Subscribe
Share
Share
Embed
This month, Tony and Giles consider President Trump’s tariff announcement out of the US, and discuss the fallout from so called ‘Liberation Day’ (01:37). Our experts go beyond the headlines and consider the market reaction to this geopolitical news both globally and here at home (05:45).
Valeria and Tony discuss recent trends in open-source generative AI and what the arrival of DeepSeek on the scene might mean for the industry. We consider how the adoption and acceleration of this technology may affect markets and geopolitics in the short and long term (14:10).
This podcast is brought to you by TrinityBridge – trinitybridge.com – and features Tony Whincup, Head of Investment Specialists, in conversation with Giles Parkinson, Head of Equities and special guest, Deputy Head of Equity Research, Valeria Moore.
We’ve created this podcast to set out possible approaches. Please do not view it as financial advice, or its content as investment recommendations. Just because an investment or an investment strategy has performed well in the past, does not mean it will continue to do so. Our predictions are based on information that is currently available, however events and markets can and do change rapidly.
Creators and Guests
Host
What is Insight Talks?
Our in-house experts share our views on the current market conditions facing investors. Brought to you by TrinityBridge.
We've created this podcast to set out possible approaches. Please do not view it as financial advice or its content as investment recommendations. Just because an investment or an investment strategy has performed well in the past does not mean that it will continue to do so. Our predictions are based on information that is currently available. However, events and markets can and do change rapidly.
Tony Whincup:Hello, and welcome to Insight Talks with me, Tony Wincup. Standing in today for my able, but I'm sure you've guessed very absent colleague, James Tullock, who's off somewhere in deepest, darkest France having a holiday. So good for him, good for me that I'm here. Lots to get through in the next twenty minutes or so, and to help me, a few trusted colleagues with me to unpick what's going on in the market. Valeria Valeria Moore, good to see you.
Valeria Moore:Hi, thanks for having me here.
Tony Whincup:Pleasure. Good to have you in the studio. Deputy Head of Equity Research here. And Giles, Giles Parkinson, good to see you. Hey, Tony.
Tony Whincup:As always, a bit of a stool wall. It's a bit of the furniture, I think I'd say for these podcasts. A great contributor. So great to have you alongside us as well. So in a few minutes' time, Valeri, you and I are going to be talking about, you know, what all of the kerfuffle was about when we learned that developers in China had produced a new generative AI tool that's called DeepSeek.
Tony Whincup:Reportedly, was really good. We're going to dig into that. We're going to ask what it means for the adoption of AI going forward. We're going to have to have a look at the markets because it's been a bit crackers, I think, for the last few weeks. But, Giles, let's start with a t word.
Tony Whincup:Tariffs, Trump, tremendous, troublesome. You've got it. I mean, there's a lot of t words I guess we could use. I don't think it's hyperbole when I've jotted this down just to repeat it. You know, Lenin famously said there are decades when nothing happens, and then there are weeks when decades happen.
Giles Parkinson:I mean, wow. What did we get last week with liberate Liberation Day? Just remind us. Yes. Liberation Day or as we might come on to, liquidation day.
Giles Parkinson:So, look, let's just set the scene here, Tony, and remember where we come from. So tariffs, they're an import tax that a country levies on goods coming in. And The US, before Trump came in for the first term in office, they had a weighted tariff weight of around 1% on all their imports. Now after the time that Trump left, that number had risen from 1% to 3%, mostly through targeted actions against China. Now, of course, as we've talked about in this podcast before, Trump made no secret on the campaign trail that tariffs were a key policy platform for him.
Giles Parkinson:And after winning the election, after being sworn in earlier this year, earlier in April, we got Liberation Day, which was the main unveiling. So there'd been a certain amount of on again, off again, Tony, in terms of individual sectors of the economy or individual countries, but this was really the across the board main unveil of the tariff policy. And what do we get? Effectively, a minimum 10% rate across the board, and then slightly higher rates for some countries. But then if you add it all up, and some of these numbers actually stack on top of each other, Tony, based on what have been mentioned before Mhmm.
Giles Parkinson:That, we believe, gets you in something like, on a weighted average basis, something in the low twenties as a percentage tariff import tax on US goods coming in. So that number's gone from two and a half three three percent to around 22 and a half percent. Yeah. Different different order of magnitude entirely, isn't it? Sort of dwarfs what we've had before.
