Getting Paid More

Consolidation is a regular news item in the semiconductor industry, and has been for years, but most of those deals have been relatively small.

What’s changing is the amount of consolidation involving big companies, fueled partly by a massive M&A fund in China, partly by an arms race in preparation for the IoE, and partly by the kind of thinking that if other companies are doing it, it’s dangerous to be left behind.

Back in 2000, when Texas Instruments bought Burr-Brown for $7.6 billion to pump up its analog portfolio, most industry watchers were stunned by the size of the deal. Compare that to some of the deals being proposed today, though, and the TI purchase looks rather modest in retrospect—Tsinghua’s $23 billion bid for Micron, Avago’s $37 billion bid for Broadcom (which notably is $4 billion above Avago’s market cap and more than six times its annual revenue), NXP’s $11.8 billion bid for Freescale, and Intel’s $16.7 billion bid for Altera.

There are three assumptions driving this latest round of deals. At least two of them are correct.

1. More resources are better. It takes scale, both in terms of expertise and R&D, to win market share at advanced process nodes. That’s where the biggest, lowest-power, fastest chips are being developed for mobility, server and networking. With NREs approaching $150 million for some of these chips, not including verification, masks, tapeouts and packaging, one misstep would be fatal for a midsize chipmaker. It would take multiple missteps to kill off a much larger company. And presumably there would be more experts around to help make sure nothing does go terribly wrong, and to be readily available to make changes in order to guarantee customer satisfaction.

2. Time is on the side of bigger companies. It takes staying power at high-value older nodes that are used for automotive, medical and other regulated markets because the development and the subsequent approval process can be drawn out over a period of years. Being able to offset those investments with other business units, until revenues begin pouring in from the new efforts, is essential.

3. Less competition will improve leverage in deals and improve average selling prices. Historically this makes sense, but it’s also the big gamble in this round of consolidation. Consolidation has worked reasonably well for surviving companies in the automotive, health care and financial services businesses, for example. But in semiconductors, the question is not whether consumers will pay more, but whether companies such as Apple, Samsung, Google, and a slew of startup companies building devices for the IoT will perceive there is added value—or at least stronger leverage on the part of chipmakers—to boost ASPs. And whether new markets will develop, either involving those systems vendors or new companies, that can generate the kinds of volumes seen first in PCs and then in smart phones.

There are other risk factors, too. Deals of this magnitude are always difficult to digest. There are business process and culture integration issues, and there is enormous pressure by shareholders to begin generating revenue quickly. There are synergies to find and build (they do take time to implement, even when they are obvious), and there are redundancies that cost money to eliminate. And any blip in the economy, whether it’s China’s stock market gyrations, Greece’s seesawing bailout plan, or the U.S. Federal Reserve’s decision to tighten capital and raise interest rates, can have a huge impact.

This is a very high-stakes poker game, with huge antes and potentially all-in calls. But as with all high-stakes games, not everyone can win. What remains to be seen is who made the best bets, how long they will last—companies don’t stay on top as long these days as they did even 25 years ago—and what the fallout will be for the rest of the semiconductor industry.