Podcast #192 Digital Machinations

 TRANSCRIPT

Ken: Good day, and welcome to episode 192 of our Momenta Digital Thread podcast series. Mike Dolbec and I are glad you joined us today. Today we're pleased to host Andrew Obin, Managing Director, Bank of America Securities Equity Research, covering the multi-industry and industrial software sectors, including Aspen, Carrier, Dover, Eaton, Emerson, GE, Johnson Controls, Rockwell Automation, and Train to name just a few. Before joining Merrill Lynch in 2002, Andrew worked at Lehman Brothers, covering the aerospace and defense sector. Andrew has consistently been ranked among the top analysts recognized for the quality of the stock pickings and estimates by Forbes Financial Times and the Wall Street Journal. Andrew, welcome to our Digital Thread podcast.

[00:01:27]

Andrew: Thank you, I'm glad to be here. Thank you for including me.

[00:01:31]

Ken: We're quite pleased to have you as well. We call this the Digital Thread podcast, and as a build-up to getting into the core of our discussions, we want to understand your digital thread. In other words, the one or more thematic threads that define your digital industry journey.

[00:01:48]

Andrew: I've covered industrials for 25 years at this point. It's interesting, but the industrial sector, where we started to see the emergence of digital and software moving the needle, was first to ag and construction. Companies like John Deere were smart farming 15 years ago, including Caterpillar and Komatsu. Of course, for the larger diversified multi-industrial company Predix, the right concept, timing, and execution were now almost a decade ago. But I think industrial IoT- Internet of Things, whatever you want to call it, is on people's radar. What's been happening is that service is key to companies within my coverage, and software is becoming increasingly important. I think the issue that my company is facing in implementing these digital initiatives; it's what's fascinating to watch. It's as much cultural as technical, and that's where much discussion is. Within the sector, the next step is the Rockwell-PTC tie-up; you started seeing these big, large, diversified industrial companies paying a lot more attention to software. You have Fortive starting to buy up industrial software companies. I raised my hand internally and asked for more coverage, and I was given more coverage of industrial software companies. In addition to covering these large multi-industrial companies increasingly focusing on software, I'm also covering pure industrial software companies. That has been my journey, I think, over the past couple of decades.

[00:03:29]

Ken: What a great place to be. We're big fans of this intersection of OT and IT, and it's the companies you've mentioned. Especially some of the pioneers like Caterpillar, we think, set the level, if you will, for some of the other industries out there. Maybe we'll do a quick level set for the audience in terms of your area of coverage. We mentioned the multi-industry sector and the industrial software sector. What are your key themes and some of your key companies in this?

[00:03:58]

Andrew: Multi-industrials; these companies used to be known as conglomerates. Also, large US industrial companies survived the onslaught of Japanese competition in the '80s, so a lot of focus is on automation. The history is back in the '60s, there were new ways of running companies and allocating capital, which was more efficient. These companies proved to be more defensive in the face of what was happening in the 1970s, geopolitical uncertainty, and inflation. These companies excelled in the period of low industrial growth in the US over the past 20 years. They focused on operational excellence in a period of slow growth. We're seeing big ones breaking up, so big ones would be Danaher, 3M, and General Electric. But what's interesting, there is a new generation coming up: companies like Amatec, Fortive, and Dover. It's a very dynamic sector; many things were happening. The companies I cover tend to have diverse portfolios, probably more focused than in the past. Still, these companies tend to be good at operations and capital allocation. Over the past decade, they've paid much more attention to software and services.

[00:05:13]

Mike: Thanks, Andrew. This is Mike Dolbec here, and it's nice to meet you. I'm a veteran of the GE Predix efforts, so thanks for mentioning that. Given the broad breadth of what you follow, what are some key macro factors that drive performance amongst the companies you cover?

