Lead — The desk sees the emergence of the AI economy as a transformative force akin to historical technological revolutions, framing it as a pivotal driver for asset allocation strategy. Per the full note by UBS, a framework has been developed to help investors navigate the complexities introduced by AI across various sectors. This comes at a time when productivity metrics and GDP contributions from AI are becoming increasingly quantifiable and critical to investment decisions, suggesting that the AI trade could lead to substantial shifts in portfolio dynamics.
What the desk is arguing
The desk posits that we are on the brink of a significant economic shift due to AI, warranting updated investment strategies to leverage this evolving landscape. Jason Draho and Paul Hsiao from UBS highlighted that this transformation parallels investment opportunities historically tied to major revolutions, emphasizing that understanding AI’s impact on productivity and market trends is crucial for informed asset allocation.
Supporting evidence stems from the observation that AI is already altering productivity figures and contributing to GDP in measurable ways, presenting investors with the need to assess if certain sectors are becoming overheated. UBS's analysis encourages scrutiny of metrics that quantify the effect of AI, which could guide traders on whether to favor or avoid specific asset classes in the near future.
Where it sits in our coverage
Our consensus target for the relevant currency pair, USD/EUR, stands at 1.075, with a range of 1.04 to 1.12. Specific firm targets include: - jpmorgan: 1.10 (target for Mar26) - bofa: 1.04 (target for Mar26)
This view aligns with jpmorgan, which sees the potential for upward movements due to AI-driven productivity gains. The desk's outlook is positioned slightly above the lower bound of the consensus range, reflecting a more optimistic view compared to bofa's bearish stance.
How other firms see it
Group-aligned firms such as jpmorgan suggest optimism around the AI economy, advocating for increased exposure to sectors poised to benefit from this shift. Conversely, firms like bofa express caution, indicating a more conservative approach to the unfolding AI narrative.
As we consider this AI-driven shift, pairs like USD/EUR may reflect the broader dynamics, particularly as central banks begin adjusting their policy levers in response to emerging economic conditions influenced by AI advancements. Watch for movements in sectors directly linked to machine learning and automation, as these will be pivotal in assessing the AI economy's overall impact.
01The AI economy represents a significant shift that necessitates fresh investment strategies.
02Measurable impacts of AI on productivity and GDP are becoming critical factors for asset allocation.
03Investors should focus on quantifiable metrics to assess potential overheating in AI-related trades.
04This transformation parallels past technological revolutions, highlighting its extensive implications across markets.
Market implications
Traders should closely monitor the USD/EUR pair as AI’s influence on productivity begins to manifest in economic metrics. An upward move past 1.075 could signal renewed bullish sentiment, while closely following quantitative assessments of AI's economic impact will be vital for position adjustments.
Risks to this view
A reversal in this call could be prompted by unexpectedly poor performance metrics stemming from AI investments or regulatory pushbacks that hinder commercial AI applications. Additionally, if central banks pivot towards tightening in light of inflationary pressures, it may adversely impact the perceived growth trajectory of the AI economy.
ubs
Hi everyone, Dan Cassidy here. Welcome back to Top of the Morning on the UBS Market Moves podcast channel. Joining me today here in studio, glad to welcome back from the UBS Chief Investment Office, Jason Draho, Head of Asset Allocation for the Americas, as well as Paul Hsiao, Senior Asset Allocation Strategist for the Americas, to talk about a recent report from the UBS Chief Investment Office, The AI Economy, A Roadmap, which outlines a framework for how investors can navigate the AI economy.
We will also cover how to assess the economic impacts of AI, along with a look at portfolio construction through an AI lens. So with that, Jason, Paul, it's great to have you both here at the table. Thank you for dropping by and for spending some time today with our listeners and our clients.
Thanks for having us. Great to be here. So excited to talk about the report.
I know this has been a long time coming and many of our listeners, our clients are very interested in this very topic. So let's get right into it, Paul, can you explain for us the origins about this report and how it's different than AI related content that's out there? Absolutely.
