The Semiconductor tide is not lifting all boats
The Depressing State of the Hybrid Semiconductor Manufacturing Companies
The sun is shining down from a clear blue sky on the fertile semiconductor soil, full of healthy chip crops ready to be harvested. According to the Semiconductor Industry Association (SIA) and its faithful sidekick, the World Semiconductor Trade Statistics, this will be the largest harvest ever.
The Semiconductor market is expected to grow 19% in 2024 and 11% in 2025. (To achieve these numbers, Nvidia Server Sales and double-count HBM must be included—but who wants to spoil a good story?).
The large consultancy companies are jumping on the wagon and are trying to outbid each other and the self-proclaimed prophets in the industry.
Even though most companies subscribe to one or several of these oracles, they fortunately do not use the market forecasts for anything strategic. In my time in corporate, we always looked at WSTS and went “Fascinating” (in the British meaning of the word) and did more of the same.
In the good old days, when the tide lifted all ships, market growth numbers were a meaningful input for timing investments and retrenchments, but this is no longer true.
This upcycle is driven by price increases, profits, and lousy market categorisation. It benefits only a few companies in the AI space while their supply chain is scouting for an upturn on the horizon.
Semiconductor companies and the supply chain need much better and deeper market research to make strategic decisions about the size and timing of their investment.
While I enjoy following the AI boom, it is nearly meaningless for most of the industry, which is why my methodology is to analyse sections of the market and follow the supply flow up and downstream. While I track the size of markets, I am much more interested in the dynamics: What is changing?
This is the first strategic question: What is going on?
Neighboring the fertile AI fields is a desolate and barren Tolkian landscape—the land of the Hybrid Semiconductor companies. They are still waiting for an invitation to the party.
The state of the Hybrid Semiconductor Market.
The Hybrid semiconductor companies are still in trouble, and there is no immediate way out. While this is no surprise, it is painful, and the cracks are starting to show.
The collective Q4-24 guidance was negative 5.2%, and the result was slightly worse: revenue declined 5.7% compared to last quarter.
The Q1-2025 guidance was equally depressive. The collective guidance was down -7 %, which means the downward trend has lasted 10 quarters since the last peak.
While the CxOs bravely performed their rehearsed tap dance in front of the broader investment community, which had to endure the prepared remarks to get to some substance, they are starting to sound like a broken record.
I am not blaming the CxO’s; they are just playing their part in the staged show and are expected to say:
“We did better than average” (In average, you didn’t!)
“Our people performed well” (It is their fault!)
“We delivered a result above guidance” (We set mediocre targets for ourselves)
While I might sound harsh, I also listen to investor calls in the upmarket, and they are like an Academy Awards show where the CxO’s can enjoy the limelight. The brutal truth is that most of the company's performance is down to the market situation and some good bets.
And the markets sucks - for both good performers, average performers and poor performers. Listening to the calls gave no hope for optimism in Q2-25 either. The Hybrid companis are stuck in a hole.
It will be no surprise to my readers that I care less about the adjectives about the quarterly performance and more about the numbers. So, let us get to the numbers.
Regarding revenue growth, only ADI was able to beat last quarter. The analog company was up 5.7% over the previous quarter, but with -6.5% guidance for Q1-25, the spot in the limelight is not going to last. In general, ADI had the most optimistic market comments and saw an uptick in China’s EV companies and their share of products for the data center.
On the other end of the scale, Renesas had a brutal quarter with a decline of close to 17% QoQ while their YoY growth was -21.6 %. Renesas guidance was -0.5% for Q1-25. The Japanse had two explanations; one was currency, and the other was a significant reduction in distribution inventory that caused less sell-through. This sounds pretty wild to me.
An investigation of Renesas’ two main distributors did not reveal anything. WT Microelectronics have not reported Q4-24 yet, and Hagiwara reported flattish inventories. The two companies are the only ones that account for more than 10% of Renesas's revenue.
Also, Microchip had a brutal quarter with a 12% decline QoQ and 42% YoY. Steve Sanghi is back in the saddle and swinging his machete. Factory personnel are sent on leave, and radical changes will be made to the company over the next period.
Lastly, Infineon also had a tough quarter with a 15% decline but will claw some back in Q1-2025 as the only hybrid company with a positive guidance of 5%. Besides parking the decline at currency fluctuations and inventory corrections peppered with positive remarks about the future, Infineon did not enlighten us much.
Fortunately, the Hybrid Semiconductor companies have a solid business model that is now significantly under pressure.
Since its peak nine quarters ago, combined gross profit has declined by 36%, and this trend will continue for some time.
From a market perspective, both the automotive and industrial markets are down significantly. While the automotive markets attempt a correction lifted by the EV business in China, the industrial market continues to tank.
The Chinese market might need the advanced automotive power semiconductors from the Hybrid leaders, but it doesn’t need them for industrial manufacturing. This is most likely being absorbed by an entirely Chinese supply chain.
From a product perspective, this is confirmed.