Robotaxis, tech earnings and 24-hour trading

0:05Pushkin.

0:30Times, and Pushkin. My name is Aidan Ryder. I'm filling in for the delightfully British, Katie Martin, and their grotesquely American, Rob Armstrong. And today, I'm joined by John Foley, the head of the Lex column. Also delightfully British. Yes, and I guess I'm also grotesquely American. Well, John, we have a couple topics to talk about today. 24-hour trading, tech earnings, and robo-taxis. Let's start with 24-hour trading. Why in the world is this a thing? Yeah, so what's happening is that the London Stock Exchange, we reported, the FT reported a few days ago, is considering extending its trading hours.

1:07At the moment, trading in London happens basically between 8 o 'clock in the morning and 4.30 p.m. What a shonda. How terrible. I know. Imagine that. But it sort of does feel a little old-fashioned, I guess. I mean, it's also true in the U.S. The U.S. trading hours are like, now it's recognized, 9.30 to 4 o 'clock. and that hasn't really changed since the 80s. Yeah, we're lazy over here. It's just 9.30 to 4 o 'clock, not 8.40. But the world has changed and everything is crypto and crypto is 24 hours. Yes, and this is, I think, the key thing. Every time we talk to crypto bugs, you try to get them really talking about why do they love crypto?

1:40And one of the many, many reasons they'll throw at you, which I'm not sure is so valid, is that, well, obviously we're all constrained by the limited hours of NASDAQ or LSEG. We need to have 24-hour trading. But who actually wants this? I mean, what's the point? So there is, trading has changed, right? Retail investors and investing from your cell phone, the GameStop meme stock saga showed that there is an army of traders who are doing stuff from their phone after dinner in the middle of the night whenever they want to. So the idea that trading should only happen in this kind of, not even full office hours, right?

2:14It's not like nine to five. In London, it's like eight to 4.30. That trading should only happen then does seem a little bit antiquated. Now, as it turns out, you can already, if you are a retail investor, there are lots of ways that you can trade 24 hours. If you use Robinhood, if you use Interactive Brokers, if you use Schwab. How is that trading? If the market is closed, how is that actually 24-hour trading? Because what they do is they're using basically dark pools, so off-exchange trading platforms where you can make a trade. You can basically sell your stock. They don't clear until the next day.

2:44And you may not get the absolute best price because you're not getting all of the information that you get during the day with all of that liquidity, all those institutions trading. You don't have what they call the tape, you know, the record of all the trades that happen during the day. That only goes through official trading hours. So it's not quite as good as the real thing, but you can get almost as good as the real thing. So what the exchanges are talking about now, and it's not just the LSE, it's also NYSE has also applied to extend its trading hours. It would be having official trading hours that go on, if not fully 24 hours, then 22 hours in some cases.

3:16So New York is also applying for this. It's not just London. Yeah. So in the US, there's a new exchange called 24X, which has been given approval to trade not quite 24 hours, but more or less 24 hours. The New York Stock Exchange has also applied, and so has NASDAQ, to trade much longer hours. They will all have a kind of break. A lot of the time we talk not about 24-7, but 24-5, so five days a week. And often it involves a stop for like maybe two hours overnight for people to reset their computers, install software, whatever it may be. But this is not just a UK thing. And actually, interestingly, even in the 80s, when NYSE last changed its trading hours by like half an hour, people were talking then about the inevitability of 24-hour trading.

3:58So Dick Grasso, who was the CEO of the New York Stock Exchange, always said in the 90s that they would move towards 24-hour trading. It just hasn't happened for various reasons. In the 80s and 90s, they were all running high on cocaine and, you know, they could go all night. Also that. But also the reasons to not do it in the 80s and 90s were in some ways much more compelling because it was all about people. So it was a very people-based business. And one of the concerns, even when they moved the trading hours, they talked at one point about moving the New York trading hours back by half an hour.

4:27So staying open for an extra half hour. And that created a lot of consternation over the idea that it would create overtime charges for companies that had to pay people who process trades when the market closes. So what's the solution now? In theory, you have more algorithmic trading. You could have desks across the world trading. Is that the idea? Yeah. Well, I guess it's a less people-intensive business. But there are still lots of problems with this. And it is ultimately still a people affair. When is your tech person going to sleep if the market is 24 hours? Evidently never. Right. You have to have software updates.

