, hey, here's my idea, or I can build you the Euro Stoxx 50 in a different way that you have less risk, so if you're quite sophisticated, or I can build the S&P 500 with any tech stock or any oil and gas stock.

‍

Stefan Wagner: 11:02 

Or already the stocks that you overexpose because you might be a person who actually got through their job a lot of stocks in one single company. I think all the digital options give so much more. You can personalize the wealth management. Absolutely.

‍

Christian Kronseder: 11:18 

Now, for the index providers, that's the last item. Index providers go through a slow but steady change from product-centric to client-centric. And this is if you're not product centric anymore, then you need to have some way how you can communicate with a client. So, this is where platforms like ours and others come in, where you suddenly can roll out to the client’s stuff that you don't need it to. 

Euro Stoxx 50 is Euro Stoxx 50, S&P 500 is S&P 500. But if you suddenly have a client, say, I'd like to have a bespoke version of that, or I'd like to have a totally different version of that, you suddenly find yourself in a squeeze because you don't have enough people to do the back testing and you don't have any way. 

But if this becomes more and more the norm, then you need to have a platform. And this is where the discussion suddenly starts. And you have interesting discussions like, oh, I'm index provider X, my rulebook is so complex, you will never ever be able to put it into software. And like, okay, well, it's a nice challenge to have, but that's a bit of a weird argument not to use those platforms.

‍

Stefan Wagner: 12:23 

I think if you look particularly at the very long-term and traditional index providers, they stack technology on top of each other. And there's a lot of effort to maintain this. So, I understand their challenges. And so, you know, I think sometimes they need to look at people like you and bring that capability in-house again or use your software for it. I agree.

‍

Christian Kronseder: 12:45 

That's one of the bets we're making. Absolutely.

‍

Stefan Wagner: 12:48 

So, is that sort of what you think is the biggest opportunity when you look at the market for you?

‍

Christian Kronseder: 12:53 

The biggest opportunity for us is, and I'm just on a really interesting call with a company that characterizes their clients using behavioral science. So, the capabilities we have that we can create based on certain information portfolios, I think that's for us the biggest opportunity. So, what we're discussing with them is like, so I have these buckets of archetypes of investors, but I can't create the portfolio for them. 

So, we can take this and either use a little bit of AI magic or not, depending on how they want to have this, and then create portfolios that are part of these behavioral traits these clients show. So, the ability that we can do this R&D thing is actually bigger than we thought in the beginning, because you have multiple entry points.

‍

Stefan Wagner: 13:46 

Would it be possible to give an example of how somebody behaves and then how that doesn't translate in what you want to do?

‍

Christian Kronseder: 13:55 

Yeah, so one of the archetypes I have, so you client, you are somebody who can still make more risk and like growth stocks. but you're not into blue-chip growth stocks, like the MAC-7 or something. So, this means this person is more interested in small and mid-sized entities, SMEs, small-cap, mid-cap. So, we can just take out of our data lake we're having. So, we propose two or three alternatives in the portfolio. We say, okay, so we'd like to have exposure to the US, exposure to Europe, if we do everything equal weighted, this is this is one of the problems you have, you get a back test, you get a fact sheet. Within like 20 seconds, you have everything on the on the table where you can show this to your client, say, hey, this is how it looks like, what do you think? So, you can have this workflow discussion or you're in the flow of discussion with a client who says, okay, now I understand what you want. These are your traits; this is the portfolio that might fit your traits.

‍

Stefan Wagner: 14:53 

Now, you've seen probably hundreds of people trying to design their basket, their strategy, but what would be your advice when designing something like a rule-based investment strategy?

‍

Christian Kronseder: 15:08 

Don't over-fit and don't over-optimize. It's very, very seductive and attractive to optimize your portfolio to the level, oh, I have a massive, massive outperformance vis-a-vis any benchmark I've ever seen. But don't forget, you're optimizing against the regime that's in the past, and you have no idea what's coming in the future. So, leave a bit of slack. for your trading strategy or your index, because otherwise it's going to fail. That's one of the key components. And if you do not really understand why you're doing this, then please don't do it. That's the other thing we've seen. So, I have optimized the CAPM and it's like the variance is like crazy and everything. But then if you run this against real data in the wild, then mostly it fails.

‍

Stefan Wagner: 15:54 

If you don't understand the reason why it works or why you're looking for it. Yeah.

‍

Christian Kronseder: 15:59 

The other thing that we found, and that's a bit outside of the index business, is that when we're talking to people that are not really into it, one of the questions that really goes on is, is indexing really profitable? And the answer we always have is like, if you have a Bloomberg terminal, just check out NDI and their profit margin. Then we don't have to discuss anymore. So that it is a niche, it really is a niche already in a niche. So, it's difficult to explain to somebody outside of this area to really why this is sexy. So that's the other thing that a lot of like we see that as a misunderstanding. So that comes over to build but also to comprehend.

‍

Stefan Wagner: 16:43 

Yeah, makes a lot of sense. People more and more, you know, AI is part of that, what helps is dealing with more and more data. But in your experience, does more actually data improve the result? Or how do you sort of best get the best possible result?

‍

Christian Kronseder: 17:03 

So, when building an index, you have to have a plan. I think more data as such don't really help you. So, if you are not really clear on what do you want to achieve with it, so is it a certain exposure you'd like to create? Is it a certain improvement of other indices you'd like to create? And I'll give you an example. I rarely see an index that is weighted based on per capita GDP per country. But this makes a lot of sense because you could assume a country that has a high per capita GDP should also have super profitable companies. 

