Artificial Intelligence - and the machine learning algorithms that underlie it - is showing up pretty much everywhere these days, so it's no surprise that there's now an AI-powered ETF.
EquBot's founders, CEO Chida Khatua and COO Art Amador, join Let's Talk ETFs to explain the inner workings of their AI Powered Equity and International Equity ETFs.
They believe placing artificial intelligence investing strategies into an ETF wrapper offers investors the best chance of obtaining persistent alpha.
And while there are reasons for skepticism, the performance of AIEQ and AIIQ versus relevant index benchmarks makes this a story worth watching in the coming years.
This article includes a full transcript of the podcast that was posted last week.
Editors' Note: This is the transcript version of the podcast we published last week. We hope you find it useful.
Sign up for Let's Talk ETFs on the podcast platform of your choice to make sure you don't miss an episode:
Jonathan Liss [JL]: For reference purposes, this podcast is being recorded on the morning of Tuesday, December 17, 2019.
My guests today are the co-founders of EquBot, CEO Chida Khatua and COO Art Amador. Chida brings nearly two decades of experience in artificial intelligence in machine learning. Most recently he spent 18 years as the Director of Engineering at Intel (NASDAQ:INTC). He holds several patterns in hardware, software, and technology. He was honored by Business Insider as one of the Top 10 people transforming investing and is one of the Top 100 people transforming business. He holds an MBA from the Haas School of Business at UC Berkeley, and a Masters from Stanford in Electrical Engineering.
Art Amador brings more than a decade of experience in the investment management industry. Most recently he spent eight years serving as a vice president at Fidelity Investments, where he was responsible for over $1.3 billion in assets. In 2012, Art was recognized as Fidelity's Number 1 consultant in the U.S. He holds a professional designation of certified financial planner and has an MBA from the Haas School of Business at UC Berkeley. Welcome to the show gentlemen, it's a thrill I have you both here.
Art Amador [AM]: Thanks for having us.
Chida Khatua [CK]: Thank you, Jonathan. Thanks for having us.
JL: Yes. So, before we dive into today's topic, which I think we're going to go with something like the AI revolution comes to ETFs, does it actually work, we'll see what the editors do at that, but so, I think your very different professional backgrounds are probably one of the biggest strengths of your firm, EquBot, and I'd love to find out more about how you both ended up in the ETF space. Chida, you come from a pretty pure tech background working for the likes of Intel and having a serious engineering education, whereas Art you come from a pretty much straight up financial background, but really on the planning side, not the fund creation and management side. So, just curious how you guys ended up starting this company together.
AA: Yes, Jonathan. So, there was actually - there's also a third co-founder name Chris Natividad. When we met him at UC Berkeley, at that time he was managing $30 billion in a multi-asset class portfolio for Apple (NASDAQ:AAPL). And really if you think about how this whole thing started, it can really be traced back to UC Berkeley, and you classroom these questions, it was actually during our asset management and hedge fund classes that Chida noticed that the amount of data was actually exploding at an exponential pace, and we were having a who two of the asset management industry come into our class and talk about their information advantage in a very narrow field or particular strategy. It was really then that Chida recognized that we could take advantage of the changing information landscape and transform the data into investment insight and deliver this to the larger investment public and also along the way we teamed up with IBMs (NYSE:IBM) wanting to super charge our efforts.
JL: Nice. Okay, you met at Haas and rest is history so to speak.
CK: Yes, pretty much yes.
JL: Cool. Okay so let's start from here. I mean quants have been in the stock selection game for several decades now, power of big data, I mean the whole history of modern portfolio theory goes back to the advent of personal computer really and the ability for buy, I'd say the Ads, anybody just kind of sitting at home or sitting in a computer lab at their university or research facility to [dump realms] of data into computers and see what the results spread out until they come up with strategies that work and it's produced things like smart beta or factor investing, and so you've had quants taking the fact that we have increasing amounts of data available and putting those into play with very specific strategies. What does artificial intelligence bring to the table that is new here. Broadly speaking, how does this approach offer additional chances for Alpha for risk-adjusted outperformance.
CK: Let me barge here. If you think about the traditional Quant investment solutions, it functions almost exclusively on the structure data being like a financial data and market data. The Quant funds on smart beta products like break down as a result of rigidity of rules over a period of time. We see this in a perpetual discussion of like at the core process value or the small versus large cap, momentum versus quality, but if you look at the marketed product time horizon, every specific Quant strategy or a smart beta strategy, it kind of fails. That means the rigidity of rules kind of fails over a broader time horizon.
