Source: Content compiled from Semiwiki by Semiconductor Industry Watch (ID: icbank),

After two years of decline, the market seems to be very optimistic about the trend of semiconductors this year. So, if so, what kind of recovery do we expect to see? What impact will China have on the recovery of mature market chips? What will storage chip recovery look like? Will we go back to foolish spending?

Now let's guess.

Recovery, not as expected

From the performance of semiconductor stocks, you may not know that the semiconductor industry has been in a downturn for over two years, but this is the reality.

The stock market seems to always be a leading indicator of future performance, but on the other hand, stocks have been expensive throughout the entire downturn cycle, and it seems that expected recovery is always delayed. The current question is, will 2024 ultimately become the recovery that everyone is looking forward to?

So far, the signs look good, but certainly not as great as we say, and it will never return to the exciting days of crazy consumption and expectations.

After the shortage caused by the COVID-19, the huge amount of money invested by the industry to build capacity has obviously exceeded expectations, leading to a downward cycle caused by overcapacity that has lasted for more than two years.

We believe that considering the duration of the economic downturn, chip manufacturers may be a bit conservative in terms of capital expenditures.

We see that TSMC expects its spending to be "flat" in 2024, while projects such as Arizona have been intentionally delayed or slowed down.

TSMC's decision not to purchase High NA EUV lithography machines from ASML will also control its capital expenditures.

Intel's spending is at a reasonable level, but far from overspending, and appears to be more selective in terms of technology rather than production capacity.

We certainly do not expect Samsung's memory spending to rebound, as the memory capacity is still offline and has not fully recovered to 100% utilization. We see that Samsung's main expenditure is once again driven by technology rather than production capacity.

Capacity expenditure falls short of expectations

Importantly, the semiconductor industry is not just a single supply and demand capacity driven cycle.

The second cycle, although not as large as the production capacity cycle, is a technology cycle. Obviously, we have gone through technological nodes and new factories, which have created separate waves of expenditure while driving overall production capacity.

We expect that the majority of expenditures in 2024 will be technology driven rather than capacity related, so the magnitude will be lower.

Intel and TSMC are both investing in technology. Samsung and other memory manufacturers must keep up with the technological changes, even while keeping production capacity away from the market. They need to keep up with the pace of technology to maintain competitiveness based on Moore's Law, which drives the fundamental cost of memory business.

Essentially, although technology spending has changed, it remains almost unchanged, while capacity spending fluctuates greatly.

We will lower our expectations for a significant increase in production capacity expenditure in 2024. We believe that there will not be a significant potential increase in memory or logic demand in 2024, which will drive comprehensive capacity expenditures.

Although artificial intelligence remains the recent focus and driving force of the industry, it alone is not enough to fully restore normalcy to the entire industry.

High bandwidth memory is certainly great, but it is far from enough to absorb all the excess memory capacity, especially because it needs to be re equipped to convert capacity into high bandwidth production. Memory manufacturers must be careful not to exceed HBM memory requirements, which may be further limited by AI logic chip capacity and availability.

We still need a broader macroeconomic recovery to drive demand for personal computers, servers, and wireless devices, which undoubtedly dominate the majority of the market.

China will become an influencing factor

It is currently unclear what impact the $40 billion semiconductor equipment and tools purchased by China in 2023 will have on the chip manufacturing market.

Obviously, they haven't all gone live and been productive yet. The question is, what impact will they have after they go online?

As China hopes to invest in equipment and all new wafer fabs to gain market share, there are signs of price weakness in the mature foundry market that China plays an important role in.

$40 billion is a very large amount of equipment, and it may double as it is not a relatively expensive cutting-edge device, indicating that $40 billion represents a greater capacity increase as it is primarily located in the backend.

It clearly does not include large items such as EUV tools worth $150 million or even expensive DUV immersion tools. Therefore, this is a very significant increase in production capacity, as it is all concentrated on low-cost and mature nodes.

For this reason. We are concerned that second tier contract manufacturers such as Gexin and Lianhua Electronics may catch up and lower prices in China to gain low-end market share, while TSMC may face pressure to reduce pricing to maintain market share. Compared to mid-range contract manufacturers, both Chinese factories and TSMC have significant cost advantages.

The main way to avoid this situation is to try to lock in the business of customers who, for some reason, do not want to do business with China. GloFo has done a great job in this area, but the vast majority of chip customers only care about price, pricing, and delivery time.

China may be one of the biggest factors affecting the recovery speed of the semiconductor industry in 2024. Although it has no impact on the leading advantage, we need to remember that the vast majority of semiconductor units are targeting mature technologies that China has already served this large market and can and will affect it.

Who will be the winner? HBM is the highlight!

We believe that as we enter 2024, semiconductor companies will experience more differentiation in their performance, as not all stories will follow the trend.

We still love ASML's stories. They are one of the few true technology monopolies in the market. The story of High NA EUV's launch will present a positive news flow, which will mask China's restrictions.

We believe that TSMC is the main beneficiary of the artificial intelligence revolution and the recent demand from Apple and Intel. They are very cautious in their spending and are better able to avoid the impact of competition from China's trailing edge. Without a doubt, they are still the best chip manufacturers in the world.

Samsung's situation is mixed, as its OEM products still cannot fully reach TSMC's level, and due to still low demand, memory may slowly recover. The pricing of memory has broken away from the long-term bottom, but there has not been a strong rebound yet. It feels more like production capacity constraints have ultimately had an impact, rather than a return to strong demand. If this is correct and memory prices are better due to offline capacity retention, then it won't be a super recovery.

Although limited, HBM is still a highlight. We may be more inclined to view SK Hynix as a pure memory company rather than Samsung, which has performed poorly in contract manufacturing.

Overall, this earnings season will be positive for chip stocks as we expect many management teams to discuss a brighter outlook for 2024, although this is still more of a hope than a reality.

The dream of artificial intelligence remains one of the biggest drivers of recovery prospects, and so far, artificial intelligence has not encountered any major obstacles that have led to its slowdown.

Compared with chip manufacturers, the recovery of equipment spending will be slower, as the demand for capacity from memory or general foundries is still not high (excluding Chinese spending)

There is still significant uncertainty in geopolitics. The tension continues to boil, but this may be secondary rather than the most important.

As the S&P index returns to historic highs, these stocks still feel overbought. Perhaps this is just a price to earnings ratio expansion that everyone is willing to believe in, rather than the excessive prosperity that investors are worried about.

I think we will find out if the earnings season can support the stock market recovery.

Original link: https://semiwiki.com/semiconductor-services/341161-2024-semiconductor-cycle-outlook-the-shape-of-things-to-come-where-we-stand/

Disclaimer: This article is reproduced or adapted online, and the copyright belongs to the original author. The content of the article is the author's personal opinion. Reproduction is only intended to convey a different viewpoint and does not represent the company's endorsement or support of that viewpoint. If you have any objections, please feel free to contact us.