The Problem with Electric Vehicles: They're Cannibals

Mike Smitka

I'm working on a paper for a conference in Torino on disrupters in the auto industry. Contrary to most, I'm not convinced that battery electric vehicles (BEVs), autonomous vehicles or "Mobility 2.0" business models will be disrupters. That's not because I believe these technologies are unworkable. Indeed, I expect BEVs will one day dominate. I define "disrupter" narrowly. I argue (but not here) that existing car companies will dominate, so that the transition will only be disruptive for the portion of the supply chain devoted to fuel delivery and engine components. That's not a small footprint. Even there the impact won't be disruptive, because BEVs will diffuse slowly.

we should look to 2030 for BEVs to go mainstream

The market for hybrids is an example to which we need to pay heed. None of the full hybrids has sold well – with the exception of the Toyota Prius, to which I return below. That is true whether they are in the form of a Chevy Volt with a range extender, or a "traditional" hybrid such as the Honda Accord. As the name suggests, hybrids have two complete drivetrains that must be made to work together. That's not clean engineering, and it adds cost – hybrid vehicles are inevitably priced $3,000 or more higher than comparable regular vehicles. If someone drives 15,000 miles a year, then at $4 per gallon the fuel savings come to only about 125-150 gallons or $500-$600 per year. So unless they are subsidized, either by the government or through rebates by the manufacturer, then for most people a hybrid is not a good value proposition.

Subsidies aren't hypothetical, but in NAFTA, the EU or China, each with 20+ million sales, it's easy to exaggerate their potential. Initial pilot projects have largely run out – those in China, only just expanded to cover hybrids and not just BEVs, are set to expire in 2020. Governments (Congress in the US) rationally aren't eager to spend money to make hybrids a competitive option: at $2000 per vehicle (not $3,000 – lower fuel costs have some benefit!) and 10% of the NAFTA fleet, the requisite subsidies would come to $4 billion per year. That's not going to happen, not in the U.S., not in China.

The key point is that not only did hybrids not succeed, they will not succeed because the cost differential is fundamental to the technology: a downsized internal combustion engine (ICE) costs almost as much as a larger one, while batteries, power controls and electric motor can only add cost. The bottom line is that we have good evidence that when it comes to fuel efficiency, consumers won't buy vehicles unless there's a value proposition.

The exception is the Prius. After a slow couple years, it became the car for the Academy Awards. Was it an economy car that someone would buy because of the savings? No, as per above, there were none. And it was not an economy car; when I last checked it came in 5 trim levels, but don't try to find a base Level I vehicle! Instead it was a trendy statement of environmental concerns, bought at the fully loaded Level V. For that a Honda Accord Hybrid provides no benefits, as it's visually indistinguishable from the regular gas-guzzling version. [I have notes in my office and will fill in names/dates later.]

BEVs face the same challenge as hybrids. They also face a moving target. Over the past 20 years carmakers have added turbochargers, electric steering and other motor-driven functions that eliminate the cost, weight and parasitic losses of always-on hydraulic systems. Start/stop alternators are now diffusing, vastly improving the efficiency penalty of in-city and rush-hour driving. As 42V systems roll out, alternator-motors will be beefed up to provide power boosts at cruising speed and capture power through regenerative braking. Add in better combustion control, and ICEs have another 15 years of efficiency gains ahead of them, and likely more. Even though batteries are falling in price, BEVs will remain niche products.

BEVs are helped today by direct subsidies in China and elsewhere (a $7500 tax benefit in the U.S.). Those face the political-budgetary limits noted above. More important are indirect subsidies as "compliance vehicles," most obvious under California's zero-emission mandate. (The California Air Resources Board is a de facto global standard setter, as the state's 2.1 million unit market is too large to ignore.) That indirect subsidy may expand: as the US moves closer to implementing its 54.5 mpg fuel standard – with parallel measures in other major markets, including China – it may be in the interest of OEMs to expand the number of BEVs they sell, to offset the sales of larger non-compliant vehicles, even if consumers are not immediately interested.

Every BEV sold will mean an ICE that is not sold. How will this cannibalization play out? With the easing of the fleet-wide US fuel economy standards of the original CAFE mandate, companies no longer need to offset fat-margin light trucks with sales of small cars. That is, the latter can now be sold – or not sold! – on a stand-alone commercial basis, as long as each vehicle doesn't overflow it's footprint-based bucket. Money-losing BEVs will thus replace the sales of positive-margin vehicles. The larger the BEV, the bigger the gap: margins in general increase with vehicle size, but the bigger the battery pack, the greater the cost penalty of BEVs.

In other words, until battery prices fall enough to make them cost-competitive with comparable ICE vehicles, each BEV sold will lower OEM profits. That is, as stand-alone products BEVs have to be able to generate the same gross margin as the vehicles that they replace in consumers' garages. My educated guess is that batteries won't reach that break-even point for a full decade. Worse, if those battery systems involve new chemistries, instead of representing incremental improvements on today's lithium-ion cells, the validation and vehicle redesign process could easily add another 5 years. By then the global market may approach 140 million units, so a lot of battery capacity has to be added.

Diffusion will be speeded if driving habits co-evolve: small commuter vehicles with modest range incur a smaller cost penalty at any battery price. They don't sell today. I believe that too will change, but I expect that to be a function of cumulative exposure – it will be an "experience good" in economic terms. Hence my choice above of the term "co-evolve." So my expectation is that we should look to 2030 for BEVs to go mainstream. Only then will a car company^* dare to offer at least one core model with only electric drivetrain as options.

* Boutique players such as Tesla are irrelevant to the industry as whole. New entrants must hurdle barriers that today's big players have already overcome, and face little or no advantages in technology. However attractive their product, they simply can't grow fast enough to make a dent in the global market.