Whereas the artificial intelligence strategies are dynamic. It is like, it is driven by the market forces, and the current event that is happening along. So, by combining the structure data, which is like financial and market data with unstructured data, which is the news, the different events, social media, social media post or the overall market sentiment and even the global macro environment we complete a very powerful and very dynamic strategy and that's what the AI essentially brings like before we didn't have the capability to combining this world out of financial of the structured information suite, the unstructured information. Under the advent of a technology, we can able to combine it together and that's the power of AI, it's a more dynamic driven, more unstructured improving of the unstructured data.
AA: And just to emphasize a point there that both of you made, it's been said that 90% of the data in existence today was created over the last two years and we think two years from now we'll saying the exact same thing and we think about the data growth. It's actually coming from the unstructured side. Right. Think about how much tweets have impacted our portfolios this year. So, really having the ability to combine that structured data that is kind of that is historical quantitative approach and the unstructured data and the dynamism of news articles and social media really helps us understand not just what to trade, but when to trade it.
JL: Yeah, sure. That's interesting and I assume there is, and we're going to get into the specific strategy and as much as you can explain, you know because I'm sure there is some black box element here that is difficult to put into words, but I assume there is some machine learning component here also which tries to improve on this strategy over time almost in real time to some extent.
CK: Yes, absolutely. Actually, the core of the engine is the machine learning. Otherwise there is no other way you can really combine the soft structured data with a structured data. So, both on structured data and also unstructured data] human use and the media sources, we could be able to learn and it is stock insight. And not only learn [indiscernible] insight, also our models want to use the insight in a most effective way and they learn from every day trade. So, I guess we can talk about it on specific examples.
JL: Yeah, sure. So, you guys have two funds currently trading the $114 million AI Powered Equity ETF (AIEQ) and the $4 million AI Powered International Equity ETF (AIIQ). How exactly does the stock selection process work here and if it would be possible for you to kind of go under the hood and just walk listeners through how the approach works, what the screening process looks like, how the actual selection process works out that would be I think very instructive?
AA: Yes, so Jonathan, let me start with the kind of high-level concept and then Chida you can go ahead and get into more of the specifics. So, the two ETFs are designed to simulate the thousands of research analysts and traders working around the clock, right. The technology is speaking over a dozen different languages and runs around the clock for sharing investment insights seamlessly across sectors, industries, and even geographies. I'll let you to kind of break down exactly how that works, as far as the selection process.
CK: As Art explained, our system kind of mimics thousands of equity analysts working around the clock. So, like similar to equity analysts being a due diligence on a company our systems look like broadly three key areas. So, our investment IT is kind of designed to mimic this quantity model than measures these three key areas like being the financial health, the management themes, the second being the management team, and third being the how - the new - the [indiscernible] really impact the profit of the - appreciation of the stock price as per - like 15,000 global companies, that is how many companies are, they are currently looking at it. So, the companies are looking at all the different metrics. We learn and feel the best that we feel the insight on a daily day-to-day basis for different time horizons.
So, the insight coming from all the three different areas, the financials, about the management team, how they are executing and also how the events are really impacting about the company profitability or the stock plus appreciation. So, then our model also [indiscernible] all the insights do some actionable items into the portfolio and that - creating that sort of items based on the portfolio is based on the investment mandate that is being given to the system. So, you can think about [indiscernible] their own unique standard and the system is able to use what is the best actionable item that can they build to need this investment mandates.
AA: Yes, the unique design right of the EquBot platform provides like a clear image of what's going on the market and it's, I get the analogy, AI is a tool that kind of allows us to see the market in high-definition, right, by aggregating all the data and being able to kind of see across the different industries and geographies et cetera. As a person maybe like an individual research analyst that's kind of [cobbling] data together who is maybe working with a few other analysts who really don't know what's in each other's heads, and the view that they put together on the market is more of kind of like a pixelated image, right?
So, if you think about it, right, everyone wants their investment manager to have an information advantage, right. They want the manager to have more market data, right, not less. And really that's what our system is about. It's having that information advantage and using the insights to make those decisions.
JL: Okay, sure, but so, this is not like, some kind of a high frequency trading system, I assume the holdings that are in the portfolios are - what's the typical holding time for those positions, I seem it's a relatively long holding period, is it not?
