ETF Edge - Can AI Beat the Market? The Role of AI & Investing 2/22/23

Episode Date: February 22, 2023

CNBC’s Bob Pisani spoke with Art Amador, Co-Founder and COO of EquBot, which runs the AI Powered Equity ETF (A-I-E-Q) in partnership with ETFMG – along with Bill Studebaker, President and CIO of ROBO Global. They delved intothe AI craze and the role artificial intelligence can play in the world of investing. Chat GPT has become all the rage these days, but what about using AI to pick stocks that could potentially beat the market? Our panel of experts weighed in on the debate.In the “Markets 102” portion, Bob continued the conversation with Art Amador from EquBot. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

The ETF Edge podcast is sponsored by InvescoQQQQ, supporting the innovators changing the world. Investco Distributors, Inc. Welcome to ETF Edge, the podcast. If you're looking to learn the latest insights on all things, exchange traded funds, you are in the right place. Every week, we're bringing you interviews, market analysis, breaking down what it means for investors. I'm your host, Bob Pisani, just back from a wonderful week in the Yucatan in Mexico. I recommend it. Today on the show, we're talking about the AI craze in the role.

All artificial intelligence can play in the world of investing. Chat, GPT has become all the rage these days. But what about using AI to pick stocks? Turns out, we've won one around for a long time. We'll talk to two experts in the business. Here's my conversation with Art Amador. He's the co-founder and COO of Equibot, which runs the AI-powered equity ETF. The symbol is A-I-E-Q in partnership with ETFMG, along with Bill Studebaker.

He's the president and CIO of Robo Global symbol there. R-O-B-O. Art, let me start with you. We're trying to figure this out. The investing world is trying to wrap its head around ChatG-PT, but you've been using AI for years to pick stocks. I think since 2017 you've had this E-T-F. So tell us what's the similarity between you and Chat-G-T-T,

what is the difference? And is Chat-G-P-T some kind of watershed moment? Yeah, absolutely. So chatGBT is an incredible milestone, right? The ability to generate AI. The difference, though, is in the knowledge base and how they're, and so for example, right, chatGBT knows a lot about a lot of things, right? But if you ask it to put together a portfolio or anything highly sophisticated or technical, it's unable to do so. So that's really one of the main differences is we've focused and applied AI specifically to understanding the critical signals in the marketplace and went very deep, where chat GP is very broad in nature.

But the interesting thing is at the heart of both what we're doing, we've kind of attacked the problem in a similar way from a technology standpoint. We've built a convolutional knowledge graph, which is what chat GBT is using, which is a graphical database that unconstructured and unstructured data and connects relationships between different concepts, entities, and events. And we're using these deep learning models to make predictions. And then these predictions, right, are measured. And then the information flows back to the powerful knowledge graph,

which is probably one of the most powerful tools in AI at this time. And so it will continue to get better and improve over time. Yeah, it's a good point. So, Bill, let me turn to you. chat GPT hit something like 100 million users in two months. And I'm just trying to figure out what is the effect going to be on businesses, not just finance, but business in general. We had talked about this earlier.

The iPhone launched an enormous app business worth hundreds of billions of dollars. Is this going to happen with ChatGPT? Is it going to launch a multi-billion dollar businesses or is it going to lead to a bunch of layoffs? I think this is definitely not a fad. And I would argue that this is, as you just said, it's kind of an iPhone moment. And I think as the field of AI continues to advance, we're moving into kind of unshared territory. And artificial intelligence is certainly the buzzword in the market now. It really thanks to a wave of sort of viral applications such as Chad GMBT, which have captured, you know,

the public's imagination. And this tool and many others demonstrate really the quantum leap that's taken place in the last few years in AI and what it's capable of. When we launched robo nearly a decade ago, we had a view and vision that we were on the cusp of ubiquitous automation, but fast forward to 10 years later, you know, we couldn't be more convicted. And this wave of innovation, I think, is set to hit the world of business and really redefine, you know, what's possible. As Art kind of talked about the most transformational companies are those that are going to, you know, be able to achieve, you know, sort of real-time AI. And so, you know, all AI is powered by information data.

and real-time data is the most valuable of all. So if you have the ability to synthesize and understand the data, you're going to be pretty successful. So, Art, I'm going to go back to investing. What's the implications for investing? So you've got this AI-powered ETF. It picks stocks on a proprietary model. But when I'm talking to you over the years about this, it uses pretty conventional signals, it seems like fundamental signals, technical signals, headline frequency. I do that. So what does AI do for investing that human analysts can't do?

