In this episode of Markets in Focus
Now that the era of “growth at any price” has ended, newly converted value enthusiasts are questioning margins and the true return on their investments. Tariq Siddiqi, CFA, Senior Research Analyst at Eagle Asset Management, refuses to follow the crowd and keeps growth profiles at the forefront, pinpointing companies best positioned to capitalize on persistent secular trends. For Siddiqi, these opportunities include cybersecurity, the shift to cloud computing, the “Internet of Things" (IoT), and the growth of artificial intelligence.
Subscribe where you listen to podcasts
The most rapid monetary tightening in over 40 years is still filtering through both the economy and the stock market. With rates rapidly moving higher, the era of “growth at any price” is all but over, and investors are reprioritizing profitability and reasonable valuations. Last year there were not many places to hide in sectors like information technology and communication services where some of the most expensive stocks were concentrated. Nearly every company was thrown out in the rush to exit higher-duration assets. But as a result, that has created what I would say are some very interesting long-term opportunities.
A secular growth trend that has been firmly in place didn't suddenly cease to exist. The need for things like cybersecurity, the transition to cloud computing, the internet of things, the growth of artificial intelligence, or AI; there's tremendous innovation that is still taking place and there are some very good companies that are exposed to the growth of these themes. But for the past year, tech has largely been viewed as a monolith and generally been treated as such by investors. I would argue that this is far from reality, and there are still some really interesting areas that investors should be looking at in the future.
This is Markets in Focus for Raymond James Investment Management. I'm your host, Matt Orton, and I invite you to join me and my colleagues as we discuss the latest trends and developments driving the markets. Visit us at marketsinfocuspodcast.com for additional episodes and insights.
So given this backdrop, I want to welcome back Tariq Siddiqi, who's a senior research analyst at Eagle Asset Management to help explore some of these secular growth themes, new technologies, and how he's looking at them from an investment standpoint. Tariq, thanks for being here today.
Great to be here. Thank you.
Diving right in, I would say enough ink has been spilled on the weakness within information technology last year, so we don't need to do that. And I think then it's worth maybe discussing some of the key secular growth themes that you're paying attention to right now, which are particularly relevant I'd say given your focus on small and mid-cap information technology.
Yeah, thanks. I think it's important to talk about various themes that remain ever important, and in some cases have only become more relevant over the last year. So despite the volatility in the market that you talked about, we continue to follow some very important themes. Markets will do what they have to do in terms of various macro factors and everything else. But what keeps me excited day in, day out about technology investing are these broader themes.
In our group, we follow quite a few themes, but I can certainly discuss some of the ones that are very top of mind for us. So first, let's talk about this thing called artificial intelligence. AI continues to be a transformative technology across various industries. We're seeing things like chatbots and I'm sure we'll talk a little bit more about ChatGPT; image recognition uses AI; predictive analytics is really working with AI. To be fair though, AI has been in use for quite some time. It's not something that just came over in the last six months. The various algorithms that Google and other search engines have been relying on are basically AI. Something as ordinary as your spam filter in your email is a form of AI, and Siri and Alexa on your consumer devices are using AI to understand human speech and then respond back. Automated decision making and things like games or even self-driving cars. AI is being adopted in almost every industry. So that's one major theme that will have implications for the next few years.
Moving on, Internet of Things (IoT) is a very important, impactful theme. The IoT market is rapidly expanding with more and more devices and sensors connected to the internet and sharing data back and forth. I believe that companies that are developing and selling IoT solutions for things in your smart homes, smart cities, and various industries really offer significant growth potential.
Another major theme is cloud computing. It has become essential for businesses of all sizes now. Companies that provide cloud-based infrastructure, software, and platform solutions are growing in popularity. This is a theme that — quite frankly — has been important for a better part of my entire career, but remains as relevant as ever. The benefits of cloud usage versus installed, on-premise, old stuff, let's just call it, are just so overwhelmingly in favor of the cloud camp now that less and less use cases make sense for remaining on-premise. Cloud computing is even conquering the last areas, such as top national government level security projects. Core banking software is now going cloud, so a lot of really good growth for cloud computing.