AA: Yes, the objective of both the funds right is to identify companies that have the greatest potential for price appreciation like over the next 12 months, right. So, you're looking at a several month time window, and generally what we've kind of seen too is there are some core holdings that have been in the portfolio since inception and kind of continue to be there for a longer period of time where there might be some other positions kind of along the edges, you picked up from there is more like satellite positions and that do kind of change more and more frequently and I know a little bit later we're going to probably talk a little bit about some of the turnover within the portfolio, so happy to get more of the details there, but really this, you are absolutely right. It's not about high frequency trading, it's to identify these opportunities that are accumulating over months.
JL: Sure. So, even though everything is being tracked and you're looking at these myriad sources and inputs both as you both describe today, I think in particular the traditional fundamental and financial data balance sheets and earnings statements and all that kind of stuff and then you're looking at things like social media and other kinds of things that are kind of unfolding in real time that really is just to build up a longer term picture of the different holdings that doesn't mean that your portfolio is actually trading off of a tweet or something like that?
CK: Yes, you are right. It is …
JL: No, I just wanted to clarify that for listeners, that's all.
JL: Cool. So, love to just get this question out of the way up from before we dig in further here to this specific approach and then to just AI generally, and how it could potentially work with the system as dynamic as financial markets. So, if your approach genuinely consistently delivers alpha why rule it out in an ETF wrapper and collect just 50 years, 60 years, how many basis points? Why not open in as a hedge fund and collect 2 and 20 or I mean I know very few funds are collecting 2 and 20 of these days, but I don't know 1.5 in 12 running this kind of a strategy instead why rule it out is in low cost ETF, specifically?
CK: Well, so, the financial access is as a part of our mission, a part of EquBot missions and you can choose that inspiration of the financial access from all our rules like UC Berkeley Haas School of Business. I say, UC Berkeley spirit you can say. So, when we started, we did have options to use a technology for a selective in terms of a putting in a wrapper of a hedge fund, but right from the beginning, we wanted to touch more lives in the technology and giving financial access to as many people as possible. So, we chose the best possible route instead of hedge funds like what is the best way we can really reach out to as many people and ETF has been [indiscernible] similarly all the future product that we are working on is more of a bringing access and bringing exposure to every individual that are interested in using say as a tool for the currency products.
JL: Nice, yeah, no I like that. That's I mean, I had a sense that there is some obviously everybody wants to make some money in the markets, but that there was some altruistic purpose here also in terms of launching product of this sort in a low-cost exchange-traded fund wrapper and not just kind of pocketing it all for yourselves. So, kudos for that. Back in October 1996, Renaissance Technologies founder Jim Simons gave an interview to Financial World Magazine where he said the following when asked how he had essentially discovered Alpha. "There is not such thing as the goose that lays the golden egg forever, the system is always and leaking and we keep having to add water to keep it ahead of the game." And I think what he was getting at there is that financial markets are a highly dynamic system, maybe one of the most dynamic systems ever constructed by humans, where - and the system is so dynamic that measurable characteristics like a company's financials and book value are in a constant state of interplay with characteristics that are at least at present and for the foreseeable future and possible to measure.
So, things like mass psychology, herd behavior, external black swan events that are inherently unknowable until after they have occurred and so any AI that is narrowly focused would seem destined to come up short in this kind of a framework. How is your system designed to handle the extremely dynamic nature of markets?
CK: Yes. Jonathan, you are absolutely right. So, the key thing is that the design of the system is critical for the AI platform to succeed in long-term. So, a rare platform - a platform which is focused on a single aspect of a market are specific limiting aspects of market led to a simpler mechanism like thinking about, for example the AI system only looking at the market sentiment or only looking at the [indiscernible] is bound to miss the overall market. So, the way we designed the system is to mimic as our current process really works, like what will be a grand scale.
Looking at all the possible vectors in the market and at the same time what AI enables us to kind of sharing the learning across - like we talked about our system [indiscernible] research analysts, so imagine you have this [indiscernible] analyst, but at the same time there are [indiscernible] learning instantaneously across all of them. So, that is the technology that kind of brings us, so it could look at all the aspects of the company, about the finances, about the market, about the events and about the management. And then finally what is the best thing that it can act upon.