Yeah, so you're absolutely right. So the signals are in some ways traditional, right? So it's signals across fundamentals, macro, technical, news, right? But the critical thing is to understand which signals are critical for which companies. So for example, GameStop is a top 10 holding. Right. So for game stop, a critical price driver might be something like news and social sentiment, right, and headline frequency, where another top holding would be city, for example, and it uses, you know, maybe the critical drivers are inflation, right, and interest rates. And so we're trying to understand, right, what are the critical drivers for each particular company? Now, the beauty is, right, is that all of these different companies are connected in very complicated

ways through different value change, right? They're customers. There's different themes like Metaverse and AI. And so what AI can do is it's more efficient, right? It's faster. So instead of having 10,000 research analysts and them together trying to put together the picture of the market, right, what AI can do would be like 10,000 virtual analysts and what is known by one analyst can be known by all analysts instantly.

So, I mean, this is an important point you're making here. And maybe Bill, you can respond to this. He just said 10,000 analysts instead of 10,000 analysts, you got, you got, you know, chat GPT or AIEQ. So there's hordes of people making stock recommendations on Wall Street, picking stocks. Is this a threat to their livelihood? Yeah, I don't know, Bob. I think it's going to take time for this to play out. I certainly think that if you're able to remove emotional human decision-making, it probably is going to be

additive over the longer term. I mean, active management struggled because people sell what works, they sell what doesn't work, and they're stuck with what's in between. So the extent that we can kind of eliminate some of the noise and, you know, take the emotion out of investing, I'm not an expert on this, but I think that there's probably a lot of opportunity in the years to come as the technology gets more and more refined. And obviously, the data can sort through various, you know, correlations and obviously there's certain factors that work in the market and then certain factors that don't. The important thing is not to look in the rear of a mirror when investing, but to look forward.

So the extent that these technologies can help you sort of navigate what's coming, I think is going to be very disruptive. Yeah. I want to drill down a little bit more on how you pick stocks or how your AI pick stocks. And I look at your largest holdings here, Novavax, AI, AIG. Las Vegas Sands, Advi, I don't particularly see a common thread, but maybe that's the point. It's not a tech fund. It's not necessarily a momentum funnel, although it uses momentum.

Explain how this program picks stocks. Yeah, absolutely. So it's more of kind of a bottoms up approach, right? What we care about is the critical signals for each one of these, you know, individual companies. And so we extract these signals from various data sources. millions of global news articles, social media posts, research reports, TV transcripts. We combine this with other kinds of data, economic data, market data and financial statements on over 5,000 publicly traded companies. We then use all this data and signals to make price predictions on the entire universe, right, to see where the stock is going.

And then the 140 names that have the highest potential of appreciation or the highest expected returns, those end up going into the into the portfolio. And then as we walk forward in time, right, those price predictions are measured against the actual returns. And then the system then learns from those predictions so that the subsequent predictions can be even more accurate than the previous ones. So this all sounds great, but how do I measure this? I mean, what benchmarks do I measure it against? Do I measure it against the S&P 500? How do I know you're doing better or not better? that are, you know my point here? I do.

So it's a total market approach, right? So it looks at 5,000 different companies. And it's having an exceptional day today, and it's outperforming the broader market, you know, in 20, 23 here. But the thing with AI and machine learning, Bob, is I think we're not looking at it the right way, right? You're talking about past performance, right?

The beauty of this is that we're just in the first inning. These systems, right, AI, machine learning, it's continuing to evolve, right? It's going to improve. So whether we're talking about chat, GBT, right, your iPhone, AIEQ, its best days are still ahead of it. But we're in an arms race, right? I mean, that's sort of the problem I have with this.

It used to be, you know, we used to try to find the best analysts 30 years ago, and now everybody's got the best analyst. Information is very evenly distributed. So suppose we get AIs that pick stocks, they're all going to be, There's an arms race of all AIs. I guess the question is, I'm going back to, why don't I just buy the S&P 500? I'm not sure what I'm getting when I do this. I love the idea around the concept of doing this because the AI can see stuff I can't see.

I made fun of this earlier with you about, you know, I can pick stocks. I can do fundamental and technical analysis. But what I can't do is surf through millions of pieces of data and find the signal in the noise. I might likely miss that because I can't analyze as much. I'm just trying to figure out where does all of this sort of go. If we all have an AI, stock picking assistant in the next 10 years, is that really going to lead to anybody's outperformance or does it, what are we looking for at that point?