Another area that we are looking at is cybersecurity. With the number of cyberattacks that are happening, it remains a very crucial theme for us. Regardless of what the macro environment looks like, management teams continue to highlight the importance of cybersecurity in their budgets, and it usually remains like number one, number two in their budget. The CIO, the CSO, the CFO, they all agree, we have to have security. Of course, there are times of accelerating spend on cybersecurity and then, perhaps times like now, when there's some digestion period. But cybersecurity remains a high priority spending item for most companies.
Another important theme is renewable energy. Climate change and sustainability are important issues, and renewable energy is a potential investment theme for investors looking to support a cleaner energy future. Companies that develop renewable energy solutions and infrastructure — or components that play on those — are all interesting investment ideas for us. And related to that, EVs, electric vehicles, are gaining popularity as more and more consumers and governments seek to reduce emissions from transportation. Companies that develop electric vehicles, charging stations, and related technologies offer significant growth potential over the next decade. And governments continue to offer subsidies and other incentives for adoption of these cars. So that's a really strong one, too, for the next couple of years into the decade. These are all the themes that we follow on a regular basis.
That's no shortage of themes that you're following right now, but I think it highlights how we started this off: that there's a lot of really powerful secular growth themes. And just to build on this a little bit, Tariq, what do you think the market is missing with respect to some of these growth themes? Because the baby seems to have been thrown out with the valuation bathwater and the selloff that we had last year. Which of these themes are investible right now?
I feel like all of these have very interesting areas that you should be looking into. Really what it comes down to for us is that in these kind of wild gyrations, these debates that we have on a daily, weekly basis about what to pay for something and how to utilize it — how to utilize it in our portfolio — that's something that continues to happen on a daily basis.
Sometimes we see euphoria in these themes, right? And sometimes everyone gets so disappointed or so focused on macro issues that these themes seem to be forgotten. The same exact investors that believed cloud computing companies can grow 30, 40, 50% in perpetuity and never have to show any profits have all of a sudden become value investors asking deep questions about the true payback on a cloud investment, what their margin is going to look like, and what the payback is. They couldn't care any more about what the growth profile is.
So our job, as investors here, is to remain somewhere in the middle and try to zig when the masses are sort of zagging. I believe that these themes remain relevant, all of them remain relevant in some way, shape, or form, and our job is find the companies that are best positioned to take advantage of these large thematic waves. And try, of course, not to overpay for them.
That's great. And so maybe two themes to dive a little bit deeper into, at least for right now: cybersecurity and artificial intelligence. Cyber is one that's always discussed with — I would say — really, really broad brush strokes, but maybe you can help break it down. What are the key segments on which companies actually focus? What are they actually delivering to an end client, and are there any parts of the IT (information technology) market that are tangential beneficiaries to growth here, like semiconductors, or something like that?
Of course. Of the two, let's start with cybersecurity first. This is a complex area. There are quite a few areas of discussion, but in aggregate, cybersecurity is basically every aspect of protecting an organization and its assets and employees from cyber threats, right? Given how every organization of any size now is connected with the internet to accomplish their tasks, the addressable market or the number of customers is almost every company. And in dollar terms, something like $150 billion is spent on cybersecurity annually, and it’s growing at two to three times the rate of GDP (Gross Domestic Product) globally.
And I don't really have to explain the consequences of not securing one's networks, right? Small companies have gone out of business due to cyberattacks, and large ones have lost millions of dollars to hackers, or paid millions in fines, or hurt their business reputation, and in some cases suffered all three. So really cybersecurity is very, very important.
Within cybersecurity though, there are major buckets that we look at. The first one — the big one that I will talk about — is network security. This is where the net attacks happen on the network. So what you want to do is secure the data and the access control: who can get in and out of the network, where they're allowed to go, which assets they can have access to, what type of data they can see and manipulate. This could be the general firewall that's on the perimeter of the network, as well as down to the individual hardware assets or even down to individual files on that network. You have things like intrusion prevention in here, antivirus, sandboxing. You can also add in network analytics, threat hunting, and response technologies.
After network, the next big one is cloud security. As I discussed earlier, cloud is everywhere now. These cloud security solutions are really controls, like policies and services that help protect an organization’s entire cloud deployment. Whether that's the applications, the data, the infrastructure, et cetera, all of this has to be protected against an attack. Some of the large hyper-scale providers have their own security solutions that come with the cloud offering, but usually they don't have the expertise and the in-depth knowledge of security issues that third-party individual cybersecurity providers have.