Giles Parkinson:I mean, I
Tony Whincup:don't know about you, Giles, but watching Trump there in the Rose Garden with that kind of game show scorecard, I just didn't get it. I mean, from an economist point of view, you're a trained economist. Does it make sense
Giles Parkinson:to you what he's to do? Yeah. So with that trained economist hat, the way that these numbers were calculated, what they did was they looked at the bilateral trade deficits and divided that number by the total imports or exports from that country imports into The US and then halved it. Now, I'm gonna borrow this from someone else, it wasn't my own pun, but that's almost like going to a doctor, Tony, and asking to be diagnosed based on your height divided by your birthday. I mean, in that sense, from a raw
Tony Whincup:from a raw
Giles Parkinson:economics point of view, doesn't make a lot of sense. Yeah. Now, look, we might have a slightly hard edged conversation around, look, what is the administration trying to do with these tariffs? Reshoring of jobs, raising tax revenue Serious stuff. Right?
Giles Parkinson:Yeah. Yeah. Yes. Quite. But I think one of the things which has maybe spooked markets a little bit is just the execution of a policy which probably wasn't great in and of itself.
Giles Parkinson:I mean, just the last sense is, in terms of that, I'm seeing in the Rose Garden and the list of countries actually, Tony, the problem is, it wasn't even countries. Some of them were just economic entities. So, example, Gibraltar Yep. Part of The UK, Reunion, part of France. And so the slightly alarming thing, Tony, is that it seems as if some of these some of these groups were picked up just because they've got their own Internet domain rather than them being standalone trading So, we've the market is, I think, putting a question mark not over so much the principle of tariffs.
Giles Parkinson:I think Trump had been very transparent about that these were going to come more around the expectation, that 22 and a half percent number. I think investors were expecting maybe something in the 10 to 15 range, but also the slightly chaotic and haphazard fact fashion in which it's actually landed.
Tony Whincup:So, Giles, this is serious stuff. I mean, just in terms of the markets, we've just had the fifth worst two day period on the S and P 500 since World War two. I mean, down 10 and a half percent. I mean, they're big, big numbers. We're recording this just before The US has opened today.
Tony Whincup:We've had the weekend. I think China was closed last Friday after the announcements. Just bring us up to speed on where we're at now.
Giles Parkinson:Yes. As you mentioned, the S and P is down meaningfully. Futures are the points I came into this room to talk to you now. They were also down again forecasting for this afternoon. But it's not just The US market.
Giles Parkinson:This is very much a global phenomenon. So, yes, those Asian markets, those were closed last week, have opened down meaningfully come Monday, and also European indices as well. So this is very much a global event. That's equities. What's going on in bonds and fixed income?
Giles Parkinson:So sovereign bonds are very well bid. Investors are treating those as a flight to safety. I'm sure we might have a discussion in a few minutes time, Tony, about sort of the future path of of interest rates. But in the very short run, investors are looking to sovereign bonds for safety and security. So the price of those bonds has risen, the yields have fallen.
Giles Parkinson:And just lastly, terms of currencies, one of the quite interesting dynamics actually we've had in the last month or so is actually the US dollar has been quite weak even as tariffs have materialized. And traditionally, the country that introduces tariffs actually sees its currency strengthen.
Tony Whincup:Mhmm.
Giles Parkinson:The reverse of that has happened in anticipation of Liberation Day, and it's continued since. Although, again, the US dollar does tend to trade risk off as a stronger currency when investors get scared. And just this morning, we have had some first signs of US dollar strength that I haven't seen for a couple of weeks.
Tony Whincup:Just so much uncertainty, isn't there? I mean, you've mentioned currencies. We don't quite know what's gonna happen there and how they'll move and absorb some of the frictional cost of tariffs. What else don't we know? Or what else, you know, do we know that we, you know, we don't know yet?
Giles Parkinson:So on the company side, microeconomics, we don't know precisely how different companies are going to be able to adapt, both in terms of those country companies that are exporting from countries outside The US, and also US consumers, how are they going to behave? And indeed, if you're a retail if you're a domestic US retailer, to what extent are you gonna try and pass through some of these higher import costs onto your consumers? So lots of question marks there. Of course, we've done an awful lot of analysis on the holdings that we have, and we think we've got good answers. But, again, there's an area of uncertainty there.
Giles Parkinson:The last one I is I think just in terms of overall investors. Are they actually, with a perspective of a few days and cheaper prices, certainly when it comes to equities, going to maybe treat this, Tony, as peak tariff and actually the maximum point of uncertainty? And all else equal, you know, you can either have certainty in high prices or uncertainty in low prices. And over time, the uncertainty will go away. So potentially, the stage is being set for less uncertainty given potentially this is as bad as it gets with respect to tariffs.