So, also more several things around that. First of all, the maximum amount of data is very, very important. So, we have to go through the data and understand which one is worth paying the attention to it. So, our IP is around finally aggregating the information and finding those - the information that is not relevant or finance [indiscernible] for example and the noise, separating noise from the real segments. And also, another thing is the [models are designed to] specific to companies across multiple - we have kind of different models for financials, models for management teams, models about how the influence is really acting like hiring the opportunity of information, using our AI technology. That kind of leads to the distinct advantage as the platform continues to learn from every trade and every market movement that happens.
At the same time, we feel that the people really haven't reached their investment potential by simply investing the index kind of product. The technology that we utilize at EquBot, we believe it has got a regular sense of market as a data source at significant advantage of selecting investment with high-quality of market appreciations. So, as I said, the market is dynamic. So, the more the data you can integrate and finally separating the insight from the noise and finding actionable items from that that gives you the interest and advantage over a long period of time and it is a very dynamic nature.
JL: Yeah, sure. And I guess, I mean hypothetically we're obviously nowhere near this point, but just like you have a bunch of different active managers doing battle, each one is trying to outperform the - whatever index, they are benchmarking themselves against and some of them are going to outperform, some of them are going to underperform and you end up with a overall kind of typical market return when you aggregate all of that together with all the passive investors, I assume that the concept that just because somebody is deploying AI and machine learning and has a model to try to understand and be markets will be at a point - at some point in the future where everyone is deploying these systems and so again the returns will only be as good as the system are. So, you'll have AIs that are repeatedly beating the market and AIs that are repeatedly underperforming relative to the other ones. So, to some extent, there is some level of zero-sum game here, correct?
CK: Yes and no. The market, first of all the opportunity is not static, [indiscernible] the fixed goal. The opportunity moves every time you ask them certain information the opportunity, the goal of opportunity, the [goal post] opportunity moves and our systems might design and need to improve over time and need to be more dynamic and understanding how the - where the movement and opportunities will be happening. So, we model, like whose insights to collect insight to learn on and whose new data to look at over a period of time.
So, what it enables us, it is a self-learning system. Our modules learn from every action that has been happening in the market and when all such system getting little more smarter and making better decision over a period of time, where we're traded also gives more insight on how that could have been paired out on a different time horizon. And also, on the top of line, that data is very, very important in AI type of systems and we are continuously looking at what additional data we can integrate to our system.
JL: So, I guess let's get into the funds themselves here. AIEQ and AIIQ, which are the U.S. and International versions of the funds that - the strategy that you've rolled out here. So, I guess, let's start with the top holding in each fund, it's Alphabet (NASDAQ:GOOG) (NASDAQ:GOOGL) that's the share class you are holding there in AIEQ's case and then Toyota Motors (NYSE:TM) in AIIQs case, is it possible for you just to kind of give more in the window for investors that are trying to understand what the process looks like? Is it possible to actually come up with some kind of narrative or explanation of how those ended up being their top holdings in each of these funds respectively?
CK: Well absolutely. So, the beauty of our system is it gives a certain amount of observability. Our system actually is not a black box, so we can look at different aspects of systems that I explained there are like three key important areas that system looks at it. So, you can think of our system looking at alphabet and the Toyota among those three different vectors. The three different vectors being the financial health or the financial opportunities, the management scores or the management efficiency, and the events affecting the price appreciation of those companies. For alphabet, actually it is almost everybody know alphabet, they have a sticky customer base and their market size is continuously in the growth path.
JL: Yes, one might think they have a captive.
CK: Yes. Yes, absolutely. And also, Google management is one of the most innovative companies in the world and the countries that are growing they are one of the adjacent market they can get into. So, all those things factors into a system, it quantifies what management also brings into the table. And also, the adjacent market they are continuously bringing additional market growth and increasing the customer lifetime value. And all sort of systems completely getting events that is supporting that argument and supporting the market dominance of Google is going forward. So, if you see on a risk-adjusted basis the Google comes out of a much more higher opportunity for reason to have.
JL: Before we get into Toyota here, how does - how do potential, I don't know if I would call this a Black Swan event, because it's a known theoretical possibility, but Google is one of the four companies that has come under increasing scrutiny by the regulatory authorities, by congress, and there is the theoretical possibility at least that some regulatory regime will take companies like Google and Amazon and break them up into smaller companies because of their interpretation of monopoly laws in the U.S. So, I don't want to get into the specifics of what may or may not happen because I don't think anyone knows, but I'm just wondering how that sort of tail risk is evaluated in the system because it's kind of a known unknown it's possible, nothing will come off of the investigations happening in congress, where it's possible it could materially really effect these stocks underlying prices and then business models?
