Conference Insights: Thoughts from our 2026 Tech Conference
The desk highlights the ongoing demand for artificial intelligence (AI) and its related infrastructure as a significant theme emerging from Deutsche Bank’s recent Technology Conference. Per the full note , firms such as Microsoft and Oracle indicated that the demand signal for AI remains robust and extends well into the next several years. This outlook is crucial as it influences not only tech equity valuations but also broader market sentiments, especially in economies heavily reliant on technological advancements. We are currently seeing consensus targets float around 1.075 with a range reflecting bullish and bearish sentiments on tech's impact on currency movements.
What the desk is arguing
The current business environment showcases an unprecedented demand for AI infrastructures, which has captured attention across industry segments. Per the full note, this demand is accompanied by supply constraints and an overarching optimism from management teams regarding their revenue prospects linked to AI.
Deutsche Bank’s report points out the significant conversations focused on the urgency to accelerate AI infrastructure build-out, particularly from major players like Microsoft and Oracle. Enhanced efficiencies in financing and the expected returns on these investments are critical parameters influencing market stability.
Where it sits in our coverage
Our analysis of the current market landscape shows a consensus target for the related currency pairs around 1.075, with notable forecasts from several firms: - jpmorgan: 1.10 for Mar26 - bofa: 1.04 for Mar26
This view resonates closely with the optimistic narrative shared at the Technology Conference, positioning it slightly above the lower bound of the spectrum.
How other firms see it
The general sentiment is shared by firms like jpmorgan and barclays, which also lean towards a bullish outlook on tech stocks influencing currencies. In contrast, bofa holds a more pessimistic view, advocating for caution against overexposure in tech-driven markets.
Related indicators to watch closely include the correlation between the EUR/USD trajectory and the performance of major tech stocks, as well as central bank policies that could shift due to heightened tech sector dynamics.
How firms align with this view
Aligned with the desk view
Contrary positioning
Key takeaways
- 01Robust demand for AI infrastructure across major companies
- 02Positive outlook influences tech-related equity valuations
- 03Shift in focus to financing efficiency and returns on AI investments
Market implications
Traders should monitor the 1.075 level as a critical resistance, with potential shifts indicated by upcoming earnings from major tech firms. Positioning signals may provide insight into shifts in market sentiment as they relate to technological advancements.
Risks to this view
A reversal could occur if supply constraints are resolved more rapidly than anticipated, significantly altering the investment landscape. Additionally, any unexpected central bank tightening could dampen market optimism, leading to bearish trends in tech-heavy indices.
PODCEPT, the podcast from Deutsche Bank Research, with interviews on current economic and financial topics. Listen as economists and analysts from Deutsche Bank present their views. Welcome.
You're listening to another episode of PODCEPT, the series where we discuss some of the best ideas coming out of Deutsche Bank Research. My name is Matt Barnard, Director of U.S. Equity Research here at Deutsche Bank.
Just a few weeks ago, at the end of August, we hosted our annual tech conference in Dana Point, California. The conference was a great way to kick off what will be a very busy sprint into the end of the year. This year's conference featured nearly 100 companies from across all the key tech sectors, including software, hardware, semis, internet, data networking, payments, IT services, and many more.
And no surprise, AI was the center of most conversations, with the overall tone for management teams remaining positive, both around how AI is impacting their top lines as well as their bottom line. To help us unpack some of the key themes discussed at the conference, with me today is Brad Zelnick, Melissa Weathers, Jean-Marc Oconti, and Nate Svensson. We have a lot to discuss today, so let's dive right in.
Brad, let's start with you. What were some of the key takeaways from the conference in your view? Sure.
Thanks so much, Matt. So, we hosted nearly two dozen software companies out at the event, and there were a number of interesting takeaways. First on AI infrastructure, the build-out still looks exceptionally strong.
Across Microsoft, Oracle, and CoreWeave, the message that we heard from these companies was pretty consistent. Demand remains robust, supply is still constrained, and importantly, the demand signal extends out multiple years. The conversation increasingly seems to be shifting from whether this capacity gets built to how quickly it can come online and how efficiently it can be financed and what returns providers can generate.
