2025 Predictions
“There is one sin which I have come to fear above all others… certainty. If there was only certainty, and no doubt, there would be no mystery. And therefore, no need for faith.” That’s the card played by Cardinal Lawrence in the movie “Conclave” to line up a very divided Church behind a shared and basic principle: doubt is good.
Every year, when I sit down and stare at this screen and start to think about the yearly predictions, I always start from that basic principle, doubt is good, and add a set of constraints that help me make the whole process more functional:
It should not be ambiguous or fluffing around.
It should not be about work.
There is no need to ride any hype.
While constrain number one is somehow easy to implement, numbers two and three are always something I need to juggle with, and this year I struggled a lot with number three! Anyhow, let’s start.
Ukraine and Russia will start a peace resolution process to end the conflict
OpenAI won’t go public but Mistral AI will be acquired
Big Oil consolidation: two big oil companies will start working on a merge
The return to office saga will continue to hit the headlines
Renewables stocks will struggle. The iShares Global Clean Energy UCITS ETF USD will fall to deep lows
Bonus
Elon Musk will be fired by Donald Trump
The ban on social media for minors will reach some European capitals
No single SMR (Small Modular Reactor) will start construction in US nor in Europe in 2025
1 - Ukraine and Russia will start a peace resolution process
It won’t happen overnight as Trump claimed before getting elected, but I do think this is the year when Putin and Zelensky will sit at the table. The war has now become very expensive for everybody, and the latest loss of revenue from the US sanctions on oil and gas and the termination of the agreement to export gas to Europe is going to bite into the Russian coffers.
Trump will definitely be a major actor in this, and he has said a few times he does not think Russia has been treated fairly. Whether you agree or not it does not matter. Geopolitics is a game that men like to play regardless of whether there’s a valid reason to do this or that. An example of this: Afghanistan. The US spent billions of dollars and multiple lives over the course of 20 years just to give the country back to the same incumbents that were there before. We can tell all the stories you want around that and believe in all the narratives that you can find online. The harsh-raw reality that will end up in history books can be summarised in a small paragraph that goes along the lines “the US fought a war to control Afghanistan and after 20 years they decided it was not actually worth it and gave country back to the same rulers they initially fought.”
But I am diverging here. Back to Ukraine. My guess is that Trump wants the war to be over. Yes he is aggressive, highly unpredictable and hard to trust, but he’s genuinely against people being killed. Like Berlusconi, he’s a businessman and he’s less interested in playing with soldiers and tanks like politicians do. Will he accept any deal? Of course not. It’s going to be complicated, but I think he’s ready to put an end to the conflict and bring US companies in to rebuild the country. As for everything, timing is the key factor.
You could argue that he will try to bend Europe first, that entails waiting to get a sense of how the political landscape might change after the German election and why not, also wait to check how much power Macron has still left if France public debt spirals out of control and the country gets downgraded by rating agencies. Besides, there’s also a Turkish question that will probably need attention, given the role Erdogan recently played in Syria and the changing nature of Turkey’s relationship with Russia. Leaving intentionally Israel and Iran aside.
2025 is yet another super interesting year where we all can put our geopolitical hat one and say whatever we want.
2 - Open AI will not go public nor Anthropic. But Mistral will be bought.
Yes, I know, was I not supposed to refrain from going into hyped territories? After two years from the launch of ChatGPT, it’s become very difficult not to speculate about the future of AI. Yet I won’t go into any technical aspects and want to just reflect on the future of the front-running company in the AI space: OpenAI. December 24 marked the second anniversary of the launch of ChatGPT, and it feels like a decade ago. After two years and two more versions of ChatGPT, and many other gadgets, the company recently went through 10 rounds of funding, raising almost $ 18B. And while it still dominates the AI space in terms of revenue, the bill keeps going up. Training new models is getting more and more expensive, and no company has yet figured out how to make money out of it. It’s late 1995 again and Open AI is seen as the new Amazon.
Yet they also can run out of money, especially if they want to develop something new, which requires some hardware component. As much as Microsoft seems to be willing to further invest into OpenAI, and it’s able to do that, the current market valuation might be quite attractive for other investors, especially if they think the value of the company is supposed to skyrocket after IPO. Will that be enough to force the AI champion to get listed. Not so quick. I doubt that will happen next year. As much as OpenAI is literally burning cash, the advantage they have, both technical and brand-wise, compared to the rest of the gang is still massive and enough not to pressure Sam Altman to go fund money on the Nasdaq.
The same goes for Anthropic, which recently closed another round of funding and has a deep-pocket investor like Amazon behind it. Furthermore, both companies need their investors to run their models. The synergy is too good for all.
