When AI Takes Over Shopping Carts

Andrew Percoco: Welcome to Thoughts on the Market. I’m Andrew Percoco, Head of North America Auto and Shared Mobility Research.

Tim Hsiao: And I'm Tim Hsiao, Greater China Auto and Shared Mobility Analyst.

Andrew Percoco: Today, why robotaxis may be approaching a commercial inflection point.

It's Thursday, August 13th at 8am in New York.

Tim Hsiao: And 8 pm in Hong Kong.

Andrew Percoco: So Tim, for years, robotaxis were really confined to limited pilot rollouts across the globe. You've done a lot of work over the last few weeks. We put out a big collaborative report on the robotaxi market and how it could be a $1 trillion TAM by 2040.

What makes this moment different than some of the other robotaxi hype cycles that we've seen in the past?

Tim Hsiao: We observe four things have been converging. Firstly, end-to-end AI is improving much faster. Secondly, hardware and the training costs are falling. And thirdly, more well-capitalized players can fund deployment. And last but not least, regulation is becoming clearer.

The leading operators are no longer just demonstrating the technology. They are running fully driverless services around the clock and generating commercial rides. So in our view, the questions has been shifting from can it work to who can expand operating areas, raise utilization and lower costs at a much faster pace.

So that's a very different setup versus the 2018 and 2021 hype cycles.

Andrew, U.S. autonomous miles could rise from 116 million in [20]25 to 16 billion by 2032. But still make up only about 0.5 percent of all miles driven. How can robotaxis become a meaningful business while remaining such a small part of the market?

Andrew Percoco: I would say, you know, obviously the U.S. mobility and transportation market is a massive market. So even with the rapid growth that we expect in robotaxis, it's going to take a long time to make a material impact in the overall market share of mobility.

But if you think about the profit pools in this business, 16 billion miles at $2 a mile can, you know, pretty quickly become a very significant TAM and market opportunity.

And I think, you know, fundamentally, if you think about a robotaxi business, I would argue you're better utilizing an asset... Or if you think about the, you know, car park, the amount of vehicles that are, you know, in the fleet today or in the U.S. today, they're sitting idle 90 percent of the time, right?

So you're talking about taking a smaller amount of volume and driving a higher utilization on that fleet and driving much improved economics. So yes, it's going to take time to displace the, you know, hundreds of millions of cars that you have on the road in the U.S. and displace the penetration of miles driven.

But ultimately, you know, we think that the profit pool and the opportunity in robotaxis are much more attractive for the entire value chain, as it relates to robotaxis. And I'd say there's a few things that we're watching along the way to make sure that, to your point, you know, this is not another hype cycle. And that there's real commercial kinda backbone to this business.

I'd say the first is seeing the rollouts continue to improve, and the density of the rollouts improve across the select cities that we've seen in the U.S. right now. Robotaxis are only available in a handful of cities in the U.S., so we want to see that continue to expand into more cities. But also the density of the fleet increase in the cities where they're currently present.

And at the same time the safety side is still something that gets a lot of questions in making sure that it is truly safer than a human driver, across technology platforms, right? There's various players in this market with different approaches to technology. So, I think seeing that the safety curve is starting to or continues to improve is going to be very important for the viability of this market going forward.

Obviously U.S. is very different from China. What have you seen in China? China has shown some impressive growth and utilization in some of the operators that are on the road in China. So just curious as to your perspective in terms of what you're seeing on the ground there.

Tim Hsiao: I think China shows that there's much in operations and skill challenges as technology challenges. The fleet  in China is above 5,000 vehicles across I think more than 7500 square kilometers in key cities. And some operators average more than 20 orders per vehicle per day.

So, total cost of ownership has fallen roughly 30 to 40 percent, while remote assistance ratios are moving from like one operator for like 20 to 40, even like 50 to 60 vehicles. And we think it will achieve like one for a 100. So that has produced real break-even happens, especially in some major cities like Guangzhou, Shenzhen, Wuhan – the tier one, tier two cities.

So in our view, I think in China, wider operating domains, fleet density and utilization rate, as you just mentioned, reinforce one another. So make it some more like the real commercial case. Instead of just, like trials as we saw a couple years ago.

If more value shifts towards the software, fleet operation, and the data, as well as the customer relations, how does that change the profit pool, across the auto industry, especially in the U.S.?

Andrew Percoco: First off, I think the auto industry in general is becoming, you know, more software focused and aware. You know, it's being led by the robotaxi market where the autonomous driving software and technology is obviously the most important part about getting this technology to market.

