The context
Mannerisms - Private Equity Ruins NXP
Mannerisms - Frans van Houten Leaves NXP
Van Houten resigns NXP CEO post, former Agere exec takes seat
NXP's biggest problem is the $6bn of debt loaded onto the company by its private equity owners led by Wall Street company Kohlberg Kravis and Roberts. This $6bn of debt involves interest payments by NXP of around $480m a year plus a capital repayment programme starting next year. NXP may be able to re-negotiate this capital repayment but the terms are likely to be onerous.
Can a change of boss turn round NXP? With CEO Frans van Houten being replaced by Richard Clemmer this week, the question is: Apart from the debt, what else ails NXP? See Mannerisms - Frans van Houten Leaves NXP
In response to the post we have had a number of insights from current and former NXP-ers, and others, which tell a story of diminishing morale, inefficient use of R&D resources, lack of vision and inert middle management. Here is an edited selection of their comments (with commenters identified by their chosen posting name):
The comments
Big Softie writes: "As an ex-NXP'er I have to say I always found Frans an inspirational leader. I believe replacing him at this stage will not affect NXP's future positively. By virtue of the cost cutting that has taken place, and the massive loss of competence already mentioned, what is left of the company is now set on a course that cannot change without significant investment, or divestment. Frans' weakness was in not understanding the disparity that existed between the things he championed as our leader, and what actually took place at the working level....or maybe he knew it but was unable to change it. A disparity that sadly remains as a legacy of the Philips days and culture, whereby there exists a substantial layer of ineffective people promoted to a senior level that abstract the leadership from reality and add no value whatsoever...in fact quite the opposite. Far too many undeserving people ascend far too rapidly to this level through some mystical process which appears to take no account of their lack of knowledge or competence. Unfortunately for the company, the prime goal in life going forward for such people concerns only their own progression. The company morale started to decline during 2007, and over the last 12 months hit a very low level. It is a tough challenge for the incoming CEO."
Maxi writes: "As a current employee of NXP, I personally don't think FvH did a bad job under the circumstances. He focused the investment, cut out a bunch of unnecessary management and replaced more or less every key manager in the company. What he didn't do is lay down a clear vision of where we are going, why this will beat the competition and how we are going to do it. Most dangerous is that employees are now in wait-and-see mode until they find out where they end up. We have had an office sweepstake for over a year now to predict how long the entity NXP will still exist. Even last year nobody wanted a date later than 2010."
Oswald Fulcanelli writes: "As to the remarks about the layer of ineffective managers at NXP. When Frans took over he removed the entire layer of managers reporting directly to him. It turned out that those people could leave without any negative effects to the company."
Utah writes: "As ex-NXPian, I want to share some views. I had attended leadership development programmes where Frans used to visit. I saw him very inspirational. At times, I felt quite proud of being part of such a company. Slowly I started realising that the company was being managed by a bad MT. No vision and no strategy !! Same set of power point were presented Quarter after Quarter. No major change !! The biggest white elephant was R&D. MT never did anything concrete to resolve the major problem of portfolio management. Major R&D investments were never fetching any concrete products / revenue. Yet no one was held accountable !!. "Highway to customer" never worked. There were still silos among various functions and they were fighting internally instead of taking on the external competition. Biggest flop was BU-Home !! Yet people are still pretty , flaunting flimsy wins and no one is accountable for the R&D debacles. Company is infected with politics to the core and victimisation is going on!! Big challenge for the new CEO. Need clean-up from the EMT."
Paul writes: "Frans van Houten did not 'leave' NXP, he was fired. I don't share your optimism about Clemmer. If you read what he has done he is an M&A guy. He is certainly sent in by KKR to save what can be saved and sell what remains as fast as possible. He has worked in semiconductor but as a finance guy, not as an engineer. This is not somebody who plans to move to the Netherlands for a long time. He has a job to do and will finish the job in 6 to 12 months. To Big-Softie, One of the reasons I left Philips Semiconductors (even before it became NXP) was Frans van Houten. I did not find him inspirational at all, all his statements came straight out of management handbooks, I never detected a true vision".
Tim writes: "Having started to work with NXP in the last few months, I can see some changes from the old Philips Semi days, but too many of the 'old guard' still remain...and more importantly, the old thinking. NXP sold their wireless business 'at the peak' and got a good price. Their auto business is in pretty good shape, but if you keep selling the crown jewels, you're left with the box (which is usually discarded, along with the instructions!). Private Equity companies cannot operate in a declining (semi) market, the burden of the debt placed on NXP is crippling - Clemmer is a money man, a KKR man, (a hatchet man?), and as such I don't see any future for NXP."
