OneWedge

I read today an article on McKinsey’s automotive blog which made me raise my eyebrows: in extreme synthesis it says that making an EV is much more expensive than making a same category ICE car.

The conclusion I agree with, but I have doubts about the cost breakdown (in McK’s parlance “cost walk”) which is included in the article:

Specifically, I think there are items which should be corrected:

Value of ICE content to remove

The value of $3,000 seems grossly understated, in view of the removal of the following subsystems:

• engine block
• distribution system (crankshaft and valves)
• cooling system (radiator, coolant circuit, pump and filter)
• exhaust system (mufflers and pollutants segregation)
• gearshift and clutch
• lubrication system (circuit, pump and filter)
• fuel system (tank, circuit pump and filter)

Battery size

I believe that a 50 kWh battery size is insufficient for a C-segment car, which should at least carry 60 kWh, allowing for a 360 km realistic range. As an example, in the very popular CUV segment which is increasingly replacing the C-segment, the Hyundai Kona is available with both a 39 and 64 kWh battery pack; despite a significant price differential, some 75% of sales are of the higher-capacity pack.

This may abate as the charging network becomes more ubiquitous, but we must remember that refueling an EV will always take 30 minutes vs. 5 minutes for an ICE car, causing people to always retain a preference for higher ranges to avoid the inconvenience of intra-day charging.

Difference in direct costs due to lower volume

I believe it’s reasonable to assume that ALL electric / electronic EV-specific content suffers from volumes which at the moment are abysmally lower than ICE-specific content. This apparently has been taken into account for the cost of batteries but has not been factored in the cost of power electronics and e-motor(s). Additionally, no mention is made of the complex regen system which ensures a portion of kinetic energy is recuperated when slowing down [EDIT: a reader suggests it’s not that complex, after all]

Difference in indirect costs due to lower volume

Again, this seems significantly understated: essentially an EV is designed from scratch without the ability of carrying over previous development efforts: as Volkswagen is proving, EVs need brand new platforms which do not exist today. Developing these platforms requires new skills and competencies which must be recruited, bought or developed, leading to huge indirect costs.

A revised cost walk might look like this, leading to a much higher cost differential between ICE and EV:

Obviously my numbers are only rough estimates (and so are, I suspect, McKinsey’s) but at least overall they agree more with what we see in list prices. If it were true that a medium-spec C-segment EV costs $35k to make, the Hyundai Kona retailing for $50k would carry a $15k gross margin – which I doubt; my estimate of a null GM seems more realistic and would explain why manufacturers are so reluctant to ramp up production of EV: they lose money every time they sell one!

As for the second part of the McKinsey article, I found it a bit verbose to simply lead to a rather obvious conclusion. The fix to make EVs profitable is composed by two measures:

• cheaper batteries – halving the cost of the battery pack
• higher volumes – eliminating the two $4k cost corrections and driving down by $2k the cost of electronics

The aggregate value of those measures reduces the building cost by $16k, bringing it within spitting distance of the benchmark ICE cost. The rest is minute process and supply chain optimization.