Last week, on August 4th, Caterpillar released its second-quarter results, and even though expectations were already quite high, the company managed to blow past consensus estimates and deliver numbers that were simply impressive, comfortably exceeding my expectations and absolutely justifying a renewed look at this business.Over the last year or two, the 100-year-old maker of bulldozers and mining trucks has increasingly become an investor darling, with shares up about 110% over the past year and 50% YTD. The reason is that Caterpillar is heavily tied to the AI infrastructure boom, not being the party that sells the compute, but supplying the energy solutions for these power-hungry data centers. As hyperscalers are building faster than utility grids can add new transmission and generation capacity, CAT’s massive reciprocating engines and gas turbines, long supplied and built primarily for oil and gas customers, have become one of the few proven, deployable solutions that offer on-site power.As one of the few reliable suppliers of these reciprocating engines and gas turbines, CAT is seeing booming demand here, driving significant growth in its business, with a backlog now stretching toward 2030.However, unlike so many other great stocks that offer exposure to the AI boom and infrastructure built out (like the hyperscalers, neo-clouds, chippers, etc.), what I like a lot about CAT is that it is much more than just an AI infrastructure story, and the Q2 results once more reaffirmed this. Sure, CAT is hugely benefiting from explosive hyperscaler demand for its engines, but the beauty here is that roughly 2/3 of CAT’s revenue still comes from other segments. Additionally, the same large-engine platforms are fungible across data centers, oil & gas compression, and mining, so a cooldown in hyperscaler capex doesn’t strand the asset; it just gets redirected to another end market that’s currently running its own independent upcycle. In other words, CAT’s business can still perform even if hyperscaler CapEx drops or AI demand disappoints, thanks to its outsized exposure to independent upcycles in construction and mining.I’ll get into all of this in detail later, when discussing the company’s outlook and the durability of current growth rates, but my point right now is that this makes CAT an excellent way to benefit from the AI boom without fully relying on the durability of that specific upcycle.Bottom line: I like this AI play and am quite bullish on CAT’s medium-term prospects and the long-term durability of the business, with or without AI. Yet, the whole AI beneficiary narrative is no secret, and CAT stock reflects this. As alluded to before, shares have been a strong outperformer over the past year, and have become quite expensive, trading at the highest valuation in well over a decade earlier this year. Of course, the big argument here is a significantly improved outlook, and Q2 gave us plenty of evidence that the re-rating is more than just multiple expansion chasing a narrative.Back in January, I wrote a Caterpillar deep dive, which can be found here, but following the very strong Q2 results, which easily surpassed my expectations set in my January deep dive, and a 50% share price gain YTD, there is ample reason to take a renewed look at this business. In today’s post, I'll break down the Q2 results and developments before updating my financial forecast and view of the stock.Without further ado, let’s delve in!Welcome to InvestInsights — an independent equity research publication rooted in long-term, buy-and-hold investing, publishing actionable stock/equity research reports weekly!📈 You’re reading my latest [FREE] stock analysis. If you like this analysis, make sure to like & subscribe to receive much more like this, weekly!Subscribe nowFinancial & Performance ReviewDelving straight into the Q2 results, CAT delivered its strongest top-line growth rate since Q3 2021, with Q2 revenue of $20.5 billion (a quarterly record), up 24% YoY. That beat consensus estimates by a solid $1.3 billion and marks the 5th consecutive quarter of accelerating growth, as CAT has seen a steep recovery over the past year, with growth accelerating another 2 points vs. Q1. Of course, given the data center tailwind, a steep recovery was expected, but this continues to surprise to the upside by a good margin, also coming in ahead of management’s expectations.Crucially, this strong, accelerating top-line growth isn’t driven only by data center power demand, but by strong end-market demand across all three primary segments. In fact, outperformance came mainly from strong growth in Construction Industries and Resource Industries, while Power & Energy was broadly in line with expectations.Furthermore, sales to users in all three segments were up strongly. For reference, CAT mostly works through dealer networks. Back in 2024, when growth turned negative, this was primarily a reflection of dealers bringing down inventory, forcing CAT to sort of undership market growth, but as demand now sharply improves, dealers order ahead of market demand, so thanks to these low dealer inventories, CAT is able to grow sales faster than the broader market. Take dealer inventory in Construction Industries, which was up $400 million YoY in Q2, which was stronger than anticipated, reflecting strong end-market demand.Breaking down the performance by segment, let’s start with the one currently getting most of the attention – Power & Energy. For reference, this segment covers engines, turbines, and related systems used in oil and gas, power generation, marine, rail, and industrial