I would say that fundamentally, things sound very good listening to executives from the various companies, but there is healthy investor skepticism just around NIMBYism derailing large data center projects, questions on where financing will come from, jitteriness in the bond markets, and whether this might be peak cycle or not. Just maybe next on cybersecurity, it really seems to be living up to its reputation as a beneficiary of AI adoption. Companies described a stronger spending environment as AI increases network traffic, expands attack surfaces, and enables threat actors to operate at machine speed.
Fortinet, just as an example, by the way, they see enterprise AI increasing requirements for network capacity and segmentation, while its opportunities to secure AI itself is still relatively nascent. We also had CrowdStrike report blowout results that same week, benefiting from AI-related demand, which to us is a strong signal for cybersecurity stocks and one we think can be a canary in the coal mine. Finally, just on applications and SaaS, we thought the discussions were definitely more encouraging than the prevailing disruption narrative would suggest.
Customers are moving beyond experimentation toward scaled AI deployments, but data readiness, governance, security, and measurable ROI are all prerequisites that these customers need. This is creating an opportunity for incumbent platforms rather than simply making them obsolete. I'll also call out an important Salesforce announcement that coincided with our conference, which is a new joint product offering with one of the Frontier Labs that was well-received by the markets and reminds us that systems of record will likely be durable into the future and that the Frontier Labs have an interest in working with established SaaS leaders, at least for now.
So bottom line, great event. I feel like we learned a lot, and in summary, we expect demand for AI infrastructure to well exceed supply for at least the next couple of years. Cybersecurity is seeing some very real AI tailwinds, and SaaS apps are likely more durable than at least a lot of the more bearish software narratives out there.
Yeah, there's a lot to unpack there. I mean, the demand outstripping supply theme, I think, is something we'll probably discuss quite a bit today. That was pretty prevalent across most of our conversations.
But more specific to you and cybersecurity, at times it's been viewed as defensive within your software space. But now with AI and some of the threats that you talked about, is this making this group more of a growth category again? And how long do you think that can last?
I think there is a lot of secular tailwind from AI benefiting the cyberspace right now. And I think it manifests and will continue to manifest in a number of different ways. CrowdStrike happened to be the first company out in front.
And why is that? That's because the use of AI today, we're all talking about agents, but today's AI is writing prompts in browsers on endpoints. As a leader in endpoint security, it would make sense that they will see it first.
But there's much to come. We see a tidal wave of demand in both securing AI, as well as using AI for better security. Now that we're in an era where the adversary is able to operate at machine speed, making use of the latest frontier models, and we're seeing that's a great effect.
One more question I want to ask. If you think of, if we're here in a year's time recording this podcast from the 2027 Tech Conference, of these three themes, I mean, it's AI demand, supply demand, cybersecurity, and the durability of SaaS overall, what do you think will be the most important theme that investors will be talking about in a year's time? I think it's going to be AI infrastructure, and I'm sure that's going to lead into what you're going to hear from some of the other speakers on this podcast as well.
Because we have visibility, we look at these company backlogs, we can see the tightness in the supply chain that's going to last for years to come. And this is a generational build out with TAMs to the order of trillions and trillions of dollars. And I think there'll be healthy debate around it, but I think that will be at the forefront of relevance in a year from now.
Yeah, that's a great segue into Semis and Melissa. So Melissa, obviously, semiconductors are the core of what we're trying to build here from the AI space going forward. What were some of the key takeaways from the conference in your view?
Sure. Thanks for having us do this. On the Semi side, I would say the most notable takeaway I had from the conference was the sheer difference in outlooks between corporate management teams and investors.
It was pretty amazing. You would be in the room with a Semis company, with the CEO, CFO, and they sound very, very constructive and frankly excited about the growth that they're about to see in the coming quarters. We see new fabs are getting built at historic levels.
We're seeing very strong demand for chips and memory is a whole different discussion by itself. So in the meeting rooms, the tone was very, very positive. Then you walk out of the meeting rooms and the sentiment from the investors themselves was a lot more, it was very different.
I would maybe even characterize it as dire. Now, the reason why investor sentiment differs so strongly from the corporate management teams, I think it stems from two pieces. Brad talked about one of them.