[00:05:34]

Andrew: Sure. Investors that follow my company focus on two metrics. First is organic growth, and the second is a return on tangible assets. For folks who know a bit about finance, this metric excludes the impact of M&A and focuses on cash flow. It's organic growth and free cash flow. Investors will focus on the big macro factor- I cover the industrial focus sector, so it's US Capexcapacity additions. The US tends to be the most profitable market. It used to be that China was a huge focus, but China has slowed, and of course, global industrial production. But right now, I would say it's the US, China, global industrial production, and capacity additions. M&A is a big focus on big themes for companies in my coverage, making the portfolio more profitable and less cyclical, so they tend to look for better quality industrials as acquisition targets with larger service components. But we've also done some work, and it's a surprisingly large focus on software and healthcare.

In some cases, it's to a point where they become something of a healthcare or life science company, like in the case of Danaher. At the same time, you also have a parallel trend of larger companies getting smaller that I discussed. We think that for the next decade, the big theme is the fracturing of the global supply chains, which means you must replicate the supply chain in multiple areas. As a result, you are getting reshoring. Of course, in the US, if you look at the numbers, the big driver of reshoring is semiconductors. There were many secondary and tertiary implications across the broader value chain.

[00:07:21]

Mike: Great, thanks.

[00:07:22]

Ken: You mentioned reshoring, which we believe is quite an important factor- also, at least in Europe, referred to many times as resilience. Especially, of course, since the pandemic. To what extent do you see this creating a renaissance for manufacturing in the US?

[00:07:39]

Andrew: I am probably one of the most bullish people regarding reshoring and thinking about it. To think about it, you need to take a step back. Strategically, where it started, the US is facing an issue of key electronic components only available in China. It's not just- you can't make these parts in the US, but you can no longer source them from allies. I think a big eye-opener for me was visiting Hon Hai Foxconn up in Wisconsin. They described their thinking that, over the long term, the US government would be unwilling to source components from China. The view is that the US government is a big enough customer. The US government is not particularly sensitive to price, so it would make sense over time to move large chunks of the supply chain to North America to service the US government as a customer. It has profound implications because I don't think people appreciate that tech has been the biggest driver of hollowing out US manufacturing over the past 20 years.

Let me give you some numbers. In 2000, the value of the US durable goods manufacturing sector was 1.7 trillion a year, of which 400 billion was tech. Today, we are at 1.6 trillion, with tech only 200 billion. These are constant numbers. What effectively happened is that over the past 20 years, incremental semi-capacity has gone outside of the US. Of course, Apple is the other big factor. Looking at the rest of US manufacturing, it hasn't done anything exciting. Still, the reality is the rest of the US manufacturing base has been fairly steady, with aerospace and chemicals being a big area of growth. That's just the framework here. If you look at capacity additions in the late '90s, tech was the last big driver of US capacity additions as far as manufacturing goes. In perspective, the total value of US manufacturing Capexis 500 billion. This number includes everything. Within that plant and equipment is about 200 billion of that, right, was pure equipment, some are over 100 billion. Industrial software is another 140 billion.

Interestingly, US manufacturing companies today spend more on software than equipment. Think about this 500 billion number, 200 billion of plant and equipment, the sandbox where the companies I cover play. The value of the current list of announced semi-plants is getting close to 100 billion. On top of it, you should think that many semi-plants will drag part of their supply chain. We went to a big industrial show, the IMTS, a couple of weeks ago in Chicago. Starting to hear from, for example, an Asian linear motion manufacturer, something that you need to make semiconductor equipment, bringing some of the sub-assemblies to the US and thinking about adding more capacity. Because it's a lot easier to serve this Capex boom in the US, it's the reverse of the process playing for the past 20 years. It would help if you also recognized that these semiconductor fabs are 80% Capex. If you have a 200 billion base of plant and equipment- so appreciate this 100 billion, it means 80 billion of Capex.

The US military is in the driver's seat, on top of the semiconductor boom driven by national security concerns. We started talking about it. On top of it, you have less outsourcing, and there is data; you can see that the amount of intermediate goods exporting to the US has been flat since Trump's tariffs. You did see that the Trump administration policy has made a difference in outsourcing. It didn't cause people to bring more stuff to the US because it did pause outsourcing, or it's been relatively flat, and on top of this, you have a stimulus. You can easily see the rate of US manufacturing Capex effectively doubling or tripling versus where we have been over the past 20 years, so we are poised to see US manufacturing Capex grow probably at the fastest rate in two or three decades. I think what's going to be very different is we see US industrial market as the fastest-growing industrial market in the world, which is very, very different from anything we've seen in the past couple of decades.