4:57You have to have some kind of latency in case things go wrong. What if you're the CEO of a company and do you not have to wake up at 2 in the morning because your share price moved by 15 % in the middle of the night? I mean this seems like a terrible idea to me, especially in New York, right? We already have the most liquid, deep capital markets here. The benefit of being in New York is you get to decide when the world is truly trading. So while there might be trading going on in London or in Asia at different hours of the day, it makes more sense for them to go 24-hour trading to align with the New York cycle.

5:27I can't believe that is the most American-centric thing. No, it's true. I'm not as American-centric or, shall I say, grotesquely American as Rob. But that is the benefit of being in New York, right? Why would you want to give up that privilege? No, I mean, it is true. There are other challenges, too, with extending trading. You have to extend the kind of regulatory oversight. You have to extend clearing. And there is talk about clearing, moving to a 24-hour business as well. And also, a lot of the big institutions don't really want this because they say, like, it's fine for a retail investor who wants to buy a couple of shares of AMC at four in the morning.

5:59But if you're T. Rowe Price, are you really going to be able to do a massive trade? Is there going to be enough liquidity to do that trade in the middle of the night? Probably not. And one of the things that there's various research that has shown that having kind of compressed or, you know, nine to five or eight to four or whatever is trading hours does kind of focus liquidity into a window. and in particularly you get a lot of liquidity at the very beginning of the day and at the very end of the day. I see. So this would just be dispersed liquidity, which has its own risks. It has its own risks.

6:24And then on top of that, you need regulators and other people staying up all night to make sure things don't hit the wall. So this is a very early proposal. We've reported that the LSE is talking about it. This is part of a bigger process that the LSE is going through, that London is going through, as it tries to work out how to be more competitive on a global level with other exchanges. It seems like having only, you know, beleaguered and not sleeping people in London is not necessarily the best way to be competitive to me. But I digress. I agree. Obviously, as a journalist writing about companies, do I love the idea of events happening, you know, companies releasing profit warnings at 1 a.m.?

7:01Not so much. Not at all. But then again, if that is the way the world is working and if more stuff becomes automated and if other markets, you know, currencies, crypto are 24 hours, maybe this is the way things are going to go. Well, at the very least, this like knocks the legs underneath crypto of why they think they're special. right? It's not just your 24 hours. Okay, next topic. So this week, we have big tech earnings. What should we expect? Who should we expect to report? So tech is a really interesting one at the moment. We've got Alphabet and Tesla coming on Wednesday. Tech earnings, like all earnings, used to be about what happened to earnings in the last quarter.

7:33Then when we moved to this AI era, it became about how much of the company's spending. Everyone was competing to spend more and more money on future AI infrastructure. And And we've now got more than$300 billion of tech spending by the big firms this year. The market then started to slightly freak out about this, saying, like, you know, what if this doesn't produce concrete returns? So now what investors are looking at is less what's happening with earnings. They're looking at guidance. Obviously, they're looking at, like, signs of what's going to happen in the future. But also what's really interesting is they're looking for detail on how and whether AI is producing actual financial outcomes.

8:09I remember over the last two years, it seems like, you know, earnings are important, but you have to really, really overshoot the runway to actually be rewarded by this market. Right. You have to show that despite all the CapEx spending, you are actually having some other profit, etc. So it sounds like it's fair to say people are kind of getting off of that, not even focusing on the earnings at all and just looking at how much of your code base is done by AI. Yeah, they're looking at what's the AI actually doing. So if you think about Google, Alphabet, Facebook, and Meta platforms, these are kind of advertising businesses.

8:39So what they're now under more pressure to tell us is how AI is helping them to sell more advertising at higher prices. And they actually are starting to do it. Meta in particular, Meta has been talking a lot about how advertising is getting more effective and how click-through rates are rising because of the AI that it's deploying. Google has come out with these little snippets like that, you know, Gemini, its AI model is now accounts for about 30 % of its new code. Also, they've talked a lot about AI overviews is a huge thing at Google. You've seen if you search for certain topics, you now get like a blurb at the beginning instead of just getting the links.

9:12But you get an answer, basically, an AI-generated answer. For a while, there was some concern that that was going to like take attention away from ads. But actually, Google said recently that it's getting as much money for an AI overview result as it did from the traditional links. That's the kind of stuff investors really care about right now. So investors are focusing on how these companies, not only are they successfully building AI for consumers, but actually turning it back in on themselves and generating revenue from that. To commercial outcomes. It's interesting. I spoke with someone yesterday, and as you mentioned before, where guidance is actually almost the most important at this point, right?

