Dan: Hello and welcome to QuantBeats. I’m Dan Hubscher, managing director and founder of Changing Market Strategies and your resource for all things quant. My co-host is Quantpedia CEO, and head of research, Radovan Vojtko. Rado, please say hello.

Radovan: Hello, hello everyone.

Dan: Rado is a former portfolio manager of over 300 million euros and several quantitative funds. You can check out his full background and mine on the QuantBeats website, or from episode one of this video podcast, which you can see on the QuantBeats YouTube channel by clicking the Q logo in the corner of your screen. And today we welcome our two guests from Quantmatix, CEO, Liam Boggan. Liam, please say hello.

Liam: Hello everybody. It’s a pleasure to be here

Dan: And head of product, John Mulligan.

John: Hi everyone. Nice to, uh, nice to meet you all.

Dan: Now just before we get to the full intros and start to find out how a former fundamental equity researcher and a Delta One trader set out to build a quantitative filter for a conviction. I need to show you some quick disclosures that are on the screen now (watch the full episode for the disclosure). since I’m a registered representative, will be giving you ways to contact me at the end of the episode and you can contact me with any questions.

Now let’s get to it and dive deep into the world of quantitative finance. So seasoned investors, aspiring quants or anyone interested in the financial markets can gain invaluable insights into complex quantitative strategies, increased knowledge of market dynamics, and a peak into the future of algorithmic trading.

So to kick us off, I’m gonna ask Liam and then John, to give us these three things in just one minute each. And then Rado will take it from there. So Liam, if you could please kick us off.

Liam: Okay. So who I am, Liam. So I am the founder of Quantmatix.

I’ve been around the markets since God was a boy. I started off my career as a fundamental guy. So I was always a fundamental guy, but I got to work with my best friend from college who was a technical analyst. Right, you know. And we married fundamentals and technicals, and while I was a portfolio manager, we outperformed significantly. So that’s kind of the background of why I’m interested in that. What we do now? Quantmatix is a global multi-asset market timing tool that works like GPS. So we cover thousands of assets across all countries, industries, sectors, single stocks, crypto, fx, commodity, bonds, et cetera. That’s what we do. We analyze those with a huge big data analytics project with ultimately to serve users with actionable ideas in an easy to digest format.

And had we worked together? John. So our third partner who’s not on the podcast, had worked with John previously. John could basically tell us for himself. So John joined our merry band when there was just two of us. He became sort of like employee number one. Now, director, partner, and key member of the team, three months after we first launched the project back in 2021.

John: Yeah, my background is as a Delta One trader working in structured equity finance and prime brokerage at a large South African investment bank. And we were based outta Dublin. I was there for about 14 years. You know, I suppose learned about the financial markets in that time. When that role came to an end for various reasons, the desk got closed down, but I was looking for a different challenge.

I didn’t want to go do the same thing I’d been doing for, you know, the previous 14 years. And I was really interested in two things. I was interested in digital asset space and I was interested in the FinTech space. It was around the COVID time as well. So it was a, you know, a tricky time for movements.

But I was exploring, those two areas and I got chatting to Paul Chu, who’s an ex-colleague of mine. And he introduced me to Liam, and we sat down for a while and we had a conversation. And I really liked what, you know, the guys were doing. I liked the philosophy that Liam had. How he’d come to that philosophy through his career and how he wanted to build something that was a scalable solution for the buy side. Yeah, I suppose, I came in as Liam said, after in month two or three. And I took on a role as the head of product and managing the development of the product out from the base level of the automating the mathematics to ultimately now managing a data research quant team, which consists of three people and our development team.

How do we work together? Well, we do a bit of everything together. You know, I think I’m also the head of marketing. That’s the nature of being in a startup. So we’re very collaborative. We’re in the office every day together. Where we can be. We’re really proud of what we’ve built over the last three years and we have really looking to push and push forward now with some really exciting projects.

And that’s my story.

Dan:Thank you. Very good. So, because you guys both mentioned, let’s set the stage for the audience. Just before I hand over Rado. When you’re talking about product, both of you, are we talking about signals? Are we talking about funds management or are we talking about both?

John: It’s both. I think initially when we set out.

Liam’s vision for the product was to create this scalable solution. To assist the buy side with timing. We were doing that through big data analytics, presenting data at an individual instrument level up to a sector index, global macro level on what was happening in the world at any given time.

