BYD vs. Tesla China Race

If it’s expensive and cumbersome, and remains that way for too long, it won’t last. Except perhaps in places where there is no practical choice.

BYD news today:

a concerted worldwide push has become the single most important strategic focus for China’s biggest EV maker, four sources familiar with BYD management’s thinking said.

Besides a drive into some European markets already underway, BYD spent much of last year conducting a study on how to set up a U.S. distribution network for its latest electric models, two of the sources said.

They described the study as advanced and serious, with specific recommendations from Detroit-based consultancy Urban Science on how many outlets in each state and city BYD would need, as well as formats for the brick-and-mortar stores.

Tense US-China relations is also a reason to hit the pause button. I would also think that the China demographic issue is figuring into BYD business calculations. There will be a serious reduction of workers going forward which millions less middle class consumers.

I think that when/(if) a family gives up their 3rd or 2nd car in favor of MaaS it will be freeing up a space on the driveway or the street in front on the house, not a spot in the garage.



Most projections I’ve seen estimate that autonomous driving will add at minimum $5K to the cost of the car. That strikes me as being optimistic, so I think private ownership of self-driving cars will be mostly limited to the same demographic currently buying Teslas.

I also believe that for most households, the bulk of the mileage and wear/tear on the car occurs from the work commute.

Combine those two observations and one can see a pretty big market for robotaxis during the morning and afternoon commutes. I envision a monthly or even yearly subscription service where one can reserve regular pickup times in the AM and PM during the work week. There could be levels of service depending on how many people you are willing to share a ride with. The advantage over existing rapid transit is door-to-door service. That’s a big one.

In this scenario people will still own vehicles but primarily for recreation and taking the kids to soccer practice. Personal cars will have lower mileage and therefore last longer. The need for a household to own multiple vehicles will be significantly reduced.

Then the same limitation would likely apply to the MaaS service - it would be limited to that same demographic.

MaaS companies have to charge enough to cover their costs. Most of the costs of owning a car are - not surprisingly - the cost of driving it around. About 80% of the total cost of car ownership for the typical driver are the costs that are generated by driving. Depreciation, mostly. If a car wears out after 200K miles (or whatever the number), then that car needs to be replaced every 200K miles - whether it’s privately owned or being driven as a taxi.
There’s also maintenance, repairs, fuel. Whether the car is a taxi or a private car, those costs per mile have to be covered. And if it’s the same type of car in both scenarios, the minimum cost will be the same as well.

There are some aspects of car ownership that are fixed (like finance/cost of capital, and maybe insurance to some extent). And you can save the costs of storage (ie. parking). But on the flip side, a MaaS car has expenses that a privately-owned car does not. Mostly deadheading between riders, but also (perhaps) greater cleaning costs, some increased insurance/liability costs, and the need to pay for storage/charging real estate somewhere.

TL;DR - the cost that MaaS companies charge for their service is unlikely to be massively cheaper than the cost of owning the same car yourself for a typical driver, and in areas where deadheading will be significant (outside of dense urban centers) those costs may be greater.

MaaS companies have to charge enough to cover their costs. Most of the costs of owning a car are - not surprisingly - the cost of driving it around. About 80% of the total cost of car ownership for the typical driver are the costs that are generated by driving. Depreciation, mostly. If a car wears out after 200K miles (or whatever the number), then that car needs to be replaced every 200K miles - whether it’s privately owned or being driven as a taxi.

Al, in reality, I would say every taxi we had (mostly Chevy Luminas and Ford vans back in the day) went well over 500,000 miles with the maintenance program we installed. If the cars were not involved in wrecks (I don’t know a cab company anywhere with un-totaled cars) they will last through focused attention to tires, batteries, plugs, belts, and oil changes, filter changes, quarter panel repair, etc.

200 miles plus per 12-hour shift = 400 miles plus per day. (My average as a Supervisor was 225-250 miles per 12-hour shift.)

