Intel’s Foundry Spin-Off Plan: Another Strategic Misstep
On Friday, Intel shares spiked 9% after Bloomberg published rumors that the company is weighing options to stop the cash bleed, like spinning off the foundry business. I believe that this would be yet another misstep in Intel's mismanagement over recent years.
I already ranted about Intel two years ago, so let’s go for another round.
Intel is losing
Once the undisputed leader in the global chips industry, Intel managed to fall from grace over a decade of mismanagement. The company preferred to put marketing people in the CEO role instead of engineers and to prioritize financial engineering via buybacks and pushing margins instead of innovating. This turned out to be a critical mistake. Intel has lost a lot of market share in its core CPU market to AMD, from 80% to 63% in a few years. However, Intel is still the market leader.
The data center segment paints a much grimmer picture. We can see that Intel dominated the market (note that there was a segment accounting change, so Intel’s graph changed colors between 2019 and 2020). Intel’s data center revenue declined almost 50% since 2020. Meanwhile, AMD almost caught up in just four years, and Nvidia has seen unbelievable growth in its data center segment, growing to $81 billion in revenue in 2024. Intel is losing, so they need to find another way to stay relevant.
I partnered with Finchat.io!
In my last portfolio update, I announced that I had partnered with Finchat.io, a great data aggregator, provider and visualizer. While there are a ton of services out there providing visualization of reported numbers from the income statement, cash flow statement and balance sheets of companies, Finchat goes further and offers data on thousands of specific KPIs and segment data, like I used in this post to compare data center revenue for Intel, AMD and Nvidia. This is highly useful, too, and I’m expected to start this partnership. If you’re interested in the service, consider using my affiliate link to get 15% off a subscription. There is also a free plan to check out if you like the service.
The foundry business
In February 2021, Intel appointed Pat Gelsinger from VMware as the new CEO. Gelsinger started his career at Intel at 18 and became a lead architect for the 4th generation 80486 processor in 1989. Finally, Intel is led by an engineer again! Maybe we’ll see a comeback of the giant. Intel had ambitious plans to start a foundry business to compete with TSMC and Samsung and become the third-largest player in the foundry business. Prior Intel already manufactured its own chips, but they now want to do the same for customers. The problem was that Intel needed to catch up because TSMC and Samsung were ahead on high-end processes.
The idea was great, especially after the pandemic and the following supply chain crisis, when companies started prioritizing geopolitical dependable suppliers. This means that customers would prefer chips produced in the USA compared to Asia. Furthermore, governments worldwide introduced large incentives like the US CHIPS Act to stimulate investment in infrastructure. Texas Instruments, one of my holdings, is using a similar strategy. However, they do it from a position of strength, not weakness. While Intel needs to turn its ship around and find new growth areas, TI invests in its current business segments and uses government incentives to build low-risk capacity for the next decade’s growth.
We can see that CapEx spending increased rapidly to build up capacity. We can also see that the ramping of the fabs has been very costly, generating operating losses of over $17 billion so far. This, alongside losing market share in other segments, has eroded EBT to a point where they are just breaking even, compared to over $20 billion between 2018 and 2021. Furthermore, Intel used financial engineering to increase earnings. Examples are their joint ventures with private equity firms to finance these deals, but selling a stake of the venture and changing the depreciation schedule for their machines from 5 years to 8 years to lower depreciation and raise earnings artificially. Without those tricks, EBT would have already been negative.
Building up a foundry business is time and resource-intensive, and you need to reach a large scale to do it profitably. Even though Intel is spending massive amounts of money (some of it is from government incentives, and they also did some joint ventures to reduce capital intensity), they still trail behind TSMC and Samsung’s capex spending, which are both at a much bigger scale already.
Ready for More?
By becoming a paying subscriber, you can read all of my writing on businesses, valuations and investing. Don't miss out on the opportunity!
Subscribe now and get a 7-day free trial!
Does it make sense to spin off this segment?
Intel has seen its business decline rapidly in recent years, and that, of course, spooked shareholders. Intel’s shares trade at prices seen in the last century. By spinning off its foundry business, Intel would become less capital intensive and operating income would become positive again (no more $5b operating losses from scaling the foundry segment). However, this would basically admit that they blew tens of billions of dollars in a venture they are leaving to rot. Realistically, the foundry business couldn’t sustain itself unless it took on obscene leverage, got a very large cash reserve from Intel (something they don’t have right now at $31 billion in net debt) or took more venture capital financing. Otherwise, the spun off company would go bankrupt or be bought out by a competitor. Of course, Intel could just load the spin-co with lots of debt and leave it to die, but I doubt this would go through without issues and help shares. Intel can’t just solve this problem by spinning it; in my opinion, this is an excellent example of the flawed direction of Intel’s management. Building a foundry business requires dedication, time and lots of cash. Without the foundry, Intel would just be left with its design business, which trails competitors, and they’d lose their last edge with internal manufacturing know-how. I’ll stay out of the company no matter how cheap Intel gets. Short-term thinking and frequent restructurings is no way to run a semiconductor company.
Lots of parallels with Boeing. Iconic manufacturing company in an strategically important industry that's collapsing slowly over the years due to terrible management. US government won't allow both companies to collapse. If bankruptcy becomes an issue, there will be bailouts. Hard to invest under such circumstances.