9:42There's still a ton of uncertainty over tariffs, growth, all the big questions for the market and the economy. And oftentimes, we're relying upon these companies to tell us how tariffs are impacting them so far, how they expect them to impact them in the future. So it seems like guidance is particularly important to that investor. And it sounds like you're saying that's important for everyone. I think it is, yeah. How can it not be when you're talking about AI, which is like a long-term investment plan? I mean, it does exist, but it doesn't really exist in the way that we think it will in a few years' time.

10:09So Google earnings per share from the last quarter, sort of who cares, unless it's an enormous miss. even if it's an enormous, you know, if they massively exceed what people are expecting, it's not that important. What we're looking at is where is this 300 plus billion dollars investment going and what are the returns going to be? The thing that we've said over and over again when we've been writing about this and, you know, we've talked to the companies about this too, like they're not giving you any sense of what returns on investment they expect from this money. They're just expecting us to believe that those returns will be generous.

10:38Believe AI will turn into something. But at the same time, they're saying like, if we don't do this, the risk of being left behind And it's better to basically waste money in order to remain at the forefront of AI. Build giant data centers for almost no purpose. They're literally saying it's better to do that and lose a ton of money than it is to lose out. Yeah, but now the market is saying, okay, well, let's see what you're actually using all this investment for internally, not just how many people you've reached with your chat GPT or name or model. It's very interesting, especially on the big tech earnings.

11:06As I said, over the past few years, not only did they have to do well on earnings, they had to super overshoot for the market to be at all happy with how much spending they were doing on AI. If you look at their earnings per share growth year over year by quarter, I mean, it seems like these companies are slowing down. Part of that is because of normalization, right? You have a year ago, you had so much spending and, you know, X amount of earning that was so great. This year, it feels like less. I'm just curious. I mean, should we expect lower earnings? Or your point is nobody even going to be paying attention?

11:33I think these businesses are huge, right? They're huge and they're kind of mature in some ways. So you can't forever have like 30%, 40 % growth rates. I mean, if you're like Palantir, for example, if you're in a really hot sort of software of sector, then maybe you can. But Google advertising is not going to grow at 20%. I think Google's expected to grow advertising revenue about 7%. Which is still very good. Yes, pretty good. It's twice GDP. More than most of S &P 500. Cloud computing is growing faster, more like in the 20s. But these businesses are so huge that they can't defy gravity forever in that way.

12:04Yeah. So tomorrow we get Tesla. Yeah. And, you know, we said that's part of the Mag 7. But their performance and what they've been doing so far doesn't really put them in the Mag-7. I mean, they've had negative earnings per share growth over the last few years. They've had a slew of issues related both to their CEO and to the company itself. One thing that seems to be leaping out to people is what is the future of robo-taxis? And you have some thoughts. Yeah. So robo-taxis are a big part of the Tesla valuation story. They are not a big part of the earnings story. So think about Tesla when you're looking at its earnings.

12:33You're seeing the earnings from a carmaker. and a carmaker that has relied until now largely on credits, EV credits, which are going away. So you will see earnings reflect that. Its valuation reflects things that Tesla is only just starting to even make. So RoboTaxis is one. Optimus the robot. I said on the last episode that those robots creep me out and I stand by that. They are basically murder bots. But those are what the valuation tells you. It tells you that people think Elon Musk is going to create this incredible RoboTax business. I've heard someone say that Tesla is a, you know, it's a company that has the revenues of a car company, which is always good, but not amazing.

13:13And then it has the VC potential of Elon Musk, right? What are the other things and future projects we're going to put all this car money into? And RoboTaxis has been the big one. It seems like Tesla's RoboTaxis are falling behind. I mean, Waymo is out there picking up people on the streets of San Francisco as we speak. RoboTaxis aren't out. Are they falling behind? So the thing about robo-taxis that I think is interesting is that the idea of robo-taxis is super simple, right? It's that you call a car, it shows up, it takes you where you want to go, and there isn't a driver in it. It's just Uber with no driver.

13:44It's so simple to imagine. It's really hard to implement. It's so hard to do. And it's hard because the cars themselves are hard to design. Waymo is by far the leading robo-taxi provider now. It's everywhere in San Francisco, not so much elsewhere. They are going into other cities gradually. But their cars are expensive to make. They've got all kinds of crazy stuff on the roof. Yeah, the spinning thing. Like radar, the spinny thing. Yeah. They're also like, they're fighting the fact that a lot of people don't trust cars with no drivers. And San Francisco is a very tech forward city. People have got comfortable.