The first would be peak cycle fears where we've already seen really strong growth, especially in the memory space. And so there's this fear that you will start to get a reversal if there are cracks in demand, if supply comes on too fast, then how much goodness do we have left? The second piece I think that investors struggled with is, is the good news already out there?
So it's great that fundamentals are strong and likely to continue to be strong, but stocks priced in a lot of that in the first half of this year. So you've seen valuations compress in the semis group over the last month or two. Investors are really struggling to see, hey, where can I get positive earnings revisions from here?
Now, there are a couple of interesting themes within the conference that I think corporates are starting to shift towards, mainly margin expansion. So if we know the top line is strong, then what are the names where you could see some company-specific margin expansion? So a lot of focus there.
Outside of the AI, I mean, memory was top of mind for everybody. There are some concerns on where is true HBM demand going? What are the solutions that could break the memory wall?
So I think that's something that people are monitoring. And then outside of the AI theme in the more cyclical part of semis, I think corporate tones still sounded very constructive, looking pretty good on the industrial semi side, automotive also doing okay. But certainly, I think most of the focus for all these conversations was centered around where AI can take the top line for this group.
Yeah. And as Brad said, and then you clearly saw it with most of your conversations, the durability of demand is a big question mark. So what do you think investors are looking for over, let's call it the medium term, to understand that this durability in the revenue growth is going to be there and persist over the next, call it one to two years, which is, I think, where the management teams feel they will be?
Well, in the semis group, unfortunately, we have a lot of scar tissue from cycles past. So depending on the investor you talk to, I think the longer they've been in the industry, the more jaded they are about where the durability of the cycle could go. Some of the interesting pieces that we're following is, for example, in memory, the contracting terms have changed.
We're seeing longer term contracts that average duration of a contract is pushing three to five years in memory. That's historic. It's never happened before.
Now, in semis, the scar tissue leads us to believe that maybe those contracts aren't worth the paper they're written on. And so there's some skepticism about how enforceable are these? What are the downside terms?
And so I think continued clarity on how that contracting is changing, you're starting to get some support from semis investors. If you can show that gross margins can profit at a healthier level than past cycles, that's what investors are looking for. Unfortunately, the only way to prove that is by going through the down cycle itself.
So it's hard to disprove the negative. But at least from our conversations with the management teams, it seems like they're being very constructive about how they're running their business, keeping an eye on the cycle while still wanting to be able to support the kind of growth that Brad's companies are looking for them to provide. Right.
Never a dull moment, for sure. It'll be interesting to follow this. Jean-Marco, you know, you now are recently covering the data networking and hardware space for Deutsche Bank here in the United States.
So it's great to have you live on the stocks. And you were obviously at the conference as well. So same question.
What were the key themes that you picked up from your conversations? Yeah. Thank you, Matt, for having me.
So I think I would definitely echo what Melissa was saying about corporate sounding upbeat and staying on theme of what also Brad mentioned around the sort of supply and demand imbalance. I would say that starting with the data center build outs, particularly on the optical side, it's clear that the bottlenecks continue to be the supply of indium phosphide and the ramp of laser chips as there's limited availability. Of course, we have an evolving architecture, which meets a very strong supply and demand imbalance.
I would say here, particularly for optical circuit switches, they particularly have appeared to be ramping up with more customers demand, as well as CPU and NPO progressing well, which is adding up to a new market. These themes have been benefiting both momentum and coherence. Now, on the networking side, it's clear to us that there is a very strong demand for back-end AI networking, given the push for scale up and scale out from hyperscalers.
Scale across is also a big theme for names such as Cisco, as there are power requirements that are hitting a wall. And thus, we need to see that data centers need to be connected from one to another. And this is particularly helping companies like Cisco.
I would also say that there were some interesting themes around campus refresh that appear to be happening with some level of urgency. There are concerns from enterprises around cyber and quantum, which position networking appliances as a priority into their budgets. Finally, on the EDA space, our conversations make it clear that fears around AI potentially replacing chip design giants aren't quite founded, as the developments from companies such as Cadence in making sure they lead the race are strong, with historically better EDA growth and all-around tailwinds, which helps them capture more of the market.
