Chip delivery time is further shortened, except for some MCUs

According to research by Susquehanna Financial Group, chip delivery times shortened again in August, averaging 26.8 weeks, slightly lower than the previous month's data, indicating that the global shortage has further eased, but some semiconductors are still difficult to obtain.

Chris Rolland, an analyst at Susquehanna Financial Group, said the shorter wait time between ordering and delivery reflects slowing demand for some technology, namely phones and personal computers. But some markets remain overheated, with orders coming in faster than chip makers can ship.

"We believe the trend of repeat ordering and inventory building is not over yet," he wrote in a research note.

Rolland said that supply chain disruptions caused by the epidemic have led to an unprecedented chip shortage, and it would be "healthy" for delivery times to return to 10 to 14 weeks.

Currently, supply of some power management components, microcontrollers and optoelectronic components remains long, and companies including Microchip Technology and Infineon Technology are still struggling to fill those orders. However, other chip manufacturers have suffered from falling demand, including Nvidia and Intel, which are highly dependent on the personal computer market.

The chip industry has often struggled to balance supply and demand, in part because the process of manufacturing components takes months. Semiconductor makers now serve a wider range of customers, including cars, appliances and industrial equipment.

The current semiconductor industry has been running so fast and focusing on speed that it has not yet realized that the foundation supporting the industry has disappeared, that is, demand has fallen and will fall further.

We have been talking about the industry being in a down cycle for months. Memory prices are falling (usually one of the first signs) and inventories are rising and lead times are shortening. More importantly, demand for semiconductor-rich electronic devices is falling.

Yet, some semiconductor and semiconductor equipment companies are still reporting solid earnings, in some cases record earnings. That makes it very difficult to talk about a down cycle when you're still making tons of money.

The speed at which the industry operates has created so much momentum for the industry that gravitational cognition has been delayed.

In many cases, the industry relies on backlogs or non-cancellable orders placed near cycle peaks, even though the product is in stock or readily available. In other cases, customers are so scared (like the automotive industry) that they continue to order even when they already have enough, simply because they don't want another shortage.

Semiconductor equipment is the worst offender in this regard, as no one dares to go offline waiting for lithography tools, lest there be another shortage.

We have seen cases in the past where crates of semiconductor equipment have been piling up on the receiving dock because it couldn't be installed fast enough or there wasn't room. In one case in the past, there was a parking lot full of crates.

Wafers sit in the OSATs’ lanes, waiting to be packaged and tested. Manufacturers like Micron start bringing products to market to support pricing.

Given the absolutely tremendous momentum the industry has had for a few years now, it’s not unreasonable to think we could be dealing with one of the biggest cases of overcapacity the industry has seen in many cycles.

Many say the industry is being more cautious in spending than in past bad cycles, but equipment order rates over the past year or more have been anything but cautious.

We are still in the early stages of the down cycle as not everyone agrees, acknowledges or recognizes reality. We are concerned that processor demand for hyperscalers, data centers and memory in consumer devices has not yet fully capitulated. Aside from supply chain issues primarily related to COVID, which have been relatively minor, we have seen little financial impact on device manufacturers. Therefore, we are still in the snowball phase and the problem has not yet grown to snowman size to encompass the entire industry.

Many so called analysts are still very bullish or buying a lot of stocks or becoming more aggressive as valuations slide. From a stock perspective we are not hit yet, we are far from capitulation.

Maybe the bell that signals the bottom of the cycle is the last of the bullish analysts to capitulate (ignore those who never change their ratings…).

We think TSMC is still one of the more defensive names among foundries or chipmakers in general. They are so far ahead that they can control and dominate pricing, and thus all other foundries in the market. Other foundries are under the pricing umbrella of TSMC.

When TSMC runs out of capacity or raises prices enough, chip customers are forced to go elsewhere to make their chips, even if TSMC is always their first choice. The bottom line is that TSMC's overflow business goes to competitors. When TSMC has excess capacity, a lot of that business goes back to them, leaving those further down the foundry food chain with much lower utilization and profits.

In semiconductor equipment, ASML is always the last piece of equipment you would cancel, given the crazy long lead times. Most process tools, such as deposition and etch tools, are more of a "turnaround" business, where you can simply reorder a canceled piece without much delay.

It is unclear how tool and chip shipments to China will fare amid a deterioration in relations. Given that China has been the largest customer for most equipment companies, this means it is an important variable that will look worse in the short term. While the impact is small today, it could have a significant impact when equipment companies scramble to order or need to find new homes for canceled or delayed products.

Lam noted that they have billions of dollars worth of unfinished tools that are shipped to customers incomplete while they wait for parts. This situation is completely different from ASML, which provides complete but untested tools to get them to customers faster.

What happens when all these tools are finished and installed???

This reminds us of the situation in the Chinese LED industry where there were many MOCVD tools sitting in crates, unused.

As we mentioned in an earlier note, the timing of the chip bill is nothing short of poetic. Micron may cut capital spending in half, and Intel has already announced it may slow down Ohio and other projects.

With careful financial analysis, are we going to end up in a "use it or lose it" situation where chipmakers feel forced to spend money where they wouldn't otherwise. Basically throwing free or cheap money at the industry even if it doesn't need it because we already have excess capacity (although probably not in the right countries).

Given that circumstances have changed significantly since the project was conceived, we may need to recalibrate the CHIP Act.

Overall, we now see much more downside beta than upside beta in the semiconductor sector. It is difficult to come up with a few variables that could significantly break out to the upside, and most seem to be the extent of downside potential.

We see no good reason to participate in a value trap. The last thing we want to hear is some analyst saying a stock is trading at a 52-week low and has a high dividend. At this point, I certainly don't care about the dividend play when there is a significant risk to my principles, with no offsetting gains.

Of course, macro uncertainty is a large part of the problem, and it doesn't look like the macro problems are getting better anytime soon. Semiconductors remain the tip of the economic spear and will be greatly impacted by any macro fluctuations.

Another question is how long or how deep we won't have any good sense of what's going to happen? Can overall demand for chips keep this short and slight slowdown going? Which way will all the variables go down?

Long-term demand seems absolutely good, but things could get a lot worse in the short term as we don’t think we’ve hit the bottom yet.