14:15And it's taken them a while. Got comfortable with the idea of getting in a car. You know, we all know people who've done it and love it. Yeah. That's great. Everybody who does it is obsessed. But turning this into a really big market is really hard because you've got the technology barrier. You've got the cost. You've got the trust barrier. And the regulatory barrier. The regulatory barrier. But mostly the cost. Because at the moment, the cost of owning a car in the U.S. is about a dollar a mile, roughly. The cost of a rideshare in Uber is maybe$2 a mile. The cost of a Waymo, the cost of a robotaxi, is about$3 a mile.

14:45So you're not going to take a robotaxi when a car will do because it's just more. You're not going to take a robotaxi for your commute to work because it's just too expensive. Unless you're trying to get really good reading done. Right. Well, unless, yeah. I guess by another point you would hire somebody. Three times more than you've had. So this is why Robotex is struggling to develop scale. Even Waymo, about 1 ,500 cars, it's probably not making anything like a profit for quite a long time. So where does Tesla fit into this? So Tesla, if it can get it right, Tesla could actually crack this code.

15:14Because Tesla can make cars potentially quite cheaply. It's talking about making like a$25 ,000,$30 ,000 RoboCab, which is about a quarter of the cost of a Waymo car. As opposed to Waymo, which is buying or partnering with other car companies. With Jaguar, yeah. Tesla makes their own cars and they can, you know, make it special. Tesla thinks it can do it without all the spinny stuff on the roof. Oh. Which is unclear whether that's true or not. Less expensive tech? Cameras are less expensive. Cameras and AI rather than radar and LIDAR, which is specialist equipment. Interesting. If Tesla can do it cheaply, say if they can get to that$1 a mile, then theoretically the sky's the limit.

15:48That's when you start taking a Tesla robotaxi to work. And that's when you decide to not bother owning a car or not taking your driving test, in theory. In theory. So Tesla, if it gets that right, but there are so many ifs, is the big threat in robo-taxis. You can see that now showing up. Like Uber, for example, which did a deal with an electric vehicle maker, Lucid, last week to buy a bunch of their cars. Uber has got to be freaking out about the possibility that Tesla successfully cracks robo-taxis. Yeah. But, you know, I don't know how much we should actually expect from this earnings. I mean, a lot of Tesla earnings talk about robo-taxis without giving you real numbers.

16:22I mean, I remember - You'll get nothing in these earnings at all. Very distinctly, they put out, I think it was a chart for robo-taxis where it's like, long chart, like what robo-taxis could do, low chart, what others can do. And there was no actual numbers on the chart. Right. Or like when he said the total addressable market for the optimum was every human being on the planet. Yes. So Elon Musk is great at saying things like this. He has said about robo-taxis, he said that Tesla could have 99 % of the market or will have 90 % to 99 % of the robo-taxi market. He may be right. If you're buying Tesla share, if you own Tesla shares, it's probably because you think he's right, which is why Tesla is a trillion dollar company and why analysts and investors are not attributing really any value to Waymo with an Alphabet at the moment.

17:05Got it. So Alphabet's not getting that much of a boost from Waymo, even though it's right now the leader. It's the leader. But if Tesla gets this right, it will pretty effortlessly take over. But again, you either believe in Elon Musk or you don't. And that's always the question. With that, we'll be back with Long and Short.

17:30This is Long and Short, the part of the show where we go long things we like and short things we don't like. John, do you have a long or short for us? I do, Aidan. My short, I'm sad to say, is bank holidays. Oh. France's prime minister has suggested getting rid of two bank holidays. And you remember that President Trump has also expressed concern that there are too many days off in the U.S. And I fear that our beloved, beloved bank holidays, which, of course, in Europe, we have many of them, might be going away. It's so funny to me that, you know, this government of France has struggled with likability.

18:02And their first budget solution is to get rid of people's holidays. Right. You'll have people working longer, being angrier, less productive, I suspect. Yeah. Well, I am long measurement. There's a great story in the FT today. Nigeria's economy is now 30 % bigger after they recalculated GDP. As we learned in the first quarter this year in the US, which had a GDP contraction because of some strange stuff around imports, calculating GDP is weird and wonky and especially hard on emerging markets. And I'm just kind of along the conversation. Great. Who doesn't love measurement, Aiden? Yes, I always keep a ruler in my back pocket.

18:38Well, with that, we'll be back in your ears on Thursday. Until then, stay sharp out there.

19:13I'm Aidan Reiter. Thanks for listening.