Most taxis were depreciated over 5 years in our fleet.

Do the math. Taxi companies look at 200,000 miles as “break in” miles.

Meanwhile, back at the $TSLA Cult compound, things have gotten very frosty with people who once owned Teslas, or, who aspired to own one. Buried in the thread is news about who might be coming back to Twitter, which means if he does, Musk will tie him around his neck as his own personal albatross. If he does, I’m outta there for good and will head to as well as hundreds of thousands of others.


Ever the pessimist.

I see two types of services, the first being rides on demand and the second rides by subscription where rides are by appointment and guaranteed (e.g., pick me up at 7AM at home and 5PM at work M-F). The latter can have tiered pricing with the levels differing in the comfort provided and the number of shared riders. One should be able to offer something attractive to most of the population.

I question that but no point arguing between what are essentially guesses. Let’s suppose you are right. The average person will likely only be able to afford something along the lines of a self-driving Corolla for $40K. The alternative will be using a robotaxi van custom designed for comfort and roominess. I think many will choose what is essentially a chauffeur service in a more comfortable vehicle over ownership, all other things being equal.

I think the cost of deadheading is overstated. There is no requirement that robotaxi companies must meet the rush hour demand. Companies will put out the number of robotaxis that maximizes their profit.

1 Like

No way.

Just this morning I had to adjust my schedule when my daughter missed the bus. I had to leave early and of course take a different route. For anyone with kids, especially in the burbs, we don’t want to be dependent on a transportation service that cannot easily adjust to our changing schedules.


It’s not pessimism - it’s just economics. The overwhelming majority of the cost of owning a car is the cost-per-mile expense of operating it (depreciation, fuel, tolls and the maintenance/repairs that result from the car actually being driven). MaaS companies will have to charge enough per-ride to cover their cost-per-mile expense of driving the car that distance. Which means that the price that a MaaS service charges can’t be all that much cheaper than the cost someone would face to provide the same amount of transportation using their own, self-owned car.

Sure, if someone barely uses their car or has an extremely high cost of parking, the above won’t apply. But anyone who drives 10-15K miles per year outside of a dense urban metro is unlikely to be able to replace those 10-15K miles using a MaaS provider for materially less than the cost of owning their own similar car. Not because of any choices the MaaS makes or not, but simply because the cost of driving that car 10-15K miles in a year is fixed irrespective of whether it’s shared or not.

We’ll see, but it’s pretty significant in existing shared-transportation systems. In Uber and Lyft, deadheading is estimated to be about 40% of miles travelled.

That’s really high. Again, the costs of car ownership that aren’t per mile (finance costs, some amount of insurance) are no more than 20-30% of the total cost. And MaaS doesn’t eliminate those costs, just reduces them. If a MaaS car is running more than 20-30% deadhead miles, it will probably cost more to operate the MaaS car than it would to just have everyone own their own cars.

MaaS companies can buy cars in bulk at considerable savings thereby reducing depreciation losses compared to individual owners. Maintenance/repair costs will be similarly reduced (tires and wipers bought in bulk) and probably done in-house. Fuel costs can be minimal for MaaS companies with their own solar/wind system.

Ideally, MaaS companies will find ways to increase ride-pooling, which I realize has not been popular with current ride hailing companies but would be the simplest way to reduce costs. This could be achieved if the AVs used for MaaS are custom designed to provide some privacy between riders.

But regardless, I think you are missing the main point. Consider a four member household with one parent working full-time and the other part-time with kid duty. This ordinarily means a two car household. But if the full-time worker reserves a robotaxi from 7AM-9AM and 5PM-7PM five days a week, then that eliminates the need for one vehicle. Going from two vehicles to one is a substantial savings even if the robotaxi charges several times the cost of ownership during the hours of use.

I anticipate that a self-driving EV MaaS fleet will be connected and controlled by an AI to find the most efficient distribution of vehicles.


