We have not written a full-blown financial, equity review of a company in some time, but this idea has been percolating in our heads for years. For all the activist investors out there – this one is for you.

There are a lot of companies making analog semis in the US today, aside from the mega cap names and a healthy roster of mid-caps, there are hundreds of tiny one out there as well. Put simply, we think there is fertile ground for further consolidation in the space. A leading example of this is in the RF space – Skyworks and Qorvo. Put simply, the market for RF products has done well over the past decade, consolidating down to five vendors today (not counting China where there are a further dozen, a topic for another day). But that five is probably still too many. Qualcomm has already shaken the market up considerably, and while its easy gains seem to be behind them (another topic for another day), there position in the market is powerful enough that over time they are likely to make more gains. There is just not enough room for Qualcomm, Qorvo and Skyworks to all be competing for the same power amplifier and tuner sockets in mobile, an end-market which has largely saturated.

Both Skyworks and Qorvo have been aware of the problem for years. For at least a decade they have gone to great lengths to talk up their non-mobile businesses, and highlighted these areas as targets for growth organically or via M&A. The problem is that despite their glaring need to diversify neither company has made any progress in those efforts. In 2014, Qorvo generated 82% of its revenue from mobile products, today that figure is around 89% depending on how you want to classify IoT products. Skyworks does not even break out its segments, instead providing this handy numberless-pie chart which seems to indicate about 60% of their revenue is mobile related, roughly the same level as ten years ago.

As much as the products for mobile and other markets carry the same names – power amplifiers, etc. – these are very different businesses. To simplify, mobile products are high volume and low margin, the non-mobile products are low volume and high margin. Despite some overlap in product development and operations, we know from first hand experience that the two sides of the companies are in a perpetual state of tension. The non-mobile products need a lot of upfront R&D work, designing chips that are not quite custom for each customer, but fairly close. They have to contend with the mobile side which has to design a much smaller set of products. The mobile products also move very quickly, typically for just a few smartphone cycle-years. The non-mobile parts can sit on the shelves for years. If you add up the lifetime margin accretion of both types, the math pencils out pretty favorably for the non-mobile side, but it can take longer than the tenure of the average cost accountant for the non-mobile products to fully realize those gains. As seen by the unchanging share of revenue, the mobile side often wins the tug-of-war for limited design resources.

We think something has to give. We propose that the two companies should merge, creating economies of scale needed to stave off Qualcomm in the mobile market. Then, they should spin off the non-mobile side of their combined businesses into a separate company.

For scale, these are their revenue numbers. We pulled 2025 revenue from Yahoo! Finance consensus. Qorvo breaks out their mobile and non-mobile businesses, we included their “CSG” segment revenue in the non-mobile as those products sell like the non-mobile side. For Skyworks we just assumed 40% non-mobile based on that handy pie-chart above. For both companies we assumed the mix does not change over the next two years, as it has not changed over the past ten.

After combining, we then split off the non-mobile business. That yields the math below. We valued these companies on Enterprise Value to Revenue multiples. We compared the non-mobile side to ADI, NXP, On and TXN, and the mobile side to MTK and QCOM, which yields the figures below. Today the two companies have a combined enterprise value of $28.2 billion, after our proposed series of transactions the mobile and non-mobile NewCos would have a combined $38.4 billion enterprise value. So $10 billion in value creation, just like that.

Of course it is not going to be that simple. Both companies do a lot of business in China, so regulatory approval for the combination is an open question. And integrating two companies in a merger is always hard, then splitting off half the combined entity makes it all incredibly complex. This would take several years to accomplish.

The key to all of this is splitting the mobile and non-mobile sides, but some will argue that the two sides enjoy considerable synergies from the combination. The non-mobile side will have much smaller unit volumes, which means it will struggle to get the best pricing from foundries like Global Foundries or Tower. But we think the benefits far outweigh those types of dis-synergies.

The mobile side benefits from better economies of scale and better volumes at the foundries. They need this to remain competitive against Qualcomm, and may even allow them to afford some R&D to blunt Qualcomm’s technology capabilities. Customers will not like losing a vendor, but those customers are all sizable enough to still maintain significant negotiating leverage, and we have spoken with a few who would actually like to see this kind of deal to make the non-Qualcomm option stronger. The non-mobile side benefits from having a management team, organization structure, and sales force focused on selling these products. While they would likely lose designers to the mobile side, the combined teams would be larger in aggregate than they are today, better positioning them to grow the business.

All of this is just our rough math, but we strongly suspect that the real model on this deal is even more compelling.