So sticking with the changes, the technological changes within the data center and as it relates to optical starting to replace copper, I think that's a theme that probably has some legs and something we heard many times at the conference. What are the key hurdles to enable that to happen? Is that a big rip and replace?
That clearly is something to get the speeds we need to have AI achieve the goals that everyone thinks it can. What are the key technical hurdles to get that to happen? Yeah, I mean, like the copper versus optics debate has been one in the space for quite some time.
I think physics really does most of the arguing here. We're going to increase the bandwidths and as we increase the bandwidths, that requires physics dictate that signal loss happens. And so the distance between how many accelerators you can connect actually shrinks.
And so eventually we'll get to a point where we need optical. And so right now we're seeing companies like Lomentum Coherency, really strong demand for the laser chips because of the shifting in architecture, both within so scale up in scale out and scale across. And so as we're shifting towards optical, there is a need for laser chips because there is a need to laze.
And we all know that silicon cannot laze. And so although we've seen some developments in silicon photonics, which help modulate the light, the actual light, you still need the laser chips to send the data. And so for me and for us, the argument is that the copper bull will eventually hit.
It's just not hitting quite today. But in about two to three years time, it will be an imperative discussion and it will be one where the optical names will likely be continuing to see tailwinds. Okay.
Very interesting. Nate Svensson, let's turn over to you. The IT services companies at the conference, and there was a number of different conversations around those and how AI is impacting their businesses, which is clearly the center of a lot of debate among investors.
And you also had a number of your payments and FinTech companies. So maybe you can unpack some of the key themes from both of those subsectors you cover. Yeah, thanks, Matt.
Thanks for having me on. Yeah, we were very happy to host 13 companies across both payments and IT services at the conference. So I'll touch on some high level themes across both and then wrap up with some brief commentary that we got on macro conditions across both sectors.
So maybe starting with payments, the recurring thing that kept coming up in all of our meetings and fireside chats was the push that our companies are making to move beyond payments and into higher value recurring revenue streams. So among the companies that attended our conference, management teams were increasingly focused on value added services, software, financial services, and of course, AI to drive the next level of growth for the industry. Some examples of this include the card network.
So that's Visa and MasterCard highlighting continued momentum across transaction tokenization, cybersecurity, and fraud prevention as AI starts to impact how payments are consummated. Separately, point of sale companies like Global Payments are increasingly viewing software as their lead product when going to market rather than simply acting as a payment processor. PayPal is another company that I would highlight given the company's push into more traditional banking and lending services through their Venmo and Buy Now, Pay Later offering.
So you take all that together and our payments companies are looking to get out of just being payments processors and starting to add more value in helping companies run their businesses and deal with new emerging threat vectors associated with AI. In IT services, unsurprisingly, every single conversation that we had focused on AI. But it was interesting to us that it did feel like the debate around AI and IT services is starting to evolve a little bit to focus on how these companies can reposition themselves to benefit in this rapidly changing environment.
For example, Cognizant highlighted that AI is increasingly helping the company win and expand client engagements, although higher productivity commitments and larger pricing concessions are becoming embedded in contracts and pushing the industry more towards outcome-based pricing models. Genpak, another company that attended, similarly argued that agentic AI can expand the scope of work they perform rather than simply acting as a tool to reduce headcount. Finally, Accenture emphasized that enterprise AI adoption still requires significant investment in data, security, governance, platform modernization, et cetera.
But the company is seeing strong demand for their cybersecurity solutions, similar to what Brad was talking about earlier, as new AI-related fraud vectors are impacting all of their clients. Very quickly, I will touch on macro conditions because, for better or worse, our sectors are very heavily influenced by what goes on in the macro. And interestingly, the broad message across both groups is pretty much stability versus what we saw coming out of the second quarter.
So in payments, what that means is consumer spending remains resilient. But there are a couple items worth monitoring, namely a slowdown in e-comm spending in Germany and some tariff-related impacts in continental Europe. In IT services, stability means that there is still limited discretionary spending.
So there's not very many IT budget dollars being allocated towards services, providers, which we think pushes out a recovery or an acceleration in growth for the industry until 2027 at the earliest. So all in all, not much change in macro trends. But that's a good thing for payments, not so much of a good thing, maybe more of a headwind for IT services companies.