I’m not missing the main point - I just think you’re wrong about there being substantial savings.

Almost all of the cost of owning a car comes from using the car. The cost of fuel, obviously (no fuel cost when it’s not being driven). But also the cost of “using up” the value of the car itself (depreciation) and moving it towards maintenance and repair. And most of the cost of insurance is covering the expected cost of what could happen while the car is moving.

Almost none of the cost of ownership is incurred when it’s just sitting there. Finance costs, absolutely. Some small fraction of insurance cost as well. But those aren’t big fractions of the overall cost.

So for any person who uses their car more than a trivial amount (say, 8K miles per year or more), the overwhelming majority of the cost of the car is incurred to get them those 8K+ miles of travel. Very little of the cost of the car is attributable to the 95% of the time it’s just sitting there.

Which means that such users are not going to get massive savings by switching to a MaaS operator. The MaaS operator will have to charge them (at least) the cost of providing them 8K+ miles of travel - and that cost is going to be pretty close to what they would have incurred if they had owned their car themselves. There might be some savings to the MaaS operator (finance certainly, wholesale prices on service and fuel perhaps); but there will be costs to the MaaS operator as well (capital costs for vehicle storage and service center, additional insurance expenses). But neither of those buckets is going to be more than a small percentage of the overall costs - so there’s not really much opportunity for any significant savings for someone who actually uses their car for more than a few K miles per year.

There are such people, of course. They’re just not a materially large segment of car owners.


*N.B. Electrification changes the above a little. EV’s have higher upfront costs (higher cost of capital) and lower operating costs (lower fuel and maintenance costs). That will shift the relative proportion of finance costs to the overall cost of the vehicle. But not likely enough to matter.

1 Like

Nice try, but not true.

The majority of the TCO (Total Cost of Ownership) is due to the fact one HAS TO OWN the vehicle in the first place.

$30k to $40k vehicle discounted retail purchase price PLUS ongoing insurance cost PLUS the cost of an ongoing safe place to store the car when not in use (at both ends of each trip). Those are the major costs of ownership. Fuel and maintenance are relatively small costs compared to the first three listed costs. Plus, the first three costs are mandatory–or a rational person won’t reasonably own a vehicle.

MaaS eliminates the first three upfront expenses in terms of cash outlay. Then, the vehicle used for MaaS is significantly less expensive than a purchased vehicle because the company has a uniform vehicle that is built in volume–so lower unit cost (think “fleet pricing”). Then, the company does not use inexpensive items that wear out quickly–because maintenance is a significant ongoing cost to the business. It is less expensive to “design in” higher-grade material (where appropriate) because the TCO to the company is less over the service life of the vehicle–which is intended to have a 15-20 year service life, longer than most consumers keep their vehicles.

An EV would be far less expensive to operate AND to build. At the moment, the cost of batteries is fairly high because the technology is “not quite there yet”. In some area, there is already an equivalent cost of operation between gas cars and EVs because the cost of gasoline is high and the area is generally warm all year (CA, OR, WA, NV, AZ, NM, TX, OK, LA, MS, GA, FL, KY, TN, SC, NC, VA, WV, DC, NJ, DE). States such as OH, IN, IL, AR, MO, NE, etc are “on the edge” of being warm enough over the entire year to not need a “winter-ready” vehicle. They are warming not-so-slowly and may become EV-ready for the EV manufacturers once the states become “warm enough/long enough”.


No, it doesn’t. That’s what I think gets lost in these discussions. Those costs aren’t eliminated. They just get shifted to the MaaS provider. So they have to be reflected in the cost that the MaaS provider charges the customer.

If you own your own car, the amount of vehicle cost you experience in a given year is the depreciation caused by your use of the car. If you “use up” 10-15K miles worth of the useful life of the car, that will be reflected in the car being worth some amount less at the end of the year than at the beginning of the year. That cost doesn’t disappear if you shift to MaaS - however much of the car’s value gets “eaten up” by the amount of miles you ride is the same, whoever owns the car. Which means the rider’s going to have to pay for it either way.