[00:13:07]

Mike: Yeah. You're right; that's eye-opening. As we were going from the large macro level down further and further and focusing on issues and segments, I just wanted to let you know that we've enjoyed your past digital machinations series and the top five trends that you speak about within that series. What are the top five trends you're tracking in that digital machination?

[00:13:34]

Andrew: Good question. Thank you. I'm going to list them, and then we'll go into a little more detail. I think it's perhaps a bit of a longer list. But first, I would say it's accelerated automation and digitization in response to labor shortages. Second, companies within my coverage are more proactive on software in M&A. The next one is you have Microsoft and AWS building out their software stacks and offering an alternative to these existing systems. The fourth one is that the industry is starting to embrace Cloud SaaS. I think the potential here ties into Microsoft and AWS. Still, this idea is that perhaps you could finally achieve this goal of scalability. I think it remains to be seen how successful we'll be, but it leaves there this golden promise. Finally, we think the oil and gas industry is very exciting and could materially accelerate its software span in response to the energy transition and a more careful approach to capital spending. Let me go into more detail. The first one is accelerating automation and digitization in response to labor shortages. I think what's happening is that you have demographics in the US where just a more experienced labor force is retiring. The younger generation doesn't want to work in the factories. Part of it is a skill shortage.

Still, I would also say it's the fact that for the past 40 years, US industrial employment has been on a downward slope, so if you're thinking about a lifelong career, manufacturing might not be at the top of your list, and companies are desperate now. But the reality is, I think people have recognized that it's just going to be hard to replace the labor force; that's exciting. On top of it, things coming back to the US will be, in many cases, a high value add. It was a natural place for a lot of automation. Still, we are seeing- you should consider automation as a multiple of industrial production and Capex. Industrial production and Capex are accelerating, and we have this demographic issue in the US labor force, forcing companies to adapt faster and faster and faster. A friend ran a printing and label company and said, "We spent on automation, like drunken sailors. We cannot convince people to come and work in our factories." On top of it, of course, if you look at the structure, the simple relationship that we have observed between labor costs and automation is that any spike in labor costs tends to follow by automation standing two years after. Multiple things are coming together, so I think that market is scorching. A second company's being more proactive on software and M&A, and you're just starting to see it. Predix was the first one, it didn't quite work, but I think it was the right idea.

Fortive followed by trying to buy these more traditional industrial software companies, more in discrete verticals that related to manufacturing. But then you had the tie-up between Rockwell and PTC. We think that the industrial logic for the tie-up has changed over time, but I think a lot of people have been paying attention. Then, of course, you have Emerson-Aspen; if we were to guess, you would see a lot more of this. What you must understand- and I'll talk a little bit about the geekier aspect of finance- is that software assets look incredibly good inside industrial companies. The reason is I described industrial investors' focus on organic growth. Industrial software companies tend to grow faster than industrial manufacturing, and industrial investors also focus on return on capital, excluding amortization. This is important because software companies are very capital-light regarding pure return on invested capital. Hence, as you add software revenue, it suddenly makes underlying industrial assets look a lot better. You bring up the average quite materially, which drives the multiple industrial names, so it tends to be various enhancing. On top of it, I would also add that industrial software companies tend to grow quite a bit faster than industrial companies on a cash flow basis.