CK: So, king of [indiscernible] in terms of Google, we rely on training our model on the past data. There is absolutely like no 100% way to predict [indiscernible] designs and how they are going to be played out. There is enough information's out there in the past and see how Google has been done like, Google actually is much better spotting in terms of withstanding this economic weakness and different hysterical event that has happened. It is actually much more resilient stocks or company to kind of readout the black swan event than any other companies out there. So, the system also looks at those things. At the same time, there is no absolute 100% crystal ball you can say that you can predict everything, but you learn from the past historical data what is the best options are there among all the rest of the market, right. So, Google definitely comes out as a better than many others.
JL: Okay. And you are working in probabilities obviously not in certainties. Okay, so Toyota was the other one and what's the story there?
CK: The Toyota is a power house in terms of financial opportunity. It is being going strong and recently there is more of a market flattening, but at the same time, the way we see the curve that'll diversify. And trying to take [indiscernible] there and the worldwide automobile market. Thought of investing in [indiscernible] and taking step in many other different ventures. And if you think about the Toyota, the management efficiency is one of the best. The Toyota management team has mastered in terms of our personal efficiency. At the same time, Toyota also is great in person innovations.
You must have heard the Toyota partnering with the sub-vendor developing of [not thriving]. So, as I said, our system longs from the historical data and how - based on different events, how the financial opportunity has changed, how the management have really acted on and how they influence the performance of the company and how the different events have really reflected in the past. So, factoring all the three into the picture, Toyota definitely is a much better in terms of opportunity value going forward, among all the international positions that we have. So, the system could able to allocate higher positions for the Toyota.
JL: Okay cool, yeah, I know it's great. So, these funds are both described as actively managed funds. The methods that you are using seem like at least on the surface and I'm sure there is a lot I'm not understanding here, but it seems like on the surface this would be right for some kind of a rules-based index strategy. Albeit is one that makes changes fairly actively, but so active generally at least in the way most investors understand and imply some sort of human intervention, which I think is kind of the opposite of why you guys are going for. So, you're trying to remove some of that human behavioral bias from active management. So, I'm just curious how you're using the term active and why this is not bundled in some kind of a rules-based index strategy?
AA: Yeah Jonathan, you're absolutely right. So, we want to remove the human buyers from the investment process, right, because that's actually a benefit to the shareholders and really allows the system to make data driven decisions and then learn from those decisions. So, if you think about it, a rules-based methodology applies a variety of restrictions, right, rules by definition. So, we wanted to, we chose to go active because we wanted to demonstrate the platform in an uninhibited way right that's not structured constraints. Now, that said, you are also right and that we do have and we have constructed indices using AI that can also be extremely effective approach as well. So, this was more of a choice really to demonstrate the power of this system by not putting a lot of restrictions, but this can also be done effectively in index approach.
JL: Interesting and I guess if there is real machine learning going on, it would be hard to set the rules of the index. I mean there is very dynamic indexes obviously indexes that switch between different asset classes when a bunch of different conditions are hit and you know suddenly go from you know being long S&P 500 futures going into fully into one year treasuries or something, but those are still rules that have to be stated in advance. So, I guess, this keeps things open ended in terms of just letting the system kind of go where it will and not having to stipulate those rules in advance, but just to be clear, there is no human intervention here, correct?
AA: That's correct.
JL: Okay. So, you're getting the benefits of indexing in terms of taking the behavioral biases out of the equation. And then hopefully, the advantage of active management that allows a manager to sometimes identify an opportunity and say, you know what, I know there is rules we generally follow, but this is just too good to pass up. So, you're kind of hoping to get that best of both worlds in terms of that?
AA: Absolutely. That's really what AI brings, right, and I think it's kind of blurring the lines between active and also rules-based strategies. It's not so black and white anymore, it's more of a gray area, right? And when you mentioned the human bias thing, I - we really don't want to inject any human bias at all, because just as we as humans learn from our experiences, right, so does the machine, right, it learns from its experiences as well. So, it's important that the system continues to learn and to grow, so it makes better and better decisions, right, and can improve performance on a continuous basis going forward.