Thanks, Dave. And sticking with the IT services, you mentioned a bit of a model shift towards outcome-based pricing for at least some of the companies. That seems like a new theme.
And if they were to be successful in moving to that type of pricing model, how would that impact margins and profitability overall? Yeah, I think the key part of your question is if they are successful, right? Historically, these models have been headcount art models.
So they are charging their clients X dollars per hour per employee, paying those employees Y dollars per hour, and then capturing the spread on those two. And now what we are seeing related to the commentary I made that we heard in our meetings with Cognizant is that the pricing concessions and efficiency demands that clients are demanding because of AI are only increasing. And so what clients are going to these services providers and saying, hey, this project that we were running with you previously required 100 people.
Now because of AI, you only need 40 people. So we're only going to pay you for 40. So you're seeing huge, huge concessions on deal renewals across the entire space.
So what you're hearing from the IT services providers is this intentional shift towards outcome-based pricing. What they want to do, again, to the gist of your question is change how these contracts are priced so that then the IT services providers can maintain their revenue growth and actually protect or even expand their margins. I think the big question becomes is will there be pushback from their enterprise clients?
The RFP teams, the procurement teams, the legal teams, they were not born yesterday. They want to capture the most value that they possibly can for themselves. So we think there's going to be a decent amount of pushback from services clients to the likes of an Accenture, a Cognizant, a Genpak, the other companies that were at our conference.
But the IT services providers are going to continue to push for this. So I think this is a massive debate within the industry. Will this shift be successful?
We're a little skeptical at this point in time, but it's becoming a bigger and bigger talking point across the industry. And I guess it remains to be seen how successful the services companies will be over time in making that shift. Great.
And just on the FinTech side of your coverage, you talked about the companies trying to move to more value-added services and away from just the simple transaction-based fees. One of those clearly is agentic commerce and payments. What are the themes discussed around that, and is that something over the near to medium term that could impact your companies?
Yeah, no, it's a good question. And you had asked, Brad, in one of your follow-ups, what themes do you think will persist when we're talking this time next year, post the 2027 conference? And maybe I'll answer it in a little bit of a different way.
I was surprised at how little agentic commerce came up in our conversations with corporates and how few questions investors were asking of the management teams that were there, just given how much time we spend with clients talking about the topic. So it was a little bit of an interesting disconnect on how much focus was actually being placed on that. And I think if you had asked me the same question as Brad, I would have said, hey, one of the themes that we think is actually going to be bigger and more magnified when we're talking 12 months from now, we think it's going to be agentic commerce.
I think we're pretty, I don't know if bullish is the right word, but I think we're pretty optimistic that consumers will start using AI to make purchases online in the future. The problem is that we haven't seen a ton of adoption to date. I think a lot of the LLMs, frontier labs are trying to work out how those models are going to work.
I think they're going to rely a ton on the card network. So again, that's Visa and MasterCard to set the rules for how agentic commerce transactions are ultimately consummated, who is responsible for preventing fraud, how are chargebacks handled, all of the things that the card networks set the groundwork for when e-commerce was becoming a big thing or when mobile e-commerce was becoming a big thing. So we think those two names in particular are pretty well positioned to help set the foundation and the structure for agentic commerce going forward.
And I think when we're doing this podcast next year, there's going to be a lot more focus in 2027 than what we heard from investors and corporates in 2026. Okay. Thank you.
Thank you, Nate. And thank you to everyone. That was a great conference.
I enjoyed being there. A little warm for Southern California, but it's okay. Still very good weather as it always is there.
So I thank you all for your time today. And if you'd like any more information on anything discussed here today, please reach out to your Deutsche Bank sales representative. You have been listening to Podsept.
Podsept, the podcast from Deutsche Bank Research. This podcast has been produced by Deutsche Bank and may contain research as defined in Method 2. The information discussed is believed to be reliable and has been obtained from public sources believed to be reliable, although Deutsche Bank makes no representation as to its accuracy or completeness.
Opinions, estimates and projections discussed constitute the current judgment of the speaker at the time of recording. They do not necessarily reflect the opinions of Deutsche Bank and are subject to change without notice. For further important information, please visit research.db.com.
Sources & References
How we cover this story