InvestingApr 20 2020 - Updating My 'Short' Thoughts On AMD
Summary
Having revisited the facts, I feel much more confident in my strategy that utilizes far-out-of-the-money options.
In the first article, I had underestimated some of the monopolistic power that AMD has in the x86 7nm space.
The rise of the stock offers the opportunity to extend the expiration in better terms than before.
COVID-19 mega-catalyst can make things evolve faster and before the options’ expiration.
In June 2019, I published “Lofty And Cyclical AMD Is 'The Big Short'”, thinking of AMD as a unique short candidate. At that point, the stock was trading at around per share, leading to a valuation that I believed to be shockingly pricey. In the following months, the stock continued surging, peaking at around per share. As I write these lines on April 16, the stock trades at around , having recovered from its recent COVID-19 lows.
Generally speaking, a stock rise can be a disaster for a short-seller. On top of risking the loss of more than 100% on the invested capital, the position is becoming larger and larger, leading to excessive and uncomfortable exposure. Unless “stop loss” is triggered, uncontrollable price spikes have the potential to destroy the whole portfolio.
Value investors hate “stop loss”, because they would ideally close their position only if “margin of safety” disappeared or became negative. They can change their minds and take the loss, accepting that they were wrong, but can’t easily afford retreating just because market prices caused them liquidity or other risks.
Fortunately, for my strategy that utilizes far-out-of-the-money options that are a very small part of the portfolio, things are much easier. The maximum loss is anyway small and limited to the tiny price of the option, and the stock price spike does not create a big problem. As I had explained in the first article, the negative side with options is that the short time window is not always enough for the events and the fundamentals to unfold and the market distortions to correct.
On the one hand, the clock is ticking, bringing the time closer and closer to the option expiration date. On the other hand, the rise of the stock offers the opportunity to extend the expiration in better terms than before. That means that skyrocketing can be actually a blessing and not a catastrophe. If you have indicated a bubble, and it is a bubble, it is better for you for the price to become even crazier. This is an awesome part of being a put option buyer, compared to a classic short-seller.
In a few words, the worst that can happen to you is for the future to prove your assumptions wrong. The second worst—the market not to adjust towards a reasonable valuation in time. The best that can happen is for the stock price to drop significantly, so that put options become in-the-money. Equally interesting is for the stock price to reach crazy highs, bringing put options lower so that if they are bought, they can extend the strategy and promise greater profits later in the future.
To keep the strategy on, you have to first re-evaluate its validity and ensure that the value and prospects of AMD have not changed against you, being ready to admit errors in the thinking process. Ego can hurt your portfolio and your returns. Equally, changing your mind just because the market stayed or became more irrational on how it is pricing the stock is the worst you can do as a value investor. Admitting your wrongs is wise; retreating to the consensus thinking and getting swept by the market is wrong.
Reassessing AMD
AMD continues capturing larger market share, while Intel (INTC) has not solved its shortage problems. According to Mercury Research facts republished here, in Q4 2019, AMD’s overall x86 unit share reached 15.5%, compared to 13.3% that we saw in the previous article based on Q1 2019 figures. In servers, the share increased from 2.9% to 4.5% in this period.
For 2019, revenue increased to .73 billion, from .48 billion in 2018. Net income was 1 million, almost stable compared to 2018. To be fair, 2019 includes 6 million of loss from the redemption, repurchase, and conversion of debt.
The increases in the market share do not surprise me. Remember, I was saying, “I cannot disagree that AMD might produce superior products, at least for some time, and have its market share reach or exceed 20%”. In the same sense, the financial performance does not indicate anything extraordinary, despite the debut of the 7nm CPUs.
What seems to be more severe than I initially thought are the issues that Intel faces with advancing to the next nodes and the linked shortages. Intel admits that its process likely won’t catch up to AMD until the end of 2021. Also, what I have learned—thanks to your valuable comments—is the significance of the chiplet approach that AMD follows.
Additionally, while I keep my opinion that AMD’s success is mainly due to a technology that it does not own (TSMC’s 7nm), I admit that I had underestimated the importance of AMD being the only serious player that can exploit this technology in the x86. This is critical, because the whole ecosystem of desktops, laptops, and servers is well-tied to the x86 architecture.
Intel should either catch up with TSMC (TSM) or retreat and become fabless. Otherwise, AMD can for long stay as the only serious x86 player that can utilize the increased transistor capacity that TSMC offers. VIA — the third of the three x86 players — needs time to get into the game and catch the same opportunity. Zhaoxin, a joint venture between VIA Technologies and Shanghai Municipal Government, aims to release its 7nm x86 CPU in 2021, when AMD will have already moved on to TSMC’s 5nm.