JL: Okay. So, digging into the specifics here, are there a set number of holdings, or does it really just depend on how many strong buys, I'm putting that in scare quotes? Because that's kind of a Wall Street South Side thing, but whatever your ranking system is there. Is there a set number of holdings, or is it really open ended in terms of what the model recommends in any given time?
CK: Yes. It is literally investment mandate. So, at the same time, they explained the system prior to we invest in mandate. So, for both ETF, so one of the investment mandates is to have a similar risk as a broader market like for ETF AIEQ it is a broader - same level [indiscernible] broader U.S. market. Some tries to vary the holding count based in the [indiscernible] volatility. So, as a system, our market uncertainty increases. We tend to see the holding count increases, but in general often maintain exposure for AIEQ between the 75 to 150 companies, of course, for long-term market appreciations. Yes. And then the same goes for AIIQ also. It looks at the broader volatility across the international market and try to find the best way and meet it - meet that goal.
AA: So, Jonathan, this kind of touches on one of the questions you asked before, right, which was how does it handle kind of black swan on forcing things, right? So, for example, when the fiscal cliff happened for the first time, right, that was kind of, it was a first-time kind of thing, right? So, you can't go back and say, well, when has the government shutdown before, right? So, the system will try to diversify, right, and spread risk as much as possible, right, in situations like that. So, going back to the Google thing, right, so it's invested in Google, because it's got high confidence in that the past events are predictive right of the future events and you can relate that obviously back to the stock price. So, that's really kind of how it handles those black swan like events.
JL: Got you. Okay. But we're, again, we're not talking about something that can say, what, U.S. equities are just terribly overvalued, and there's only 75 that look reasonable, but we're going to keep some amount of the portfolio in cash or three-month bills or something right now. It's - it is - there are rules in terms of it being committed to be close to 100% invested in the asset class it's covering, correct?
AA: Not exactly. It can't so, for example, you bring up a good point, right market is at all-time highs right now, as of talking today. So, some of the opportunities - the opportunities there might not be as great as, say, for example, in March of 2009, right, when the markets kind of sitting on an all-time low, where the opportunity looks significantly more, so when there's less opportunities to invest in. And so, it does have the ability to go into some cash, right, so that I can wait for opportunities and we've seen this in the past, where issues with trade war kind of flared up, right? The portfolio diversified it, put a little bit in cash. And when the markets sold off a little bit, we saw a little bit of a correction. We saw it take advantage, right, investing in some of the higher beta names like Zendesk and so forth. So, it does use cash as a tool, but we don't expect it to ever have any high amount of cash. I think the highest amount of cash we've seen so far has roughly been maybe about 4%, is that right, Chida?
CK: Yes. That's correct. Yes.
JL: Yes, that's interesting. So, there - but there isn't any prescribed amount like it can only go to X amount theoretically that the system could go higher than that?
CK: Not really. There's - yes, the - there is a mandate to be fully invested, but we use the [indiscernible] as a tool to wither the opportunity - the availability, opportunity and also seeing the market conditions one of the best way to [indiscernible].
AA: One of the objectives, Jonathan, right, for the U.S. fund is to match the broader volatility of the U.S. market and the International one is to match the volatility of the broader international market. So, because of that kind of higher objective or priority, you're not going to say going to cash with a 100% cash or something like that with the funds not moving. It's not ultimately trying to like time the market and totally get out of the way of downturns. It's trying to continue to maintain equity exposure, because really, when you think about it in the role of a portfolio, it's really meant to be a core holding and act like has similar level risks of say, like the S&P or total market, but then deliver excess returns and the international one same thing on that side.
JL: Gotcha. Okay. So, investors will have to wait for you guys to roll out funds and other asset classes to get that, that full rotation, asset class rotation strategy going with AI, let's say for the time being, at least?
AA: And I think that's really kind of how we envisioned expanding our services, right? It's kind of going beyond the asset classes and geographies and really into the launch of AIEQ and AIIQ was to appeal to the largest audience and play a bigger role in the portfolio more of a core holding than, say, a tactical kind of positioning.
JL: Yes. I know that makes a lot of sense. So, what does the annual turnover look like for these funds? And is tax management built into the process here?