Yes, and not AI is created equal, right? And people have different approaches on AI, right, and how to use it. different technical structures, right? As I mentioned, right, our technical structure using things like knowledge graphs and deep learning models can be very different than the way someone else kind of approaches the problem, right? So you want to make sure when you're investing, right, that you're investing with people that have done this before that have deep expertise and deep knowledge in this space, right? So that way you're getting the best AI possible. And that's why we've partnered with some of the best companies out there like IBM Watson to

ensure that we've got the best AI and that we're unlocking these investment opportunities for investors. And I love the ETF format because anybody can buy it, right? Yeah. Right. And how often does AIEQ rebalance? So it's data dependent. It's an active fund and it can rebalance daily. And I would say most days it probably does. Okay. Bill, let me dive a little deeper into your Robo ETF. You This tracks a global index of companies involved in robotics and automation. How does the ETF pick stocks, or how do you pick stocks? And where does AI fit in right now with robotics and automation? I guess we kind of still do it a little bit of the old school approach.

I mean, we sort of pride ourselves on being industry experts. We have a team of seven PhDs in our team, some of the world's renowned thought leaders and entrepreneurs that have built businesses, and they were very helpful in the early days of helping us create the ecosystem to invest in robotics and AI. Because when we started this 10 years ago, we recognized that there was some pureplay companies, but there wasn't kind of a GICS classification system for them. So we had to go and produce that.

So we came out and defined a company as either a technology or an application. So the technology is what makes the economy system work. The computing, the AI, the processing, the sensing, the actuation, and the integration of those technologies. And then the use cases, so where is robotics being deployed? But we knew it was an industrial manufacturing, you know, how about into health care, into ag, into security and surveillance, you know, into health care and others.

And as the use cases of all, those sectors would obviously evolve. And so we go out and we basically invest in what we believe are the best in class companies. We do essentially a proprietary ranking where 50% of our ranking is to identify companies that have high revenue purity. And the other 50%, we do a deep bottoms up analysis by looking at their market share leadership, their technological leadership, and understanding where and how they're investing. And that's a difficult component to just capture it through an AI because you have to understand their businesses and where they're investing their R&D, look at where they've invested in buying companies and what their RIs have been, and what their track record is for integration, et cetera.

And so we then apply an ESG screen to this, and that out pops a investable universe. And we take the top names from there and then put it into an ETF wrapper. You know, your robotics. So robotics was, I mean, it's been, I'm a science fiction fan. I grew up reading science fiction. I still do. and the robots are stealing our jobs, stealing our souls.

This is one of the oldest tropes in science fiction. There are stories going back to the 40s about this. I guess does this somehow have more currency these days, have robots stealing our jobs? It seems to me if robots are stealing our job, they're doing a pretty bad job of it. There's four jobs than people are at the moment. Can you address that old trope?

The robots are stealing our job and would suddenly hit a singularity? You took my line, Bob. I like it. I think it's true. If you look at the companies and the countries that have the highest utilization of automation, guess what? They have the lowest unemployment rates. I mean, if robots are stealing a lot of jobs, you know, go figure. We have 11 million unfilled jobs listings in the U.S. If you look at the density of robots globally, there's going to be about 3.4 million industrial robots installed by the end of this year. that is in comparison to a global economy that employs about 500 million in manufacturing employment. So we have a long ways to go.

If you think about the number of data scientists and people that are trained in AI globally, it's a de minimis figure. I think it's going to take a long time for this to happen. More importantly, what's coming out from robotics and AI is better productivity. I mean, we went into the pandemic and came out of the pandemic with 1% higher, with 6.5 million workers. Okay? That is the definition of productivity.

And I think you're going to see a tsunami effect in terms of prices coming down as a result of deflationary pressures from these technologies. And I can't predict everything, but I have high confidence. This is going to be very additive to our economies globally and, importantly, just generating new growth. So you think AI will create deflationary effects. Did I understand you correctly? Yeah, because look at what's happening.

I mean, you know, we're at undeniable inflection point because of the price or the computing capabilities and the cost of computing, which is plummeted. It's now creating an array of use cases that a few years ago was only Elon Musk science fiction. So it's all about, you know, lowering costs and providing more ubiquity of service and product, et cetera. And, you know, you look at what's happening. in e-commerce. I mean, there is an arms race to create better fulfillment and have lower inventory shrinkage and everything. Everything results in, you know, basically having more product

out the door at a lower cost. And so inevitably over time, I think these are very powerful technologies that are very disruptive. In terms of, let's go back to robots, not worry about AI. Is robot deployment accelerating? Is there a way to measure that? You know my question? Are we reaching some point where suddenly there's been a big increase in the use of robots in the last few years? The International Robot Federation releases data yearly about the deployments. And it's, I would say it's somewhat measured. I mean, it's roughly around 15%. I mean, No one wants to go out and add too much capacity too quickly and then find out that maybe the industry is a little more cyclical than they might have thought. I mean, these are clearly, you know, heavy intensive capital deployments.

