Tech Thoughts Newsletter – 18 November 2022.

It was another interesting week in the markets – with lots of volatility and fragility in tech. For us, there was some good news as Infineon, Applied Materials, Cisco and Palo Alto (and ASML last week) – all core holdings, all raised guidance and beat. They’re benefitting from structural demand in their key markets, but also show very resilient business models and returns profiles, which is what we continue to look to own in the portfolio.

Key takes:

Auto semis showing no signs of a slowdown – Infineon raises guidance and returns are going higher

• Infineon (owned) gave us a surprise pre-release on Monday afternoon ahead of its scheduled results. It beat the quarter impressively but most importantly for us raised its 2023 and mid-term targets around both growth and profitability.
• We’ve spoken before about the growth in auto semis (and in particular the power components that Infineon supplies into electric vehicle powertrains) – not only is that helping its top line growth, but it’s also driving Infineon’s return on invested capital higher- which is important for us as shareholders.
• Infineon announced its largest ever capex plan – a €5bn 300mm fab in Dresden, Germany – which will mean continued cost advantages over competitors. Interestingly the release stated that there will be funding through the European Chips Act. Again, political will is helping semis right now (we talk further down our letter about the new Taiwanese Chips Act). It also follows several trailing edge fab announcements – we know there is a significant structural shortage that needs to be built for, which Applied Materials (below) also spoke to. 
• Infineon also noted design wins in Silicon Carbide at Stellantis and a Japanese OEM – and we think it’s also won Tesla’s fast-charging onboard charger in the Model 3. Silicon Carbide is a market with just 5 competitors, extremely limited by materials supply – that means we expect good pricing power and for that to be one of the drivers of Infineon’s higher ROIC over time.  
• ROIC is also being helped by its cost-efficient 300mm fabs as well as natural operating leverage from the top line growth.
• Staying on Autos, this morning, Ford’s CEO said that demand for auto chips will outstrip new supply.
• Auto semis continues to be a long term attractive market we’re happy to be exposed to (we also own NXP).

Cloud and data centre spending also remains robust. We think the market is overly concerned that US data centre spending will see a dramatic slowdown – we disagree.

• Nvidia (owned) was perhaps one of the sets of results that investors were most nervous about this quarter, partly after the very big surprise negative warning last quarter.
• We’ve seen a very very large correction related to gaming and crypto (which is still hard to disaggregate – despite Nvidia’s best efforts, its gaming chips can be utilised and repurposed for crypto mining). Looking past this short-term crypto/PC and gaming weakness, really the key for this set of results was its data centre business – which is really the long term secular driver of Nvidia.
• There have been mumblings around a US cloud/hyperscaler spend slowdown and investors are looking for signs that that market is the next shoe to drop.
• Happily, this set of results just didn’t show that: Nvidia’s data centre business is still growing 30% yr/yr, and the guidance implies this should continue to remain pretty robust, with higher value, more performant products driving demand.
• Nvidia will continue to be driven by (1) AI workloads increasing (2) workloads shifting from CPU to GPU – in some cases this can be 20x more power efficient, which we know will be an increasing feature of the cost component for cloud businesses, but also given that they are all very motivated to cut their carbon footprint (noting COP27 this week).
• Cisco (owned) beat the quarter and raised guidance for the full year. We mentioned in our portfolio changes last week that we had increased Cisco into results, given our view of still continued robust demand in the sector despite market concerns. For Cisco the slight uncertainty into the quarter was also around supply.
• This set of results spoke positively to two things (1) they’ve done a good job with redesigns to navigate that supply, especially given the China lockdowns over the past few months. (2) order trends look good, which along with the guidance raise, shows continued demand for Cisco’s products and rebuffs some of the concerns around share losses.
• We’ve spoken before about not expecting the dramatic slowdown the market fears in cloud and AI. We think both Cisco and Nvidia speak to that. The reality is there isn’t that much capacity in US public cloud/hyperscaler compute, and if there is a pause in spend it likely relates much more to waiting for new higher performance (and higher value) products – like Nvidia’s Hopper, memory upgrades (DDR5) and platform transitions (PCIe 5). Cloud continues to be a very robust secular driver and a market where you keep seeing tech driving upgrades around performance (because performance/power is so key to the ongoing TCO) – tech leadership driving demand, which is good for the long term.

But there is something going on in China cloud spend…

• We would acknowledge though that as Nvidia commented, there does seem to be some softness in China data centre spending. We can see that from Tencent and Alibaba’s reported capex – which has come down quite substantially from prior years.
• To be honest we’re still struggling a bit to understand what’s really going on here: whether China has overbuilt its cloud services – it is absolutely true that there has been a lot of spend over the past several years, and we have limited visibility on capacity utilisation of cloud services in China. 
• Or, whether it’s temporary messaging from China big tech – the China internet industry message seems to be around disciplined investment, and politically given they’re cutting staff, perhaps they need to be seen to be slowing spend elsewhere.
• A final speculation is whether China cloud spenders are slowing spend in order to prioritise China designed chips. China has built an impressive fabless ecosystem – Alibaba announced that it had put its in-house designed ARM based CPU into production for use in its data centres. So there may be politics at play here too (an ongoing factor in China – see below our comments around gaming).

