Samsung Q3 2021 & Semis – Cycle economics

Some semis are value stocks.

• Samsung fleshed out its preliminary report by reporting another very good quarter supported by strong chip pricing thanks to the current shortage.
• Despite signs that the cycle may be peaking, Samsung has already sold off quite some distance and is beginning to look attractive again on valuation.
• Q3 2021 revenues / net income were KRW74.0tn / KRW12.06tn ahead of estimates of KRW73.0tn / KRW11.5tn and operating margins expanded to an excellent 21.4%.
• Semiconductors once again underpinned the numbers generating nearly two-thirds of all the operating profits of the company with operating profit margins of 38.1%.
• Despite Samsung’s massive capex spending on semiconductors, the company still generated KRW9tn in net cash flow in Q3 2021 underpinning its status as a financial powerhouse.
• Unfortunately, the outlook is very uncertain as the shortages persist for many components although pricing for some parts of the memory market does appear to have peaked.
• This could be lower demand for memory as a result of other component shortages holding up device production.
• Either way as the dominant player in memory, this has not hurt Samsung meaningfully and its guidance for the coming quarter and for 2022, pricing is not a concern.
• Samsung’s valuation is also not challenging as it is trading on 12.0x 2021 PER and 10.4x 2022 PER with earnings growth of 24% expected over the next 3 years.
• Qualcomm trades at a similar level when one takes growth into account as its somewhat higher PER is because it is expected to grow faster than Samsung.
• MediaTek on 14.5x 2021 PER with 3-year growth of 42% is also very interesting.
• Hence, I don’t think that there is much to choose between the three of them.
• However, while we are in the upswing of a supercycle currently, there is a strong possibility that the top is not very far away.
• Semiconductor companies are building fabs like crazy and when the logistics situation sorts itself out, inventory building is likely to ease meaning that demand for components may suddenly turn down.
• Every time this happens it is always slightly different in its cause and its manifestation, but the end result is always the same which is a secular rotation of money out of the sector.
• The defense against this is a long/short strategy where one goes long the value end of the market balanced with a short on the expensive end of the sector.
• Qualcomm, MediaTek, and Samsung fit the long end of this trade while one could consider AMD, NVIDIA, a semiconductor industry ETF or index for the short end.
• Intel is a special case on its own as the shares have fallen hard and look set to fall further leaving investors with a binary choice.
• This decision is essentially based on whether one believes that Pat Gelsinger will pull off Intel’s revival.
• This choice is also complicated by the possibility that Apple has demonstrated that the x86 processor is now obsolete which would mean that Intel becomes a foundry.
• I am very far from holding this view, but it is a possibility and it won’t be until 2022 that we see an x86 processor on TSMC’s 5nm process that one will begin to have an inkling of just how much TSMC has to do with Apple’s current superiority.
• I think I might get another shot at earning Intel at $44 but with the risks that Intel is currently facing, I might just lose my bottle again.