Since WW II corporations in the U.S. have gone through several "periods" in which the nature of the corporation was redefined. First we had simple companies that produced mostly industrial products and the bare consumer necessities.
The first noteworthy change was the conglomerate period in the 1960s where acquisitions of any kind added value. High PE (price-earnings multiple) companies purchased low PE companies but the added earnings plus cost savings were erroneously valued by the market at the higher PE ratio.
Then in the 1980s the first private equity firms appeared to do LBOs (leveraged buyouts). In the early days the PE firms were able to buy undervalued companies, strip out and sell unnecessary assets and earn a nice profit selling the remaining company.
As the undervalued assets disappeared, the private equity firms switched to doing rollups. Buy a platform company in an industry and then buy multiple companies in the same industry. This strategy produced scale, economies of scale and administrative cost savings.
Today we see a trend toward companies themselves (with no help from Wall Street) outsourcing operations to reduce costs and focus on the areas that are critical to creating value for the customer. More on this theme in this article from Danielmiessler.Com, "The Future of Renting vs. Buying".
What all of these examples make clear is that the concept of a corporation changes to address the opportunities to extract additional value. As Ronald Coase made clear in his Nobel Prize winning work, corporations are formed as a means to reduce (transaction) costs. While this logic was originally applied to understand integration efforts in corporations, starting with the LBO phenomenon we see corporations basically decoupling functions and assets to lower costs. This trend toward decoupling will only accelerate in the years ahead as we become comfortable with using IT to manage and control a wider and wider range of remote third party functions.