And what about Nvidia? Nvidia is the cash king of the moment and it’s showing more and more interest in integrating down the value chain and not remain just an infrastructure player. An acquisition might speed up that. And who is the other player in the market still relatively small that could benefit from being integrated into Nvidia ecosystem? My bet is on Mistral. A French small company that did get some attention from Microsoft at the beginning of 2024 as a way to diversify from their big bet on OpenAI but that amounts raised is a fraction of what OpenAI and Anthropic have raised to date. And Mistral decision to move to Palo Alto might be read as a sign the company will try to appease some company in the US. That’s why I think that by the end of 2025, a major deal will be announced, and Mistral is at the top of my list.
3 - Oil industry consolidation: 2 major oil companies will start looking at a possible merge
After last year’s failed prediction, and the new info acquired, I am willing to go against the main principle my first boss once told me: never bet against oil. Global oil production is set to expand even further in 2025 and while demand is expected to keep growing, the market should remain well oversupplied. And this even without considering a change into OPEC’s strategy. Yes, Israel might ruin this by launching an attack against Iran, with oil facilities being a possible target. Furthermore, the US has just sanctioned more Russian oil tankers and while it probably was the last move the Biden administration did, it’s very unlikely Trump is going to reverse that. And Trump is also keen on getting the dollar back to lower value, which is generally bullish for oil, so are you really sure you want to be that bearish Matteo? If I sum up all the bullish and bearish factors, the global oil outlook does not look bullish at all. What does that mean? It means lower oil prices and a renewed fight for market share, which will inevitably penalise oil companies that have a larger exposure to new investments.
And I want to be very bold. My main suspect here is BP. The British company is the one that is currently underperforming the rest of the industry, and it’s probably the main company that could benefit from stronger ties with another major. And on the other end of the line, I see two main companies: Chevron and ADNOC.
I know it’s a bold bet. So many things need to line up to make this something concrete that it’s highly likely I get this wrong. Yet I think there’s going to be increasing pressure on BP in a bearish environment. The fact that BP has a brand-new management, a management that has been with the company for decades, does not suggest this is something viable from a professional perspective (you don’t become the CEO of a company you’ve worked for years and then give up the role for a greater good). So it might not happen this year but I think the management will be forced to look at this option, and the new Trump administration might actually see a deal like that with a US oil company involved something they would likely back and help navigate usual regulatory hurdles and antitrust concerns.
4 - The return to the office saga will continue to hit the headlines
In a month time, it’s going to be 5 years since the region of the world where I had just moved to started doing something defined at that time as “surreal” which was just the cue of what was about the change the whole world. 5 years in which we went through the big resignation, digital nomads becoming the new trend and offices and workplaces inevitably going through all sorts of changes. After 5 years we can see few trends consolidating: on one side, big companies have all started to re-introduce some form of office presence and more and more companies, over the course of 2024, asked their employees to fully go back to the office, which now goes under the acronym RTO (Return to Office). Other companies have fully embraced remote working and countries have started to compete to attract the new nomads. Is 2025 the year when we will all go back to the office? I doubt.
Shocks like Covid happen at once but then to go back to normality takes much longer. And any shock changes the old definition of normality. Cos that bell curve has shifted, tilted, twirled and flipped. And when it’s about the workplace, that bell curve has definitely changed. My guess is that big companies will try to reintroduce as much as possible some form of return-to-office policy even in 2025, while the rest of the market will still remain in a hybrid or remote mode. And the issue won’t be settled for few simple reasons:
Hybridisation, or flexibility, is a new competitive advantage that small and more innovative companies will keep playing to attract talent away from big companies that can pay higher salaries. The more big companies force employees back in, the more small agile companies will benefit from it. And this is not something that finds a balance over a couple of years.
RTO is effectively a measure that helps companies tackle efficiency gains indirectly: it's way better to deal with quitting and resignation waves than to lay off people. And in the era of AI, higher operating costs, expected productivity gains, and higher margins, big corporations will likely opt to tackle costs in multiple ways.
Demographics: I learnt last year that you need to play the demographics card in any argument since fertility rates across the globe are declining, the water is contaminated and penises are shrinking. How does that influence the RTO wave I have no idea but it’s definitely going to have a big impact. You just need to wait for it.
So what’s the prediction? Expect more companies to mandate a return to office policy in 2025 even if they don’t have enough space for everyone. But also expect more companies to keep leveraging hybrid workplace policies to attract talents at a cheaper salary. We live in a world of digital twins, high connection but low connectedness, where a person spends on average 5-to-6 hours a day staring at her/his phone and where you honestly wonder whether those 5-to-6 hours are the same whether you are commuting to an office or not. But I am digressing.
5 - Renewables stocks will struggle -> The iShares Global Clean Energy UCITS ETF USD will fall to deep lows
2024 market yet another record year for renewables globally with new capacity additions that the IEA forecasts to reach 666 GW, 30% up compared to the record year registered in 2023. A trend set to continue: it’s a one-way road only. Renewable costs, solar in particular, keep falling down and in most regions, they are cheaper than fossil alternatives for generating electricity. Last year, at the COP, countries have agreed to triple the renewable capacity installed globally by 2030 and China solar overcapacity is not going away anytime soon.
