Hello everybody, and welcome to the Ingram Micro fireside chat at the Goldman Sachs Communacopia + Technology Conference. I have the privilege of hosting Mike Zilis, CFO of Ingram Micro. My name is Kat Murphy, and I am the analyst covering Ingram Micro and IT distributors here at Goldman Sachs. We have about 35 minutes for today's session, inclusive of analyst Q and A. Perfect. Cool. To get started, I want to talk just more broadly about AI demand. It is driving significant uptick in demand for both data center hardware and for edge devices. How does Ingram Micro's role in helping customers change when we introduce this net new vector of demand? Does your role look different than it would in more of the traditional fulfillment or distribution capacity? Is this pulling on your core competencies as a distributor with global reach? Yeah. I think the whole AI phenomenon for sure, to answer the very end of your question, is absolutely pulling on core competencies and what we do as a company with the broad technologies we have of 1,500+ different vendors, 165,000 different customers around the world, that breadth is sort of key. I think what we are seeing in AI right now is kind of two different things. One that we have been calling out a lot in our earnings calls as an area that more than doubled year-over-year, which is GPU and AI infrastructure. From a product perspective, this is the GPU and high-end CPU itself, but also technology with AI embedded in it, AI capabilities embedded in it. That could be networking, server, storage. PCs is a little bit different animal, which I can touch on, but that is not in those numbers. Really what this is large enterprise, and this could be large sole enterprises, it could be hyperscalers, it could be neoclouds who are building out data centers. We are participating in that, but a lot of it is more fulfillment, so it tends to be lower margin, gross margin business, but it is very accretive from an operating margin because there is very little cost to serve, and we are not stocking this product, so it is turning very fast, very efficient. The growth of this business is not the sole driver, but a decent driver of the fact that our ROIC year-over-year is actually up 240 basis points in Q2. It is very good business. Now, the other side of the spectrum is really much more nascent, which is more investment into SMB. We are seeing a lot of infrastructure as a service investment there, because SMBs are not going to build out their own data center. They are going to be more firmly planted in kind of hybrid, off-prem, and the like. We see that happening. We see SMB, and we are driving education programs that we call Enable | AI through SMB to help them understand where AI is going. But it is a far earlier end of the spectrum there, where POCs need to become more relevant and repeatable, where we will see some of that motion really pick up. I want to touch on both of those opportunities, but first, thinking about the role of IT distributors, both in addressing the fulfillment of some of these large AI projects, educating your SMB customers, you are also helping them navigate higher input costs. What is the role of Ingram in sitting between both your vendors on one side and your customers on the other in digesting these very significant increases in component costs across the board? Yeah. Anything that is memory related is seeing ASP increases for sure, and some more pronounced than others. I think we are starting to see the rate of increase lessen. We are still seeing ASPs go up, but to a lesser degree in some of those categories in more recently as we exited Q2, for instance, and into the new quarter. There is still ASP increase. Now, what we can do is a handful of things. One, our breadth allows us to get product. So even where it is less scarce because the OEMs are putting less of their production capacity on lower-end spectrum of devices, where there is availability, our breadth allows us to garner access to that. So we can help service the customer. The breadth of the OEMs we have is also a benefit, so even if product may not be available or the ASPs have gone up to such a degree with vendors A, B, and C, we may still have vendors D, E, and F that are still playing in that space, and it could be a viable alternative. So that is where we can really work with those customers. But as ASPs go up, we are not eating that cost. It is a pass-through for us. And in some cases where there may be limited buy-in opportunities to stock inventory ahead of price increases, there could even be a modest benefit from that perspective. Gotcha. One more high level, and then we will dig more into the AI fulfillment opportunity. In navigating some of these dynamics around input costs in particular, how are you seeing the ways in which your vendors are navigating the channel changing? Are you seeing an opportunity for a growing distribution TAM or a consolidation at the high end for distributors with a breadth like your own? I think it is. I think we are seeing some consolidation. It plays across a number of different fronts. Certainly, again, the breadth is intriguing to a vendor because through one relationship they can access, in our case, 200 countries around the world, 90% of the world's population. That's an intriguing value prop. I think the other thing that we see play into the vendor relationships is business practices. How we operate, how we think about Foreign Corrupt Practices Act, how we think about how product is handled in an ethical and appropriate manner. It may not be the way a small, local, or sub-regional player may operate, so that too is a value prop that we're going to, more often than not, be on the positive side of any of those sorts of decisions. So that's worked well for us. Before we talk more about some of the business segments within Ingram, I want to talk about Xvantage. For the purpose of this conversation, can you explain to those who may not be as familiar what Xvantage is and how it is creating a competitive moat for Ingram as you navigate all of these different dynamics