Tony Whincup:And just very briefly, it's been really interesting just watching different approaches from the the the all out kind of bombastic, well, you do 20, we'll do 20, we'll take you head on to the more subtle kind of, well, we're in negotiations, or we are publicly telling you we're gonna drop our tariffs if you we'll only drop ours. I think we've had Vietnam, for example, do do that. What are you making of this kind of realpolitik that we're seeing playing out here?
Giles Parkinson:So I touched on just now, Tony, how actually the the lev the amount of tariffs have come in some something in the low twenties. That's certainly way out of investor expectations. But, actually, if you sort of zoom out and look at, you know, where have bond yields been in the last few years, where have equities been for the last few years, actually, you could make a case that there's not been that much of a market reaction, and I appreciate the way that I'm saying that just after the very sharp moves we've had. So as much as one can ever judge these things, Tony, I think there isn't market expectation that, actually, these tariffs are there, but they're not going to be permanent. And maybe, actually, they're walked back by, you know, as much as 50% half over the next few months or so, so in quite a short time frame.
Giles Parkinson:It really is over to other countries. I mean, Trump and the administration, I think, have been very clear signaling that they're open to deals, they're open to negotiation. And, yes, different countries around the world, different economic blocks like European Union, example, have signaled has signaled everything on that spectrum between outright retaliatory tariffs. We've seen that from China towards, you know, maybe in the short run, a more negotiated outcome.
Tony Whincup:There's a few fragments of optimism amongst that given Trump's unpredictability, isn't there? So there's, I mean, interesting times just to watch how all of this plays out. Just a footnote on The UK's position. You and I were talking just a couple of weeks ago about the spring statement and then the autumn budget and how the headroom keeps getting eroded. Well, this is just it's been another dent, hasn't it, in sort of Rachel Reeves' coffers?
Tony Whincup:It's a sticky wicket for her to bat from, I think, coming into the autumn budget.
Giles Parkinson:Yes. Tricky one. So in terms of the moving pieces and Reeves', you know, famed fiscal headroom, so that buffer had been eroded coming into that spring forecast largely as an as a higher interest rates for UK government borrowing costs, but also slightly weaker economic growth. Now she took actions, as we touched on, mostly on the spending side in order to restore that fiscal headroom. But again, there's an interesting setup for next autumn, Tony, which is all else equal, these tariffs which The US has imposed on Britain are gonna be negative for growth.
Giles Parkinson:Now the OBR, actually, in that spring forecast, they did come out with several scenarios They did. Yeah. Around tariffs. And again, there's a lot of uncertainty around, you know, how much this actually filter through. And as I said, at the moment, the stance of the new government is negotiation.
Giles Parkinson:So it's not clear if these tariffs are actually gonna last on Britain at all. But all else equal, and the OBR's analysis is consistent with that, that's negative for growth. Although, actually, given the fall in gilt yields, that, all else equal, is positive for government interest costs. So so there's puts and takes, but, yes, it's a difficult task ahead for the Bank of England. All else equal, they're probably at the margin more likely to cut interest rates just because growth is weaker, but then that fiscal board is very much back in Reeves' quarters to how she's going to handle these trade offs of, on the one hand, a small windfall from lower borrowing costs, on the other hand, a weaker economic growth, less tax revenue coming in.
Tony Whincup:Okay. Just one final question before we get to to Valeri. I've been waiting patiently by my side for for for a chat. Joe, it was nonfarm payrolls last week. I'll admit, I was a bit confused.
Tony Whincup:Now it was a stronger print, I think 228,000 jobs, last posted. I mean, I'm kind of on that, is it good news or bad news? What does the Fed need to see just to, you know, to put a floor under this and start to cut rates, ease credit conditions in the private sector that might offset some of the friction that we are now gonna see from tariffs?
Giles Parkinson:Yeah. So I appreciate tariffs have grabbed, rightfully so, a lot of the headlines in in in the last week, but my concern over the last few months has actually been just high interest rates. The Fed went on hold in December. It's not cut rates so far this year. And all else equal, it wants to sit on its hands because inflation's not convincingly back to 2% targets.