And I think that’s where Liam mentions that concept of market GPS. Where are you now and where are we going? That was the concept. It was to build this scalable solution that would act as an overlay and an objective overlay for typically your fundamental portfolio managers and traders who had their existing investment process. I suppose, where that’s developed to is a place where we kept getting asked a lot of questions about when we were trying to get trust in the product. Well, if it’s so good, why aren’t you running money yourself? Or why aren’t you creating strategies from the data that we have? And ultimately that’s kind of what has occurred over time.

And we got an opportunity last year to start looking into the that sort of more systematic strategy space, which has been a really exciting opportunity for us.

Dan: Okay. Thanks for that, John. All right, Rado, can you take us from here?

Rado: I will probably start with the philosophy. And I have a question for Liam.

You said that you were fundamental guy and now you are quant, or at least you manage the quant you have in your company. For me, it’s like interesting journey. So why you move from the fundamental to the quant world and how you did it, and what do you see as an advantages? That’s what is interesting for me.

Liam: I suppose it’s actually very simple.

So at the start of my career, I was a proper fundamental guy and I’ve had some various sort of fundamental jobs in my life, right? So I was a fundamental guy in terms of like being an equity analyst and a portfolio manager. But one thing which struck me very early on in my career is that fundamentally good companies are fundamentally good for a very long time.

But what I learned when I was managing money right at the start of my career is that how you actually outperform is by taking money off the table on your favorite stock when it’s run up too much. And possibly buying a stock that you hate when you think it’s on the turn. Right.

And it’s all about timing. Right. You know? So I said fundamentals don’t change. Bad companies stays a bad company. A good company stays a good company. Sometimes it changes. But ultimately, the time horizon for fundamentally good companies or fundamental research is much longer than the day-to-day price action in the market.

And so it was an understanding, the connection between those two things. It was like lifting scaled up my eyes in terms of my career. Like at one stage though, I was broken for like 20-odd years, but I had a job working as head of research for a top ranked research firm in Dublin. Right. For one of the brightest, smartest guys that I’ve ever met. And we made a point of almost like not publishing research. The analyst teams used to go marketing with their spreadsheets, and they used to take P&L balance sheet

Line by line was how we actually did the marketing, and that had its place. But ultimately I think people’s time horizons have shrunk and investors appetite for sustained periods of underperformance when everybody’s got prices on their phone, right,has collapsed. So you need to get timing right. And so that’s kind of been the driving force behind will say that all of the fundamental portfolio managers that we’ve been servicing, that we continue to service at Quantmatix, they’re all really good at their jobs, but they eventually, they underperform because they get the timing wrong. ’cause they get emotionally involved in the stocks that they like. They don’t sell their favorite stocks at the right time. You know, and it creates all sorts of problems. So like, so the purpose of Quantmatix was to give an objective, big data on emotional tool, which just simply has no ax to grind collaborative tool, work with people as opposed to against them.

That’s the philosophy.

Rado: You discuss the data a lot and you mentioned the big data. So what do think is the most important dataset or what are the dataset that you are using? And I mean, when you build the signals, which dataset you think are important, et cetera, et cetera. So how do you get the value out of the data?

John: I suppose we try to ensure that we get good quality data from our suppliers. So, you know, we’re using, high quality vendors like ICE Data Services. Couple of other, again, high quality providers because we want good quality data feeding into our data research team. When they’re doing their analysis and coming up with new product solutions for our clients.

There’s a vast amount of liquid assets that we can look at, and we really, our focus is on liquid assets and our focus is on momentum and trends. So a lot of what we do is price. We look at volume data, also. We’re not going down to the intraday level. Our target market was always slightly more longer term.

So because of that it’s a little easier to handle. So we’re typically dealing with from four hour data up, rather than going down to sort of tick data level. From that then, one of the key things that we try and do in our analytics is we’re looking at those themes, the relative changes in different groups in the market.

And one of the things that we’re proud of our ability to be able to do is go down into granular levels of groups that we look at to get sentiment of what’s happening either in an index or an ETF. And that gives us a sort of an insight I think, that differentiates us from other people. We typically try to do things on an equal weight basis as well, rather than a weighted, it’s just a preference that we have I suppose from our experience and from some of the research that we’ve looked at, because particularly in equities, things get tilted a lot to the nature of some the giant caps these days.

And also if you look at the tilt towards the US globally. You know. But yeah, so we’re very conscious about getting good quality data input into our models, but also then being able to produce good quality granular data that we can succinctly deliver to our users. Because at the end of the day, our users are busy people.

Rado: As you mentioned.I mean, it’s mainly the momentum and reversal and some other signals. And when we look at the momentum and reversal and we have one individual market. I will select, let’s say US equities, or I don’t know, US equity index S&P500. What is like the most important driver for that market? So it’s like mainly signals that are coming from that particular market itself? Or it’s also important, I mean, to get the signals from the all of the other market because at the end the individual market is influenced by all of the other markets and all of the assets all around the world.