applications. Demand here is shaped by global energy investment cycles, electricity consumption growth, and the need for reliable power in remote or industrial settings. It is less directly linked to construction cycles and provides a degree of earnings balance within the broader group.Historically, it has been one of CAT’s lesser-known segments, but it has rapidly grown into its largest operating segment by revenue in recent quarters amid unprecedented demand driven by data center energy needs, making it an investor focus.Sales here hit $8.2 billion in Q2, up 17% YoY, largely in line with expectations, driven by higher sales volume and favorable pricing. Sales to users were up a very strong 33% YoY, amid unprecedented demand, driven by 72% growth in power generation (thanks to data center demand), 6% growth in sales to oil and gas customers, offset by a minor decline in the industrial end-market due to a decline in engines sold in the marine application.Regarding power generation, it is worth noting that CAT is a global leader in turbines and industrial engines, with a market share estimated at around 40% in the U.S. The crucial dynamic, as noted earlier, is a timing mismatch between AI infrastructure buildout and the grid’s ability to keep pace. Utility-scale transmission and new generation capacity can take years to permit and construct, while hyperscalers are trying to bring gigawatt-scale data center campuses online in a fraction of that time. On-site power generation is the bridge, and CAT’s reciprocating engines and gas turbines, technology proven over decades in oil & gas and industrial applications, are among the few solutions that can be manufactured and deployed at the scale and speed this buildout demands.Besides, even when grid energy is available, these data centers are likely to purchase generators for backup power if the grid fails, given uptime requirements and redundancy standards. So, one way or another, CAT benefits from hyperscaler investments.And this is driving incredible growth in this segment today. Yet Power & Energy was CAT’s slowest-growing segment in Q2, largely due to supply constraints as CAT ramps up production capacity amid this incremental demand. Second up is Construction Industries, now its largest segment by revenue, which includes machinery such as excavators, bulldozers, loaders, and compact equipment used primarily in residential and non-residential construction, infrastructure development, and rental markets – primarily the well-known yellow heavy equipment vehicles. This segment is closely tied to global construction activity and public infrastructure spending, making it more economically sensitive, but it also benefits from Caterpillar’s strong position with large contractors and equipment rental companies.Q2 Construction Industries sales were up 35% YoY to $8.3 billion, driven by higher sales volume and favorable pricing. Strong sales volume was driven by a 22% increase in sales to users, the sixth consecutive quarter of STU growth, as reflected in the $400 million increase in dealer inventory. This ordering ahead of demand drove the strong outperformance in this segment.Looking at regional construction performance, North America was especially impressive, with a 50% YoY sales jump, driven by very strong rental fleet loading and equipment sold into both nonresidential and residential construction. Meanwhile, STU was also positive in the EAME, driven by strength in Europe and Africa, offset by unsurprising softness in the Middle East amid the ongoing conflict. Finally, Asia Pacific saw healthy growth in China, offset by softness outside China, keeping performance slightly below expectations.Crucially, this segment also significantly benefits from the data center build-out, as these are heavy construction projects. Hyperscale campuses require massive site preparation, earthmoving, foundations, access roads, drainage, and ongoing expansion. That means sustained demand for excavators, dozers, loaders, compact equipment, and rental fleets, often over multi-year build schedules. Especially in North America, this has driven strong growth in nonresidential construction, and the outlook is favorable. Meanwhile, management indicated that residential construction remains relatively weak by historical standards, leaving significant upside over the next few years. In other words, despite strong growth, this industry is seemingly nowhere near a top.Third up, let’s discuss the Resource Industries segment, which is by far the smallest for CAT by revenue. This one focuses on equipment for mining, quarrying, and heavy earthmoving applications. This includes large mining trucks, electric rope shovels, underground mining equipment, and related services, used to extract commodities such as copper, iron ore, coal, and aggregates. Crucially, the performance in this segment is heavily influenced by commodity prices and mining capital expenditure cycles, but it also benefits from long equipment lifecycles and a substantial aftermarket opportunity as mining customers prioritize uptime, rebuilds, and parts availability.Resource Industries revenue increased by 20% in Q2 to $4.6 billion, also driven by higher sales volume. Sales to users were up 17% YoY, in line with expectations. The primary driver of growth was mining, along with heavy construction and quarry & aggregates.What is driving growth here is still a classic replacement cycle, not relying on data center demand. You see, the current installed mining fleet is old, and as commodity prices remain above certain investment thresholds and utilization is