After that, we have endpoint security. This is the security that's basically securing the PC that is sitting in the office, or on the remote campus. And now, of course, over the last couple of years we've had the whole work from home situation. Protecting whatever that PC is doing, or even that smartphone that can access an enterprise's network, or just that email — anti-phishing, anti-ransomware — it’s all part of this whole endpoint security bucket.
Moving on, we have IoT security, as I talked a little bit about in the theme. This is an emerging issue: as a growing number of devices can talk over the internet and share data back and forth, it immediately increases the risk of hacking those devices. Usually IoT devices tend to have lower-level software code, so hacking them, in a way, is harder and less fruitful. But as their capabilities increase, and the number of devices increases, IoT devices are becoming an entrance into the main networks.
Of course, there are a lot of good and not-so-good companies in the space. Technology changes rapidly, as threats evolve, and the companies that have a great point solution today can easily be left in the dust by a better solution tomorrow. Or the threat evolves to another vulnerable area, and the budget dollars shift. You have to remain very much on top of this theme.
So even within the theme, it's fairly complex. When you look at and identify investment opportunities now, are there companies that provide all of these different services? Are there companies that specialize in, say, just network security or cloud security? Do you look at those types of companies differently?
No. I mean, what we have are companies that often provide a bundle of things that you can have. You can have some of the cloud stuff in there. You can have network stuff in there, IoT in there, different kinds of things in there. And then you've got point solutions. So you do have to look at them differently. The ones that are more platform based, the ones that can do sort of all of everything, sometimes they win because a large enterprise is like, “well, I'm going to standardize on your platform and we'll figure out how to do the small little things.”
But then you also have best of breed, like if you need something in network intrusion, that's your one small problem, you want the best of the breed there. Well, then the platform guy is probably not going to be good enough for you. You go for somebody else.
So that's how you have to understand: okay, which bucket do you fit in? Are you more of a platform guy? Are you more of a point solution guy? You don't want to be in a situation where you're the bad point solution guy, or you're a platform that's missing some big chunks. That's how we try to figure out the buckets within the investment field.
No, that makes a ton of sense. And then tangentially to that, are there other plays within information technology that build off of the theme, that might not be direct investments in the theme, for example?
I would think there, I mean, of course within cybersecurity there's certainly a lot of other things that would come along. You definitely need to spend a lot on the infrastructure to make sure you can have visibility into your network. So you can have the software, you can have the cybersecurity solution, whatever the different things that are there, but you have to have the infrastructure, you have to make sure that, hey, we do realize there's a threat here. We do realize there's an intrusion here. We do realize there's something. Okay, how do we counter that? How do we make sure that we have the kind of remediation tools — if you will — to be able to attack that? So that's related to cybersecurity, some of the things that you have to spend on top of the cybersecurity solution directly, to make sure that I can address that theme properly.
Let's move on to artificial intelligence, because that's another big theme that you've mentioned. It's been in focus a ton lately, and it's clearly much broader than just ChatGPT, which is taking the world by storm right now. What are some of the key growth areas within AI, and how should we think about it beyond just IT?
AI is definitely the buzzword at the moment. The number of management teams talking about using AI has just exploded in the recent earnings calls and the conversations that we have. But before we get there, let's sort of define what AI is: it is basically letting the computer decide, based on certain inputs “on its own,” quote unquote. But it can be a bit simplistic. What AI means in the current scenario is using algorithms trained on massive amounts of data to come up with pattern recognition, make those decisions, and basically make judgments like a human might.
In order to do that, AI requires a foundation of specialized hardware: a semiconductor chip called a GPU, a graphical processor that is really, really focused on training that algorithm on that massive amount of data that we're talking about. You have to have the software for writing those training machine learning algorithms. So we can feed it all that text and train the algorithm and can create a chatbot that can access questions that mimic how a human might, because it has seen a similar word pattern among the millions and millions and millions lines of text it has actually read. So feed it a ton of images, it starts creating images better than those millions of images. Feed it a lot of other kinds of HTML code, it will start to spit out HTML code based on what it has seen among millions and millions of lines of codes.
So AI can be great at a whole lot of repetitive tasks that a human might waste time doing, but it has its limitations, based on how creative it can be. So the most immediate impact, as I just said on the semiconductor chip side of things, GPUs are going to be important, memory's going to be very important.