When you look at the companies that supply and the actuation, that's what makes the robot move with sublimitur accuracy, semiconductors, et cetera. There's a lot of supply chain that's tied in that also has a very measured. approach to their capacity and growth. So while I wish that the industry was growing 50% per year, that's really not the case. You know, I would say, you know, somewhere in the low to mid-teens is a good way to sort of think about the growth. So, Art, what's next here? I mean, what I notice here is chat GPT is going to keep getting better.

But what else beyond that? So out there is quantum computing, for example. Is that going to be commercially viable, in your opinion, in the next few years? And where would that fit in? Well, things like this always happen a lot sooner than you think. And we did partner with IBM, who's made some great strides in this space. And quantum computing is, you know, the, it's possible that it may unlock or solve some of the world's most complex problems. And when you think about the markets, right, the markets are extremely.

complicated, right? There's all of these very complicated relationships between different companies and events, right, that are occurring. And something like Quantum could unlock many possibilities for investors. And so, you know, we're already kind of planning and thinking about how we could apply something like that to the investment space, right, and really unlock these investment opportunities for investors. And it gets me really excited, right? There's been some incredible milestones. You had Watson Jeopardy Challenge. You had Google DeepMind. Now we got ChatGBTGBT, generating language.

Quantum might be the next watershed moment. Yeah. I think it's very important to point out about ChatGPT. There's been a lot online mocking ChatGPT,

which I find amusing. So it's very typical to somebody post someone and say, look, I ask Chat GPT for, you know, the five best restaurants in Bologna, Italy, and one of them came back in the wrong city, and the other one didn't exist. And ha, ha, ha, ha, isn't this stupid? And the thing that I think it's important that people don't seem to get, these mockers, is these systems learn very fast. And the next iteration, 4.0 or 4.5-50, isn't going to make those mistakes. And while we're getting older every minute, it's getting smarter every minute. So it's quite remarkable to me, I think, and people

are going to be shocked when they see the 4.0 come out and even more shock when the 5.0 comes out. I don't know if it's a thousand times more powerful or 100 times more powerful, but it's certainly going to be more accurate than the current 3.5 version. And I think people have not really thought through the implications of something that's, you know, 10 times more accurate and a thousand times more knowledgeable than something that existed a year before. And I think people are going to be quite, quite surprised. And those kinds of mocking a hat, you see, it didn't give me the right restaurant, are going to go away sometime in the next year.

I don't know if you have any thoughts on that, guys, but am I on target or am I off base? The beauty, right? So if you chat, GBT, I typed in, what is the future of investing, right? And it gives me the response. And then I either like the response or I dislike the response, right? And then after I've liked or disliked it,

it says, well, what would have been a better response? So people are mocking this, right? There are millions of users. Those millions of users are actually training the system to make better responses into the future. It's absolutely incredible. And you're right. This is going to advance very, very quickly before anybody even, before anyone, a lot faster than many people think. Yeah, and it's listening to people.

And it's learning from people. I'm hoping, I don't know how to do this to write like a blue one. for chat GPT, stupid things that people said to chat GPT. I'm quite sure there's hundreds of people who have fed their tax returns to GPT, asking it to find some tax loophole or some way it can reduce its taxes. Like, talk about, you know, stupid ideas. I think I'm going to feed my tax return to, you know, an artificial intelligence system. I guarantee you this has happened. I don't know how to write that blooper reel, because I don't know how to get access to that information,

but I'm quite sure that it's going to happen. Now it's time to round out the conversation with some analysis and perspective to help you better understand ETFs. This is the Market's 102 portion of the podcast. We'll be continuing the conversation with Art Amador from Equibon. Art, thanks for continuing staying with us.

I was fascinated by our discussion and I'm wondering if the chat GPT is some kind of singularity. You know, a man I've known for 20 years, Ray Kurzweil wrote a book 20 years ago called The Singularity is Near, talking about a point at which essentially artificial intelligence becomes so advanced. It's essentially far more advanced than human capabilities. Ray is still around and still making predictions on this. But chat GPT has become such a phenomenon that I'm wondering, are we overhauled?

hyping this, or is something really singular happening here that truly is a watershed? So if we're saying, can AI do things that humans, people are incapable of, I would say absolutely yes. Now, if we as humans could freeze time and put life on pause and spend the time, reading news articles and, you know, increasing our own knowledge base, then maybe, but that's not humanly possible. If we're talking about AI becoming self-aware, I think that's something totally different. And in that instance, there needs to be ethical standards set up as it relates to kind of developing AI. But this is happening, and it's happening really fast. Yeah. Five years ago, about five years ago, I don't know exactly when, I made a prediction because these kinds of things were around already, these kinds of talk about this happening. And I said that in five years or so, 30% of the adult population will name a personal digital assistant, an artificial intelligence, as its best friend. Now, it seemed a little wacky at the time.