AA: Yes, great question. So, the turnover is completely data dependent, right, and the dynamic nature is an asset of the fund. Now, that said, the turnover that we've seen over the past 12 months in both funds has been roughly about 130%. So, it's important to remember that the fund places a higher priority on return in risk over taxes, but still the fund is tax aware, right? So, it is taking some things into consideration, but not - it's not going to let the taxes wag the risk or the performance of the fund. The other thing that I do want to mention is, we're very excited and also optimistic about this new ETF rule, right, which will help not just our two ETFs, but all active funds tax management, right, through the custom basket process.
JL: Yes. I know that should definitely help. So, yes, you're not looking at your end distributions or anything in that sort?
AA: There was a distribution that was issued last year, but this year, no, we're not anticipating any.
JL: Okay, nice. And I guess, not that you would allow, let's say, a theoretical taxable event to wag the dog so to speak, and determine what you were going to buy or sell, but I guess, if all things are equal, let's just say theoretically, that your risk threshold, let's say, you're getting close to crossing it for, let's say, a holding like Toyota, and maybe you're close, but you're not quite there yet, but if you sell it now, then you can ensure that there won't be any taxable event on that position. So, maybe, you kind of just move the goalposts too tiny in terms of tax harvesting and not wait since it's close enough anyway. So, maybe I don't know is there could be some kind of like a tiebreaker or something if all else is equal?
AA: You kind of touched on it a little bit. So, right, the objective is to optimize the taxes, right? But just when you think about priorities, it's below the risk of the performance. So, it's like, hey, if you can lower the taxes, right, without negatively impacting the performance or the risk in the portfolio, right, in the overall opportunity within the fund stays the same, then move forward and try to lower the tax bill in the fund.
JL: Sure. And that's, I mean, this is the beauty of exchange traded funds. I have to imagine that these strategies in a mutual fund rapper without in-kind transactions and now the new ETF rule be much more difficult to actually run this in a tax-efficient manner in that kind of old school mutual fund rapper. So, the ETF is really a godsend in terms of that with it sort of fund is it not?
AA: Yes, absolutely. And then it also goes back to liquidity, right, daily transparency, so it's a more effective tool. And then it goes back to the initial thing, which is the access component, right, because anyone can buy as little as one share of the ETF and now with fractional trading, you can even buy less.
JL: Yes. Yes, absolutely. So why just curious here, because AIIQ has gathered assets at a nice - sorry, AIEQ, the U.S. fund has gathered assets at a really nice clip, and then AIIQ, the International one, which is a little newer, obviously, very little in the way of assets, do you have any insight into why that would be the case? Is it - do you think it's just because it's a U.S. audience, and so they're looking for that core exposure kind of a home bias thing, or is there something else that play here also?
AA: Yes. I think it's a couple of things. I think you hit the nail on the head there, but I think back at my days at fidelity, helping families make investment decisions and thinking about their portfolios. It's difficult to get U.S. investors to invest in international markets. Everyone's got a home bias and as they should.
JL: Especially after the last decade, right?
AA: Yes. You took the words right out of my mouth, that's - it's even harder, right, because every year you have those consultations with these investors and you tell them to put something international, but then you're after a year for the last 10 years. It's underperformed, right? It has underperformed in U.S., right? So, there's that aspect of it. There's also the aspect of AIEQ was the first ever AI-powered ETF, so I think there was a lot of pent-up demand that, that fund received. And then - and so I really think it's those two things.
However, at some point, international will come back and outperform the U.S. And we know AIIQ is in really well-positioned to capture those flows, because if you think about the performance of AIIQ, it's been nothing, but exceptional. So, this year, it's about 29%. And it's beating it's benchmark by more than 7%. So, it's - it couldn't be doing any better. So, it's really well-positioned, and we're seeing a lot of interest from institutional investors, as well as retail investors and having some very constructive talks. So, I think it's a matter of time before we see some larger inflows into that fund. But getting back to the home bias thing, I would always kind of anticipate that AIIQ will - would be a larger fund than AIIQ because of the way in which U.S. investors allocate to their diversified portfolio in the home bias attached to that.
JL: Yes, sure. No, I just had a similar conversation with Ed Lopez over at VanEck because similarly with their Morningstar Wide Moat Funds, they have a U.S. and an international and the U.S. one is similar to you guys. It's like the same kind of multiple, like $15 or $20 in the U.S. fund for every $1 in the international one, and similar strategies, similar solid performance. So, yes, it's definitely something about the whole home bias thing, but, of course, because none of us have crystal balls. If you don't have the fund already created when that secular shift happens, then you don't have the ability to capture the assets. So, we -we've seen so many times in the ETF space of bond sits there with the low amount of assets and then overnight, money just kind of floods into it as things change. So, yes, I'm sure AIIQs will come - it's time will come as well. So, looking at a future, what other sorts of products does EquBot have planned?