that you just outlined? Yeah. It's a great question, and I could probably spend an hour on this, but I'll try to answer this in two minutes. Xvantage is our platform on which we operate. It is a proprietary platform that we started investing in in 2022, has more than 400 machine learning models embedded in it. We have seven patents already approved, another 35 roughly that are pending. So very proprietary technology, but at its core, what Xvantage is trying to solve is to bring a B2C experience into a B2B world, which isn't always very easy when you think especially about the fact that we sell what we call advanced solutions, where there's oftentimes, on average, more than six solutions, more than six different products or services that make up ultimately the solution to the end user. That is six different vendors, six different providers, different pricing programs, different rebate structures, different terms and conditions, and we are automating all of these different elements. We have also layered in machine learning for what we call Intelligent Digital Assistant, or IDA, which is enabling not only our sales force, but this platform is used as the sales platform for our customers, where IDA is helping to prioritize the sales motion. It is pulling out a lot of administrative and back-office costs that are now automated, used to be manual, and it is allowing our sales force and that of our customers to be more proactive rather than reactive. These are really compelling case studies where in the timeframe since we have reasonably deployed this, which right now we are in 22 out of our 57 countries around the world. We should be fully deployed in the next year and a half, roughly, to get it out to the remaining tail of countries. But we have deployed it towards our largest countries first. So it is meaningfully deployed already. The first phase was to pull out OpEx, become more efficient, and that allowed us to remove $200 million in annualized OpEx out of our business back between late 2023 and into early 2025 timeframe. And that does not need to come back. So we are far more efficient on the operating expense line item and the leverage story. It has also allowed us to grow above market. We have been growing double digits now for six straight quarters effectively, and have been growing faster than market ostensibly across most of our markets, especially in markets like Asia-Pacific, where automation and capabilities like this can be more critical because they are differentiators from what others are offering, especially local and sub-regional players. And now what we are getting into now is more using the pricing models, dynamic pricing optimization, and AI learning to build more of a margin play. So that is a little bit earlier stages. That is kind of just starting now, where we can try and drive more of that profit, that gross margin accretion from that perspective. But it has been a very good story for us, and that stickiness of the system, including the fact that, again, our customers are using this as their day-to-day platform, absolutely is creating a moat for us. Maybe sticking on that gross margin accretion opportunity, can you talk about what in a perfect world is going to drive that? Is it selling more in each individual sales motion to your customer? Is it deeper penetration within a certain vertical of customers like SMB? What is going to help that margin accretion story as Xvantage becomes more penetrated? Yeah. Some of the things you referred to are just pure mix. We make more profit off of a small and medium-sized business because we add more value, bundle more services around it, versus just moving product. We make more profit on advanced solutions than cloud, because again, it is more bundling, it is more complex. There is engineering skills and other things that come into play, and therefore distribution gets paid for it versus just selling a PC or a notebook or a mobile device or something along those lines. Those are all beneficial. Then I would just think about what machine learning can do. Here is a really simple example where we often have notable rebate programs and back end as far as how we sell. With the AI that is embedded in Xvantage, we can be analyzing not only demand pipeline, but also inventory in the system, ability to procure, and be able to look at opportunities to toggle sales to drive to enhance that profit of a back end that might be attainable in the next handful of weeks if we move in a slightly more aggressive manner to drive more profit for the business. That is just a very practical single example of which I could give dozens as far as where we are now starting to toggle more towards that as we have more scale and operating capability. That is very helpful. I want to switch gears back to the advanced solutions segment and talk to some of those GPU sales and AI infrastructure projects that you called out, mentioned that they doubled year-over-year in the most recent quarter. Can you discuss what Ingram's role is? You alluded to kind of more of a fulfillment role, but really what your right to win is in addressing some of these opportunities with large enterprises and neoclouds as they are making these purchases of servers and compute at the high end. Yeah. I think some of the things I mentioned earlier are really the driver that is allowing us to win, which is scale and reach. But we cultivated that first by building the relationships with the key OEMs in that space. So many of those that I mentioned earlier, like networking server storage that just has AI computing capacity embedded, those were always relationships, although we continue to build those in strategic manners. But the true GPU players, that is a little bit of a newer element. We have always had some level of CPU and components, but it has never been a main focus, whereas this is a slightly different kind of opportunity. But at its core, what we are providing is logistics to get product from point A to point B. We are offering credit. But these are large enterprises where we are not necessarily doing