Intel reported earnings last week, and the Street did not like the numbers, with the stock down about 10% since the print, while the SOX Semis index rose almost 10% over the same period. The company painted its disappointing guidance as a function of cyclicality in some of their end markets, but we think the results point to a deeper problem in how the company needs to communicate.
The root of this problem is best encapsulated by this comment from CEO Pat Gelsinger on their recent call:
“I’ll just say, Intel Foundry, every quarter from here until the decade and we’re seeing improvement”
Intel Foundry will take until 2030 to be meaningful. We cannot really judge the company’s long-term prospects until that process is complete. We will probably have a clear idea before then, but we imagine that by the time we reach the end of the decade many investors who care about Intel today will have new jobs or even be retired. So how does the company message before then?
Working in its favor, the company appears to be on the right trajectory. First, their manufacturing process seems to be in really good shape, with both 18A and 14A on track, potentially allowing them to recapture the title of process lead from TSMC, with the added bonus of what we think is a meaningful cost advantage. Second, judging from their comments on the call, they seem to be regaining share in client and data center. This all points up and to the right. If the company has now put out all the bad news, it is possible they have a great set-up with each incremental earnings between now and 2030 adding to that story or rejuvenation.
Of course there are problems with this. The manufacturing process improvements will not really be felt in the market until next year. Regaining the title of process leadership is further out and again is not meaningful in its own right if the company does not have the right products in place. Which leads to another issue, the path between now and the next 23 quarters is not going to be straight up and to the right. They will have to deal with the vagaries of the market, inventory cycles and product dynamics. There will be good quarters and bad ones, perhaps very bad ones.
In order to bridge this gap, the company needs to prioritize credibility above everything else. If anything, the company has lost this with the Street more than anything else in the past decade. And as much as the current management has made huge strides in delivering on promises, the gap still exists. Credibility is what matters most now, more than growth, more than technology.
Here, we worry the company’s history and culture may stand in its way. The company still has some of the bad habits it developed when it was the market leader. Chief among those is assuming that investors believe their comments about the market. The Street and Intel seem to have highly divergent views of current conditions, especially in the data center. Most investors now take it as a given that Nvidia is the dominant provider in data centers. And while we can debate all aspects of that position, it is hard to argue with Nvidia’s market share. Any commentary about AI needs to be attenuated through that lens.
More broadly, the company needs to align with some basics of human comprehension patterns. Roughly speaking, bad news hurts twice as much as good news feels good. This is why the old investor relations trick of sandbagging works so well – dump all the bad news out there at once, then string out the good news over a long time period. This is more than underpromise and overdeliver, the timing matters a lot. We think this is why the company seems to be struggling so much right now. They gave us a bunch of good news six months ago, but then the last two quarters have been painful. This is the opposite of the what they should be doing. We do not mean to be overly critical of the company, or their comms team, the market is tricky right now. That being said, we do see some critical areas for improvement.
Chief among these is the organization needs to get better coordinated. From the outside, every quarter looks like the finance, marketing, product and operations teams all seem to be singing from different songbooks. For instance, the company held its Live event last month, and while that was not open to the public, the company did put out a lot of positive news around that, which raised expectations on the Street, only then to crush those expectations with their earnings. Intel needs to have a multi-quarter plan that ties all its functions together, with joint marketing meetings from all stakeholders. Do not market heavily ahead of bad earnings, make sure that product stories jibe with what the Street is expecting. Of course, this is challenged by the fact that Intel has to simultaneously speak to multiple audiences – customers, partners, investors, employees, etc. Live was a partner event, so the company needed to put its best foot forward to get that audience excited enough to design new products and promote them to their own customers. Still, there should be some nuance, especially in the way the company communicates publicly around such events.
And while we think the company is perfectly capable of doing all this, we want to emphasize that this is an important problem. Some may argue that the company’s stock will ultimately be determined by fundamentals – if those take a while to develop then eventually the stock will catch up. But there are consequences to a poor stock price. For one, outsiders – like customers and partners – are aware of Intel’s stock price and will grow more cautious if they see the stock hit repeatedly. There are also more looming threats, with the stock at current levels it is vulnerable to takeover. Counter-intuitively, these risks get worse as the company’s manufacturing process improves. We will explore this more soon.
Overall, we think Intel is looking increasingly better, so now the challenge for them is to convince everyone else of that.
Leave a Reply