Having said that, I insist—and have reinforced my opinion—that the game won’t be easy for AMD. Despite the shortages, Intel has initiated a price war, as expected. AMD doesn’t have a lot of freedom in pricing, and subsequently cannot achieve and maintain extreme margins. Even if Intel won’t be able to cope with TSMC, AMD is not alone and Zhaoxin x86’s potential competition is not the only long-term threat.
In the previous article, I had made a reference to (QCOM) Qualcomm’s Snapdragon 8cx ARM processor. A few months later, the (MSFT) Microsoft Surface Pro X came with an SQ1 ARM processor, based on Snapdragon 8cx. Despite some compatibility issues, Windows® 10 on ARM is an example of how the architecture can slowly gain ground in laptops.
The mobility is even higher in servers, where we have a bunch of 7nm ARM attempts to besiege the market, claiming superior performance compared to Intel and AMD, all using TSMC technology. Again, those solutions are not suitable for customers that are tied to x86, but they clearly showcase that AMD is not the only chipmaker that can exploit Intel’s weakness over TSMC:
-
Graviton 2 by Amazon with 64 cores.
-
Altra by Ampere (founded by former Intel Chairman Renee James) with 80 cores.
All this shows that if I should be very optimistic about the prospects of a company, that would be TSMC and not AMD. Under the condition of an attractive price, I would prefer to buy the underlying foundry (TSMC) that produces decent net margin, return on equity, and has a history of steady growth. Especially if the next few quarters were terrible thanks to the coronavirus situation and its stock went south, TSMC might become a great cyclical bargain. Its current market capitalization of 2 billion gives a P/E of 23x, which does not give me adequate “margin of safety”. If you are not an advanced investor, my video can help you understand the importance of P/E.
TSMC’s Revenue & Net Income of TSMC during the previous 20 years.(Source: Finbox)
At this point, let me take the opportunity to summarize my initial perception for these companies. I see TSMC as the great company at a so-so price, Intel as the great company with long-term risks at a low price, and AMD as the fragile company at a terribly high price. TSMC has the prospects, Intel offsets some of the risks because of the lower price, and AMD is pure speculation with extremely high risk at this valuation.
If you bet that TSMC’s density will kill Intel, you are better off to buy TSMC. If you think that Intel can keep up with transistor density, buy Intel. If you think that x86 will be in trouble because of ARM, worry very much about Intel and AMD. We can’t know what will happen for sure, but it seems that in any case, AMD is the worst choice to buy.
Revisiting the Scenarios & COVID-19 as a Catalyst
Reviewing my initial assessment, I would take into account what I had underestimated the previous time—the significance of AMD’s monopoly in exploiting the TSMC technology for the prevailing x86 architecture.
Basically, I would increase the probabilities of the dream scenario from 10% to 20%. The scenario is based on assumptions of AMD dominance that could bring billions of earnings and justify a market capitalization even several times higher than the current. I won’t make any effort to put an upper limit to the range, because no matter where it lands, this scenario would leave put options worthless. The magnitude of this extreme is important only for buyers of the stock and classic short-sellers who risk getting burned. For put options buyers like me, only estimating the probability of this scenario makes sense.
Consequently, I would reduce the probabilities of the bad scenario from 60% to 50%, which remains the most probable. After this adjustment, the good scenario (30%) and the dream scenario (20%) would now have a combined probability of 50%.
The bad scenario, which assumes that AMD stays the same company as in the past, has a fair value in the range of .69 to .23 per share. The good scenario, under which AMD keeps an improved market position and becomes a different and sustainably profitable company, gives a fair value in the range of .23 to .86 per share. The dream scenario, as discussed, has untapped upside potential. To learn more details about those calculations, please read my first article on AMD. The close price of the stock was .95 as I am writing this is on 16 April, showing how ridiculously high-priced AMD is—provided that my estimations are right.
The scenarios are based on long-term assumptions, which are not affected by the current turmoil, and thus have allowed me to be fair and decrease the probability of the bad scenario. However, we should not forget that the short-term risk has significantly increased, and we have a mega-catalyst present against the “bubbly” AMD stock.
In my first AMD article, I explained that generally I am agnostic about time, and I just suspected some catalysts for a correction in the stock as the reversion of the chip cycle; an Intel price war; the market gains and the profit margins not matching the expectations; a recession or a crash; and trade wars that could affect TSMC.
Compared to the above, the COVID-19 situation is a mega-catalyst and a factor that is bringing a lot of risk to the chipmaker. A slowing economy can produce chip oversupply and trigger the reversion of the chip cycle, making AMD report much uglier results than the expectations. A year ago, I was thinking of it just as a possibility. Now, we have a cause and tangible developments suggesting that a recession has started.
The first graph below shows what happened to AMD’s revenue and earnings during the previous two recessions. The second one highlights that the market capitalization went below revenue. AMD’s current revenue is .7 billion or 5.75$ per share.