Semicap equipment – after ASML last week, AMAT beat shows again the structural drivers winning over memory cyclicality (Micron)

• AMAT (owned) – beat the quarter and guided above expectations and really showed structural strength in logic/foundry/IDM outweighing the cyclical memory market.
• It came as a bit of a surprise given AMAT cut its guidance 3 weeks before the quarter ended – What seems to have happened those 3 weeks in between is that the China export constraints had a little bit less of an impact, and then supply chain eased to push it on a bit too. It makes us think that perhaps the cuts (from all the US players) around China may have been somewhat premature.. And we know from SMIC last week that China wants to keep spending.
• Looking ahead they’re seeing good growth ex China, certainly beating market hopes which are still hovering around flat qtr/qtr across the sector..
• Their comments on the secular drivers of their business and the broader industry are well worth pulling out around tech leadership and semi content increases:
  “They’re all racing against each other for power, performance cost, their relative competitive position…. Consumer-driven markets are clearly softer, while the automotive, industrial and power markets remain robust. Those investments are underpinned by large inflections, including the transition to electric vehicles, accelerated adoption of industrial automation and growing demand for renewable energy solutions, especially in Europe.”
• While they noted they are seeing memory pushouts, they also have a record level of backlog, and are still supply constrained.. This is all contributing to our view that the risk dynamics in the industry are shifting and short term revenue estimates to us look overly derisked
• Their services (tool maintenance) component too was strong and helps with the cyclical profile of returns – services is now well over a quarter of total revenue.
• AMAT’s good news followed a Micron warning and guided capex cuts mid-week that had put some doubt back into the sector as memory cyclicality reared its head again – Micron announcing it would reduce wafer starts by 20% and further cut capex.
• We’d note the stark difference in the short and long term capex outlooks – long term, the company is still committed to building out its US domiciled manufacturing footprint, related to geopolitical will and the CHIPS act. The company is still committing to investing in EUV (tools provided by ASML) and we still see long term investment in memory tech upgrades as a necessary given the rising competitive threats in China.
• Elsewhere in semicap, rumours of a Taiwanese Chips Act (after we wrote last week about Japan). We’d note too that Infineon’s fab announcement was related to the European Chips Act. It isn’t totally clear to us that lots of government subsidies are ever helpful to industries over the long run, but we think we’ll start to see it have a meaningful impact in order books and revenues over the short and mid term.

Cyber security – Palo Alto benefitting from spend consolidation

• Palo Alto (owned) results were great – Q1 billings growth of +27% yr/yr beat consensus (+22%), and they raised the top end of the guidance range – not by a huge amount, but good in the market context of broader softer commentary (from the likes of Fortinet).
• The growth is being driven by customers increasing their spend within Palo Alto. We think it’s executing the land and expand strategy – and increasing its platform offering (it announced another acquisition today) amazingly well – and it’s probably benefitting from some of that consolidation of spend that both its legacy and niche peers are being pressured by – below a quote from the CEO Nikesh Arora, talking to that:
  “Cybersecurity deals are getting more scrutiny, suggesting deeper and longer reviews of transformational projects. New conversations that include payment terms, the discounts are causing deal cycles to elongate… Given the increased scrutiny and return requirements, the silver lining for Palo Alto Networks is that we are having more conversations around consolidating platforms than we’ve ever had before. We think customers are less likely to purchase newer security products. Instead, they will continue to consolidate towards like-for-like capabilities from fewer vendors.”
• Importantly too, Palo Alto’s revenue upside is driving profitability and cash – generating $1bn FCF in the quarter. Again we focus on those businesses that can really drive the leverage that we’d expect them to – and note that this is sometimes missing in software, with some businesses overly dependent on sales and marketing expenses.
• For us, Palo Alto is the standout software stock to own in the cyber space

Finally – A number that surprised us this week.. When China wants children to stop playing video games, it happens

• We don’t own any China stocks in our portfolio, but one number which stood out to us this week in Tencent’s release was this:
  “With the implementation of our world-leading Minor7 protection program, we have become fully compliant with China regulations, and fostered a healthier industry environment. Time spent from Minors decreased by 92% year-on-year.”
• 92%!! It’s an astonishing number for the definitive market leader in China gaming- and it does show just what a different environment China is for businesses and indeed investors. There is undoubtedly a huge amount of innovation in China tech – not only the internet companies – it has also grown a very large fabless ecosystem around ARM. But for us the difficulty is in assessing the political risk which can translate into significant earnings and returns risk – as has been seen with Tencent.

Staying on gaming – Embracer reported and lowered its guidance – we don’t own it, and see significant concerns around cash flow and ongoing high development costs excluded from adjusted numbers

• Like many gaming peers, Embracer has been hit by the macroeconomic situation leading to weaker demand.
• What’s really concerning to us digging into the numbers though is (1) whether underlying sales are really growing ex FX impacts; and (2) the cashflow – Embracer capitalises its development costs which means headline adjusted numbers look relatively better, while cash flow – which really shows us the underlying health of the business, continues to be weak and leads us to question the valuation of Embracer’s business.

Portfolio

• Last week we increased our positions in Cisco ahead of numbers – which, as we expected, was very solid and the company raised their full year guidance.
• The second position we increased was TSMC – we just thought the valuation of a world leading company had become too low. Gladly enough it seems like Warren Buffet’s team thought the same.

This week, we have not made any real adjustments to the portfolio.

For weekly insights on the latest market updates, please subscribe to our Tech Thoughts podcast.
For more information about Tech Thoughts, please visit https://www.gpbullhound.com/tech-thoughts/.

We provide investors with access to category leading technology companies, globally. Our assets under management have a total value of more than €1bn, and our limited partners include institutions, family offices and entrepreneurs. Learn more about our funds here.

Enquiries
For enquiries, please contact GP Bullhound’s Public Funds team at public.funds@gpbullhound.com.

About GP Bullhound
GP Bullhound is a leading technology advisory and investment firm, providing transaction advice and capital to the world’s best entrepreneurs and founders. Founded in 1999 in London and Menlo Park, the firm today has 12 offices spanning Europe, the US and Asia. For more information, please visit www.gpbullhound.com.