I’m just interested to see how you see that distinction between the, just looking on individual markets and looking on the, I mean, the whole universe of all of the assets that, how they influence each other. So.

Liam: That’s a very interesting question. If I can just take that one. So, at the start, Dan used a phrase, which I love. Which is just like filtering for conviction, right.

You know? Yes, we generate scores and signals for individual instruments, right? But we generate scores and signals for individual instruments over multiple timeframes every day, right? And not only multiple timeframes, but then we combine that by taking support, dynamic support, and resistance levels over daily, weekly, and monthly timeframes for every instrument, every day also.

And so, we do this for. Every, it’s like we look at liquidity, like we’ve got about 10,000 instruments on our system, right? So we look at 10,000 instruments by three timeframes. We basically are producing charts or data on 30,000 charts a day, right? So we have a huge amount of data.

So we look at the US equity market, you know, and we can see all of the 504 stocks that make up the S&P500, right? But also we have all of the sector, so we’ve got the sector ETFs and we’ve got the individual components of the sector ETFs. Like basically we can see the tide turning before anybody else can.

So we have industrialized a process of doing analysis, which allows people to see movements. And thematic ideas are created routinely in our system. So, but we also said we look at the whole world. So it is about looking at everything in context. So we have a report which looks at like the global macro, so we can overlay all of the industries across the world at the same time.

And then we can look at all of the individual stocks within each index. We can look globally at all of the sectors. We have 120 metals and mining stocks across the world. So when we see signals in gold, silver, copper, palladium, whatever, we can then see which of the stocks are moving. Like right now. Right now, and we actually published something at this last week.

You know, we see US energy stocks tend to move before the oil price, right? We saw 20 something energy stocks, 29 energy stocks move positively last week. We have another 20 that triggered signals yesterday like so it is seeing the tides turn, generating ideas thematically. And then it allows like a fundamental person or a technical person to then try and screen from our system what are the statistically most significant signals so that they can then use that to make a decision.

Dan: Yeah, that’s an interesting point. And let’s, let’s continue teaching people about the US equity market. Welcome to the US equity market, everyone. It’s a complicated place. Some of the people listening to this already know this, but I think that there are gonna be some people that don’t. That fact of life in the US equity market, there’s no single source of the truth, right?

Pricing on the same instrument comes from multiple values. So how are you dealing with that?

John: Cause we’re using and taking end-of-day data. Generally, we can get price data from the various different exchanges from our provider. So we will have multiple lines of the same stock from different exchanges.

Usually a lot of the price arbitrage takes care, most of those price differences. But obviously there are volume differences across the exchanges. Particularly when you look at something between the NASDAQ and the New York Stock Exchange. But we’ll look at all the individual lines as well as a consolidated line.

Dan: Yeah.There was a day when the national best biding offer was it, but these days there isn’t. But I guess if your timescales are long enough it matters less. Just before I hand back to Rado. I’ll thank Liam for being so gracious. But for the record, that filter for conviction, you didn’t get that from me. I stole that from you. So thank you for giving me that one. But that was your creation. So Rado, please go ahead.

Rado: Liam, I like your answer. So. About small tides and how you can see, I mean, how the big tide is forming up from those small tides that are forming in the market. I mean, I like that. Now the question is, what is the role of AI?

I mean, everybody all around the world right now is talking about AI. I cannot open newsletter or anything and without seeing the AI there. So how do you see the AI in all of that? I mean, what do you think, how it’ll evolve and how will it impact the service you provide.

Liam: So maybe I’m the less technical. So. And I know that John fill in the technical side.

Right. You know, because it’s gone beyond me. But I mean, essentially, so we hired our first senior AI engineer this time last year. Right. You know, so we use supervised machine learning. We’ve been doing that from the get-go. We are absolutely aware, you know, that there’s so much fake data in the world that what we’ve been really clear about is that essentially we rely on data that we can trust and we are not going to allow AI find data leap to erroneous conclusions as it can.

Right. And we’re also not going to help feed somebody else’s model. So what we are doing with AI is that we are.. There’s something like 12 and a half million calculations in every single score that we produce for every instrument every day, right? So it’s way beyond the capability of any human to do any of this stuff, right?

