Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
The Macro Brief – Global tech trends
619 Plays2 days ago
From the latest innovations to the investment strategies of big tech giants, Nicolas Cote-Colisson, Head of Global Tech Platforms Research, outlines what to watch out for in the tech sector this year.
Click here for appropriate Disclosures, including analyst certifications, and Disclaimers that must be viewed with this podcast:https://www.research.hsbc.com/R/101/CzkHXhS
Stay connected and access free to view reports and videos from HSBC Global Investment Research follow us on LinkedIn https://www.linkedin.com/feed/hashtag/hsbcresearch/ or click here: https://www.gbm.hsbc.com/insights/global-research.
Recommended
Transcript
2026 Tech Sector Dominance & AI Investment
00:00:03
Speaker
Hello, I'm P.S. Butler and welcome to The Macrobrief, where we look at the issues driving financial markets. And in this episode, we're focusing on one of the sectors set to dominate in 2026. No surprise, it's technology.
00:00:16
Speaker
We'll be looking at the key trends to watch from the AI value chain and the latest innovations to the investment strategies of big tech
Interview with Nico Cotcolisson on Tech Trends
00:00:24
Speaker
giants. And to do that, I'm joined from Paris by Nico Cotcolisson, head of Global Tech Platforms Research. Nico, bonne année.
00:00:32
Speaker
Merci. Thank you, Pearson. Hello, everyone. Great to have you ah for this discussion. Very topical. ah So maybe we could start, Nico, by recapping on CapEx budgets and financing needs. 2025 was a year of ever increasing CapEx numbers. And then at some point, people started to say, how is this all going to be financed? And you published some work at the end of last year looking into this. So maybe give us a quick rundown on that.
00:00:58
Speaker
Yes, well, you're right. 2025 was an incredible year for investment. i mean, if you think about capital expenditure for big tech, it almost doubled in 2025.
AI's Impact on CapEx and Financing
00:01:09
Speaker
And we even saw a few tech companies passing the $100 billion dollars CapEx bill threshold last year. Now, AI has been a fantastic catalyst. I mean, all these large language models have required and still require a lot of compute power.
00:01:27
Speaker
And for that, you need data centers, you need chips and many other elements. Hence the big increase in capital expenditure seen in 2025, but also foreseen into 2026. Now, you you mentioned the financing of all these capex and rightly so.
00:01:46
Speaker
What I would say is that unlike the previous tech innovation booms, the large listed tech companies have a lot of cash on their balance sheet. So the recourse to debt is pretty limited. And overall, we expect the funding to come from the current free cash flow generation.
00:02:06
Speaker
Also, I may add that looking at the capital allocation strategies out there, there are still very significant share buyback programs in place. So if investment a attract good returns, then the share buybacks could also be reconsidered in theory.
00:02:25
Speaker
Now, for those New York AI companies developing large language models, it's fair to say that they have very ambitious computing ah power plants and they are not yet fully financed, but they generally have solid stakeholders to get them through.
00:02:43
Speaker
Okay, so taking that into account, let's take a look at some of the risks and opportunities for 2026.
AI Demand & Infrastructure Bottlenecks
00:02:51
Speaker
Firstly, bottlenecks and capacity constraints. Is that likely to get in the way of continued AI momentum? By that I mean, you've sort of answered the question that there is the cash is there, but can it actually be spent yeah at the rate that people anticipate, given perhaps some of the constraints on availability of chips and the time it takes to build data centers?
00:03:12
Speaker
Of course, allow me to start with just one comment. Regarding ai we think that we are at the very beginning of a mega cycle. I mean, AI will penetrate every production process with the aim to boost productivity gain. So the appetite for investing in AI infrastructure, it will remain high, but so high that bottlenecks are appearing. Now, if you listen to the largest cloud providers, they are all mentioning cloud capacity constraints. You have rising demand.
00:03:44
Speaker
It's not going to stop. And we do not expect this narrative to change in 2026. Actually, one very large cloud provider told us client wants infinite capacity. I quote them. An important bottleneck is energy.
00:04:01
Speaker
You may have noticed that compute power is now measured in gigawatts. It's no longer in how many bytes you may proceed. so This tells you a lot about how energy hungry are all these processes.
00:04:15
Speaker
But about that, think about the lead time to get a new gas turbine. I mean, it takes about four years or more, and we are not expecting small nuclear reactors to come to the rescue before 2035. So you understand that energy is a very important bottleneck.
00:04:33
Speaker
Drilling down on ah on this energy constraint, is that going to be a factor a big factor in 2026 or is there room enough for the growth in 2026? There will be growth in 2026, but will all these companies be able to answer the big demand? No.
00:04:50
Speaker
Capacity constraint will remain in 2026. A bit too early to say for 2027, but look, high demand, constraint capacity, it's not a bad business model, is it?
00:05:01
Speaker
Very true. Okay, so for the next question, we're going to need some jargon
Chips in AI: ASIC vs GPU Metaphor
00:05:06
Speaker
busting. What do you mean in your recent report when you say ASIC chips are in vogue, but there's still room for GPUs?
00:05:15
Speaker
I accept it can be a bit triptych. So allow me to use a kind of a metaphor to compare these two types of chips. So on one side, you have the ASIC chips, stands for application-specific integrated circuit, And the other one is GPU for graphical processing units. So this is the hard part the ah the hard part of the explanation. Now, think about the GPU as worker who is ah kind of a jack of all trades.
00:05:47
Speaker
no They can do a little bit of everything and anything, but not necessarily a master of one specific field. So you can give them a general problem.
