Intel stock price
Intel (NASDAQ:INTC) is undervalued because of the cyclical problems in semiconductors. Second Q revenue was down 3%, with the PC business up 1% and the data center business down 7 %. However, this misses the importance of the 10 and 7 nanometer chips. The earnings in the next four years will be driven by these two projects. In the first half of 2019, spending on these chips was $6.9 Billion, more than 20% of revenue with no shipments. In the fourth quarter, 10 nanometer shipments will be significant, but lower yields are projected to lower the gross margins by three to three and a half percentage points. Like any new product, the yield will improve as more experience is gained with the manufacturing process. The growth from these new processes will make Intel a strong buy.
Demand for the cloud led to three years of data centric growth and increasing stock prices. The first quarter earnings led to a sharp decline. The stock partially improved then dropped after the second quarter announcement followed by a pickup to the level before the second quarter earnings announcement. Revenue was down 3% vs. prior year, with PC chips up 1% and data center down 7%. PC market share declined in PCs because of a lack of capacity for older designed, low-priced products. The data center includes the rapidly growing but small Mobileye and Internet of Things. Without these two segments, data center was down 10%. The cloud is still growing rapidly, but the cloud providers are not adding new data centers. Rather, as new capacity is needed, they are dropping in additional processing capacity into existing centers. The growth of the cloud has lowered sales to enterprise and government data centers.
10 Nanometer Program
Intel formerly shrunk size taken up by transistors and other component every two years to double the component density. The shrink from the current 14 nanometer chip to 10 nanometers was delayed by several years. The 10 nanometer was more than just a 2 X increase in density from a smaller die size. Its design also increased the chip density, so the chip density increased by 2.7 times. In addition, the design facilitated the next shrink to 7 nanometers. By the time management realized that this approach was way too ambitious, Intel was too far down the road to change. They delayed the introduction yet again and proceeded. The 7 nanometer will have a 2X increase, so that chip will have 5.4 times more components per square area than the current 14 nanometer chip. The higher density increases processing speed and lowers power consumption. Therefore, the chips are faster, which allows more features and longer battery life in a laptop.
Competitors rushed ahead and introduced 7 nanometer chips that compete with the 10-nanometer chip. However, the upfront cost of shrinking both lengths and width is huge in this size range. The competitors’ 7 nanometer chip shrunk on one dimension is roughly the same as the Intel 10. This high upfront cost of shrinking on both dimensions gives Intel a long-term competitive advantage.
With the 10 nanometer process delayed, Intel has been reworking the 14-nanometer chip architecture to wring out more performance. They have been remarkably successful in doing this. The combination of the upgraded architecture with the 10-nanometer size has produced substantially improved performance.
Tests of PCs with the new chips were held for selected trade press in a hotel conference room. The 10 nanometer process was used in the “Ice Lake “chips with the “Sunny Cove” architecture in the normal I3, I5 and I7 ranges with integrated Iris graphics. The i7 is an upper range, but this will be the top end for this range. The reviews were good with excellent graphics and higher speed.
Most of the chips in the i7 and other ranges will be the old 14 nanometer process for some time to come because it takes time and money to change. In 2020 and 2021, more mid-range and budget versions will be introduced in the 10 nanometer process. In 2019, 11 different chips will be produced.
The illustration below goes into the details of the chips. For example, in the upper left, the cores are listed in an i7. Four cores are used with fewer cores in the i5 and I3 ranges.
Production of the 10 nanometer began in the second quarter for inventory. The PC chips will begin shipping in the second half to provide PCs for the holiday market. The first models were shipped in the third quarter in small quantities.
The new 10 nanometer process yield, as with any new die size, will improve as manufacturing gains experience. In the fourth quarter, the gross profit is projected to decline by 3.0 to 3.5 percentage points from the 60% gross margin. With higher volume of 10 nanometer production, 2021 gross margins could be around 57% even with much improved yields. Management has said that 2020 gross margins could be between 60 and 57 percent. The yield improvements are inherently uncertain. The large number of chips made further complicates yield. For example, none of the chips built in 2019 are for desktops. The laptop chips come in U versions and Y for ultra-thin models.
New Product Development
The PC chips are simpler than data center chips, so they were introduced first. Given the importance of the data-centric market, the development of these chips is being accelerated. The first server chip samples have been provided to server manufacturers with production beginning in the first half of 2020 with ramp-up in the second half. Therefore, new product development and manufacturing startup cost will be unusually high. In 2021, the first 7 nanometer PC chips should ship. This will be followed by the next two years of developing chips on the new process.
Semiconductors have a reputation for cyclical performance. Intel’s current P/E is 12 compared to the S&P 500 index of 21. The P/E is 40% below the S&P index. Most of Intel’s products are proprietary. They have market swings such as the recent slowdown in data center chips. The high growth should raise the P/E ratio over time as well as increase earnings. Revenue in 2020 will be higher, and earnings, despite all the products, will be up. Intel should recover its previous high of $59 per share by the end of the first quarter. It will move up from there, as its growth potential becomes more evident.
Management emphasizes that the 10/7 nanometer process will have an excellent return. The products have to sell well for this to be true. This will not be a smooth, dependable next few years. Neither the CEO or CFO was comfortable with the gross margins, so large quarterly swings can be expected. However, four years from now, the results should be much higher than the 2019 guidance of $4.10 per share. Intel is a strong buy.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.