It’s also way beyond the capability of any human to actually analyze all of the trends and things that are going on within the system, right? So we have a massive amount of stuff. So how we are going to use AI is to actually basically keep our own like locked box, right, of good quality data and use an AI tool, which we’re helping develop ourselves, to actually sort and deliver insights from the stuff that we can’t see ourselves. Right. You know? In order to serve to our customers a better, faster way. I mean, I think ultimately AI is going to change the world. And if you take Quantmatix like we can get to serve up to you. It can show, you can filter down from the S&P down to, you know, like the four stocks that are interesting today in about two seconds, right?

So I think that the implication for the investment industry is absolutely enormous, right? You know that there is, it’s ripe for disruption. And the cost base of these large institutions that’s going to change. Like you would not with 2025 tools like Quantmatix, you do not need to have several thousand portfolio managers do it or research analysts doing all of the research.

This, and a tool like ours can highlight what’s worth looking at, both positively and negatively. And telling you how to spend the resource that you have more wisely. So I think that there’s, it’s definitely not going away, but there’s huge risks to letting it drive your process because you can get it wrong.

Right? So we’re just gonna use it to serve our purposes, not to drive our purpose.

John: Some of the things that we’ve done today, Rado, you know, we’ve used a lot of, you know, supervised machine learning, which isn’t new stuff. You know, this is stuff that people have been using for 20, 30 years.

Rado: Yeah, yeah, exactly.

John: And it’s all just bundled into the concept of AI now, you know? But, so we’ve been using a lot of the more traditional methods in terms of machine learning to identify features that feed into our high conviction signals and things like that. We’re working on something really interesting at the moment on sort of a continuous exposure ranking. Using kind of gradient boosting machine learning LaMDA models.

And then the other sort of thing that we’re really interested in is the use of generative AI and sort of large language models to summarize the data that we have because. In fact, one of the biggest challenges for us is we’re not gonna put a mathematical equation into a large language model.

Right. But we do produce a lot of really good outputs that we have to find a way to deliver to our users. Summarizing all the data we have in very short paragraphs. ’cause people just don’t have the time to sift through all of the lovely screens that we have. The lovely visuals, you know, sometimes they just want to go, here’s my portfolio. Where am I exposed? Where do we see these small tides shifting? And then be able to use something like, you know, a generative AI model to produce a simple paragraph that gets delivered to them daily. That they can read on their mobile phone on their way into work. It’s that combination of some of the more traditional machine learning that’s been around for decades and also exploring a lot of the academic research that is constantly, you know, coming out and looking at new ideas that we can explore.

But then there’s also this the newer AI, the generative AI where. But for us, that’s all about, you know, summarizing data and making it easily accessible for people.

Rado: Yeah. I like that idea. So it’s basically.. So the using the LLMs like an interface between the vast amount of results and the humans that do not have the time to read all of those vast amount of the results.

John: Exactly.

Liam: Exactly.

John: But ultimately we try and do it as kind of Liam said, you know, we don’t want those LLMs reaching out into the world and just pulling out anything, so we’re trying to get it to summarize our data. And the way that Quantmatix looks at the world.

Rado: Yeah, yeah. And here is another question.

So, I mean, we discussed a lot about how to produce the signals, what is important. We discussed a lot about the data, but now here is the question. So let’s say that I have 5 or 10 great picks. And I want to buy those picks because, I don’t know, I see that the tide is coming. And question is how to structure a portfolio out of them.

So there’s a different kind of the question. It’s the portfolio management question. So I mean, how should I place those 10 ideas into the fund? How to manage those 10 ideas into fund and how to know that, okay, right now I need to select those two ideas and to remove them and I need to take two new ideas. So, how to do that?

John: Yeah. At the moment we act, I suppose, the technology partner for a strategy. So we use that term assets under technology. There’s effectively a strategy driven around the high conviction signals that we generate. We look at sort of two timeframes, a short term timeframe with a duration of trade of approximately 10 days. Medium term which is longer.

It’s probably 60 to 70 days. You know, and what we’re doing? I suppose is, the investment manager is trying to be as systematic as possible, as rules based as possible in terms of, you know, we identify the system identifies a signal at close of business on a given day or at the close of a week, and then that allows the investment manager to prepare for execution the following day.

And there is a level of discretion there in terms of if they feel the market’s gonna be really strong that day, they may, you know, accelerate an order and look to, try and do something early on in the market. Or they may just TWAP it, VWAP it.

Liam: Within all of the signals that we generate every day, we have a particular high conviction set of signals that are statistically very powerful with really interesting characteristics going back over 15 years, one year, six months, whatever timeframe, they’ve got the same characteristics, right? So we understand what these signals are, right? So in terms of like just choosing, you have your five fundamental stock picks.