00:05:58
Speaker
And they have a wide range of general skills to solve it. So for example, take a GP for general health issues. Now on the other end, think about ASIC as a highly skilled specialist in one domain.
00:06:15
Speaker
So they can do their particular specialty very well and fast. But if we are going to use a medical metaphor again, imagine you have back issues. You must be better served to if you seek out an orthopedic specialist rather than a GP.
00:06:33
Speaker
So if we bring this metaphor into the AI world, The GPUs are jackable trades and they can do a bit of AI training, a bit of inference and other types of computing like a GP doctor.
00:06:47
Speaker
But more and more companies are now using ASIC chips to do very specific AI tasks like inference, which is a part of the AI pipeline, which gives you the answer to your specific questions super fast.
00:07:01
Speaker
Very clear. I think I understand it now. Okay, but now, ah to be clear, ASIC versus GPU. We believe that no matter how many specialists, the ASIC, you may add to your AI data centers for inference-level computation, there is always to be a lot of demand for generalists, the jack-of-all-trades GPUs, as they offer more fungible computation from a cloud computation andrast infrastructure perspective.
00:07:28
Speaker
Okay, so you you and I have to admit have been around long enough to remember previous technology booms. And ah in any boom, the next phase is
Market Consolidation and Specialists
00:07:40
Speaker
rationalization. So how do you expect this to play out both in terms of market participants and monetization?
00:07:46
Speaker
Yes. Oh, look, as we just discussed, capex are huge. So this become a high sunk costs economic model. So those who can monetize will remain, others won't. And there won't be much of a middle ground there. But for the strongest players with the best infrastructure and models, it could be very transformative.
00:08:08
Speaker
And you think that there is, ah just reading your report, that there'll be a few major players, but there will be room for some of the smaller specialists? Yes, you're right. So on the long tail, we will see many specialists, but on the front end, it's going to look like more an oligopoly market than anything else because of this sunk cost characteristic.
00:08:28
Speaker
So I was also interested to read in your report about smart
AI's Evolution and Smart Glasses Revival
00:08:32
Speaker
glasses. I thought that that that had been something that had been tried and left to one side. But you talk about the evolution of AI into consumer devices. And do you think that AI will help smart glasses stage a comeback?
00:08:44
Speaker
Definitely. It's a very different world from the one we had 12 years ago when the Google Glasses first appear. AI is this time opening a world of opportunities.
00:08:55
Speaker
Think about AI. It can turn a pair of glasses into a very powerful form factor. but Just think about the real-time voice conversation with AI, the in-frame camera for contextual information. i mean, these features open potential for a right range of service. Now, think about the smartphone then. i mean, the smartphone has been around for almost 25 years, first offer offering voice and then digital services, but smartphones, they now face a risk, if not to be replaced immediately, at least to be displaced.
00:09:32
Speaker
And then in addition to the smart glasses, we understand that some AI companies are also considering new hardware to interact with AI, So expect the coming years to bring many surprises.
00:09:45
Speaker
I look forward to seeing you with it wearing a pair of smart glasses, Nico. Final question.
AI's Dual Impact on Productivity and Jobs
00:09:50
Speaker
Just today, there's a headline in the press about the mayor of London, Sadiq Khan, warning that London could be dramatically impacted by job losses resulting from ai productivity advances. It's not a new debate. Others have have warned about this. And yet elsewhere,
00:10:05
Speaker
it's been noted that in professions such as the legal profession, the competition for junior talent remains red hot. So is AI a revolution for the job market or is it an evolution?
00:10:17
Speaker
I see. Well, technically, AI is an evolution because It has not been invented last week or last year. It's actually the result of decades of progress in mathematics, increased computing power and the massive data sets that are provided by the internet. So it's a continuation of the digital automation that began in the 20th century. Now, socially and economically, it's a revolution, essentially because of its speed and the scale.
00:10:49
Speaker
Now, we talked about AI being a mega cycle earlier. We think that the gross productivity gains can be massive and academic research is already measuring it.
00:11:01
Speaker
But this may come with externalities, including the issue of workforce displacement. So striking the right balance between productivity gains on one side and job market on the other side is definitely a challenge. But history shows that technology usually creates more work than it destroys.
00:11:21
Speaker
And if you allow me ah to give you the bottom line and allow me also to quote a top exec at a very large tech company, AI will likely not take your job, but a person using AI might.
00:11:35
Speaker
Very good point to end on, Nico. That's been fascinating. a quick tour d'horizon, as we would say in French, on the AI world. and We look forward to having you back as no doubt there will be some more significant developments this
HSBC Insights and Resources
00:11:48
Speaker
year. But for now, thank you very much for joining us.
00:11:50
Speaker
Merci. Thanks for having me.
00:11:57
Speaker
Nico Cocolisson there on the big tech trends to watch this year. Now, it's only mid-January, but it's already been an eventful few weeks for global markets. If you're an HSBC client and would like to stay up to date on all our latest research, then please download our app.
00:12:12
Speaker
Just head to Apple's App Store or Google Play and search HSBC Research. And don't forget to check out our sister podcast, Under the Banyan Tree, where hosts Fred Newman and Harold van der Linde put Asia's markets and economics in context.
00:12:27
Speaker
This week, they're talking all things China with our chief economist, Jing Liu. You can listen wherever you get your podcast. And if you've got any questions or comments, then you can get in touch with us at askresearch at hsbc.com.
00:12:42
Speaker
But that's it from us for now. This week's podcast was hosted by me, Piers Butler, and produced by Tom Barton. Don't forget to like and subscribe to The Macrobrief wherever you get your podcasts. So until next week, thanks very much for listening.