Long-time subscriber, Wilson M., asked me to take a look at mobile chip designer, Qualcomm QCOM. I did some background research and pulled data into the Framework Integrated model, and am posting those materials here.
Qualcomm is in the midst of several major changes. Its primary market – mobile telephony – has matured and is slowing, so the company announced it would acquire European chip firm NXP for $47 billion ($36 billion to buy the stock and another $11 billion to retire the debt).
The announcement, made in October 2016, represents the largest semiconductor industry acquisition in history, and we think might presage greater industry consolidation.
NXP is a leader in semiconductor chips for the automotive market during a time when semiconductor content in automobiles is surging. It is also a leader in semiconductor chips for the Internet of Things (IoT) – the move to include logic circuits and Internet connectivity to appliances and machines. The company’s most recent 20-F (an annual report filed by a foreign company listed in the US – equivalent to a 10-K for a domestic company) is here. NXP’s Investor Relations site, which has past quarterly announcements, is here.
Qualcomm is the most important company in the market for mobile Systems-on-Chips (SoC) and other chips used in mobile telephony. SoCs, like Qualcomm’s Snapdragon chip, incorporates several functions (e.g., WiFi, Bluetooth, signal processing, and application engine) onto a single chip, enabling greater miniaturization and compatibility. Qualcomm generates most of its sales from Chinese, Korean, and Taiwanese manufacturers of mobile phones that in turn produce finished phones for companies like Apple and Samsung. As such, most of its profits are held overseas, so the acquisition of NXP – a company based in Europe – represents a tax-friendly transaction to QUALCOMM, which will not be hit with taxes to repatriate funds to the US to make a domestic acquisition. The QUALCOMM presentation announcing the acquisition is here.
The move into IoT might come at a good time for Qualcomm by virtue of its status as one of the most hated entities within the mobile telephony ecosystem. Qualcomm is, at its heart, an intellectual property company, and its innovations and patents underlie the CDMA and OFDMA technologies that enable 3G and LTE (4G) mobile capabilities. One of Qualcomm’s divisions sell Snapdragon SoCs and other chips (which Qualcomm designs, but contracts out for manufacture); another of its divisions sell other companies the right to use its CDMA and OFDMA patents. In other words, if you want to make a smartphone, you have to pay Qualcomm something – whether you use its chips or not.
Apple does not like to be over a barrel, but as long as it wants to keep making iPhones, it is. It used to buy modem chips (the chips that allow mobile phones to connect to WiFi) solely from Qualcomm , but last year, placed an order for these chips from Intel for half of its iPhone 7 order. However, even after changing suppliers, Apple realized that it was still paying a hefty licensing fee to Qualcomm for its patents.
At about the same time, the Korean trade regulator found Qualcomm guilty of anti-competitive practices, and Apple took the opportunity to file suit as well. It claimed that the requirement to pay the licensing fee even though it was buying half its chips from Intel was unfair. It stopped paying its suppliers the fee due to be passed through to Qualcomm , and Qualcomm countersued.
Part of what rankles Apple so much is that the Qualcomm license is associated, not with the value of the chip sold to the phone maker, but to the value of the phone sold to the customer. Apple contends that people don’t by iPhones because of the great modem technology, but rather because of the design and functionality of the device itself. Qualcomm’s counterargument is that the modem technology provides benefit to the user of the phone that is proportional to the overall value of the phone. Would you buy a nicely designed iPhone if you knew it wouldn’t be able to connect to WiFi or communicate over 4G networks?
The case is ugly, but there is no one in the mobile phone world who would be sad to see Qualcomm taken down a peg or two. Here are some articles about the dispute:
- Apple files $1 billion lawsuit against chip supplier Qualcomm (Reuters)
- Apple Alleges ‘Mounting Evidence’ Against Qualcomm (Bloomberg)
- Qualcomm CEO Expects Out of Court Settlement With Apple (Fortune)
The issues raised here bring up some interesting modeling questions when valuing Qualcomm.
Qualcomm’s organic mobile phone revenue growth is tied to the growth of the proliferation of mobile devices in general, and the use of CDMA and OFDMA technologies in particular. This tailwind is offset by a regulatory headwind related to Qualcomm’s licensing practices. The Apple case brings up the possibility that other large phone makers might revolt against Qualcomm and keep it tied up in the courts for some time (holding back Qualcomm revenues all the while).
Qualcomm’s revenue growth will also be influenced by the NXP transaction. NXP generates about half the revenues of Qualcomm, so the first fiscal year to incorporate NXP’s results will show on the order of 50% higher revenues than the previous year. After merging, NXP’s organic revenue stream will need to be added to Qualcomm’s, but there is some uncertainty there as well. Might it be possible for Qualcomm to extend its intellectual property model to the automotive and IoT space as well? If so, NXP’s organic revenues in the future will look different from those of the past.
Both Qualcomm and NXP are chipmakers, but their businesses differ a great deal in terms of profitability. Qualcomm is “fabless” (i.e., contracts manufacturing out to a third party like Taiwan Semiconductor) whereas NXP builds many (or all – I don’t know NXP’s business well enough to say for sure) of its own chips. From an Owners’ Cash Profits perspective, Qualcomm’s legacy business is bound to be much more profitable than NXP’s legacy business. The new company – Qualcomm + NXP – will have a profit profile that is a combination of both of the legacy businesses.
In addition, there are likely going to be integration costs and the payment of severance to people who hold positions that are eliminated because of the business combination. This will likely push down near-term profitability of the new Qualcomm, but future profitability will rise to the extent that Qualcomm’s management can effectively incorporate the NXP acquisition.
Clearly, investment level for the year during which the NXP transaction closes will be higher by $47 billion – a significant increase over the trailing five-year average of $1.5 billion.
Investment level for the new company will be related to the investment level of each of the legacy firms, but Qualcomm may be rather less likely to spend on investments for a year or two while its management sorts out the myriad of details related to the NXP acquisition.
The acquisition of NXP represents a very large bet on behalf of the management of Qualcomm that the Internet of Things will be a rapidly growing field in the medium-term. We believe that Qualcomm’s medium-term growth will hinge on the success of its acquisition of NXP and its ability to meet the demand for greater connectivity.
We have entered historical data into the model but have not made forward projections for any of the valuation drivers. Qualcomm’s financial statements are pretty clean and easy to get through (though the descriptions of the technology is difficult), but I am happy to answer questions about forecasts during Office Hour sessions.