Then it’s a matter of saying, is there a high conviction signal? Which you’d act on immediately, otherwise you would be looking at the other signals in terms of like the dailies and weeklies on our sort of general signals, which are, you know, they work pretty well. The high conviction signals work very well.

Right. You know, so what our fund only does basically, it only trades very high conviction signals. And the characteristics of these is they basically, they get you in at the inflection and they get you at the top, right? And so you can build, sometimes you get a lot of positive signals at a positive inflection that allows you effectively gear up and then it gets you out at the top, you know?

And it does not generate positive signals at the top. Right. it gives you, it has this sort of, this interesting set of performance characteristics which you can rely on and look back over. And that gives everybody who looks at them confidence that there’s a, you know, you don’t get them all right.

But certainly, it helps with the timing very significantly.

Rado: I have a question here. How often it happens that you have periods in the market when you do not have a lot of the conviction? So I mean, there are just a few signals. And then for example, you have.. In the three months you have like period in the market when you have like 50 signals.

Rado: So a lot of them. Are those convictions spread out evenly. Or they are like, so they are, or probably they are not. But, uh, I’m just asking. I mean, what is your experience? What is your experience?

John:Yeah, we do see these.. we do see clustering of signals. At particular points in time.

So. An example of this would be you look at 2022. It was a particularly negative year in the US. Throughout that time we did get clusters of daily signals when the market was correcting and had short-term values. We had throughout that period with very few medium term signals, high conviction. You know, and effectively that kept us outta the market for a lot of that year, for those longer duration trades. It does happen, you know, that we see these and there’s a lot of power and we’ve done a lot of testing on the data on this for clustering of either our medium term or short term signals.

And we really take note of them and we will always advise our clients and our users as to when we see these notable periods, because usually they have a lot of predictive power.

Rado: As we are discussing the clustering. So is there.. I understand that there is definitely clustering in the particular markets, so there will be probably clustering in US equities in commodities, but do you see the clustering of the high conviction signals across the markets? So I mean that the commodities and equities, and I don’t know, crypto, they tend to have high conviction signals at the same time? Or it doesn’t happen very often? I mean, it’s just my question.

Liam: I think in general you get thematic clusters within a market or you get clusters in the market

Where on a specific date. Right. So the week ending, the 3rd of November, 2023. Right. We started off that week with a large cluster of positive daily signals that was in a huge cluster, like just across all global markets. I was just unmissable and this is a very rare event.

The dailies have, as John said, the dailies have like a 10 day duration, right? So we were kind of expecting that they would like daily signals that they would kind of fail or they just have expire. But instead, we actually had, like across our system, we had over 1100 positive weekly signals that confirmed.

So the dailies basically handed the button onto the weeklies, and that rally started at the 3rd of November, 2023. I could show you it on our system and you know, we, it was the rally which drove 2024. Right. So, and we also know that we were talking to a bunch of hedge funds and big institutions, and they were saying all the bulge bracket brokers and strategists were saying: “fade the rally, the risk reward’s not right”. And all we did was like, we showed them, we pulled up a chart of saying like, this is the run rate of the signals. This is this. You know, so. It’s just the data, right? This is a very high conviction event, so sometimes you get very high conviction events like that.

Other times. So John’s saying clustering. Back to US equities. In September of last year, we had a cluster of signals where we had three of the highest conviction signals in US Airlines and Delta, United Airlines in America. But we also had our lesser conviction signals in Alaska and Spirit and whatever.

So we had a specific identifiable cluster of signals in the US Airline space. It looked up for news. There wasn’t much by way of significant news, but the stocks rallied, you know, and the higher conviction stocks like went up by over a hundred percent. Right.

You know, so when you see a cluster, it does mean something, which we don’t necessarily know has happened, happened.

I mean, the other amazing thing, which we sort of highlight at times is that we had a positive signal at gold on the 6th of October, 2023. The night before the Hamas attack. Right. So we actually had started to see this rise of positive equity signals. The reason why we spotted it is ’cause it stood out.

Right. You know, buying gold, buying equities is not necessarily consistent. Right. As we can see now. Right. So it stood out for us. So we said our. It captures events. Sometimes we don’t know what they stand for right at that particular point in time, but you know, it’s significant and that you’ve just gotta trust it.

Yeah.

Rado: Yeah. So I mean, what I see, or what I understand from that is that when you have like a lot of those underlying ways at the end is like the stenograph. So I mean, you are looking at that and you are trying to capture the earthquake and you see that there may be an earthquake or there may be a tsunami and before, I mean the actual event because something.. Something is happening in the market and you cannot see it on, I mean, yet, but you see it on the small waves.