Yet, chickens are coming home to roost. It’s becoming more and more challenging to invest in renewables and in the western world, but not only, we are fully experiencing the harsh reality that transitioning from one paradigm to another is not a stroll over a park, but rather a walk on the wild side. The number of hours with negative prices in Europe, as in the US, has skyrocketed and it will continue to go higher. Oil turning bearish is definitely not good news for renewables. Incentive schemes are getting smaller and smaller and the energy crisis has resurfaced nuclear power as a new contender for public money and nuclear needs A LOT of money, up to the point that the Commission had to open an investigation on the new Polish nuclear power plant in Lubiatowo-Kopalino which is estimated to cost an astronomical €45 billion (something between 140 and 170 EUR/MWh even without considering costs overruns). Direct consequence of that: cash flows for new renewable plants have become increasingly uncertain and there’s a general fear that your next neighbour might ruin your party once he installs his own solar or wind farm nearby. The grid takes a while to develop and while batteries are strongly emerging as a way to tackle all of the above, we are not there yet. Please don’t ask “what about green hydrogen”?
Don’t get me wrong, the underlying growth of renewables will continue also in 2025, but my guess is that this year does not look like a good year to invest in renewables. Lots of renewable companies across the whole value chain will struggle, and the disillusion around ESG will probably start to ramp up across the whole finance world (we have already had a touch of it turning its back to Net-Zero and all that jazz). All renewables’ ETFs may continue to struggle this year, rock-bottoming to new lows.
Bonus predictions
Elon Musk will be fired by Donald Trump
It will not be an episode of the Apprentice but the collaboration starts on very shaky foundations: two very big egos, one fictional Department and a shared interest to keep it more like a PR campaign than anything else.
Now, Taking Elon’s latest political stances aside, I genuinely wonder how much he decided to get involved for purely speculative reasons as many suggest, and how much is yet another dopamine hit he craves, to prove everyone he’s able to reboot the public system as any other company (see Twitter). As much as it’s never for one reason, I think he went beyond just fame and he’s playing all sorts of games to stress-test the whole system and check what he can get away with.
But as the say goes, it’s all fun and games till someone loses it. Trump does not strike as a person that tolerates being overshadowed by somebody else, and the fact that he co-appointed Elon and Vivek for the same position is just another indication he wants to keep an eye on him. Elon, you are fired!
No single SMR will start construction in US nor in Europe in 2025
Everyone has her/his own pet that likes to play with, and I discovered in 2024 that I can’t help myself when the conversation goes into the future-is-nuclear territory. I just can’t. Too entertaining. My position? All nuclear discussions happening across all sides of the Western world are just a weapon of mass distraction. Very simple! :D However, as in everything, I am very open to be wrong here.
The new shiny gadget the nuclear industry is trying to sell goes under the name of SMR, Small Modular Reactors. The scary word “nuclear” has been dropped and the needed technology rebranding is ongoing. Yet the road ahead is still quite impervious. There’s not a single SMR currently in operation in the Western hemisphere, and the existing Chinese one is operating less than 10% of the time. It’s still a lab solution, and as much as the industry is trying to gain the needed attention to redirect public and private money away from renewables, I do struggle to see any commercial project starting construction in the next 12 months, if not years! And I do not believe for a second that the increasing demand from data centres will jumpstart any SMR construction this year.
Tighter control on social media for minors is coming and I am not talking about the US TikTok ban
This is both an easy and difficult one. To a certain extent, Europe is already tackling the issue, and the Commission is developing guidelines under the Digital Services Act to ensure privacy, safety, and security for minors on online platforms, which are expected to be adopted in 2025. So it’s happening, what’s the big deal? The big deal is how we are going to do it. Australia, which recently passed a law setting the minimum age of 16 for access to social media, puts the burden on social media platforms to implement "reasonable age-gating practices" to enforce the ban. Companies have one year to work on that.
In Europe, a few countries have already introduced measures like age restrictions or parental consent requirements but both their implementation and enforcement have proven to be quite challenging. And Europe won’t let companies self-regulate, especially social media companies that are ready to cross the Rubicon of facts and reality anytime they see it convenient. I suspect the time for online identity requirements is coming and the opportunity offered by the entry into force of the European Digital Identity Framework regulation, might create the right opportunity to tackle this. As for smoking and alcohol, here’s a widespread consensus that minors should not be on social media and something more concrete needs to happen. And if you haven’t read it yet, this book is worth some time away from social media.
A first rule of thumb
Let’s test the predictive power of the 3 main LLMs: ChatGPT, Claude and Gemini.
According to the new oracles, I am likely to get 2 out of the 5 predictions correct. I leave aside any consideration about the variety of responses I get from the 3 models and just look forward to seeing in which form these stories will morph, change, and develop over the course of the next 12 months. It’s going to be a very interesting year.
The world is moving fast, and while AI will increasingly provide us with answers for everything, it is our doubts that keep us curious, humble and connected to the profound mysteries that shape our world.
2025, Let’s roll.