So over the last couple of years, I'm sure ourselves and I'm sure listeners have been abreast in a bunch of AI related content. And it's clear to us that we're living in an AI economy. It's a technological development that we view as significant as the industrial or IT revolution, already transforming the US economy, different parts of markets.
And I think it was important for us to generate a report that provides a roadmap for investors, but how really to think about the AI trade. So not only to provide different links to where we see AI affecting different parts of the economy, from the GDP counts to productivity to markets, but also pointing towards quantifiable metrics and things to watch out for. So when we think about, is the AI trade, for example, overheated?
Is it starting to affect certain asset classes? We're putting it this way where we from a high level, we can see how the AI trade affects these different variables, but also putting a quantifiable metric in it. So think about it as a roadmap and that's a subtitle that we're using for investors to look at how AI is affecting different parts of the economy and the sort of connective tissue between each parts of it.
So with that background, Jason, Paul provided, talk to us a bit more about the framework that's provided and AI's influence. Well, you know, Paul touched on, you know, some of the many ways in which AI is impacting the economy from huge amount of investment, major energy demands, it's driving the equity markets that are helping consumption, it is having an impact on the labor market. When you hear all these things, like, well, how do you make sense of like what really sort of matters?
And, you know, we have an existing macroeconomic framework. We try to bring it back to like, you know, ways in which we understand which the market sort of trades off of. And so what we try to do is introduce a framework that can assess the economic impact from AI from three different perspectives.
And the first is to actually evaluate how is actual GDP growth being impacted and you can break that down into the different components. There's consumption is the biggest part of, you know, kind of the economy, investments, net exports, trade, we get a lot of semiconductors from other parts of the world, and then government spending. So that's the one thing.
The second is to look a little more conceptually, like, how does this actually impact the long-term potential growth of the U.S. economy, particularly what does it mean for productivity? Because long term, that's ultimately what kind of drives the economy, it's what drives kind of wealth creation, and AI is a productivity or should be a productivity enhancing tool. And the third kind of, you know, perspective is to think about then the macro regime.
And this is our framework. We define the macro regime by sort of actual growth and inflation. What are they?
You know, are you sort of in a reflationary environment, stagflationary, Goldilocks, because AI is going to drive the economy, therefore, like understanding the regime. And the reason why we have this framework, there's kind of three reasons for this framework. One is, as Paul mentioned, you know, it's kind of a roadmap to try and navigate really rapid change by contextualizing, you know, all these different ways that AI can impact.
So again, like, no one knows the future, but at least having sort of a map to sort of get you in the future helped, you know, to kind of figure out where you're going, like, that's number one. The second is that AI has become what we would define as like a macro factor, and that affects all asset classes. It's not just sort of a line item in a portfolio, but, you know, sort of impacts everything the way you think about, you know, the economy and the financial markets.
And the third is, going back to this macro regime, you know, if we were to say the macro regime is going to be reflationary, what we know asset class performance typically is behaves in a certain way. If it's a different regime, it's a different environment. Given how important AI is, if it helps us inform the macro regime, it helps us inform and make asset allocation decisions.
So we think this framework is consistent with the way we already approach asset allocation, but also tries to contextualize all these different pieces. So investors are not just like randomly sort of like, you know, pointing out things like, you know, we can actually sort of distill it down to something that is consistent with making sort of portfolio allocation decisions. And we'll speak a bit further about that later in the conversation.
Paul, I know in the report, you mentioned an AI flywheel. What is this? And what are some speed bumps that you may foresee?
Certainly. So Jason mentioned how we have a lot of investment with the idea that AI will be a productive force, not only for the overall economy, but for profit margin for firms. And CIO is estimating that in 2030, we think that AI related revenue globally could reach as high as 3.1 trillion.
