From a ecomonic/management point of view a totally vertical manufacturing structure (ie entirely in-house) is horribly inefficient, unless you can produce on a huge scale.
This is the reason AS, ETA, PUW, Durowe, Unitas, Valjoux and all the small Swiss watch casing companies pushed into the US watch market, later had their movement being cased up in the US, then absorbed the remnants of the dying US watch industry. The dispersal of manufacturing assets made for cheaper production.
Seiko can have nearly vertical manufacturing of mechanical movements because the group as a whole is large enough. Seiko makes precision instruments, electronics, clocks and similar item, so the facility that makes tiny gears can run at full tilt year round. The same is true for the all the other components of a watch, as many of their other products share technology. They can even forge titianium efficiently, as their sport equipment also use forged Ti products (golf clubs for one).
A watch manufacturer that produces 50,000 watches a year, and wishes to make its own movements vertically, could run a gear cutter about two weeks and have enough gears for the years. Not very efficient use of a machine that cost several million dollars.
Then their is the workman training problem. Which workman is less prone to make errors due to training? One who runs a machine 8 hours a day, 40 hours a week, 48 weeks a year, or a guy that works a machine for two weeks and the next time he sees the machine is two years later?