When they start appearing.

Liam: It’s amazing that you use that terminology because we used to use that terminology all the time. Used to call it before we started call GPS. So, I suppose the specific example, so before we set up Quantmatix, I was working for a broker in London. We had like a really basic version of this product, but on the 15th of January, 2020.

I wrote an email saying: “Alert, alert. Hedge your portfolio. We have the COVID cluster”. I thought everybody knew COVID was gonna be a disaster. Right. And then the markets fell 30%. But you know, it was like the seismometer. We got the signals, we got like 250 negative weekly signals. And then 10 days later the market rolled over.

Right. And then in March we had like a couple of hundred daily and weekly signals. At the turn. So yeah, my, my other colleague is not on this. He used the a great expression, which was like: When the smartest guys in the world touched the market with that program trade, our system can pick up a footprint of what they have, what they’ve done.

It is a bit like that seismometer, so you can see what happens. You can see the dollar, we can see the gold, we can see the equities, we can see all of this stuff at the same time.

John:One of the things, Rado, I think an interesting phrase that we often use is we’re nowcasting with an expectation of persistence.

Sometimes you do see these, you see those tides, they start to shift and you see the little ripple start to build up and you don’t know why. And that’s difficult as a human because you know, you’re trying to emotionally say: “Well, I don’t know. Do I just put full faith in the data?” But ultimately what we’re doing it’s objective.

This is what is happening. It’s not disputable. The price action is the price action. The volume action is the volume action. Sometimes, we’d be in the office, myself and Liam and Paul and you know, you do sometimes find yourself second guessing things, but you know, it’s important to remember that this is objective.

It’s an interesting thing we have to deal with.

Dan: I think the seismograph comment was just meant to be because fun fact, the last time we actually had an earthquake that you could feel in New Jersey, which is where I’m sitting right now, I was on a Zoom call with Rado.

As it happened, he was looking at me like, what’s happening? And I looked at him like, I don’t know what’s happening. But, I wanna pull on a slightly different thread. And, Liam and John, you both mentioned it, we’re dancing around something here, but John actually did a really good job in kind of channeling how clients think about this and the way they ask it. I might’ve scared you off because we can’t talk about specific fund products and performance on the podcast, but we can talk about what you do. So to put it in John’s words or the client’s words, if your signals are so good, why aren’t you trading on them?

Do you have fund products that are trading on them? You can say yes, no.

Liam: Yes. We now have fund product, which is trading on the signals, and the beauty about this answer is that it was a subscription paying client who ran his own fund, who performed, who found the signals to be very useful, who approached us this time last year asking if he could run a real money trial, just specifically trading.

So he ran a special situations fund and he thought it’d be a good idea to see if he could run a real money trial trading our highest conviction signals. And that trial in the third quarter of last year was so successful that the broker that he was working with, that a prime broker he was working with approached us.

And asked us if we would partner with them. So they are the regulated entity. They have created this Bahamas based hedge fund and we are supplying them with the technology. So John was saying assets under technology is that part of what we do now. Yeah. So it’s. And frankly. That came down the track probably three or four years faster than we had anticipated. We used to say. Why are we trading money ourselves? It’s because we poured our life savings into building the platform, right? And it is our trade. But actually it’s much easier to say we have a fund and people just go, oh yeah, okay.

And we can actually dynamically show people on the site you can like drag a date slider the performance of what the signals were, et cetera, et cetera. So, yeah.

Dan: Thanks Liam. Thanks John. So we can’t get too much into the specifics of that, but I thought it was important for the audience to know that you’re doing both things.

So welcome to the wonderful world of running signals and running a fund all at the same time. It’s a lovely task. So, Rado, you’ve done these things, you know, what do you think about that?

Rado: It’s a complicated to run the fund because, I mean, the running the fund is not just by picking the signal. It is also about building the portfolio and combining all of those signals and I mean, doing the correct money management and the risk management, et cetera, et cetera.

So it adds, I mean, additional layers of things that we as a fund managers need to think about. So it is not just about the signals, but that’s the beauty of that. But there is another thing, and I cannot get it out of my mind, so I know that I’m switching from the question to the question. I like the discussion about the earthquakes.

And how those, I mean, small ways are like earthquakes. I know, I know that I’m returning back to what we discussed but..

Dan: I’m getting PTSD here. But go ahead. Go ahead. This is a good one.

Rado: Yes, I know. But I have here one theory. And what I think is that, what is the difference between the markets? The nature is that the markets are driven by the human.