AMD’s Revenue & Net Income during the previous 20 years.
(Source: Finbox)
AMD’s Revenue & Market Capitalization during the previous 20 years.
(Source: Finbox)
The Stock Market Reaction and My Experience with Options
What is unbelievable is that despite these unprecedented moments of the lockdown, the stock has proved to be immune, which is an illusion of safety. The past performance and the stability of the price do not safeguard you. Considering that the underlying business is affected, the higher the price of the stock, the higher the downside and the risk.
In February, the whole stock market seemed to be robust, until it was not. This should become a lesson for AMD shareholders. This total irrationality of the market makes the short potential of this stock much more interesting. A recession is coming, and this stock remains ridiculously overvalued, being close to its highs.
The irrationality is bigger in options, where we generally have two types of distortion that derive from the fact that the market idiotically values the options according to the price and the volatility of the underlying stock. The market accepts that under the market efficiency hypothesis, the market value is the fair value, and that volatility is risk. It trusts the underlying security’s price and fluctuations, rather than evaluating the actual business value and the real risks.
If AMD rises to 0, the option market won’t care if the stock is a bubble or not a bubble. It will anyway think that is very improbable. If the stock becomes very volatile, it will think that this has become more probable.
My initial plan was to have enormous gains in the event that the put options became in-the-money. What I had not factored in is for the put options to skyrocket just because the VIX and volatility went up. I came into a big dilemma—should I take those significant earnings or stick with the plan? It is like you have bought insurance for a cyclone, getting 10x your money if it passes through your town. When news is erupting that the cyclone might pass, the insurer tells you, “Take 2x in hand and let’s cancel the contract”.
I initially had AMD 10 puts for Jan ’21. Later on, using the logic of extending the strategy—and since the stock was skyrocketing—I added AMD 15 puts for Jan ’22. For the latter, the market was offering me 100% profit when VIX was trading at record highs. Very similar to the example of the insurer’s offer and the cyclone. I decided to stick with the strategy and reject the “insurer’s offer”.
We don’t know yet how long the COVID-19 turmoil will last. My guess is that the virus itself will be something short, and I don’t give a lot of attention to the very short-term effects, like the increase or decrease in the demand for personal computers. What can be more serious than the virus itself are the effects of our reaction—and, I am afraid, our overreaction (lockdown)—to the threat. The problem is that the economic outcome of the lockdown and the fear may constitute the triggering cause for the eruption of other hidden economic problems, bringing a major recession. In that case, the chip industry will feel the pain, and as I have said, this seems to be a mega-catalyst for AMD stock to collapse.
Considering the above, I found it wiser to stick with my plan. Anyway, I was—and still am—invested in stocks for the long term. Keeping a small amount of put options in an extremely overvalued stock is a good hedge and insurance that I didn’t wish to remove from my portfolio. I said no to the temptation of cashing in some “volatility profits” and destroying the strategy that aims options to become in-the-money, bringing huge profits at a time of distress.
As expected, the options lost their gains when volatility went down this month. This gave me the opportunity to buy additional shorter-term AMD 25 puts for Jan ’21, recognising the effect of the coronavirus mega-catalyst. If the stock went down to , those would be anyway the first to sell, letting the rest run. So, I picked a shorter expiration date this time.
Going against companies is not what I generally do, as I am a holder of stocks. I see buying put options of AMD as insurance for bad events. I don’t aim to predict the market, but I suspect and find it very probable that most market participants have seen nothing yet. When you see all those bubbly stocks more than 50% down, you will discover what a real crash looks like.
Conclusion
Having revisited the facts, I feel much more confident in the strategy I described in detail in my previous article. Since then, not a lot has changed in terms of the assumptions for assessing the long-term value of the company. I am admitting, though, that I had underestimated some of the monopolistic power that AMD has in the x86 7nm space. Because of that, I increased the probability of the dream scenario from 10% to 20% and decreased the one for the bad scenario from 60% to 50%.
The big change is the coming of the COVID-19 mega-catalyst. Combined with the shocking immunity of the stock so far, it makes the strategy more promising than ever. It can be extended in better terms and with a better probability of success, because things may evolve faster and before the options’ expiration.
For very experienced and event-driven investors, this is the best “Big Short”, with enormous earnings potential and a mega-catalyst present. However, my personal style of investing accepts this to be only a small part of the portfolio, as a hedge. Mainly, because as an investor, my mind is long, and as a human, my heart is long as well.
My best wish is for everyone to be safe, find a way to beat the virus, revive our economies, and see my put options go to zero.
Disclosure: I am/we are short AMD. 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.
Additional disclosure: I have bought AMD 10 puts for January 2021, AMD 15 puts for January 2022 & AMD 25 puts for January 2021
0
0
0
0
0
0