The only cost that actually gets eliminated (for the most part) is the cost of finance. Assume you have 10 people who drive their own cars 10K miles per year, and the cars have a useful life of 200K miles. Every 20 years, those ten people will (collectively) buy 10 cars. Now, if they do a MaaS arrangement with the same cars, they will all “share” one car that’s driven 100K miles per year. That car will have to be replaced every two years - so over time, the total cost of buying cars is the same in either arrangement. The difference is that in the first scenario there’s continuously 10 cars’ worth of investment tied up in their ownership every year, while the MaaS arrangement only has one car being owned at any one time.

That’s a real savings - but it’s the only one (except for those folks who have to pay significantly for parking). You don’t save much in insurance, because we would expect the insurance for a car that drives 10x the miles (as in the above example) will cost somewhere close to 10x the premium. The MaaS operator also has to provide for an ongoing safe place to store the car when not in use overnight, and unlike private car owners (who overwhelmingly live in single family detached homes), they don’t have “surplus” land in a front setback that can be used for that purpose with little additional marginal cost.

There’s no free lunch, and no free rides. The MaaS provider is going to have to recoup the entire cost of the vehicle through their charges to their customers, just like if those customers bought their own vehicles themselves. The only difference is the time scale over which that expense is recouped. That’s not nothing - there are financial cost savings to buying 10 cars one at a time over 20 years rather than having 10 cars all at once for the 20 years. But it’s likely not enough to cover the deadheading costs.

I don’t understand this. jerryab2 specifically says “outlay.” I understand your point that there is ‘no free lunch’, that costs are pushed forward to the consumer with MaaS, per use, but the outlay expenses themselves are eliminated. That’s one of the main ideas behind MaaS is it not?


I don’t think that’s right, except in a pretty modest sense.

Take insurance. While you (often) have to purchase an insurance policy on an annual basis, most insurers let you pay by the month. Other things being equal, we would generally expect to see the insurance portion of the MaaS charges over a month roughly end up being the same as it would cost a privately owned AV car driven the same amount. The same is true of the other cost, parking - most parking costs are assessed monthly, either in the form of apartment rent or monthly parking at a business destination.

Where it might show up is in the purchase price of the car - but even there, consumers (typically) have a leasing option, which allows you to spread the cost of only consuming a certain number of miles of the cars’ useful life over the time you expect to drive that number of miles. Typically 36K over 3 years, but there are often options. We would expect that the depreciation component of the MaaS charge over a given month’s worth of rides to be relatively close to the monthly cost of leasing the same AV over time.

Yes, MaaS would let you pay daily instead of monthly. But that’s a pretty modest benefit. MaaS does certainly lets you avoid frontloading costs if you purchase, rather than lease, though.

And let’s not forget that MaaS companies are not charities. They’re in it for a profit.

There may be a bit of savings on the financing, and potentially some savings in the cost of the vehicles. The MaaS company can probably buy the vehicles at a discount that regular consumers can’t get.

But those should disappear into the profits of the company.


On average, the major cost of ownership is depreciation. On average, the primary contributor of depreciation is age (it is frequently said that the moment you drive your car off the lot it has lost 10% of its value). In other words, a significant part of the cost of ownership is incurred even when the car is just sitting there.

This means that for a MaaS vehicle, the depreciation caused by time is paid for by all the users. This means that each user pays only a fraction of that depreciation cost. If MaaS use for that user is sufficient to replace ownership of a car, that is a substantial savings.

As with time depreciation that insurance cost for a MaaS vehicle is spread among many users.

Once again, the cost of parking the MaaS vehicle when not in use is borne by all users.

In short, the economic benefit of MaaS over a private vehicle is that several significant costs that are not dependent on amount of usage (i.e., depreciation by time, insurance, parking, AV software updates/connectivity) are shared by the collection of MaaS users.