While the nature is driven by, you know, the nature. So , what happens in the market is that, once the people start to see those small ripples, that tsunami or the earthquake is not caused by the nature, but the humans ourselves. So it means that we are somebody who see.. Or there are all of the other guys who see those ripples and they will start trading on it, and they basically allow signals to happen.

Liam: I’m a believer, right? So of my career, right? I’m a.. Put myself in the box of saying I’m a loose-efficient markets guy. Right? Which means I believe exactly what you’ve just said. Right? And..

Dan: It’s a big statement.

Liam: Yeah. Yeah. You know, so I do believe that what we have. It’s kind of, it’s a behavioral analysis tool, right?

I had thought that what we, that market or that price tells you all the publicly available information. Right. You know. So there was a very interesting article, a lunch with the FT, where they interviewed Eugene Fama. About three or four months ago, right, where he actually said that he thought markets were more inefficient now than they were when he wrote his efficient market hypothesis paper.

Right. But it dawned on me thinking about it, is that I was thinking about our product the wrong way. Right. It’s because and what it’s dawn to me is that the time of market efficiency is the day of company results. It’s the one time where everybody gets all the information at the same time. Right. So the rest of the time where we see these clusters of signals, we actually are kind of detecting the moment of almost like peak market inefficiency.

It’s telling you when somebody knows something, when it’s not necessarily in the news. Right. You know? So that inflection before the news, that is where our signals occur. And so, I mean, I’m absolutely fascinated with this whole topic of when people get the news, when the price, what’s in the price.

But I kind of believe that the point at which our signals are triggered, are the point at which the smart money knows something and like the most of the time it’s, those signals are not generated on earnings day. They’re generated before earnings day. Sometimes they’re generated immediately afterwards where you do get new news and everything turns.

So markets, our company is going up. Everybody’s very bullish. They announce numbers, they’re bad. You get a signal and it tanks, that’s fine. But the times where.. In the middle of like a quiet period, you get a signal and it starts to rally. Or there’s something like the other Airline stocks where the Airlines start to move and you search through all the news and you’re thinking there’s nothing which is earth shattering.

There’s no brokers produced any research, but five Airlines moved on the same day and three of them generated our highest conviction signal and the others generated another signal. That’s the academically fascinating bit switch. And it’s why we now have a academic collaboration with the leading New York University, if I’m allowed to say who it is. Actually know.

So we’re kind of John and myself, a bit like data nerds as well as being interested in monetizing the product. It’s fascinating seeing how and when the signals occur, where they occur, the time that they do the stories that they tell.

John: Just to follow on from what you said there Rado, like that herd effect, you know, that’s very well studied, but probably not fully understood behavioral aspect in finance. And I suppose we do see that. You do see these, the ripples start to build and then the herd follow in. And that adds more conviction to some of these movements.

Rado: Exactly.

John: It’s a very common thing, I guess a lot of.

What our analytics is doing it’s capturing that herd mentality, but it’s also then trying to identify the turning points when the herd stops or the herd starts to slow down. Those behavioral tendencies are so prominent in the market and I do agree with you.

Rado: I like where we are going. It seems that we are all fascinated by the understanding how the market works. This is something that is also, I mean, the reason why I’m studying the market. Because for me the research is more interesting than the trading itself. Because, I mean, at the end when you have the signal and you just trade the signal, I mean, it starts to be boring after a while. Because you are taking the signal you are buying, selling what the signal tells you. But I mean, looking for a new signal, looking for a new ideas. That’s interesting for me. I like your comment that you see maybe more important than the time around the earnings announcement is the time between earnings announcements, because if something happens there, it is in unusual time when nobody expects that. Because, I mean, the time when the new information should arrive is exactly at the, during the earnings announcement and there are three months between them. So if something is happening in the time between, that’s important. So I like that. And maybe we will take a look on this in some of our Quantpedia research papers or stuff like that.

Liam: Yeah.

Rado: So I, I like that.

Liam: With my fundamental hat on in my previous role as head of research. Right. This Irish broker, we actually had a plan, right? So we saw that, that like quarterly earnings in the US. So like European companies had ordered like interim and finals. Where they had a pre-close statement before each, and they had an AGM.

So companies used to talk to the market five times a year. Publicly make a public statement about their performance. Right. So what we used to try and deliberately do, with our research hat on, we would try and make an excuse not to like attend the conference call. Obviously. But the companies would, you know, certainly the Irish companies would try and come and meet the brokers and tell their story. We’d always try and find an excuse to meet them in the quiet period, right in the furthest away point. We had a really good process. We knew exactly the tone, what they said, and all of the questions. We were all set up around the table to grill the management was to work out there.