These large data centers are needed to train these algorithms, do all that work to make sure that the AI is actually spinning out the right answers for you, right? And then, not only do you have to train it, but you also have to fine tune it. This is a new concept. With ChatGPT, the latest version that really caught the world by storm, is not the fact that we had the big training data. That's been going on for some time. What ChatGPT’s recent version did was that it basically fine-tuned it by using human input to say, hey, no, that doesn't make sense. When you spit that answer back, that didn't make sense. Use it this way.
What it did was that it really gave it a lot more polish, if you will, to the answers coming back, and it made feel a lot more human-like. That was a huge, huge improvement. And for all of that, what you need is a lot of semiconductor usage. And then with that, of course, with every chip, you have other things in the data center that interact with that chip. You've got optical components, you've got ethernet components, you've got servers, obviously, you've got memory, you've got the whole entire data center infrastructure market, which will see an uplift from AI going up.
One of the things that we were fascinated by: Google talked about how an AI-based search query is about 10 times as intense on their backend than the standard queries that we've all been doing for years now. This intensity gap — I mean, AI-based query versus a standard query — will shrink over time as AI gets better, but it does show you that while this AI is cool, it costs a lot of money.
Along the way, you need investment in software tools to build those applications that can utilize what the algorithm is saying, sort of making sense of it in human terms. A lot more software will be sold to people who will need to make sense of what I can do for them and how they can automate various tasks by utilizing it.
One of the recent examples that I saw is basically, from a coding point of view, a whole bunch of coders. Now that initial — call it 20, 30, 40% of the time — they did it just on repetitive tasks, setting up the code, putting in the initial lines, setting up the libraries, all of that can just be done automatically through AI, via ChatGPT, or what have you. Now developers can go in and actually do the real creative aspect of it.
A lot of software tools will be sold to various end markets, to explain to them “Hey, how can I use this amazing thing” beyond the silly stuff that we've all been using ChatGPT for?
You also have to remember that other tools are required to continuously train this AI algorithm. What you have, the beauty of this system, is not the fact that just, hey, once we train it, we're done. It's a continuous system to improve the algorithm, and that's what's going to be very interesting to see: the kind of tools that will be developed to continue to improve this stuff to get better and better and better.
Yeah, that's a great point. The continued training investment that's going to be required in that. And then, I mean, just the staggering statistic that you had of an AI-based search query being 10 times more intense on the backend than a regular query. I mean, it really highlights how much will need to be invested in this at the end of the day. And I've got to ask you, before we leave the AI subject, have you used ChatGPT or Bard? Were you impressed?
I know I have my own opinions, and I haven't really used it for many great, useful cases, but how does it get leveraged beyond search engines? I know you touched on that just a little bit, but based on my own interactions, and my questions to GPT in responses, we're still a long way from robots taking over the world or robots taking over our jobs. I'd love to get your input on that, Tariq.
Yeah, no, absolutely. ChatGPT is sort of the “aha” moment in AI, right? For the longest time, as I said, we have used various aspects of AI. We've used this automation to generate responses from the computer when you're writing your email, and Gmail, and so forth. Even Outlook does that now. It tells you the next word. I mean, those are all low level AI, as it's called. But it was less than useful.
However, ChatGPT really showed how amazing AI can be on the consumer level. It's like, I guess, how the Netscape browser was back in the mid-1990s: it showed you the power of the internet. The internet was there, and it was being used by universities and research facilities, et cetera, but not something ordinary people could easily use. ChatGPT is that killer app that shows you the power of AI. So I think it's going to be very fantastic.
We have already seen instances within the coding side of things, and people writing emails. From a personal point of view, I love it. I've not used Google's Bard, and I know some issues have come up with that, but I have used ChatGPT to write outlines of things I'm working on. I would like to say that I came up with this thought process of our conversation that we're having by using ChatGPT, but it's not that good.
I have certainly drafted tons of emails, the quick emails that you need to do a lot of stuff on, ChatGPT's great for that. I've tested it by asking existential questions about life, death, and everything in between, and figured out some interesting answers from there. I've tested out the whole computer code writing thing on AI. It's strong. It's powerful. I used to do computer coding way, way, way back in the day, and this is way, way ahead of whatever I was able to do a few years ago. It's pretty amazing from that point of view.
