While it will take some time to understand the impact of the recent US Tax cuts and Jobs act on the Semiconductor industry, the opportunity for US corporation to repatriate income certainly has made a splash. Traditionally, US semiconductor companies pay very limited tax at home. The majority of the revenue and investments are made in other jurisdictions that often have much more attractive tax regimes than the US corporate tax system.
The top 25 US semiconductor companies have increased the tax provision from 2B$ in Q3-17 to 16B$ in Q4-17. With a repatriation tax of 15.5%, this indicates that 80B$ is being brought back into to the US. Almost all of it is by 4 companies: Intel, Qualcomm, Western Digital and Texas Instruments.
What to do with 80B$
The intent of the Tax cuts and Jobs Act is to bring back investments and jobs to the US through a lower corporate tax rate. Undoubtedly this is also going to impact semiconductor revenue and jobs although it is unlikely that it creates significantly more manufacturing jobs in the US as the current immigration laws make it difficult to attract the right kind of skills. There are many more reasons than the tax rate that decides where to place a manufacturing facility.
All of the top 4 semiconductor companies are in a shape where they already can fund their own investments and return surplus capital to their investors. The question is what they are going to do with 80B$ in the US? We believe that we will see a couple of significant acquisitions in the semiconductor space involving US-based semiconductor companies, not only as a result of the cash but also it will be easier to get regulatory approval for two companies originating in the same triad.
What to buy for 80B$
If you have to buy a competitor, you might as well buy one that grows revenue. Either by selling attractive products or selling in attractive markets – or even better: Both!
Currently, there is one very attractive product area in the semiconductor market. Memories have grown from under 27% of the total market to over 37% in only 7 quarters. Certainly, some of this growth is based on inflated pricing that will return to a lower level at one point in time, but the memory market is a lot different from the last memory pricing cycle. From a market perspective, the two main growth areas are Smart Phones and the Datacenter, in particular, the hyperscale cloud datacenter. Part of this growth is based on increased memory pricing and will be reversed at a point in time. What is different in this memory cycle is that the data center has been able to absorb the increased memory pricing like no other application before. While PC and Smartphones are impacted directly by consumer decisions, the memory consumption in the datacenter is an investment decision. And you can be pretty sure Apple and Amazon are not going to run out of storage space.
Although it is not the optimum timing to buy a memory company at the peak of the memory cycle, we believe that memory is now a strategic component at the same level as processors used to be and it will be a vital growth area in the future.
If you were to acquire a US-based memory company, you would end up with a very short list, or with insignificant players. Memory revenue from US-based companies is concentrated in three companies: Western Digital, Intel and Micron. We believe Micron is the next large acquisition target and we believe that it makes a lot of sense for growth challenged Intel to attempt to become a DRAM company again. This aligns well with the company signalling a strategic change and more risk-taking.
Intel has a very clear market strategy, here illustrated by the 3 Horizons framework. Defending its two core markets, PC & Server, win the emerging Datacenter & Cloud market and then pursue future bets.
Eventually, Intel has a target to become a “Data” company, even if it will be true to its manufacturing roots, most of the future revenue could be generated by data, insights and knowledge.
Micron is a good match for Intel in all of its strategic markets. It has higher growth rates (some price-driven) and meaningful revenue in each market.
Micron is also sufficiently significant to make a difference on Intel’s numbers and they are perfectly aligned with Intel’s target markets and several large memory customers like Amazon and Apple are also repatriating a large amount of cash and could create a bidding war should they decide to engage. With so many potential buyers, we believe the chance of Micron coming under offer in 2018 is high.
While Qualcomm is unlikely to make a move given the current Broadcom battle and NXP integration it is likely we will see TI back in the M&A game after many years of absence. And there is a lot to choose from.
Should you be interested in our data and insights please contact Claus Aasholm for more information