high, it triggers fleet renewal. So, as commodity prices remain strong, helped by semiconductor demand and automotive electrification, CAT should benefit.Finally, briefly worth highlighting is CAT’s financial services arm, which saw sales increase 10% YoY to $1.1 billion. This was driven by higher average earning assets across all regions. For reference, CAT earns interest income by providing loans and leases to customers buying machines and to dealers holding inventory.Overall, CAT delivered a strong quarter in terms of growth, driven by all four segments, each with independent growth drivers. Most importantly, momentum looks good in each, with none looking near a top.The backlog momentum reflects this. CAT reported a backlog of $72 billion at the end of Q2, up $9 billion sequentially and 92% YoY, with all three segments contributing to the sequential improvement. That is incredible growth, showing the insane demand CAT is seeing.For data center power generation, CAT is practically sold out for the next several years, already taking orders for 2029 and 2030. But backlog growth isn’t just driven by data center demand; the oil and gas power backlog is up 100% YoY as well. And don’t forget that these power systems require ongoing maintenance, overhauls, parts replacement, and service contracts over decades. Like heavy equipment, every engine installed creates a long tail of high-margin aftermarket revenue, which is incredibly promising.Helping backlog growth on top of volumes is pricing, with the shortage significantly driving up unit pricing. According to JPMorgan, a combined-cycle gas turbine delivered in 2031 will be 3x as expensive as one last year, with a current lead time of roughly 5 years. This is driving excellent visibility for CAT as well.On that note, let’s move to the P&L.CAT reported an operating profit of $4.5 billion, up 54% YoY, driven by strong sales growth and excellent margin expansion. This reflects an operating margin of 21.9%, up 430 bps YoY, driven by strong operating leverage from rapid sales growth, driving results well ahead of expectations. This still includes a $400 million tariff headwind from tariffs announced since the beginning of 2025, which was well below the $700 million management expected, so that’s a big positive. Furthermore, this was largely offset by IEEPA tariff recoveries of $392 million, so the net tariff impact in Q2 was nearly neutral.The net result was a very healthy operating margin and a strong YoY improvement to the highest level since Q2 2024. Given the expectation of sustained strong sales growth and tariff headwinds easing over time, I see plenty of room for CAT to improve margins from here.Further down the line, CAT reported non-GAAP EPS of $8.17, beating consensus by a sizeable $1.97 and reflecting 73% YoY growth amid strong sales growth, expanding margins, and a lower share count.Finally, Q2 FCF came in at a very strong $5.1 billion, reflecting a 25% FCF margin. YTD, that brings the total to $5.7 billion, up 119% YoY and reflecting a 15% FCF margin, a 6-point improvement YoY. Those are very strong numbers.This allowed management to return $2.2 billion to shareholders in Q2 through repurchases and dividends, well covered by FCF. That CAT is shareholder-friendly is no secret – the company retired nearly 20% of its shares over the last 6 years and has been raising its dividend for 32 straight years.Following an 8% hike earlier this year, shares now yield 0.76% based on a conservative 26% payout ratio. And not only has it shown amazing dividend consistency, but CAT also grows it at a respectable rate, with a 5-year CAGR of 8%.Meanwhile, CAT also maintains an alright balance sheet, even though I am not a huge fan of its capital allocation strategy in recent years. You see, management has been adding a lot of debt to fund its royal cash returns to shareholders, returning over 100% of FCF, which has worsened the balance sheet quite a bit.As of the end of Q2, total cash on the balance sheet stood at $6 billion, while total debt hit $45 billion. This translated to a net debt of $39 billion, up roughly $10 billion over the last 5 years. Now, I don’t deem this a huge problem for CAT given its excellent cash flows, which shouldn’t make maturities a problem, but I do hope to see more active balance sheet management from here.Ultimately, I can only say I was hugely impressed by CAT’s excellent Q2 results. The company continues to far outperform the expectations I set in January, not only thanks to explosive data center demand, but also thanks to strong performance across all three primary segments. And I really like the momentum I am seeing, with CAT sitting on a business that can accelerate further and continue to deliver strong results for years to come, with good visibility.Meanwhile, margins are improving despite elevated investments in the business, cash flows are excellent, and financial health is acceptable.Finally, for what it’s worth, CAT’s reinvestment metrics look sublime. Its TTM ROIC sits at 22% and its TTM ROE at 57%, showing management knows how to deploy cash and deliver strong ROI.With that, let’s move to the outlook!More content like this? Consider InvestInsights PremiumInvestInsights Premium is designed to help you find best-in-class compounders, identify deep-value opportunities, avoid costly mistakes, and stay focused on fundamentals when markets get noisy, with full transparency into how I make decisions over time.In practice, this gets you:6–8 Research Reports per month, 2x more than free subs do, which includes 3-4 Deep Dives!Full visibility into my personal