There's a considerable debate in my space about how expensive industrial software is. It's not as expensive as you think, given that they compound much faster. As interest rates are going up, which changes a little bit, but also appreciate how industrial investors look at the world, industrial software tends to look good. All of this is becoming a big focus for companies within my coverage. The third one is Microsoft and AWS building out their software stacks. I think that train has left the station; we shall see how disruptive these two will be. Microsoft works more with the IT department because they're so dominant on the IT side. AWS, as far as we can tell, is more active going to folks on the factory floor. The message seemingly, we understand what it is you're doing. But they're bringing a lot of expertise; they're bringing a lot of money. They've upgraded their talent and are building their different software stacks. It's not these isolated ecosystems that you've had on the factory floor, and they're starting to be disruptive. We shall see how disruptive this will be, but that's something material. We expect- just given the numbers that the industry is talking about. Generally, this should be a high area of growth, and this segues into the whole SaaS topic. Historically, industrial companies had been very conservative and favored on-prem software. Right now, I would guess SaaS is probably over 10%, but the view is that it will get over 50 fast. If you do the basic back-of-the-envelope math, it tells you that legacy on-prem software will grow maybe low single digits at best versus high single digits in the past. Of course, SaaS will be the big area of growth, and SaaS, depending on an assumption, will grow from 20% to 30%.

A lot of focus from both legacy players that tried to move to SaaS, but obviously, it's an ample opportunity that Microsoft and AWS see in the space because suddenly, the space, if you think about it, we estimate the value of industrial software probably 60 billion a year, depending on a definition. Right now, it's 10% SaaS, so it's only 6 billion. You guys know what the valuations are. But if it continues to grow, and if SaaS becomes half of it, suddenly, you have 50 billion SaaS and TAM, depending on the multiple you put- all of a sudden, it's half a trillion or more in market value available for grabs so a lot more people are interested because all of a sudden, you can scale. You can make a lot of money. Don't underestimate that as a key driver. Finally, I would focus on the oil and gas industry as big whitespace, and you should realize that historically, the oil and gas industry has been about growth. It's a very, very inefficient industry. If you look, the number of man-hours that it takes to get a barrel of oil out of the ground today versus 30 years ago has not changed, which is staggering. You don't realize it because the industry had a lot of capital to throw at growth. Between energy transition and a more careful approach to capital spending, we see a lot more focus on sweating the assets harder, necessitating more digital and software spending.

To put it in perspective- we said that the manufacturing industry spends $200 billion a year on property and equipment and $140 billion on software. The comparable numbers for oil and gas industries are 120 billion on property and equipment and 9 billion on software. The oil and gas industry spends a fraction of what manufacturing spends or what happened in manufacturing; manufacturing ended up spending so much on software because, in the past 20 years, we had no growth in Capex and capacity expansion. If you're running a factory, the only thing you are allowed to spend money on is efficiency, efficiency. Our view is if you're going to start getting growth in the oil and gas industry, they must focus on making things more efficient. As a result, we think structurally, the software spending for the industry can double, triple, quadruple- frankly, on the same overall level of Capex.

[00:23:30]

Mike: Wow, I'm smiling. That's a fascinating perspective. I have a couple of follow-up questions, if we have time and before we move on to the next one. There was a vibrant answer that you just gave - you've answered it explicitly, but it generates some thoughts in my mind. Suppose I go way back to how manufacturers are deploying software today. Do you see that historically, it's been about efficiency and not so much direct contact with the people that work in factories? One trend we're seeing now is the investment in software to augment frontline workers or, later in the stages, to augment plant manager decision makers, so software that real people use as opposed to efficiency that drives machines in some better way. Do you see that at all?

[00:24:23]

Andrew: Yes. This goes back to what I said about the labor shortage. Folks who are retiring have all the expertise. As you're trying to bring in more people- the US does not have a history of apprenticeship, and you must get- how do you replace a 60-year-old factory worker who knows this machine inside and out with a 20-year-old who did not grow up working on cars? It was his dad, right? You don't do that anymore because cars are all electronic these days. Now, we see things like augmented reality and trying to automate processes. It's a big, big deal.

[00:25:05]

Mike: Okay, thank you. I've heard the 'workers retiring' problem jokingly described as the 'Gray2k' problem.