deployment. We're not building out the data center ourself, and that's why it tends to be more fulfillment. But as I said earlier, very low cost to serve, very working capital efficient. So it's accretive from an operating profit margin and a ROIC perspective. Maybe sticking on the impacts of the mix, where low cost to serve, so operating profit accretive, ROIC accretive. Should we be rethinking about how we measure success in the advanced solutions segment as GPU deals become a larger share of Ingram's story as to not penalize the company for going after some of these big opportunities? If gross profit is the metric investors tend to look at. I would love to get to that point. Okay. I think we guide for a reason on our net revenues, but also on GP dollars and then EPS. The GP dollars is really where I would focus more. If we're growing that at high single digits or low double digits, which was the case in Q2 as an example, our most recent quarter, that's a sign of health across the business, regardless of what the margin rate is. But what's even more important is on that 12.5% revenue growth, 12% GP growth, or GP dollar growth, we were at least 2x, and in some cases even 3x, faster growth on profit. So that's where that OpEx leverage and scalability and automation come into play to couple with the GP growth, and that becomes incredibly compelling if we can drive that kind of a ratio of growth on the bottom line versus the top line. Advanced Solutions grew 13% in the quarter, but certainly benefited from that doubling year over year of your GPU sales. What does this imply for what I will call the core advanced solution or the non-AI, explicitly AI portion of the business? There is a couple factors that I would think about. One, it is not that the core or more legacy advanced solutions are weak. It is just simply what I said a few minutes ago. I think the OEMs are gearing their production capacity more towards the higher end, and that is causing the projects to gear more towards the higher end. The demand, theoretically, is still there at the lower end when the product becomes available, and we are seeing that when we are getting access to those products, and we can move them, the projects are still there. Then there is a handful of areas that do fit in the advanced solutions category that are peripherally benefiting from all of this investment, and probably the single biggest one I would call out is cybersecurity. There is more edge and endpoint devices touching data. Data is being transmitted in different ways to handle AI workloads. Data centers are being stood up. More companies are relying on hybrid and off-prem, and that too is benefiting our cloud business and infrastructure as a service in particular. So those areas are growing double digits, and strong double digits actually, on a year-over-year basis because they are benefiting peripherally from all of this. Other categories that you have talked to in Advanced Solutions outside of cybersecurity would include, we will call it traditional servers and networking storage. Anything you are seeing in trends in those particular parts of the business that you would call out? Nothing that I would say is concerning or overwhelmingly something notable. I think we're seeing ebbs and flows in strength depending on the quarter. Sometimes there could have been a large project in the prior year, so the compare looks a little bit different. But I just think all of those areas you just noted are tending to see more growth driven by AI-enabled in the higher end than the lower end right now. You mentioned OEMs gearing their own production towards the high end, but SMB demand or maybe lower-end mid-range demand's still there. What's giving you confidence that this is not demand destruction, but rather demand deferral when you think about these future projects? Yeah. I think we just haven't seen huge evidence of projects just being canceled outright or deferred indefinitely. Yeah, there are instances, but it's not notably different than what we maybe normally would experience. But what we are seeing, and this kind of went into this factor of about 3% positive impact on our revenues, were four different things in our Q2. One was just the ASPs themselves being higher. Two was maybe a bit of pull forward. That's a little bit hard to ascertain exactly what's pull forward, especially in a two-tier model like ours. But there could be a little bit of that that also drove some boost in sales. Then there's two offsetting headwinds, I guess, for lack of a better word. One is just the supply itself, the constraints. It's taking longer to get products, get them in the door. Supply chain is slower. We have more backlog than normal. Two, a little bit of the demand elasticity that you're referring to. So on the elasticity of demand, what we're seeing more of than just cancellations or indefinite deferrals, is just more selectivity. Customers may be making decisions to downsize the number of units they would have ordered, or perhaps downsize the specs to something slightly lower end to try and do their project in a more economical manner. But the one thing I think we're pleasantly surprised on is just the general resiliency that is out there in a highly increasing ASP environment where demand has stayed fairly resilient. On that three points of net benefit that you quantified for the most recent quarter, you talked about two to three points of benefit for the upcoming guided quarter. Yeah, exactly. Can you talk more about what the process is and how you weigh those various tailwinds and headwinds? Yeah, I wish I could say it was an incredibly precise process, because there are a lot of moving parts here to try and triangulate, but we feel pretty confident in that rough order of magnitude. A lot of that's based on just the discussions with our customers and what we go out and understand from the ground level up, where is demand, where is pipeline, what are the projects that are ongoing from both our customers and ultimately out to end users, and similarly, backward into our vendors as