portfolio, including allocations, transactions, and conviction levelsAn overview of target prices and ratings, updated weekly, all available in a web-based Hub.Price tag? Only $95 per year or $9.50 a month. That comes to just $1 per analysis, with the rest provided at no cost.Sound useful? You can upgrade to Premium today using the button below!Subscribe nowOutlook & ValuationAs always, let’s start with management’s guidance. Following the excellent first-half results and strong demand momentum across all end markets, management raised its full-year 2026 expectations despite ongoing uncertainty from geopolitical events.It now expects revenue growth in the mid-to-high teens. This will be supported by production capacity expansion in the second half of the year to improve product output, especially in reciprocating engines. I expect this to fuel a sharp acceleration in Power & Energy sales growth.Meanwhile, management also raised its operating profit margin, now expected to be higher than before, and, excluding the IEEPA tariff recovery, this should come in at the low end of the long-term target range of 21-25%, so likely around 21-22%, suggesting H2 margins to improve from H1. This includes elevated investments to offset most of the gained operating leverage and a total tariff headwind of $2.2 billion (excluding refunds). Considering the circumstances, that is very strong margin guidance. Finally, FCF is expected to be in the top half of the $6-15 billion long-term target range, suggesting a record FCF of at least $11-12 billion.As for what will drive this growth, data center demand will remain a key driver, but oil and gas engine demand is also expected to set new records in 2026. In Construction Industries, the outlook for North America is good, with construction spending at healthy levels, helped by strong nonresidential demand coming from critical infrastructure programs, heavy construction, and data centers. In the EAME, demand should remain solid, driven by construction activity in Africa, more than offsetting Middle East headwinds. Finally, in resource industries, demand should remain healthy amid rising demand for copper and gold and positive dynamics in heavy construction and quarry and aggregates. Crucially, key commodities remain above investment thresholds, utilization is high, and fleet age remains elevated, which is a very good backdrop for CAT.Overall, I’d say that is excellent guidance, again well above what I anticipated at the start of the year.But personally, I think the more important question is how durable this current growth and demand is. You see, CAT’s business remains inherently cyclical, tied to construction budgets and activity, commodity prices, and investments from oil giants and now hyperscalers. A growing portion of revenue coming from installed base services, by now about 40%, does offset some of this cyclicality, but it’s still a cyclical business. The current upcycle is strong, but how long will it last, especially the booming demand from hyperscalers, which is critical in the medium term?Well, management indicates that current gas prime power orders run into 2029, turbines are booked even further out, and diesel standby is booked well into 2028. Some orders across these products already run to 2030, so this engine demand looks quite durable. Furthermore, the backlog-to-near-term-delivery ratio (~59% shippable within 12 months) has held stable for three straight quarters, so even as management adds capacity, the order trend or demand keeps growing at least as quickly.So, as far as we can tell right now, this datacenter-driven engine demand looks quite sustainable, with growth likely ramping in 2027 and potentially 2028 as CAT improves production output, and I don’t see a huge drop-off before 2030. Crucially, I see the risk of order cancellations as very small, given the waitlist and the fact that these facilities will already be under construction well before then. That makes for a very strong setup.Meanwhile, the construction outlook remains relatively strong. Projections indicate that 800 million people will move to urban areas over the next 10 years through 3035, driving significant construction activity. Besides, many governments are increasing outlays on roads, bridges, rail, airports, utilities, and urban development, often tied to long-range plans for transportation modernization, climate resilience, and public works. Because infrastructure projects typically use heavy equipment intensively over long durations, this trend supports elevated baseline demand for both new machines and aftermarket services.Additionally, the long-term outlook for mined materials is exceptionally favorable due to electrification, decarbonization, and infrastructure build-out. Copper, nickel, lithium, iron ore, and other critical materials are essential for power grids, EVs, renewables, and data centers. Supply constraints and years of underinvestment mean miners must sustain heightened capital spending in the coming years to renew an aging fleet, supported by elevated commodity prices.Across the end markets, automation and electrification also function as long-term tailwinds, driving up unit prices, accelerating a replacement cycle, and growing the services pie through software.Ultimately, management itself now has the following targets through 2030:A 6-9% sales CAGR, accelerating from 4% in the prior 5 years, with all segments likely contributing.The operating margin is expected to strengthen to 21-25% by 2030, up from 15-19% today, driven by improving operating leverage, primarily from scale.Annual FCF to hit $15+ billion by 2030.CAT aims to return 100% of FCF to shareholders. This includes a high single-digit dividend CAGR, which should account for now more than 60% of FCF, leaving ample room for buybacks.Now I’ll move to my updated expectations. Clearly, I was far too conservative back in January. Following the strong H1 results and the current demand momentum, I have significantly raised my financial forecast. For 2026, I now expect CAT to deliver sales growth just shy of 18% to roughly $79.5 billion, assuming a slight moderation in the growth rate in Q2 as CAT starts lapping last year’s stronger recovery quarters, which will pose a tougher comparison. I expect Power & Energy growth to accelerate, while the other segments, especially construction, will moderate slightly.On the bottom line, I expect very strong margin expansion to persist in H2, driven by a lower tariff headwind and strong operating leverage. I now expect EPS to grow 44% YoY and FCF to come in just shy of $12 billion, consistent with management’s indication that H2 FCF will exceed H1. This reflects a 15% FCF margin, balancing much-improved operating cash flows with higher CapEx to support production capacity expansion.Looking further ahead, I expect growth to moderate in the following years but to remain in the double digits, driven by positive growth in Construction and Resources and outsized growth in Power & Energy as a result of data center demand. I do not expect a significant weakening in any segment; the lower growth rate reflects higher base comparisons in the prior years.On the bottom line, I now assume CAT to deliver consistent margin expansion through 2029 as a result of strong operating leverage and easing tariff pressure, which should allow for a high-teens EPS CAGR. Finally, I expect FCF margin to improve to 16% by 2029, allowing FCF to exceed $17 billion, driven by strong cash flows, offset by CapEx growth.All these assumptions are reflected in the updated financial model below.That then brings me to the valuation, and although the medium-term outlook for CAT has significantly improved over the last year, CAT shares have kept up just fine, up 50% YTD and 106% in the last 12 months, which means shares are far from cheap. At a current price of $859, shares trade at:31x 2026 earnings.33 2026 FCFWhile these multiples might not look extreme against double-digit sales growth and a high-teens EPS CAGR, let’s not forget we’re looking at an industrial giant with exposure to inherently cyclical end-markets, construction, mining, and oil & gas among them, even with the AI-driven diversification argument made earlier. For reference, the current 31x earnings multiple is an 82% premium to CAT’s 10-year median of 17.5x, and 33x FCF marks an 89% premium to its own 10-year median. Shares have fallen from peak multiples above 40x earlier this year, but even after that pullback, CAT still trades at a very hefty price tag relative to its own history.In other words, CAT is no longer being valued as a cyclical industrial riding a strong upcycle; it’s being priced closer to a structural growth compounder. Don’t get me wrong, I think a premium to CAT’s historical multiple is genuinely justified here, as hyperscaler budgets behave less cyclically than the oil & gas and mining capex CAT has traditionally depended on, and the visibility into 2029 and 2030 is about as good as this business has ever had.Nonetheless, we shouldn’t act as if this isn’t a cyclical industrial, something we tend to lose sight of in the midst of an upcycle. Every upcycle in CAT’s history has eventually given way to a downcycle, and while the drivers this time look more diversified and better contracted than in past cycles, the underlying business is still tied to capital spending decisions that can turn quickly when conditions change – don’t underestimate that.Ultimately, I definitely view CAT as a long-term compounder, able to generate strong returns through the cycles, as it has done for decades. Yet, the current share price and multiples are too rich for my taste. For reference, say we apply a 25x 2028 exit multiple, which is the highest multiple I would be willing to pay for 2028 profits; I calculate an end-of-2028 target price of $996. From a current share price of $858, this implies annualized returns of roughly 7%, including dividends, which is well short of my 15% threshold.In other words, this is a wonderful business with a strong medium-term outlook, but I don’t deem the current risk-reward for CAT shares anywhere near attractive. Personally, I will only consider buying shares below $700, which would require a 19% pullback.Rating + fair value: Hold — Accumulate below $7002028 Target Price: $996Implied CAGR from current price: ~7%Please leave a like if you enjoyed this post and found it valuable!Subscribe nowDisclosure: I/we may or may not have a beneficial long position in any of the securities discussed in this post, either through stock ownership, options, or other derivatives. This article expresses our own opinions, and we are not receiving any sort of compensation for it.No recommendation or advice is being given as to whether any investment is suitable for a particular investor. The information provided in this analysis is for educational and informational purposes only. It is not intended as and should not be considered investment advice or a recommendation to buy or sell any security.Investing in stocks and securities involves risks, and past performance is not indicative of future results. Readers are advised to conduct their own research before making any investment decisions.