So to get there, what we think the, what we call the AI flywheel is basically sort of a virtuous cycle for AI investment. First the bucket we're calling productivity, basically what can AI do? We've gone from really, you know, tech space sort of responses back in 2022 with the release of chat GPT to a lot more complex functions that are on the verge of, if not already replacing a lot of task work that companies use on a daily basis across the day.
So how complex can, what sort of things can AI do? That's the productivity part of it. The second part of it is the compute or really token demands.
And we think that this is sort of the currency of AI economy. So as long as an enterprise and consumers keep increasing AI adoption, they will pay for these tokens in order to use them. And more token demand is linked towards some more investment.
So these are the data centers that really are needed to provide the compute power to power all these things. And that will eventually lead to more things that AI could potentially, more complex problems that AI could potentially solve. And that's the back to the productivity part of it.
So that's the virtuous cycle that we're pointing out. Obviously that there are certain, certainly some seed buns ahead. The two that we're outlining in the presentation here, the two big ones, is first of all, energy.
Just the physical constraints on the power grid and the amount of investment. Even when we have all the investment in the world, do we have the zoning, political will, and also just the physical constraints manageable to power all this compute demand. And the second is financing.
So I mentioned that AI will require a huge amount of investment from here on out. I think we're estimating around $5 trillion from now to, or 2025 to about 2030. That's a lot of investment.
And in order to get that financing, whether it's through equity markets or debt markets, that's a lot that has already been priced in, but potentially more if you have AI adoption rates exceed forecasts right now. So that's the flywheel and those are potential constraints that we're looking out for. Jason, on the macro front, there does exist a concern out there as to how AI perhaps will impact the labor market over time.
What are you seeing on that front today? Well, that is probably the thing that's most stressing people out. The fact that AI could disrupt and take away a lot of jobs, certainly for younger people.
We're still very much early in this process. What we can see in the data thus far is that AI has had only a sort of a really modest net effect on employment. If we think about it, a monthly jobs per month that are being created, maybe in the range of 10,000 per month this year.
Keep in mind, any given month, there are roughly ballpark, like three and a half million jobs created and maybe 3.4 million jobs lost. So this is a pretty small, close to our rounding there thus far. That doesn't mean that we won't stay there, but that's where we are right now.
It's likely to get higher. Now, in order to really kind of think about it, sort of monitor it, sort of have the roadmap, you have to think like, well, how would the labor market be impacted? And in the report, we kind of go through some of the different sequences and steps.
First is like what jobs, but not only jobs, but what tasks are sort of exposed to AI. Because there are certain things that we all do that could probably be automated away. Other parts, hopefully recording podcasts, cannot be automated away.
Then the question is, well, even if the task can be as exposed to AI, will a company adopt it? And there has to be a cost-benefit analysis, and this kind of goes back to Paul's point about the flywheel, that it can be productive enhancing, there could be a demand for it, but it depends what the cost. Does it actually make economic sense to automate every single task you can?
There's also, from this adoption, it can automate, but it also can augment. So if it allows us to do data analysis quicker, put reports together faster, it actually can make a lot of workers more productive. I think the doom and gloom is focused on suddenly all these jobs are going to be automated away, when the reality suggests a lot of tasks, a lot of jobs can actually be augmented over time.
Just at this point in time, some high-level statistics that we can see is that maybe about 20% of jobs have a high risk of automation. That doesn't, again, mean that they will, and it means maybe more certain tasks, as opposed to where the automation would take place. Adoption is happening pretty quickly.
If you look at corporate rates of our using AI, I think more than 50% of S&P 500 companies would say they are, but again, it could be still a small tool, it's not completely disruptive of what they're doing, and it's only a few tasks that are being impacted. Now it's also true that AI adoption is happening faster at larger companies, and in particular in the high-tech sector. So if you talk to people in San Francisco, they'd probably say every job is going to go away.
