Analyst Robert Maire previously pointed out in ASML's earnings report that, for a long time, their view of the semiconductor industry has been twofold: one, through the two different components of supply; and two, through the lens of demand.
They believe that the current downturn's "round" of decline is primarily driven by the supply side, as the industry, after experiencing Covid and supply chain crises, recklessly expanded its capacity. The repeated expansion was obviously a hasty abandonment, and now, having "exceeded" the required supply, we find ourselves in a state of oversupply, initiating the current downturn.
Demand has also softened, perhaps with the acceleration of downward pressure on demand due to global macroeconomic concerns.
In their view, the "second phase" of the current chip cycle is likely to be driven more by weak demand, rather than by excess supply as was the case in the first half of the year.
This could be worse, as supply issues are often more easily resolved than demand problems, because if you're a chip manufacturer, you can control the supply, as we've seen in the memory market, where Micron and Samsung have reduced capacity and production to support pricing.
The issue lies in the industry's inability to stimulate chip demand. Lowering the price of automotive chips does not stimulate demand.
The decision was made based on their forecasts from ASML's financial reports.
They have long stated that the industry has always believed that any rational semiconductor manufacturer would not cancel or delay lithography tools, as they fear ending up at the end of a very long production line later, and in a worse position.
But this fear seems to have faded, as ASML's order book has shown a marked shift, with new tools being introduced and other clients being brought in to fill the gaps left behind.
Additionally, if chip manufacturers like TSMC, Intel, and Samsung are willing to delay crucial lithography tools, it certainly suggests that demand will not improve quickly. This indicates that the semiconductor industry's downturn is more severe and longer-lasting than previously thought. If you believe the industry will "bounce back," you would not delay new lithography tools. This has had a very ominous impact on the entire industry, as it masks the view of a rapid recovery.
Following today's analysis of Lam Research's financial report, Robert Maire reiterated his stance. He noted that Lam's memory business has dropped to a low not seen in a long time, with memory accounting for 32% of the overall business and foundry services taking up 46%. The larger portion of the memory business is likely in services, which has also been on a continuous decline, as the shipment of new tools to memory customers may drop to a level almost close to zero.
Lam explicitly stated that memory capital expenditures have not yet hit bottom, as they commented that memory spending is expected to "decline further." The company estimates that memory capital expenditures have already dropped by 50% as memory manufacturers continue to reduce bit growth and bit output in an attempt to support the ongoing decline in prices.
It seems quite clear that, given the trajectory and momentum of memory not recovering before year-end, and when it does occur, the recovery will be weak and gradual.
Memory manufacturers will have a substantial existing capacity that can be brought back online before purchasing new equipment, and even consider expanding or adding new wafer fabs.
In addition to idle memory manufacturing equipment from the halted wafer factory, there's a technical downsizing, allowing for increased production capacity without the need for purchasing new equipment.
In the end, due to idle equipment, declining output, and potential technological changes, memory manufacturers can easily survive for one to two years based on the current demand trajectory without incurring any additional expenses.
The bigger issue is that, under the current loss-making scenario, it may only be Samsung that has the economic capacity to survive another year.