[00:25:11]

Andrew: Yes, we met with a robotics manufacturer, and welding is one of the most- people don't appreciate it; welders make a lot of money, and good welders are very important. Historically, that’s been one of the US' competitive advantages. I met with a robotics manufacturer, and they said, "Look, the average age of the welder is 52. Next year it's going to be 53. Historically, you would not necessarily replace a welder with a robot. Suddenly, there is demand for these welding robots, small welding robots- they would do it for large. It's because these folks are retiring, and it's a high value add job. Given how much these folks get paid and its importance, it makes sense to automate. To any teenager who's listening to this, who wants to work in a factory, welding is a very, very good living. It isn't going away anytime soon.

[00:26:05]

Mike: I see. One more follow-up, if it's possible. You describe the attractiveness of software to Capex-intensive industrials. I have been on the other side where the challenge is that SaaS-based software companies would love to have- generally expect a higher multiple than industrials are comfortable paying, so there's this natural conflict. I don't think an industrial will do the multiple that Adobe paid for Figma anytime soon. How do you see that natural tension in the future?

[00:26:39]

Andrew: Look, we have a great example in my coverage of Fortive and Vontier. They believe- because they look at things in terms of cash flow. As I said, if you look at the terminal value- we've done this exercise, and as I said, interest rates do matter. But what's interesting is that industrial software companies- I'm within my coverage to trade and let's call it, 25x EBITDA versus regular industrial companies trading at 15x EBITDA. If you look at the implied terminal growth value based on conservative cash flow assumptions, what's very interesting is that industrial software companies trade at par or cheaper than these multi-industrial companies. What's interesting- I think it depends on how you view it. The art of the issue is not just internal; it's also external because, believe it or not, part of your issue is that industrial investors are not used to looking at software companies. They say they like free cash flow but don't understand that a mature SaaS company is just a cash printing machine. It's very interesting. At this point, I think I'm the only industrial analyst with broad coverage of industrial software, so it was revelatory just how much cash these companies can generate. Once you appreciate it, buying industrial software companies becomes a no-brainer. But as I talked to more established industrial analysts on the buy side, that's exactly the issue you're running into. They go, "Well, these things are trading at 30x EBITDA or 15x sales; why in the world would I pay up for it?" The answer is that you'll get a ton of free cash flow in three years.

But I think the issue is that the investment community is very siloed, and by and large, industrial investors are unfamiliar with industrial software companies. I think the issue for industrial companies is how you sell this deal to industrial investors. What we have observed, though, is that if you look at announced software deals across the board within my coverage, what's interesting, by and large though, industrial names don't go down on software acquisition announcements, which is very interesting. If investors are as negative as I'm telling you, maybe investors are starting to get it. We have seen that multi-industrial companies have been careful; they're not going for a slam dunk, large acquisition. They're trying to do these bolt-on deals. But deals are getting larger. You have Emerson-Aspen; you have, obviously, Schneider-AVEVA. But suppose you look at the company that was buying when they announced these deals. In that case, the stock is not going up. Still, in many cases, it is holding up better than you would expect, getting this "perceived negativity" on software. Maybe I'm contradicting myself, but at the same time, if you look at their reaction to announced deals, industrial investors do appreciate that there is something there to the cash flow generation of these companies. The biggest issue is not the math; it's the culture. It's how do you operate a software company inside an industrial conglomerate? That, to me, is a bigger issue than just pure math.

[00:30:23]

Mike: Thank you, that's a perfect segue to one of the questions that I wanted to ask you next, which is whether several industrials- and you just mentioned one of them, have approached their software stacks as separate entities, Emerson, and Aspen, most recently, Schneider Electric and AVEVA. Is there an ideal model of merging a hardware and software business in the industrial space or at least a formerly exclusively hardware culture, learning how to collaborate with and manage a software business and its own Capex hardware business?

[00:31:03]

Andrew: I think the answer is we don't know yet. That's why I think Emerson and Aspen- the deal's structure as we see it is to try to incubate a software company outside of large industrial conglomerates. But then, of course, Schneider-AVEVA shows you where these could go, given the importance of software to your service business- which is the bread and butter of these companies. They're doing all these deals because they see disrupted service-oriented business models. Either they will disrupt themselves, or somebody else will do remote monitoring and preventive maintenance. This is where all these companies make money. I think you have seen multiple examples. I think you've seen Emerson-Aspen and AVEVA-Schneider- it depends on how you structure these deals. The structure of these deals is very important regarding whether they will work, how much control the parent has, or the software acquisition subsidiary. You have seen Predix as an example of something that didn't work.

