Surviving the Semiconductor Industry - NXP Semiconductors

Introduction

September 12th, 2008. NXP semiconductors announced plans for a tough "redesign program that will bring NXP to a healthy financial situation and position the company for future growth". Chief Executive Officer Frans van Houten commented, "Measures include increasing the competitiveness of our manufacturing base and reduction in our work force, resulting in a leaner, customer focused company, well positioned for growth in our core businesses.” [1].

Since then, NXP has been true to its word, announcing the closure of 4 fabs - with the restructuring estimated to "affect" 4 500 people [2].

NXP has also been involved with a number of joint ventures and acquisitions, radically transforming the company. In an interview with regards to a joint venture with STMicroelectronics, the CEO can be quoted as saying, "The additional proceeds of the 20 percent stake will enable NXP to further build leadership positions through innovation and investment in NXP's core businesses. But, wireless was NXP's core business. Now, I'll have to re-learn who NXP is" [3].

It is clear that NXP's owners (a collaboration of private equity firms who bought the company from Phillips in 2006 [4]) have a very different idea of what NXP needs to be in order to remain competitive. In order to be number 1, massive change needs to take place.

This article will give the authors opinion into how the business model and strategy is adapting through the interpretation of recent events.

Private to Private Equity - Corporate Structure

Semiconductors no longer had a place within Phillips [5] and in 2006 sold the division to a series of private equity firms. For whatever reason, Phillips most likely at some point realized that separation would allow both parties to perform better because it could no longer support the development it required, at which point decide to focus on "Healthcare and Lifestyle" [4].

Phillips Semiconductors on the other hand, saw the potential benefits of remaining private (over becoming public). Private firms have certain freedoms by not having to report financial and regulative data back to the public (or in less detail). They can make and implement decisions quicker and can typically be more secretive - which is beneficial in a market where innovation wins all [6].

This is a prime example to show how the organisational structure of the company is actually restricting its capabilities. In order to advance, they must must break away and re-form; with a more focused business strategy, the capability of the company in this case increases.

At this stage NXP appears to be in a very positive position to become the company it needs to be. Separation has made it 'lighter' opening opportunities that could have blocked previously. It can now make the (potentially risky) transformations it deems necessary .

In a fabless world, who will make actually make components?

The 90's saw the emergence of a new, "fabless" business model for the semiconductor industry (QUALCOMM, Broadcom, Nvidia, ATI Technology, Xilinx and Altera are all fabless) [7].

The closure of the 4 fabs and plans to outsource all processes past 90nm heavily suggests that NXP aims to become fabless as well. But why the sudden change in attitude when it used to be considered beneficial to be able to develop them yourself [8]?

The reasons are numerous and complex, but I feel can be summarised by how sustainable in terms how long, and how much effort it requires, for companies to maintain that their process technologies are ahead of the rest. The Semiconductor industry has come to a definitive point, where due to the runner costs and huge investments required, it is no longer sustainable and the companies no longer capable, to fabricate components themselves (references 7 and 9 highlight the issues involved in more detail). With TSMC, why go against them and be slower to market [8] ?

The company NXP plans to out-source to is Taiwan Semiconductor Manufacturing Company ltd (or TSMC). The scale of these foundries is colossal - to put it in perspective: in 2006, 720 fabless companies were served by only 15 independent foundries [7] (also see reference 10 for some more impressive numbers). Therefore it will be specialised companies (like TSMC), which have the knowledge, money and support that will make the components.

Is it good business strategy to sell your key business?

It is somewhat intuitive to associate a business's future success with becoming larger. As businesses plan to take more of the market, they must expand in order to meet its ever-increasing demands. From points mentioned in this article so far - it would appear NXP is taking a different approach (as indicated by its plans to become fabless and sale of its core business).

The reason for the sale of its wireless business is difficult to analyse. Was it a key move to get out of a too-competitive-market [11] or simply to raise some money to ease its debt troubles[12]? Whatever the reason, it is clear that NXP is trying to get more out of fewer 'pies'. Sustainability, from their perspective, comes when you are number one (for wireless it was position 6th [11]) - which arises though innovation. NXP feels it can longer waste time, money and resources in areas where it knows realistically it can not be number one.

NXPs aim to be more 'sustainable' can be further supported by looking to its recent joint ventures. Both Moversa (JV with Sony) and NuTune (JV with Thomson) [13] are companies that overlap with its current businesses - it clearly has no intention in loosing its number 1 position in these markets.

The link between Ownership and Profit

Perhaps another way to interpret the "definitive point the Semiconductor industry has reached" is to say that large scale ownership is no longer an advantage for the majority. NXP is trading ownership, of its fabs and some business areas for (hopefully) more profit.

By going fabless, NXP plans to save $550million annually [14]. Additional, none liquid assets will gained: it is no longer diluted by the day-to-day maintenance and cost and no longer has to plunge millions of dollars into research (NXP has left the Crolles2 chip alliance [15]).

Conclusion

It can be seen that NXP is taking a "consolidation of its eggs into fewer baskets" strategy in order to survive. On paper it seemed like the correct thing to do, NXP is putting up its hands and admitting, "we can not be the best in everything" and is focussing on what it is best at - although not everyone agrees. Stepping down recently as CEO van Houten has been criticized as being a "conventional management type doing conventional management things" [16]. In April this year Kohlberg Kravis Roberts & Co (who owns an 80% stake in the company) announced that the market value of NXP has droped "to 10% of its purchase price", which was $8billion approx [17].

Undoubtedly the company has been placed in a very precarious position; its investors will certainly have to wait for any returns. With the company in such a state it must be temping to "sell what remains" [17] however, after such a reformation, surely this was expected? As long as its investors are willing to hold on for a while longer, sit out the economic slump, and return focus to its products and customers, there should no reason why NXP should not make a steady return.

Fin.

References
1. BusinessWire.com. 2008. Redesign of NXP Semiconductors Organization. [Online].
2. By Mark Osborne for Fabtech.org. 2008. NXP Semiconductor closing four fabs [Online].
3. By Will Strauss for wirelessdesignasia.com. 2008. ST-NXP Wireless + Ericsson: Our Thoughts. [Online].
4. nxp.com. 2006. Philips to sell majority stake in semiconductors business to private equity consortium KKR, Silver Lake and Alpinvest (press release)
5. By Frans van Houten for NXP. 2006. NXP: Creating Value [Online].
6. By Bill Medley for courierpress.com. 2007. Private ownership has its own risks and rewards for these companies. [Online]
7. By Deepa Doraiswamy for frost.com. 2006. Fabless Semiconductor Model – Shaping the Semiconductor Industry’s Future! [Online]
8. By David Manners for electronicsweekly.com. 2007. NXP going fabless after 90nm. [Online]
9. By Anne-Francoise Pele for eetimes.com. 2008. NXP's restructuring was triggered by wireless JV, world economy, says CEO. [Online]
10. BusinessWire.com. 2000. TSMC Announces Grand Opening of Fab 6. [Online]
11. By Anne-Francoise Pele for eetasia.com. 2008.NXP seeks leadership in remaining units. [Online]
12. By David Manners. 2009. NXP debt swap is successful. [Online]
13. nxp.com. 2009. About NXP. [Online]
14. By Ann Steffora Mutschler for NXP. reorgs for $550M in annual savings. [Online]
15. By Yoshiko Hara for eetindia.co.in. 2007. NXP to outsource IC production beyond 90nm to TSMC. [Online]
16. By David Manners for electronicsweekly.com. 2008. Frans van Houten Leaves NXP. [Online]
17 By David Manners for electronicsweekly.com. 2009. What are NXP's problems?. [Online]