Semiconductor companies, advanced processes and private equity: it all fits together

As process geometries continue to shrink, the semiconductor industry faces a growing—some would say a disabling—dilemma. Each new process node demands tighter integration between chip designers, library designers and process engineers. Similarly, the increasing integration levels possible with each new node have erased the distinction between chip architects and system architects, requiring the chip design team to be closer to the systems designers and the final product distribution network. These forces strongly favor the old IBM style of vertical integration.

Working directly against this trend, the increasing capital demands of the semiconductor technology, coupled with the growing maturing—and hence lower earnings growth—of the electronics sector as a whole—put intense pressure on industry executives to disaggregate. The markets reward managers who keep a draconian eye on return on not just basic financial measures like earnings and return on investment, but upon more sophisticated issues including leverage, asset utilization, and perceived volatility of future revenue streams. All of these factors push management to break companies up into small units, each of which is financially homogeneous: manufacturing in one capital-intensive business with relatively stable revenue streams, design and development in another business with less capital demands, less even revenue flows but valuable IP, and research in yet another business where it can be quietly buried and forgotten.

But the disaggregation that makes sense in the equity markets makes no sense at all from an operating standpoint, because it separates the very people who have to work most closely together. Therein lies the dilemma.

Ironically, private equity has proved to be an exacerbating agent in the problem, just when it can offer the tools to solve the underlying dilemma. Looking at the examples of NXP and Freescale, for instance, suggests that private equity owners are just as likely as public investors to break an integrated company into pieces for financial reasons, outsourcing that which can be expensed, separating businesses with different capital needs or distribution models, and dumping anything with long-term return.

This is tragic, because the same financial innovations that have fueled the growth of private equity in its pursuit of maximum leverage have produced just the right tools to deal with this dilemma. The foremost of these is called the Special Purpose Entity, or SPE. (For instance, see here.) An SPE is a unit within a corporation that has been financially separated from the rest of the business to isolate a particular pool of assets, a particular cash flow, or some such. SPEs can protect a stable cash flow from the cyclic or random variations of more volatile business lines, for instance, or they can protect a particular asset from other risks—even bankruptcy of the corporation.

SPEs are used primarily to create complex securities—loans that are secured by the assets or cash flows protected within the SPE. This allows the owners of a business to use just the right financial instruments to leverage each of the company’s key assets and cash flows.

Generally, SPEs are regarded as tools for increasing leverage while decreasing interest costs compared to traditional equity or debt financing. But in the specific case of the electronics industry, they may have another very important function. SPEs could allow semiconductor companies to remain integrated for the sake of operations, but still have access to the advantages of disaggregation for financial purposes. In the current global environment, where financial performance is a matter of survival, but adaptation to advancing technology is equally vital, this might just be a key.