And I literally just made a stab at it, I guess, but it seems to be coming true. Can't you see the moment when this happens, when chat GPT or an offshoot of it, essentially becomes a personal digital assistant to you? You talk to it. It makes your appointments. It books your doctor appointments. It books your plane trip to Florida for your vacation. It knows everything about you, essentially.

And it becomes your best friend. Is that viable or am I still sort of talking science fiction? No, it absolutely is viable. I mean, if you think about everything is so connected nowadays, we got smart homes, right? And just imagine if you allowed your assistant to know your investment preferences, right, how you do your taxes, it was able to read all your emails. I mean, all this stuff, there's privacy issues and it's, you know, it might be a little scary, but the AI could learn more about you. than even, you know, your partner in life. It would be able to make predictions and, you know, come up with things that could be very helpful, right?

And I think we need to think about very, you know, we need to think hard about how much we want to let AI into our lives, right? Because there's a tradeoff between privacy and convenience. But we're getting there to the point where AI can do all these things for us. And it's incredible from a technological standpoint. Yeah. It's also incredible in terms of how quickly it has sort of overwhelmed a lot of people's expectations. I'll give you an example. I have been a science fiction fan since the 1960s and still subscribe to science fiction magazines.

One of the ones I subscribe to, Clark's World, today, announced that they were stopping submissions to the magazine because they had been flooded with chat GPT-generated science fiction short stories. And they can't even figure out the extent to which some of them are real and some of them aren't. But all they know is there's been an enormous flood sort of auto-submissions in the last couple months that are clearly tied to chat GPT. I mean, think about the implications of that. This is just a tiny science fiction magazine. But, you know, what does this mean for literary works, for artistic works?

what does it mean for teaching in schools? It's literally come out of nowhere. We've discussed this for years, but suddenly it's very real now. That's startling to me, how quickly. That's why I keep bringing up this singularity. We had been discussing these for years, but all of a sudden, it's here.

It's in front of our face. Yeah, so I wouldn't, we just got out of a pandemic, right? The first thing that people wanted to do was go see their friends and their families and their loved ones. people are traveling like crazy right now, you know, at the Amador household, we're trying to plan our next three vacations this year. People want to be connected to people.

So I think of AI, at least in the current environment, as a productivity tool, right? It's going to make people better and more efficient at their jobs. I'll give you an example. My wife is a copywriter, right? And companies probably aren't ready yet to let an AI take something that's AI generated and auto publish it, right, without having somebody who's got the domain expertise, right? To look at that, make sense of it, and then decide if it's ready or if there's any tweaks that ultimately needs to be made to it. And you were, you know, you were mentioning this on the, on the show about people submitting their taxes, right? with these AI systems, if you want them to be highly technical, like in our case, as it relates to investing, you have to go very deep, right?

And something that's more broad like chatGBT, it knows a lot about a lot of things, but it really is still kind of lacking that highly technical content and sophistication in order to do these things. So there isn't a system yet that can do everything and do it extremely well. there's use cases, right, in which we apply AI. And that's, that's, that's where we are right now. But how the job task it does, right? That, that is going to get exceptionally better, right? Think about chat GBT. There's millions of people using this.

And they're training chat GBT responses. And it's going to get better quicker than anyone's imagining. Yeah. That's what these people who keep mocking it, uh, on the, on Twitter and the internet and think it's a dumb system, don't understand. They're voluntarily training the system. ChatGPT has 100 million users that are now training it for free.

It's remarkable when you actually think about it. I was wrong in my prediction a few years ago that 30% of the adult population would name an artificial intelligence as a personal digital assistant as its best friend in the next five years. But I'm gonna stick to my point, because I think chat GPT is a singularity of sorts. And I think I'll say it again.

Okay, I'm wrong in time. But in the next five years, 30% of the adult population will name a personal digital assistant as its best friend. And I don't think we've even begun to explore the implications of that. Science fiction has for a long time. But I think the public as a whole hasn't. But I'm going to have to leave it there, Art. We can go on for hours about this. But I'm going to have you back.

We'll talk about it some more. Everybody, thank you for listening. I'm Bob Vasani. Thanks again. And make sure you tune in to the DTF Edge podcast next week. InvescoQQQ believes new innovations create new opportunities. Become an agent of innovation. Invesco QQQ, Invesco Distributors, Inc.