And if you look at the back, it's like, it's crazy. It's crazy what comes out of that. Switzerland versus Germany, for instance. Yeah, there's a consistent outperformance in certain stocks here. So then once you know what to do, then more data might be helpful. Absolutely. And then it also depends on what you want to bring in. So, you can, of course, everything is somehow connected to each other. But does it make sense to have this entire network of dependencies? Or you say, well, there are certain risk factors I'm comfortable with. I'm sure I always use certain ones, and others I disregard. That then comes down to how much to understand about markets.

‍

Stefan Wagner: 18:15 

It matters the data you use to train an AI model. 

‍

Christian Kronseder: 18:20 

Exactly. When we talk about AI in general, there is this idea that AI is going to take over trading. I think you as a former trader probably could disagree more. Because the ability to anticipate or understand when there is a regime change in the market is something you cannot currently train into an AI because there's not too many regime changes that are there to train the AI properly.

‍

Stefan Wagner: 18:47 

Yeah, I think the challenge here is often that, you know, the investor's memory is probably 18 months long at best. Yeah, while in order to be statistically significant, we have to go significantly further back.

‍

Christian Kronseder: 18:59 

Exactly, exactly. This doesn't exist. So, in that case, more data doesn't help you here.

‍

Stefan Wagner: 19:05 

And then one of the data points is more and more obviously being also through regulations coming through is ESG, the environmental social governance. But does it actually have an effect? Is it meaningful in the investment, if you can comment on it?

‍

Christian Kronseder: 19:20 

So, we have an open platform. So, if you have your own ESG framework or an ESG provider, we can put it as a platform and they can use it as we would like to use it. Now, the challenge I see with ESG, then there's many folds. So one is, if you look at the number of startups in the ESG space, it clearly tells you there's no framework that is usable for everyone. It also clearly tells you that my E, the Christian Kranzler E is different from Stefan's E and different from Stefan's S and different from my G. which makes it very difficult for the end investor to really differentiate, like, what am I actually buying if I buy an ESG overlay on CAC 40 or an SMI here in Switzerland. 

So, from that point of view, I find it very problematic, ESG investing. I totally subscribe to the idea. We all have to become more sustainable. That's some sort of prerogative of everybody who's in business. But I am very skeptical that this is going to be really great, because what we've seen in the discussion with our buy-side prospects and clients is that they tell us, well, at the end of the day, if my ESG portfolio doesn't deliver, they sell it and they want to have a return. Yeah. So, there's a discrepancy and a gap here. So, I think in terms of implementing it, I'm very skeptical. I'm very skeptical.

‍

Stefan Wagner: 20:51 

We talked quite a bit already about AI. Is there something else, how you basically leverage it? And maybe a comment on how you think it will change the index business?

‍

Christian Kronseder: 21:02 

I'm not sure whether you know this, but I used to be a guest professor for data science for a couple of years at a local University of Applied Sciences. So, I had some sort of really deep dive into AI, and also, I dabbled in AI since 2010. It was called data mining at that point. I think where it comes in indexing is not so much in building indices, because deep learning is a black box, and you can't sell a black box product into a regulated market. This doesn't work. But where you can use it in indexing is either by helping to select and build a rationale for an index. So, it's as we have with our specialized search engine, so for thematic investing. But I think the biggest leverage you will get in AI and indexing is actually in operations. Because there are a lot of jobs, and my prime example is managing co-productions. 

That's some sort of a super tedious job, but you have to be really into it, you have to be really focused. And these are, in general, those type of activities where AI is really good at. So, you have to have a certain, and you can't do this for eight hours. So, this is a support tool to help you manage tedious tasks that require a certain mental capacity. but are very routine. So that is a clear one is corporate actions, I repeat myself. Another one is data cleansing. And the third one maybe is creating fact sheets with a narrative around it. So, we take different sources of structured and unstructured data and create a narrative on why is this particular index the best index you've ever seen on this planet.

‍

Stefan Wagner: 22:48 

Fantastic. No, Christian, extremely insightful. Thank you. Thank you very much. There are always three questions I like to ask everybody right at the end. Start maybe a little bit sort of philosophical, you know, what is sort of your definition of success?

‍

Christian Kronseder: 23:08 

Closing my laptop and be satisfied in content at the end of the day.

‍

Stefan Wagner: 23:14 

No, that's a very succinct answer. I like it. And if once you close your laptop, or maybe you still need to leave your laptop, but sort of, I always like to ask what is on the top of your current music playlist?

‍

Christian Kronseder: 23:29 

I'm a little bit hard pressed because I'm biased here. So, of course, my son's music. Yeah. But I'm very diverse. So currently, I listen to Baroque operas, but this might change next week to another playlist on ambient music. So, but this week, it's opera.

‍

Stefan Wagner: 23:48 

Now, last question in here, and I'll let you go. You know, if people like what they heard, and they would like to get in contact with you, how do they reach you?

‍

Christian Kronseder: 23:59 

By email, very easy. It's ck at all index.com. We kept it short or on LinkedIn, and happy to have a discussion anytime of the day.

‍

Stefan Wagner: 24:10 

Thank you, thank you very much, Christian.

‍

Christian Kronseder: 24:12 

Well, you're welcome, Stefan. As usual, a pleasure talking with you.

‍