I feel like it's a great starting point for various projects, but you have to work on it to make sure. You cannot use that auto-generated response that ChatGPT gives you, but rather edit it and make sure, first of all, is it accurate or not, right? That's the biggest issue. It’s that you have to make sure, first of all, that the accuracy is there, it's using the right facts, et cetera, and then you can adapt it to your personal style, and sort of present your viewpoints, and so forth. That's where it really gets interesting. It's great for silly jokes. I've used it to create a bunch of dad jokes that I try with my kids. Oddly enough, I find ChatGPT's jokes way funnier than my kids do, but that's probably not the AI's fault.
Yes, it's funny how we find interesting uses for all of this stuff as it comes up. But I'm going to do a 180 as well, because I think one area — as we build on this conversation and we think about applying all of these different secular growth themes — another area we don't talk a lot about is industrial technology. Thinking about construction, manufacturing areas: how are those parts of the economy leveraging these sorts of technologies in their development?
I feel like this is a very exciting area. We're on the cusp, I mean, I believe we're on the cusp of a very big, but slow-and-measured adoption of advanced technology tools in industrial projects. This is not a consumer area, which adopts things quickly. It's not an enterprise data center, which comes in very quickly. It is slower, but it's a big market.
We already see software being used to plan, design, and construct and build things, of course. However, as the technology is getting better and more user-friendly — especially using the cloud computing that we were talking about — it opens up access to so many more users than before. I believe industrial tech is a very exciting area.
One of the most exciting areas, at least to me, is this concept called digital twins. In effect, digital twins is a virtual model designed to reflect an actual physical item. The digital twin is a virtual copy of that physical asset. It could be something as small as an engine in a toy car, or as massive and complex as a multi-billion-dollar project for a building or a massive bridge. But with digital twins, the advantages are that you can then simulate the various environmental impacts before the project even begins. And more importantly, after the project's completion, the digital twin technology can help diagnose those problems and reduce the number of times that actual physical activity is required for that object.
And Tariq, how is that different from simulation or design software, which are things we've had for, I'll say, decades at this point?
That's a good question. Really the difference is the marrying of that computer aided design, the CAD design, the blueprints, with the real-time sensor and the IoT data coming from the equipment, or whatever the product is, right? Real-world input is fed into the digital twin so that the digital twin on the computer maintains fidelity with the real-world object that's in the field. The wear and tear on the equipment, for example, can be updated in real time. So you know which part is about to fail before it actually fails and causes damage. You want to be able to be proactive, rather than reactive, and digital twins help you do that. Simulation is basically before the actual product is built; use that for planning, and designing, and so forth. That digital twin helps you maintain that object in the real world after it's been deployed.
As we think about and talk about these major secular growth themes, I don't think we can say we've done a good job unless we touch on this: it's electric vehicles. It's a major growth engine for traditional car makers. It's given rise to a whole new group of auto manufacturers, but this growth is powered by technology and chips. So Tariq, how do you look at the growth opportunity within EVs from your technology standpoint, and how do you segment those investible opportunities?
The electric vehicles are just amazing. I know there is some debate regarding whether they're actually less harmful when you've taken into consideration the entire production process and where the electricity is being sourced from, the mining for the various rare earth metals, et cetera, and the lithium that goes in there. There is certainly some debate there. But overall, from my point of view, I’m from the camp that EVs are just better vehicles. There are things that need to get better in the vehicles themselves, such as charging time and range, as well as the infrastructure needed to support the vehicles. But over time, EVs should continue to take share from your traditional ICE (internal combustion engine) vehicles. Almost every auto manufacturer is trying to replicate the success of the leader in this space.
Besides the auto manufacturers themselves, we are looking at quite a few different ways to play this major theme. One area is in the chips that are being used in these vehicles; the semiconductor content rises dramatically from traditional vehicle to an electric vehicle. Battery management, inverters, powertrain, and motor management are all areas where semiconductor content usage rises dramatically in these new types of vehicles.
Add in things like automated driving, with sensors and cameras, better infotainment systems, and an electric vehicle can have as much as 10 times the semiconductor content of a traditional vehicle. I believe we've done an earlier episode on semiconductors that really just focused a lot on the auto market and how the semiconductor usage is in the auto market.
