When you were young and your Dad--and/or Mom--took you to a baseball game and you walked into a magical place named Wrigley or Fenway or Yankee Stadium, just after they tore your ticket (in my case it was long before bar code scanners) you were immediately serenaded with the following chant:
“Scorecard! Get your Scorecard here! Ya Can't Tell the Players Without a Scorecard!”
The scorecard was an essential part of the day’s occupation because with the exception of some well-publicized stars many fans were not very familiar with the players on the field. And the same might be said today about companies in the semiconductor industry, given what has become a record year for mergers and acquisitions.
As an observer with a podium on these pages, lately I have been asked what impact I think the recent spate of M&As will have on the industry going forward? Overall, and noting there are several different and significant forces driving recent mega-mergers, my response is that it will have less of a lasting effect than one might think. But before I say why let me review some of the major scoring plays to date, in descending order from very-big transactions down to those that are just big (all figures given in US dollars).
- In the semiconductor industry's biggest merger to date Avago Technologies, which bought storage chipmaker LSI about a year ago, has now acquired Broadcom, a major supplier of integrated circuits for the mobile phone industry for $37 billion ($17 billion in cash and $20 billion in stock). The new Avago will be the world's sixth biggest chip vendor. Avago’s specialty is developing RF components and power amps for use in smartphones. With the acquisition, Avago expects to target such market sectors as servers and data centers and wearable products.
- Intel is plunking down $16.7 billion to buy field programmable gate array (FPGA) maker Altera Corp. The transaction is expected to help Intel bolster its position in server systems (companies have been using FPGAs alongside Intel’s Xeon chips to help speed up their servers) and other equipment found in corporate data centers, as well as fuel its plans to supply chips for the Internet of Things (IoT) market. Adding Altera will allow Intel to explore FPGA/CPU combination devices based on its Xeon products that could give data center applications a speed boost where on-the-fly re-progammaing can pay dividends such as in data base management. Designers of aerospace and defense high-performance embedded computing (HPEC) also often pair Altera FPGAs with Intel processors for signal processing applications including eletronic warfare (EW) and signals intelligence (SIGINT).
- NXP is acquiring Freescale for $12 billion. With $10 billion in combined revenues the deal would make the two companies the world's ninth largest chip maker, surpassing competitors STMicroelectronics and Renesas at $7 billion apiece and approaching Texas Instruments’ $12 billion, according to financial analysts. The new NXP would be number one in automotive chips and according to the company the addition of Freescale’s computing chips to its security and wireless communication offerings will make it a stronger player in the emerging IoT market. NXP was formed in 2006 when a consortium consisting of Kohlberg Kravis Roberts & Co. (KKR), Bain Capital, Silver Lake Partners, Apax and AlpInvest Partners NV took an 80.1percent stake in Philips’ semiconductor operation with Royal Philips retaining a 19.9 percent interest.
- In December Cypress Semiconductor announced plans to merge with Spansion in a $4 billion stock transaction. The combined company will be a provider of embedded MCUs and specialized memories for embedded systems. It will have the number 1 share worldwide in NOR Flash memories and number 1 share worldwide in SRAM memories.
- In October 2014 mobile chipmaker Qualcomm agreed to buy Britain’s Cambridge Silicon Radio PLC for $2.5 billion to boosts its offerings in Bluetooth wireless technology.
Consolidation in the chip business is not new. We have seen a series of similar vignettes played out often before and usually without much wrenching industry-wide change when looked at over the long term. There are many past instances of the let’s buy each other syndrome. In offering the following few examples I was tempted to use the biblical term “begat” as found in the King James version (where in Genesis Enos begat Cainan and Cainan begat Mahalaleel, etc.), but I won’t just to maintain a level of sanity.
