Automotive foundries
Lots of stuff has happened recently -- I took the last final exams of (hopefully) my entire life, I graduated from law school, I've begun preparing for the California Bar Exam, and in the process of doing that, realized how utterly inadequate some of the Bar courses I took were.
But what I'm writing about here, though, is an idea that dawned on me about 2 weeks ago when I read that General Motors had sold the Saturn brand to Roger Penske. Now, I don't know Roger Penske, but apparently he used to be a race car driver and now owns a bunch of car dealerships. I assume he also is the namesake behind the Penske line of rental trucks.
Anyway, I own a Saturn and it's not a terribly bad car -- although I did have to fix quite a few things on it within the first 2 years of ownership, and I also have to routinely ignore a bunch of glaring faults like how noisy it is, how underpowered it is, and how its fuel economy is only okay given the power. What drew my attention with the Penske deal was that he's going to keep the brand alive, but seek out contractors to make the actual cars. In other words, Saturn will be just a brand and a retail network of dealers and parts distributors. The actual cars will be made by someone else.
This sounds unusual, but its not unheard of -- the BMW X3 is actually not made by BMW, but rather by a subcontractor is Austria (partially owned by the Canadian company Magna, which now is co-owner of General Motors' Opel brand in Europe). Until the economy turned to shit in 2008, I believe Porsche outsourced some production of its Boxsters to a contractor in Scandinavia -- Finland, I think.
All of this got me to thinking about the foundry model by which most every computer or electronic gadget is made these days. Years ago, I assume that every company who had chips or electronic gadgets to make designed it and build it at facilities they owned and operated. Starting, I think, in the late 1970s and early 1980s, specialized contract factories began opening up throughout Asia (mainly in Taiwan, but now in China too) that do nothing but reduce to reality a design that you provide them. Taiwan Semiconductor Manufacturing Corp (TSMC) is a major foundry. SMIC is another one -- I forget what it stands for, though, sorry.
For better or worse, the bifurcation of manufacturing from, well, everything else (sales, design, etc) has made electronic gadgets cheaper and made it possible for more companies (smaller ones, typically) to enter the market. From what I've heard, lots of these smaller companies are startups with only a handful of employees -- basically a company that could in no way afford to own or build its own factory.
A very similar parallel can be drawn to the Outsourcing business model that places like India, China, and the Phillipines are known for. In other words, by providing standardized and inexpensive computer programming and call center service, costs go down and the market becomes more available to more people.
And this got me to thinking: can this same foundry model work for cars? Would it be possible to trifurcate design, manufacturing, and everything else (or birfurcate into manufacturing and everything but manufacturing) so a small startup could be a player in the car business?
There'd be the obvious psychology problems, of course: cars are expensive and are kept for a long time so people don't want to spend a bunch of money on a brand that will disappear tomorrow (cough, Daewoo, cough).
You also have a sheer logistics problem. Cars today may share some mechanical similarity, but are still largely built from very dissimilar parts -- this is obviously why a Toyota Camry looks different from a Honda Accord which looks different from a Ford F-150. A single factory building all of these models would literally be filled to the rafters with parts. Organization would be hugely difficult.
On top of all this, I don't know if the economics of a foundry model for cars would even work since it would be squeezed two ways. On the upside, profit would be squeezed by having more companies playing in the auto industry (i.e. more companies making passenger cars = lower prices for passenger cars). On the downside, you'd have high costs as well, given the parts problem I just described and the massive organizational nightmare that would entail. The organizational nightmare might have indirect costs as well, such as training employees how to assemble dashboards in 50 different ways for 50 different models and the increased cost of repair/rework/warranty if an assembly mistake results.
Although something tells me that Morris Chang and a bunch of the other electronics foundry model pioneers in Asia were having this same thought/discussion about 30 years ago, and that turned out well didn't it.
But what I'm writing about here, though, is an idea that dawned on me about 2 weeks ago when I read that General Motors had sold the Saturn brand to Roger Penske. Now, I don't know Roger Penske, but apparently he used to be a race car driver and now owns a bunch of car dealerships. I assume he also is the namesake behind the Penske line of rental trucks.
Anyway, I own a Saturn and it's not a terribly bad car -- although I did have to fix quite a few things on it within the first 2 years of ownership, and I also have to routinely ignore a bunch of glaring faults like how noisy it is, how underpowered it is, and how its fuel economy is only okay given the power. What drew my attention with the Penske deal was that he's going to keep the brand alive, but seek out contractors to make the actual cars. In other words, Saturn will be just a brand and a retail network of dealers and parts distributors. The actual cars will be made by someone else.
This sounds unusual, but its not unheard of -- the BMW X3 is actually not made by BMW, but rather by a subcontractor is Austria (partially owned by the Canadian company Magna, which now is co-owner of General Motors' Opel brand in Europe). Until the economy turned to shit in 2008, I believe Porsche outsourced some production of its Boxsters to a contractor in Scandinavia -- Finland, I think.
All of this got me to thinking about the foundry model by which most every computer or electronic gadget is made these days. Years ago, I assume that every company who had chips or electronic gadgets to make designed it and build it at facilities they owned and operated. Starting, I think, in the late 1970s and early 1980s, specialized contract factories began opening up throughout Asia (mainly in Taiwan, but now in China too) that do nothing but reduce to reality a design that you provide them. Taiwan Semiconductor Manufacturing Corp (TSMC) is a major foundry. SMIC is another one -- I forget what it stands for, though, sorry.
For better or worse, the bifurcation of manufacturing from, well, everything else (sales, design, etc) has made electronic gadgets cheaper and made it possible for more companies (smaller ones, typically) to enter the market. From what I've heard, lots of these smaller companies are startups with only a handful of employees -- basically a company that could in no way afford to own or build its own factory.
A very similar parallel can be drawn to the Outsourcing business model that places like India, China, and the Phillipines are known for. In other words, by providing standardized and inexpensive computer programming and call center service, costs go down and the market becomes more available to more people.
And this got me to thinking: can this same foundry model work for cars? Would it be possible to trifurcate design, manufacturing, and everything else (or birfurcate into manufacturing and everything but manufacturing) so a small startup could be a player in the car business?
There'd be the obvious psychology problems, of course: cars are expensive and are kept for a long time so people don't want to spend a bunch of money on a brand that will disappear tomorrow (cough, Daewoo, cough).
You also have a sheer logistics problem. Cars today may share some mechanical similarity, but are still largely built from very dissimilar parts -- this is obviously why a Toyota Camry looks different from a Honda Accord which looks different from a Ford F-150. A single factory building all of these models would literally be filled to the rafters with parts. Organization would be hugely difficult.
On top of all this, I don't know if the economics of a foundry model for cars would even work since it would be squeezed two ways. On the upside, profit would be squeezed by having more companies playing in the auto industry (i.e. more companies making passenger cars = lower prices for passenger cars). On the downside, you'd have high costs as well, given the parts problem I just described and the massive organizational nightmare that would entail. The organizational nightmare might have indirect costs as well, such as training employees how to assemble dashboards in 50 different ways for 50 different models and the increased cost of repair/rework/warranty if an assembly mistake results.
Although something tells me that Morris Chang and a bunch of the other electronics foundry model pioneers in Asia were having this same thought/discussion about 30 years ago, and that turned out well didn't it.
Labels: business
posted by Andy!! at 11:55 AM
0 Comments:
Post a Comment
<< Home