Any change. In their tone of voice and their level of conviction, anything that could give us an edge. Right. You know, so that was, we actively sought to do that. So I can only presume that, again, back to where those signals occur. Those signals occur somewhere. Somebody has worked out. And who knows that right or wrong at the time. And it appears as a green triangle or a red triangle on our system highlighting. This is an inflection, this is a change.

Rado:Yeah. Yeah. Somebody figured it out and it’s, I mean, at the end it in enter price. Do you see the seasonality? In those signals? If it’s driven by, I mean, those earnings announcements, et cetera, et cetera. I mean, they are seasonal, so there can be maybe seasonality patterns in those signals.

John: Absolutely. Rado, it’s one of the things and that one of our quant researchers has been looking at. Her background is as a physicist. So she is very interested in, you know, wave patterns and things like that. You know, and has been able to identify seasonality impact in the data that we have. And it’s then trying to find a way to adjust for it, deliver that in a way that our users understand better. But for sure. We do see it.

Dan: Yeah, that’s an interesting point, John. So let’s build on that and perhaps start to wrap up. So there’s a number of use cases of this and a number of different types of users, and I don’t think we’ve really drawn those out. So there’s the whole like investor in the fund use case that’s kind of over there. You know, they just put money in it and that’s it.

But for the people actually picking up the signals and using them, do you have to be a quant? Yourself. Do you have to be kind of that, like out of the matrix robot kind of mind to use this? Or can you be a fundamental analyst, a normal human being? You know, how does that work?

John: Yeah, absolutely.

I think we always.. We were always building this product for, particularly for people who were not technically or quantum minded. Because they would have their own processes and their own ways of doing things. So, you know, I think we were always looking to act as an overlay for somebody’s already established fundamental investment process.

But it’s an interesting that you talk about use cases. We have a number of clients using in different ways. We have clients who trade rebalancing. And momentum is very important for them in the run up to looking at, you know, stocks that are going in or out of an index. So very distinct use case. Trading desks that are looking to add value to their PMs, especially in a time when trading guests are being outsourced all the time.

Portfolio managers who are looking for, again, that overlay to their typical fundamental investment process. And then you have just your active retailer. Your “protail” is that we call, who just wanna take a punt. And so yeah, there’s many different use cases.

Liam: Some of our largest clients are like proprietary trading desks, and the rationale is that they see themselves as being very good at constructing derivative trades.

They see our value is that we’re very good at direction. So actually our direction, their derivative skills, putting on spreads, call spreads, put spreads, whatever, you know, around a high conviction directional trade, that’s a use case, which underappreciated probably by ourselves. Right? You know? ‘Cause we don’t have that there.

We’ve just spent time actually just creating a product to actually go and help other people with. New colleague, Doug is his name. I think Dan met him. He brings that expertise in-house, you know, so that’s something which we’re delighted to have that we’ve got a perspective that we can actually add a bit more value in talking to clients about the.

Thematic ideas that occur from time to time.

Rado: For me, it was really pleasure to meet you guys. I do not have a question. I have like a lot of the ideas that I would like to test on data.

Dan: You’ve just given a whole bunch of ideas to the king of academic research over here is what you just did. Yeah.

Rado: I mean, just..

John: We’d love to collaborate with you, Rado.

Dan: Good. Okay. All right. So idea factory and the new outcome for QuantBeats. So let’s wrap up on that one. So thanks Rado. Thanks Liam. Thanks John for teaching us about market GPS and seismometers. I’m gonna have some nightmares tonight, but it’s okay. Be glad to see you guys again very soon when I get over it.

But for all you listening to meet Liam and John from Quantmatix, to meet Rado from Quantpedia, to join us as a guest or if you’re interested to otherwise support the podcast or support the channel, contact me, Dan Hubscher at the details here that are shown on the screen right now. And just note for the audience that any questions regarding investment offerings will be referred to an email follow-up for compliance purposes, as these videos are not intended to be about investment offerings.

Okay. So if you like this video, please do come back for our next interview with a quant manager guest. You might also wanna watch the other videos on the QuantBeats channel. There’s another episode earlier that you can find where we can have a bit of a earning season smack down with these guys. If you want different views on how all that works.

And don’t forget to like, comment and subscribe to QuantBeats on YouTube. Again, you can get there via the Q logo on the corner of your screen. And finally, you can visit our websites and YouTube channels for more shown right here. So. John, Liam, Rado, thanks for joining and everybody have a quant day.

John: Thank you.

Liam: Thank you.

Rado: Thank you.