Giles Parkinson:Now this, in my opinion, was causing a pent up growth scare, which equities wouldn't take well. So in that sense, we the market's got two things to contend with at the moment, Tony. It's got tariffs, and it's got this economic slowdown. So back on the Fed, what we needed really was some something to maybe prod the Fed into action. Unfortunately, in terms of non farm payroll report, we didn't get that on Friday, Tony.
Giles Parkinson:It was a strong report. Residential construction, a key sector that we pay a lot of attention to. Added jobs. The rate of unemployment did tick up ever so slightly, but that was more sort of a rounding from 4.1 to 4.2. And also, in the sense, it rose for very good reasons, which is because labor participation stepped up.
Giles Parkinson:So all else equal, no clear starting gun in terms of that payroll report that gives the the Fed an easy excuse to worry about the employment side of their mandate and to get back on the rate cutting wagon in 2025 just yet. But we do think, ultimately, the Fed does do that.
Tony Whincup:Okay. Good good to try and make sense of some some complicated factors there, Giles. Thanks for that. Valeria, good to see you. DeepSeek.
Tony Whincup:Yes. Good to have a little chat about that. It causes, as said already, a bit of a kerfuffle. When was that? Back in February, beginning of March now?
Tony Whincup:New rival technology from China, generative artificial intelligence. I don't I don't know if you don't if you're gonna sort of plea please me or appease me just by going through some of the basics. I'm a bit of a dummy with some of this. So just start by just explaining what artificial intelligence is and then what generative AI is, and then we can look at DeepSea.
Valeria Moore:Yes. Absolutely. Let's let's start from the basics. As you said, artificial intelligence is all about computer systems that perform task typically requiring human intelligence, like understanding a language, making decision. Then we go to general artificial intelligence, which creates a new content like a text, an image, a music, or code.
Valeria Moore:And an example is, for instance, the the ChatGPT that is a chatbot and can engage into a natural language conversation. So generative AI takes inputs such as a text, an image, a coder, and generate new content such as turning a video into a text. And how does it work? They use neutral network, which is computational models in spite by the way the human brain works.
Tony Whincup:You mean like neural networks just in in kind of mirroring how we how we might Yeah.
Valeria Moore:They they mirror the the functioning of the human brain to identify the patterns and the structure within the existing data that are put in as inputs to generate new and original content. And it can be very useful. It can shorten the software development life cycle. It can increase the level of automation to do simulations and be applied to improve the outcome of many different industries like biology, healthcare, manufacturing or even simply customer service via automation and efficiency.
Tony Whincup:Yeah. I was sorry. I should interject. I was just reading the other day about sort of the pharmaceutical industry having, you know, three three trials, you know, the phase one, phase two, phase three, phase one about safety, phase two, phase three about efficacy, and that phase two and three could be collapsed, you know, drastically in terms of years by finding out the point of failure for a drug or whether it could actually, you know, be of benefit and actually be monetized. Quite radical kind of leaps in terms of how quickly we can do stuff.
Valeria Moore:Yes. That's a great point, and it's important to to underline that thinking about biology, it's a very complex science because it it relates to reality. So by that, the artificial intelligence will will help solve problems faster in the long term.
Tony Whincup:So with DeepSeek, I mean, was it surprising for you? I mean, you're an analyst. You follow this stuff night and day. You know, was it was it better? Was it quicker?
Tony Whincup:Was it was it cheaper? Was there anything radical about it?
Valeria Moore:Absolutely. DeepSeek. What is DeepSeek? It's a Chinese artificial intelligence company that has emerged a few years ago from a university startup. So it has taken a new approach on artificial intelligence models using techniques that create a simple model out of more complex one at apparently at a lower cost.
Valeria Moore:In reality, the the cost is not lower because the the model uses all the previous models built before, and we don't know the cost of that pretraining which is creating the previous model.
Tony Whincup:So So that kind of $5,000,000 price tag that someone reportedly put on it was was nonsense, was it?
Valeria Moore:Well, of course, it's this isn't it's that that's not the case because the big unknown is the cost of the pretraining we we should we should add add it up. And overall, they are remarkable models, but not miracles. And every day, we constantly have innovation. We we head it from the sea, but we have from the Western, artificial intelligence, lab every day, and they do share innovation and knowledge, and they leapfrog each other. So that is a positive outcome.
Tony Whincup:I mean, lots of different implications from from having this, you know, new rival on on the block as it were. I guess, in just in terms of computational power, what else comes to mind as some of the the implications that you're considering now?