Fresher writes: "So a new NXP is born (again) ? After two year's getting rid of competence, how do you want to build the future ? Of course there is M&A but up to now it proved to be a way to fill the roadmaps and have something to sell. From inside, most of us are just expecting to be part of a branch that will be sold to someone else. That's how we see the future: outside NXP. Things may change for sure. But it will take time and not sure KKR can afford this. And how will NXP pay back its debt meanwhile ? Either they re-negotiate the debt now, or they sell BUs or both. My bet: automotive & ID sold during first half of the year."
Anonymous writes: "Bottom line improvements in semi conductor companies are largely contributed by optimizing the R&D cost, as the impact of other variables such as cyclicality and rising capital costs are almost uniform for all players. R&D cost in a typical semiconductor company is about 15-25% of the Top line. A 1% optimization in R&D costs translates into a 1.5 M$ savings for a 1 B$ company. NXP is a glaring example where the R&D team and Business team were not actively collaborating. In some of the Business Unit, the CTOs are mere "administrators". They are neither updated with Technology nor aware of Business / market trends. Major activity for these CTOs are budget /resource number crunching !!The biggest challenge for Clemmer (Richard Clemmer, the new CEO appointed to replace Frans van Houten) is to start from R&D team clean-up !! It is common to find only handful of projects contributing 70 to 80% of the company's revenue, while there are a few thousand projects which are being pursued. As a result, companies are unable to take up "disruptive innovation" projects since most resources are engaged in projects of lesser strategic impact."
Tom writes: "The problem is that semiconductor companies have lost the ability to charge for the value of their products. Prices are set entirely by competition rather than value - for example a Bluetooth chip is priced about the same level as two tracks from iTunes. To get prices which reflect value semiconductor companies need to create scarcity and the way to create scarcity in a knowledge based business is through patents. The alternative to enforcing intellectual property rights is that engineering will become a worthless activity that can only be carried out in low cost economies and by open-source hobbyists. There is a lot wrong with the patent system but the central issue is that it is far too hard, for creative companies to get paid a reasonable royalty for access to their inventions. An increase in patent litigation is a healthy sign that the industry has decided it is time that customers should pay what the products are worth."
Big Softie adds: "I believe there is another aspect which deserves inclusion here, and that is the disruptive effect that (embedded) software technology has had on the mass-market semiconductor business model. The unrelenting march of semiconductor integration means that most mainstream applications now consist of one large multi-core SoC device, and a huge stack of enormously complex software. In order to demonstrate and secure design-wins with these massive chips, semiconductor companies are forced into "bundling" significant amounts of software that is not accounted for in the traditional semiconductor pricing model, which is based on materials costs and processing.
The only way that OEMs and product manufacturers can differentiate their own product based on such an SoC, in terms of functionality and feature set is by customisation of the embedded software. Since most are unwilling, or even unable to undertake all necessary software development themselves, the task falls back to the chip supplier. In short, fierce competition in the market forces chip vendors into supplying their customers with bespoke solutions, which have to be maintained over the total lifetime of the customer's product. In return, the customer expects to pay a standard volume-based price, which erodes over time and total volume growth.
Whilst many semiconductor companies are perfectly capable of creating such solutions, none of them seems willing or capable to achieve the revenue from the market that such a solution demands in order to represent a profitable long-term business. A business that supplies and supports customised products over lifecycles of many years at standard high-volume prices is not sustainable; the ongoing support costs are largely unknown, and not taken into account in the chip pricing. We have already seen several semiconductor companies withdraw from high-volume markets where the huge software costs render their business unprofitable. I believe we will see more of this over the coming years….unless that is the semiconductor guys also have the balls to either say "no", or to ask their customers to pay the appropriate price for the products and bespoke service they are getting."
Hw writes: "Just an info for Softie: most of NXP profit comes from simple multimarket chips from BU MMS and NXP is currently moving out of SOC chips, investing in BU MMS activities."
Take part in the debate on Mannerisms, the blog of David Manners, Components Editor on Electronics Weekly. Updated twice daily, it is the authoritative commentary on the semiconductor industry. Sign up for the Mannerisms eNewsletter.