CK: That if you look, our system - we talk about our system is like continuously learning and also the same time it is a very highly scalable. That means, we can leave different objectives, and it will try to meet objective the best possible way. Actually, we can run like millions of investment mandates, not only investment - let different investment mandate, but also, we can run across multiple geographies. Right now, we are already looking at like 15,000 global companies and all across the world. So, you can imagine our platform can be used to run more than just ETF. So, that not only different ETF, but also different geographies. Also, you can bring different kinds of instrument across different geographies, like different financing instrument like mutual funds, stocks or product, RIA.
AA: Managed portfolios, that sort of thing I would mention.
CK: Yes. And also, we're already working with some of the leading financial institutions to build the product like that using our platform. And also, we're working to scale our platform across multiple geographies like Asia, Europe, and also our goal is to support like multi-asset class and also in future. So, there are so many different possibilities. And we are being approached for different opportunities and we are selectively choosing who to work with. And there are quite we're excited about things that we can do with the platform.
JL: Yes. No, it is really exciting. So, we'll definitely stay tuned to see what else you guys have coming up here. So, in terms of investors that want to go on and continue researching both the topic and concept of AI and machine learning and investing and then your funds, specifically, what's the best place for people to go online to be able to do that?
AA: Yes. Our website, Jonathan, is probably the best equbot.com. We also have a newsletter that, that investors can sign up for, which talks about future products, as well as provides market insights. So, that would probably be the best place to go. We're also partner with IBM Watson. They've got some training courses as well on AI and Watson, which we utilize a lot, which kind of also provide some one-on-one tutorials. So, really, I think those are probably the best places.
JL: Oh, nice. Okay, cool. And then what about social, are you guys on Twitter or LinkedIn or anything in that sort?
AA: Of course. So, Twitter is probably the best, but you can also follow us on Facebook and LinkedIn.
JL: Okay. So, what's your handle on Twitter?
JL: So, E-Q-U-B-O-T, nice. I'm now following you guys. Excellent. And when do you get the machine to actually start doing the tweeting for you, that sort of thing?
CK: [Indiscernible] why not?
JL: Chida, I see that. I'm following you. Also, I see you're on here separately. Very nice. So, listeners can go, they can follow you guys there, stay abreast of everything that's going on. Anyway, any - anything you guys wanted to leave our listeners with before we go here?
AA: Yes. I think the important thing to think about right now, right, is go back to what's changed over the past couple of years, right? And really 90% of the data in existence today was created in the past two years, right? And that growth is going to exponentially - it's only going to exponentially continue, right? So, when you think about investing, right, the game has really changed because of this explosion of data, because it is impacting your portfolio, right, every single day. And so, you need a system that can adapt to the ever-changing information environment. And so, we think AIQ, AIIQ are exceptional tools, right, to be doing that. And we're very proud of how AIIQ has really outperformed its benchmark by several hundred basis points, not just this year, but since inception.
JL: That is, that's exciting.
CK: Just to add to that point, like the - if you see that the growth of data, the only tool that really can keep up is the - by using this in learning. And the way it is trending right now, we believe that the majority of asset that direct or indirectly will be influenced by the machine learning or AI. So - and it's good time to be engaged and understand and be aware of what is happening on the Artificial Intelligence space.
JL: Yes, sure. I think that's a good place to leave it. Anyway, I wanted to thank you, Chida and Art for taking part in this conversation for enlightening our listeners about what you're doing here, which I think is really unique and fascinating and definitely look forward to following your progress over time and seeing how this goes, because it definitely seems like it could be the wave of the future and you're doing it right now. So, that's very cool.
CK: Thank you, Jonathan. Thanks for the opportunity.
JL: Yes, absolutely.
Disclosure: I am/we are long AIIQ, AIEQ. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: For Disclosures, both Chida Khatua and Art Amador are personally long AIIQ and AIEQ. For a full list of holdings, updated daily, listeners can go to equbot.com/...
Jonathan Liss does not have positions in any of the stocks or funds mentioned in today's show.