So, for example, Spansion, recently acquired by Cypress, was a former joint venture between Advanced Micro Devices (AMD) and Fujitsu. Cypress itself began in 1982 when a group of engineers left AMD to form the company. Renesas Technology was established on April 1, 2003 as a joint venture of Hitachi, Ltd. (55 percent) and Mitsubishi Electric (45 percent). Next, Renesas and NEC (established in November 2002 by a spin-off of the semiconductor operations of NEC Corp.) combined with Renesas Electronics starting operation in April of 2010.
In 2005 the private equity firms KKR and Silver Lake Partners teamed up to buy Agilent Technologies' chip arm, SPG, for $2.66 billion to form the company now doing business as Avago Technologies (yes, the same one that just bought Broadcom). In July, 2006 AMD entered the graphics market with the acquisition of the GPU manufacturer ATI Technologies for $4.3 billion in cash and 58 million shares of its stock. And as you may recall Texas Instruments acquired National Semiconductor in 2011 in an all-cash transaction of about $6.5 billion.
With all of this consolidation—current and historic-- one might be led to believe that market share would become much more concentrated among the top 20 industry suppliers. But statistics do not bear this out. The top 20 suppliers of semiconductors in 2014 represented 66.7 percent of the total industry revenue; in 2010 the top 20 represented 65.2 percent and in 2006 it was 63.7 percent. (Source: iSuppli).
A good question is why all of this activity now? Some industry watchers place the current M&A trend squarely on the fact that it is a means of growing quickly, of increasing sales and boosting market share in an era of relatively modest growth, at least by semiconductor industry standards. Other analysts say the wave of deal activity is a direct result of the rising cost of developing new chips and the subsequent cost barrier to entering and succeeding in an applications market you are not currently playing in. It can cost $100 million to produce a leading edge chip these days, which is a 3x to 10x jump over similar costs a decade ago.
When I said earlier that the current spate of mergers would not have a severe impact on the industry I am not forgetting the human element. Yes, there will be layoffs. For example, when Cypress Semiconductor merged with Spansion it triggered a layoff of 1,600 employees. What will happen to these and other employees caught in what the British so expressively refer to as “redundancy”? Very likely most of them will land on their feet primarily because chipmaking is still a very healthy business, even though its executives bemoan the fact that the industry is growing at just 5 to 7 per cent a year. Admittedly this is less than the 13 percent 20-year annual average growth of what is now a $300 billion industry, but there is no reason to believe its 40-year track record of growth will come to a halt any time soon.
According to the Semiconductor Industry Association the industry accounts for a quarter million direct jobs and over a million additional indirect jobs. Last year job website CareerBuilder said it had 3.5 electrical engineering jobs posted for every job-seeking candidate. Earlier this year Randstad US, the third largest staffing organization in the United States, put engineering and manufacturing positions on its US “Hot Jobs” list. So jobs should be there for engineers finding themselves caught in an M&A squeeze.
The semiconductor business has come away from previous consolidation periods relatively unscathed just as it has withstood hiccups such as the bursting of the Internet bubble in 2001 and the financial shemozzle of seven years ago. In fact the chip business has come back from our national financial crisis of 2008 and 2009 at a much faster pace than the overall economy. To put things in perspective consider that Gross Domestic Product (GDP) in the United States rose just 2.2 percent in the fourth quarter of 2014 and actually dipped by 0.7 percent in the first quarter of 2015. By comparison, then, if presented with a steady 5 to 7 percent growth employees and executives in most other sectors of our economy would be breaking out the bubbly.
So who’s next? Reuters reports that with its CEO Steve Laub due to retire at the end of August, Atmel is being shopped around. The news wire's anonymous sources say investment bank Qatalyst Partners has been engaged as a consultant for the $4 billion company. In 2008 ON Semiconductor and Microchip Technologies launched a hostile takeover bid for Atmel but the offer was eventually withdrawn.
There are also a great many analog chip companies out there, an estimated 150 or so making a variety of chips that among other things manage power and allow wireless communication. Those in the know in the financial community suggest that among suppliers that roughly fit the target size and product model for potential deals include companies such as Micrel, Intersil and Silicon Laboratories.
Stay tuned.