That is ultimately trickling down to personally owned cars where you're seeing more autonomous technology being deployed. Auto OEMs are able to charge subscription revenue for this software. So it expands, I'd say, the value proposition of buying a vehicle expands the profit pool for the OEMs.

It changes in some ways the cyclicality, or can change the cyclicality of the industry if you've got more kind of recurring revenues, subscription like business model versus just a hardware focused OEM model, which has been kind of the predominant focus for the OEMs historically.

I'd say the other angle, interesting angle here is, you know, as this business scales, there's gonna be a lot of vehicles on the road. There's gonna be a lot of fleets of vehicles on the road. Those need to be managed by somebody or some company, right? So if you think about, you know, the rental car industry, right? These companies have been in the business of managing fleets and renting out fleets for  a very long time. They know how to do that very, very well.

I think there's an interesting opportunity for that part of the value chain, to participate in aiding these robotaxi fleet operators, in scaling and bringing their business to market. Charging, maintenance, reconditioning, all the things that  take a lot of time and a pretty large amount of physical infrastructure.

That's an opportunity for the rental car industry to come in and leverage their existing know-how to help. And, you know, I think Tim, an important part of this commercialization process is driving down the cost structure of robotaxis. They are very sensor; heavy sensor heavy. They're very compute heavy. I think China is the clear leader on cost and supply chain. I think in China you're seeing robotaxis, you know, around $35,000 to $40,000, which is considerably lower than what we see in the U.S. today.

So, how do you think that that will accelerate adoption in China, but I'd say more importantly overseas as some of these robotaxis businesses look to expand outside of China.

Tim Hsiao: In our view, it could be a major accelerant because as we noticed that the depreciation is still one of the largest fixed costs for robotaxi. So, as we just mentioned, I think, $35000 to $45000 US dollars, the purpose-built robotaxi can lower the breakeven utilization threshold. And make it easier to finance fleets and open cities that could not support the $150,000 US dollar vehicle.

And not only in China, because globally, I think the Chinese cost deflation can be paired with the local ride-hailing platforms in the overseas market that provide demand and regulatory access. But as we highlighted in our previous, the global reports once again, we don't think the cheap vehicle is sufficiently by their self.

So in our views, on top of the competitive cost structure, registration, data localization, insurance, and local operating costs can still delay the margin curve, particularly in Europe, which we think there are still quite a lot of uncertainties.

So Andrew, as we just, as we just discussed, the lower vehicle costs help, but the operating model still has to work, right? So with operating costs expected to fall and the margin potentially moving above 30 percent or even higher at scale, what are the key assumptions investors should focus on?

Andrew Percoco: There’s a handful of key assumptions you need to sensitize to get to that 30 percent or more margin structure in this business. I'd say the first is going to be utilization, right? You need to be running these assets at a high utilization to essentially amortize those fixed costs over a larger number of miles driven.

Number two, insurance today is probably one of the largest buckets of cost when we think about this business. Insurance is, from our perspective, a big unlock for this industry as the safety, as we mentioned before, the safety data continues to improve. We think that will be a reason to, to expect that the insurance costs associated with autonomous driving technology and robotaxis will continue to decline.

It's about 30 cents per mile on our estimate, so it's very significant in terms of the overall cost structure of robotaxis. Drivers or where there's the most sensitivity around the model. Obviously, there's charging costs, there's maintenance costs. Those are, I think, fairly known at this point. But the utilization and insurance, I think, are the two biggest drivers of really getting that margin profile to improve over time.

Tim, I guess when you think about the next, call it 10 to 15 years, I think we will put out a trillion dollar market by 2040 from a TAM perspective.

What do you think the biggest markets are that investors should be watching, in terms of getting us to that trillion dollar TAM? Obviously, U.S. and China are kinda leading now, but what are the next markets people should be watching?

Tim Hsiao: In addition to the major market, as you just mentioned, the U.S. and China, in our views, I think we also need to focus on markets like Europe, the Middle East and Southeast Asia. I think their scale is underappreciated, as we highlighted in our previous report. Because if you think about that, Europe, the Middle East, and Southeast Asia in aggregate have roughly four million taxis together ride-hailing vehicles.

So even with 25 percent conversion, they imply that about one million is the L4s vehicles. The Middle East offers supportive regulators, you can tell, simpler operating environments and higher fares. And if you think about the Southeast Asia, the ASEAN, I think the market has dense demand and strong local platforms.