well. What are they seeing from a market perspective? There's quite a bit that goes into it and the effort around it. But we give the combination of those four factors for a reason, because to try and quantify each one of those individually would be quite challenging. Interesting. Switching gears to cloud, then I want to touch on client and endpoint solutions as well. First, tremendous growth in cloud last quarter, 44% growth or 54% excluding the CloudBlue divestiture. Yeah. Talk about the strength there. You made reference to the infrastructure as a service offering, but help characterize that strength and growth and how we should be thinking about growth from here. Infrastructure as a service for sure is our fastest growing category there, and for some of the reasons I noted earlier, where I think it is not just SMB, we are seeing it across all elements of customer. But that has been strong growth and we expect it to continue to be strong growth for any foreseeable future. But modern workplace, which would include things like Copilot and tools such as that, and even software, just more general software has remained solid for us across cloud. But infrastructure as a service is the standout. And for this audience, can you just maybe paint a more specific example of when you are selling or engaging in an infrastructure as a service deal, what that might actually look like for one of your VAR customers? This is consumption-based models. Think about AWS, Azure, Google Cloud, so those sorts of players, where consumption-based models are driving demand for data interchange and infrastructure and storage that is done in an off-prem or a hybrid sort of model. It is not unlike any other solution. One, we are selling oftentimes infrastructure as part of a broader project that could be bundled with hardware, operating systems, any number of different things, cybersecurity and other offerings. There is a stickiness there because of the way our relationship is between the vendor on down ultimately to the end user, and that consumption model that continues to have, while it may be volatile depending on consumption, it has an ARR kind of aspect to it. Very different unit economics too, or financial profitability for your cloud business. Can you talk about expectations for the accretive nature of cloud? Yeah. How you think this kind of tremendous growth in cloud mix should help with the company's overall, we will call it corporate gross margin? Yeah. It is very accretive. A lot of this is an accounting convention, when we sell cloud, we sell it on an agent basis, and therefore the revenue is accounted for net. To use just a really practical example, if we sell a PC for $100 that had a $90 cost of sales, our sales would be $100, cost of sales $10, gross margin, $10 and 10%. Sorry, gross profit $10 and gross margin 10%. That same cloud sale would carry a $10 net revenue, and a 100% gross margin. The only reason that is important, and I will put this more in the order of magnitude of cloud as a whole, and the fact that this is a meaningfully scaled business with investments we have made for more than a decade to build a cloud business. It is less than 1% of our net revenues because of that accounting convention. But whereas last year it ranged anywhere, depending on the quarter, from 12%-15% of our gross profit dollars, this year to date in the first half of the year, it's in the high teens because it's been growing strong double digits. This is a meaningful scale business that's approaching one fifth of our gross profit dollar generation. That's very helpful. Switching gears to client and endpoint solutions, up 12% year-over-year in the last quarter, strong demand across notebooks, PCs, components, and you guided importantly to high single digit growth for the upcoming quarter for client and endpoint solutions. You mentioned that on the earnings call that you still see room to go, room to grow we'll say, in the PC refresh cycle and optimistic around the opportunity for AI PC penetration in particular. What's giving you confidence in the outlook for the PC market overall? We can maybe touch on the pricing versus units dynamic, but the opportunity for AI PCs in particular. Yeah. I think PCs, I think there's stats out there from various sources that would suggest that perhaps upwards of 30% of equipment hasn't been refreshed yet. So that is sort of the tail that still exists, to some degree in the coming quarters that could potentially see continued growth in PCs. The AI PC is the intriguing part because still only about 30% of our PC sales, and this is also roughly the industry metric as well, about 30% of PC sales are AI-enabled. So that could actually smooth a little bit of the cyclicality. We need to see how that plays out in the coming quarters. But we do see that opportunity. But the AI PC is becoming more critical as workloads and the capacity needed to run those workloads becomes more onerous from AI and other categories. We do still see some room to grow there in the coming quarters. As you mix shift towards more premium or high-end PCs within the client and endpoint solutions business, does this change the kind of long-term lower margin for client and endpoint, higher margin for advanced solutions framework that we've typically thought about, or is it still a relatively lower margin business? I think it will continue to be relatively lower margin. There could be some of that opportunity, but as I said earlier on your more general question around ASPs, first and foremost, they are a pass-through for us, and therefore, it's not necessarily in its own right causing an uplift in our margin rate. Where we do get the potential for that uplift is more where we're bundling the services around that. So with an AI PC and different edge computing and need for more access from different devices, that does open the door potentially to bundle these solutions across a bigger project, and therefore services and other offerings could allow for that bill of materials to be priced a little bit differently and create some upward margin