Valeria Moore:That that's a good point. In fact, the the demand for for for computer, it's it's it's it's not decreasing on on the contrary. The the training cost involve more compute. And in addition, if these model piggybacks on previous large models, then the previous large models require compute. And also it is important that in technology, there is never enough compute.
Valeria Moore:And when cost decrease, if that is the case, demand increases. And and so we will have a greater demand as the total addressable market expands. And in particular, if we define the inference as inferencing output through the application. In particular, this large market will have a greater application even with lower cost or more diffusion of artificial intelligence.
Tony Whincup:It sort of seems to me that many companies are working out how they're going to monetize the AI prize, if I can put it like that. But there may be some other barriers, right? Well, I don't know. You tell me, do we need more data centers? Do we need more reliable energy sources?
Tony Whincup:Need more computational power of a magnitude that we can't have? I mean, what other barriers are there to the adoption of this technology, you know, really increasing productivity or improving people's lives or or economic gains?
Valeria Moore:Sure. It's all of the above. I we need, for example, enterprise to increase gradually the adoption of artificially intelligent artificial intelligence for for the market to to expand. We need the continuous innovation. And it's not only power, but that goes hand in hand with energy efficiency so that the data center can use the power in an efficient way and deal with the limit, the bottleneck.
Valeria Moore:So it's it's overall a gradual continuation of innovation, and we are at the beginning of of new innovation drivers.
Tony Whincup:I mean, I feel like I'm gonna take you off safe ground with my sort of penultimate question, but it just strikes me just how sensitive this stuff is geopolitically. Right? You've this sort of multi multipolar world, different spheres of influence and centers of gravity. Does that mean that we're likely to have more regional or, different variations of these technologies that will rival each other, or will will we really see universal ones that, you know, are for the benefit of all? And I say that within the context of a world that is arguably a little bit less global than it was, you know, from talking with Giles there than it was two weeks ago given what what Trump's agenda has done.
Valeria Moore:Sure. I can comment on those. From my point of view, geopolitics are often an unknown variable, and and we need stability on the other hand for investment and and innovation. And also there will be, and there is already a trend to build locally, selectively, some replication of artificial intelligence tools that that that at the expense of efficiency. But then we also have some polarization.
Valeria Moore:The US tech companies, they have added a great innovation and value. Chinese company have mastered their own areas as well. For instance, look at electric vehicles. So innovation, if you step aside from the current volatile contents and look into the future, innovation, it's it's important, and that that that will will will be what what prevails. And so then the companies that address that best will ultimately add value.
Tony Whincup:Okay. I mean, I'm gonna paraphrase. DeepSeek wasn't necessarily better. It wasn't necessarily cheaper, but but good to have a new kid on the block. And and, actually, I take that as a bit of a tonic optimistically, Valeria, that there are these long term structural trends that will be transformative, that our clients will benefit from, hopefully, at some point in time.
Tony Whincup:So thanks for that. Giles, just wanna come to you. Do you anything you wanna pick up on there as a a a final a final point the sort of structural trends that, you know, stretch beyond the the tweet to tweet kind of paradigm that we're in right now, living second to second with what Trump does or doesn't say. There are long term forces at play that we have to remind ourselves of.
Giles Parkinson:Yes. And investment managers are always dealing, grappling with uncertainty. And I think, Tony, that's, when we say actually, oh, today the the current environment is very uncertain, I think, actually, what we mean is maybe any previous certainty that we might have thought that we had in recent months actually was misplaced. The future is always very difficult to discern, particularly when we get into the world of macroeconomics and geopolitics and so on. That said, I then take great comfort in, actually, there are pockets of predictability.
Giles Parkinson:And, actually, it's much easier in my opinion to forecast the future the future evolution of a particular company, whether it's going to be on the right side of the trend that Valera is talking about or potentially on the wrong side. Actually, that's almost a pocket of certainty within sort of the swirling sea of of all of the wider uncertain bits. And in my experience, several decades of investing experience, good companies win through in the end. Very good.
Tony Whincup:Nice few words to end on. Valerio, I enjoyed that chat. Thank you, Giles. Thanks for your thoughts as always. Pleasure to have you here.
Tony Whincup:Well, that's it. Thanks for listening. If you like it, please share it or comment on our socials. You can visit trinitybridge.com for more investor insights across Prince TV or, of course, this podcast. But for now, that's it.
Tony Whincup:We'll see you next month, when your usual host, James, will be back. Thanks. Thanks, Helen.
Valeria Moore:Thank you.