This notion of shared costs is why for most people flying on a commercial airliner makes more economic sense than owning their own plane.


Age and use contribute to depreciation - which is why mileage (and the condition of the vehicle, not just its age) is such an important part of determining the value of a used car. But your MaaS provider is going to be experiencing much of that “time” depreciation as well (as you point out, the first 10% comes in the first moments of ownership, which the MaaS provider will experience more often than an individual buyer). There’s going to only be a small amount of “time” depreciation savings to spread around - much of it will get eroded by the “harder usage” increased depreciation of having lots of unsupervised non-owners using the car.

I fully concede that if people use their cars the way most people use planes - only a handful of trips per year - then MaaS will absolutely make sense for them. But almost no one does. Nearly everyone who owns a car will use their cars more than a few times per year.


Not only is the MaaS vehicle likely to cover a lot more miles in any given time period than the privately owned car, I’ll be the insurance rate will be higher for a vehicle used for large numbers of disparate riders.

Yes, it does. Otherwise, the business would not be able to survive.

Purchase price: Retail vs production cost = HUGE price difference. In addition, the retail version of a vehicle is FAR less durable than would be an industrial/commercial version of the same vehicle. Plus, the commercial version of the vehicle would have a far longer expected usage life because it was designed for that purpose. Vehicles that are designed to be produced for only 1-3 years (made and sold to retail customers) have a far shorter average useful life than a vehicle designed for 15-20 years of service with ongoing maintenance. Business is allowed depreciation on ALL its assets, including vehicles owned. Individuals are NOT allowed to take depreciation unless the vehicle is used for a business purpose. Hint: A relative was a manufacturers’ representative for OEM auto parts. His territory was Minnesota through Montana. He did 30+k HIGHWAY miles per year in that territory and he bought a new car (higher-end Oldsmobile or Buick) every 3 years or so. The dealer LOVED him. Guess why.

Insurance costs: An individual insurance policy depends on a variety of factors, but the insurance company has to cover tens of thousands of INDIVIDUALS and VEHICLES = $$$$$$$$. A commercial insurance policy, per vehicle, is far less expensive. Plus, the MaaS operator has to essentially be able to PROVE to the insurance company–AND THE GOVT–their MaaS vehicles will be significantly safer than a comparable number of non-MaaS vehicles. Otherwise, why would the public bother with MaaS? SAFER = LOWER COST OF INSURANCE (big $ savings there).

Space savings: If you don’t have to park, then you don’t have that cost (anywhere). No parking expenses. More house because less required protected parking at home for that valuable necessary vehicle. No parking tickets. No speeding tickets. No DUI/DWI costs. MaaS AVs do NOT need to be stored somewhere. The ONLY time they are NOT on the street is when they are being serviced and/or cleaned. With an EV, it can just sit (on the street is fine) and wait for a customer to summon it. While waiting, there is virtually no operational cost to the MaaS operator. Remember: “Wait time” is included in the overall system calculations. Vehicles will be used far less at certain times, so that “wait time” is included as a “cost of doing business”. Overall vehicle utilization/efficiency will be far higher overall, thus allowing the MaaS operator to offer better service at a lower price than a similar number of privately-owned vehicles.

Ride summoned similar to a Lyft/Uber, but one difference. Vehicles are positioned near where historic demand has happened, so they are usually already in the area–and it is just a few minutes to wait. Faster than walking through a parking lot, etc. There could even be dedicated fobs for each user. So no looking up stuff to call. Just hit the button and the info is sent–and you get back the vehicle ID and when it is expected to arrive. Vehicles will be located near shopping, etc–so, short wait times.

There could be other uses for MaaS. Dedicated vehicles designed to handle deliveries for stores or commercial delivery services (UPS, FedEx, USPS, local delivery for most businesses, etc). Guess how fast businesses could grow when their “reasonable delivery cost” territory is the entire MaaS network. Across town is no big deal any more.