trend. That's helpful. I want to touch on your mobility portfolio as well. Smartphones are facing a lot of the same dynamics around component costs and ASP increases as the PCs, and one of the major smartphone vendors is expected to change the cadence of their release schedule— Yep —tomorrow. Yep, exactly. How should we think about the impacts of this on Ingram's mobility portfolio? Anything to share just to contextualize where that portfolio is most oriented today. Yeah. It is smartphones that we do in a number of markets. It's probably more pronounced in our Asia-Pacific region and our Latin America region, although in EMEA and North America, we are doing smartphone sales as well. Those can be into, to a lesser degree than we used to play in retail and e-tail categories, but also into large enterprise. It kind of runs a little bit of a gamut of the ultimate end user and the channel into which that product is sold. As far as the new release schedule, we've always had seasonality. It just has tended to concentrate to right around now where releases are coming out from the major OEMs in the same rough timeframe. So it may just change a little bit of the seasonality. Perhaps it smooths it overall. We'll have to see how that plays out, but I'm not too worried about the release schedule necessarily. Gotcha. I have a couple more, and then I can turn it over to the audience to see if there is any Q and A. Sure. Shifting gears, I do want to ask about Ingram's capital allocation priorities. Sure. Across shareholder returns, some strategic investments, I know tuck-in M&A in certain strategic regions has been important in the past. Also maintaining a healthy balance sheet— Yep. —achieving eventually investment grade status. Can you talk through your priorities here today? Yeah. So they really haven't changed a tremendous amount. We've repaid quite a bit of debt from when Platinum first acquired us in July of 2021. So our debt to EBITDA leverage is actually in a pretty good place. Therefore, opportunistically, some repayment of debt could be potentially part of the play, but it's not necessarily a priority anymore, especially with EBITDA growing at the rate it is. We're going to continue to invest in the business, both organically and inorganically, where opportunities present themselves. M&A is still going to be part of our strategy where we see those opportunities. Then I think the final component is return to shareholders, and we're happy that we had a dividend right out of the gate when we went public in October of 2024, and we've managed to raise that dividend every quarter by 2.5% or more. We are still closely held by Platinum Equity, and therefore a more traditional share repurchase program doesn't make a lot of sense. We have no interest in reducing the float, but eventually that'll be another part of shareholder return where free cash flow will be used for share repurchase as well. We actually have, as Platinum has done two registered follow-on deals to reduce their ownership by another 13% in the first two quarters of this year. We participated in buying some shares from them to the tune of $105 million. We'll deploy capital that way as well to continue to reduce the overhang of a closely held company. That's very helpful. Yeah. Any questions from the audience? Maybe another one for you more on, and you touched on this when talking about the Xvantage platform. How should we think? You are helping customers in many different ways deploy AI within their own organizations. Are there any examples beyond your usage of Xvantage as to how Ingram Micro is deploying AI internally, either for operating or margin expansion opportunities, ways to make your own company more efficient? I think Xvantage is the single biggest example with the machine learning that is embedded into Xvantage. We are a user of AI in other ways as well, even with some of the different platforms and systems that we use across the rest of the business. We are looking at now, and this is not nearly the capital outlay that Xvantage has been, but we are now also modernizing some of the back office, whether that is finance systems, that is critical for my team, but also in areas of warehouse ops and other areas where we can continue to modernize and use machine learning to our advantage. So we are a user of AI agents and so forth, as well as a seller of it. Good to know. Yeah. Lastly, you are hosting a Capital Markets Day next week. Yes. Anything you can share to preview for this audience what we should expect to learn about, what part of the story you are most excited for investors to better understand following next week's event? Yeah. We are doing it, one, at the new New York Stock Exchange that happens to be in Dallas. I think we are going to be one of the first presenters at this new facility, which is a pretty beautiful setup. We will obviously be broadcasting this online, so I encourage you all to check it out. It is a week from today, Tuesday the 15th. We will certainly have broader strategy from our CEO, Paul Bay. We will spend an hour, including some Q and A embedded on there, on the platform as well, just to continue to educate the investor base around what we are doing with Xvantage. Sanjib Sahoo, our EVP of our platform business, will be speaking on that piece. We will be doing a panel with our regional presidents around what we see around the world. We are going to be doing a panel with some key customers. Then we will bring it home at the end with a bit of financial overview, including some longer term expectations as far as how we see the market over the coming years. Kind of the usual thing, but very geared and very timely, honestly, since it is now been. It will be a month short of our two-year anniversary of going public. It is timely in the sense that we spend a little bit of time talking about where we are on the strategy and how we see that evolving in the coming years. Great. Mike, thank you very much for the time. Yep. Everyone, if you could join me in giving a round of applause for Mike. Thanks. Thank you. Thank you. It was amazing.