And then for us, one specific area within semiconductors for EVs that we're focusing on is something called silicon carbide. This is a type of semiconductor that helps to charge the car faster and retain more of that charge. That's the help for that range that I was talking about as one of the things that need to get better. Silicon carbide is definitely part of the answer how to make EVs better.
Silicon carbide inverted chips are chips that can handle much higher voltage coming into the car at a much higher temperature, as opposed to traditional silicon — what they call IGBT (insulated-gate bipolar transistor) inverters — that have a low tolerance for voltage. Silicon-carbide-based inverters can tolerate a lot more voltage coming in. Plus, silicon carbide converters tend to be smaller in size, weigh a lot less, and their power density is a lot higher. So all in, you can have a lighter vehicle or use more of those silicon carbide inverters to increase the speed with which those vehicles charge.
With silicon carbide in vehicles themselves and the charging stations, there's a ton of growth coming for this product. It does require a lot of capex and some cost reductions. So you are seeing companies announce big factories to get the scale to reduce the costs, so it can be as broadly used as expected. Big factories are going up in New York. There's one in North Carolina, another one in Colorado, and one in Europe, a big announcement in Europe from Germany. They'll be serving a huge, huge wave of demand for silicon carbide chips. We are very excited about this area for the next five years or so.
Another area related to EVs, although it doesn't specifically have anything to do with the motors or powertrains, is what I was talking about with automated driving, what is called ADAS (advanced driver assistance systems). Given auto manufacturers are already revolutionizing how the car's internal system is working, they've also been adopting tools that help the car drive itself. Although there's one auto manufacturer that's taken a lead in marketing this, the technology is far, far from perfect. As somebody who has played with the system, and the beta version, and the high-end betas, it is very, very far from perfect.
For a good, functioning, high-level ADAS system, I am in the camp that you need cameras to take in the visual of what's around the car. You may also need these laser-based radars, the lidars, and a whole bunch of companies provide these kinds of tools. And then going back to the AI discussion earlier, this whole concept of using cameras, and sensors, and lidars to make sense of what's happening around the vehicles, it's called machine vision. The algorithm is making that decision on the fly.
You need a lot of processing power, in terms of the chips, and then for connecting it back via ethernet-based connectivity to various parts of the car. And it all needs to happen a hundred percent correctly, and very quickly. It has serious ramifications for a wrong decision with a vehicle going 60 miles an hour, and they are very different than a wrong answer for my ChatGPT query.
So yeah, there are a lot of very interesting areas within this new car theme. Whether it's in cool new infotainment systems, multiple displays, internal combustion being replaced by electric motors, and even within the electric motors: there are different types of electric motors and how they're being utilized. Battery technology needs to get significantly better. There are a lot of different ideas here that we've been looking at that we're involved with currently, as well as looking for in the future.
Well that's great. And there's a lot I'm excited about just listening to all of this, and I know we're running low on time. So I've got one final question, and that's to ask you, Tariq: what are you most excited about going forward, as you think about the intersection of the current market environment and the opportunity set behind all of these secular growth drivers?
Honestly, I'm most excited about technology investing as a whole. What we started discussing, the valuations are a lot more palatable now, a lot more investors sensing — and even, of course, companies — that free money days are over. We can really discern, in this new environment, between companies that are allocating capital in a reasonable way and identify where there will be returns. This kind of reminder is exactly where you want to be able to find those true innovators, in small cap, that are the ones that are truly disrupting massive markets in midcap with their business models, which will help them earn proper returns utilizing these various teams.
These themes haven't gone away. There may be ups and downs, a little bit. There are accelerations, there are decelerations a little bit, for sure. But these themes don't go away. And with the valuations and some of these things coming back, we are looking at all sorts of companies. Some of them may be profitable, some of them not profitable, but the tension remains on the business model, unit economics, and how these management teams will earn returns on the investments that they're making.
Perfect. Tariq, this has been a fascinating, really enlightening discussion. So thank you so much for your time.
I certainly thank all of our listeners for tuning in. Hopefully you all found it as interesting as I did. And until next time, take care.
Thanks for listening to Markets in Focus from Raymond James Investment Management. You can find additional episodes and market insights at marketsinfocuspodcast.com. You can also subscribe to our podcast on Apple Podcasts, Spotify, or your favorite podcast app. Until next time, I'm Matt Orton.