And of course, Euro markets definitely can't be ignored because Euro will move more slowly, because we think the regulations and the data rules would initially add cost. But the truth is, if you think about the European market, I think the taxis or ride-hailing fares are among the highest globally, even compared to the U.S. and rest of the world.

So in our view, the material margin could be more attractive. And this market, on top of the U.S. and China, in our view, can support several regional winners. So, not only limited to a very, you know, the single one or two markets.

Andrew Percoco: Yeah, it’s great Tim. It sounds like, you know, the robotaxi race, if you want to put it that way, will be won by those who can really bring together technology, and a compelling cost structure while also following the proper regulations and making sure the safety is improving at a rate that's acceptable to regulators.

So, Tim, thanks for taking the time to talk today. And thanks for listening. If you enjoy Thoughts on the Market, please leave us a review wherever you listen, and share the podcast with a friend or colleague today. 

Ariana Salvatore: Welcome to Thoughts on the Market. I'm Ariana Salvatore, Head of US Public Policy Research at Morgan Stanley. Today, the latest on US Iran tensions, talks, and the path to a deal.

It's Wednesday Aug 12th, at 2 p.m. in New York.

The diplomatic picture in the Middle East has shifted yet again.

Last week, there was growing optimism that the U.S., Iran and Oman could reach an arrangement to improve commercial passage through the Strait of Hormuz. But the two sides have since hardened their positions.  This week, we've seen some bouts of escalation, and headlines have been mixed over the past few days.

At the same time, the energy security picture remains complicated. The U.S. administration says the seven-day average of oil leaving Hormuz has risen to almost 9 million barrels per day. But traffic remains well below normal conditions, and the risks we think are no longer limited to the Strait. We’re beginning to see potential for disruption across multiple regional chokepoints and alternate shipping routes.

That brings us back to the framework negotiated nearly two months ago. The U.S. and Iran signed a Memorandum of Understanding in mid-June. It was intended to create a 60-day window for negotiating a more durable agreement. That framework addressed commercial passage through Hormuz, the US naval blockade, sanctions relief and frozen funds – as well as longer-term negotiations over Iran's nuclear program. But the implementation has proven much harder than agreeing on the framework itself.

So where are negotiations getting stuck?

First, there's the Strait itself. Iran has tied a full reopening of the Strait to a broader package that includes an end to the U.S. blockade, sanctions relief and compensation. Washington, in turn, is trying to preserve economic leverage and appears unwilling to provide those concessions upfront.

Second, sanctions sequencing: The U.S. wants relief tied to clear signs of progress, while Iran is seeking confidence that any relief is durable and not easily reversed.

And third, there’s the nuclear question: enrichment levels, Iran’s existing stockpile, and a longer-term verification framework. These are still to be negotiated. That’s likely to take longer than the 60-day time period.

So, what’s the right framing here for investors?

We think it’s not necessarily a deal or no deal binary. It’s more so a series of partial agreements, implementation tests, setbacks, and renewed negotiations. After the June deal was signed, we flagged several live paths to re-escalation: execution risk around sanctions and Strait control, a potential divergence between the U.S. and Israeli objectives, domestic political pressure in Washington, and the basic challenge of resolving core nuclear questions within such a short time frame. We think those risks are now becoming more visible, but we think both sides have strong incentives to avoid a return to a full conflict, like the type of engagement we saw back in March of this year.

Moving forward, the signposts we laid out in June—maritime normalization, access for the International Atomic Energy Agency, sanctions implementation, military restraint, and rhetoric—all remain the right trackers to watch. But expect the bargaining process itself to be noisy, unstable, and non-linear. Rather than a clean transition from conflict to ceasefire to final deal, the more likely path will have fits and starts.

So what should investors do with that information?

On oil, our commodity strategists remain constructive on prices, given the ongoing supply uncertainty and the emergence of new chokepoints across the region. Altogether, they see those constraints keeping the market relatively tight compared to the levels we briefly saw in June when the MOU was signed.

If there’s another sharp rise in oil prices, our U.S. equity strategists think that could be a key risk to the near-term outlook. Our US economists agree, but also think the Fed would need a bigger shock than markets previously expected to resume hiking. As a result, we expect the Fed to stay on hold this year.

Thanks for listening. If you enjoy the show, please leave us a review wherever you listen and share Thoughts on the Market with a friend or colleague today.