If you talk to people in other parts of the country, they'd say, yeah, we're sort of a toying with it, but it's not nearly as impactful. It does also seem to be maybe having a little bit more impact on younger workers, a lot of concern about college graduates, are they finding employment or not? So there's probably some marginal impact there, but let's keep in mind that, let's say the unemployment rate for recent college graduates, if it's 6% now, historically it might be 4.5%.
And there's other factors at play, it's not just AI that's disrupting it. So again, it's sort of the margin that's mattering, but not as a wholesale game changer, certainly not yet, and it's something that we'll be monitoring on a regular basis going forward. Well, Jason, some very helpful clarity on a factor that's top of mind for many, including us podcasters out there, so thank you for that.
So before we wrap up, I do want to revisit the point on how AI is impacting asset allocation. Jason, to hear your thoughts, and Paul, we'll close it out with you. So ultimately, we have to make investment decisions.
So this is really interesting to discuss the economy, and certainly for a lot of our clients, if they're running a business, helping them. But our job ultimately is to make asset allocation decisions. I think ultimately, at a high level, constructing portfolios in what we think is now this AI economy, it's going to require assessing how each asset class is affected, what the markets are pricing for different AI outcomes, their exposure to different AI risks.
And then ultimately, the best way to express your views on AI, it's not just through stocks, it can be across the full spectrum of asset allocation. Yeah, just to add to that, so at the end of our report, we have a couple of slides detailing how we're thinking about how AI might affect different asset classes. So right now, it's more just asking the right questions than necessarily providing point forecasts or the answers.
So for example, for equities, we know that we've had chip makers, hyperscalers have a big rally. The next potential leg for the AI trade might be the adopters. We've outlined a couple of sectors that we think have interest in adopting AI, can help margin expansion.
When it comes to rates, the questions we're asking are rates likely to go higher or lower because of secular inflation or secular disinflation, how much that's priced in by treasuries right now. For credits, are corporate bond spreads accurately pricing the credit risk to financing the AI CapEx? And that might change just given how geopolitics is playing out.
And the commodities, will the demand for scarce inputs needed for AI cause prices to go higher? And commodities have done well in recent memory so far. And then finally, I think this is also related to the geopolitical implications of AI, but in foreign exchange, will US leadership in AI perpetuate US exceptionalism and USD strength just because of capital inflows?
And we do know that the US right now is the global leader by a large margin when it comes to AI investment so far. So is the dollar linked to that? Well, Paul, Jason, from hearing this today, obviously a lot of useful, impactful use cases for artificial intelligence.
The conversation will continue, of course, we'll encourage our listeners, our clients to read further into the report, which is available for you now up on UBS.com slash CIO. For clients of UBS, please contact your financial advisor to receive a copy of the AI report directly. Thank you so much for joining us today.
We've been joined by Jason Draho and Paul Hsiao from the Asset Allocation Team here at the UBS Chief Investment Office. Paul, Jason, look forward to continuing the AI conversation with you both. Thank you very much.
Thank you. Thank you for tuning in. Be sure to visit UBS.com slash studios to view the entire UBS studios suite of podcast channels along with our video offerings, such as UBS Trending.
You can also follow us on Instagram for content highlights at UBS Trending. UBS Studios is part of the UBS Chief Investment Office within UBS Global Wealth Management. Visit UBS.com slash CIO to view the latest research.
UBS Chief Investment Office's investment views are prepared and published by the Global Wealth Management Business of UBS AG or its affiliate UBS. This material has no regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is published for informational purposes only. As a firm providing wealth management services to clients globally, UBS AG and its subsidiaries offer both investment advisory services and brokerage services.
Investment advisory services and brokerage services are separate and distinct, differ in material ways and are governed by different laws and separate arrangements. In the USA, UBS Financial Services, Inc. is a subsidiary of UBS AG and a member of FINRA SIPC. For information, please visit our website at UBS.com forward slash working with us.
For a full legal disclaimer applicable to the independent investment views produced by UBS, please visit our website at UBS.com forward slash CIO dash disclaimer.