Still, then you have Honeywell trying to replicate it organically, internally. They've had some bumps in the road. Then you have Rockwell trying to do something else. You have Rockwell far behind the likes of Siemens and ABB Schneider regarding their software capability. What they're doing is they're trying to leapfrog them with a SaaS-based MAS offering. They are betting that legacy, large industrial software players. Some of them will have trouble transitioning over time, and by being nimbler, they'll be able to leapfrog them. Initially, they tried this- cross of owning a stake in PTC, now the focus seemingly shifted to more internal-focused, M&A-driven. I think different folks are trying different things. We believe this Emerson-Aspen model balances, at least in the near to medium term, risk-reward. As I said, I have observed over the past decade that the biggest issue is not the valuation of this company. The math works. The biggest issue is ensuring that you can retain the right software engineers and that they don't quit. To us, this issue is at the heart of the Emerson-Aspen deal. Software engineers like to work for software companies. Then, of course, there's this inherent IT OT software engineer conflict in all these companies, but I'll leave it at that.

[00:34:09]

Ken: Well said. Let me ask; I know that you cover public companies for the most part but what are some of the industry-focused venture investors you're watching as well?

[00:34:20]

Andrew: I'm in a slightly different boat because I cover large companies. I think I am focused on more mature companies because of cash flow generation. We have seen many of these companies being part of private equity, so we're paying attention to a Genstar, Francisco Partners- it looks like what Mountain Capital is doing. Because we have seen a lot of deals in my sector come out of private equity, and it's just a function that goes back to this cultural issue and the ability of these companies to pay high multiples consistently, so in a way, they've outsourced the creation of these companies to private equity at a price. Then you pay a premium, and then we pay attention to you. We know many of your guests- Jon Sobel from Sight Machine, Andy Bane from Element, and Natan Linder from Tulip. I'm a huge fan of Joe Perino; I follow what he used to say at LNS. I'm a huge fan of your podcast, and I get a lot of my ideas there. And then, as I said, I also spend a lot of time talking to folks inside industrial companies, trying to figure out what they are doing. But as I said, I cover large companies that are conservative, so I tend to try to figure out what bigger names come out of the needle. That's what I do. Of course, we go to a lot of industry shows, spend a lot of time at AWS and Azure booths, and try to see who the new partners are presenting to the world. We're trying to establish relationships with these folks, talk to them, and learn from them.

[00:36:01]

Mike: Great, let's change gears, Andrew. Where do you find your inspiration?

[00:36:09]

Andrew: I am lucky to have a job that I like, and I work with people I like, so that's why I have stuck around for as long as I have. Beyond that, I try to stay focused. I have a wife; I have two young kids. I'm lucky to have smart friends, and we have dialogues with them. My wife and I are active in the local Jewish communities, so just trying to focus on the basics and staying grounded and boring in my personal life. Still, as I said, I'm lucky to do something I like.

[00:36:42]

Ken: Well, answered. I think Mike and I would also agree with our roles in professional and personal lives; it's nice to see that meshed together. Andrew, thank you for sharing this time and insights with us today.

[00:36:53]

Andrew: Thanks so much. As I said, I'm a huge fan and honored to be included on your list of speakers because I learned from many of your speakers, and I admire a lot of them, so thanks for having me.

[00:37:05]

Ken: Well, thank you for the nice plug, and I think you'll make a great addition to our Digital Thread thought leaders. This is Andrew Obin, Managing Director of the Bank of America's Securities Equity Research. Thank you for listening, and please join us for the next episode of our Digital Thread podcast series. Thank you, and have a great day. You've been listening to the Momenta Digital Thread podcast series. We hope you've enjoyed the discussion, and as always, we welcome your comments and suggestions. Please check our website at momenta.one for archived versions of podcasts, as well as resources to help with your digital industry journey. Thank you for listening.

[The End]