Semiconductors: Beware The Fadeaway

|
About: NVIDIA Corporation (NVDA), Includes: AMD, AVGO, FDN, INTC, IYW, MLNX, MU, NXPI, QCOM, QQQ, SMH, SOXX, SPY, TSM, TXN, VGT, XLK, XLNX
by: The Fortune Teller
Summary

Semiconductors, the hottest segment on Wall Street in recent years, are cooling down fast. So fast, that we might forget they rule(d) world.

When 2017 ended, I held NVDA, MU, and AMD. When 2018 ended, I only held one of those, and it's not necessarily the one you might bet on.

Soon enough, I'm likely to find myself holding no semis whatsoever. This is reflecting a view that 2019 is going to be a tough year for the segment.

Nevertheless, depending how things unfold this year, it's not out of question that I'll finish 2019 with one long position, just like 2018 ended.

This time round, though, the position I might ride on to the next year won't be the same one I celebrated the most recent New Year's Eve with.

The Fadeaway Move

"Fadeaway" is a well-known, though quite difficult, move in the game of basketball.

A fadeaway or fall-away in basketball is a jump shot taken while jumping backwards, away from the basket. The goal is to create space between the shooter and the defender, making the shot much harder to block. The shooter must have very good accuracy and must use more strength in a relatively short amount of time. - Wikipedia

Image result for fadeaway basketball

Michael Jordan was one of the most popular shooters of the fadeaway. Wilt Chamberlain, Patrick Ewing, LeBron James, Kobe Bryant, Hakeem Olajuwon, Dwyane Wade, Karl Malone, and Larry Bird are few of the greatest of all time ("GOAT") who are also known for using this move. It's no coincidence that those who brought the fadeaway move to perfection are also those making the GOAT list.

Similarly, over recent years, semiconductors have brought their fadeaway move to a perfection, creating more than enough space (aka profit margins) between them (suppliers) and their "defenders" (customers), making the shooting stock price much harder to "block".

Here is how well semiconductors have "played" during a period of 3 years, prior to Q4/2018 (i.e. 9/30/2015-9/30/2018):

Semiconductor stocks total returns during a period of 3-years, 9/30/2015-9/30/2018

A few quick observations:

1. Almost all the leading names outperformed the overall market (S&P 500). The total returns of Nvidia Corp. (NVDA), Advanced Micro Devices Inc. (AMD), Micron Technology Inc. (MU), Taiwan Semiconductor Manufacturing Co. Ltd. (TSM), Texas Instruments Inc. (TXN), Broadcom Inc. (AVGO), and Intel Corp. (INTC) were much better than that of the SPDR® S&P 500 ETF (SPY).

Only Qualcomm Inc. (QCOM) and NXP Semiconductors NV (NXPI) underperformed the index. We focused on these two companies (together with AVGO) in "three musketeers", our most recent article.

2. With the exception of INTC, all the names that outperformed the SPY also outperformed the technology benchmark - Invesco QQQ Trust (QQQ).

3. Four companies - NVDA, AMD, MU, and TSM - managed to outperform the iShares PHLX Semiconductor ETF (SOXX) benchmark. However, only the first three managed to do so decisively.

The three semis that have outperformed the most (during the 3-year time period) - namely AMD, NVDA, and MU - have starred in a few of our articles over the past 15 months. To wit:

In case you haven't figured this yet - we love semis. We think this is the past, present and future. It's hard (and frankly doesn't make any sense) to argue with the growth and potential here.

We like all three names - AMD, NVDA and MU - and we own all three.

Having said that, we do like AMD better, mainly because we see it as the better pick from a risk/reward perspective but also because we see a greater potential for AMD to shine in an already shining kingdom. Putting it differently, it will be easier for AMD to beat and perform than its main peers.

Source: Semis: It's Not Only Who You Wish To Play With, But How You Wish To Play The Game (Part I), March 14th, 2018

In this two-part article we focused on our favorite semis AMD, MU and NVDA.

At this point you (hopefully) can understand why we hold this trio and why AMD is our sunshine.

Just as the three tenors - Spaniards Plácido Domingo and José Carreras, and the Italian Luciano Pavarotti - created an incredible singing group during the 1990s and early 2000s, we believe that the three semis are creating an incredible growth group in the 2010s and into the 2020s.

No tears, no fears and one O Sole Mio: AMD.

Source: Semis: It's Not Only Who You Wish To Play With, But How You Wish To Play The Game (Part II), March 16th, 2018

As you can see, we were spot on (already in late 2017; more details/proofs hereinafter) with picking AMD as our preferred semi at the end of 2017, as well as with dumping NVDA.

Same goes for reversing our long position in AMD into a short position, just ahead of Q4/2018. This short position got closed on during Q4/2018, and we have had no position in AMD ever since. (Until this week; more details below.)

The Fadeaway Mode

Nevertheless, we remind you that "the shooter must have very good accuracy and must use more strength in a relatively short amount of time."

Semis had the accuracy and strength, but it now seems like, for many of them, time and energy are running out.

Here's how the exact same list of companies above have performed since Q4/2018 started and until now (closing prices of 3/12/2019):

A quick observation: Only four names outperformed the overall market (S&P 500), the technology sector (QQQ, XLK, VGT, FDN, IYW) and/or the semiconductors benchmark.

The total returns of the iShares PHLX Semiconductor ETF (SOXX), SPDR® S&P 500 ETF (SPY), and Invesco QQQ Trust (QQQ) benchmark ETFs since 10/1/2018 are better than those of Taiwan Semiconductor Manufacturing Co. Ltd., Micron Technology Inc., Qualcomm Inc., Advanced Micro Devices Inc., and Nvidia Corp.

Only Intel Corp., Broadcom Inc., NXP Semiconductors NV, and Texas Instruments Inc. have performed better than the benchmarks.

The "Fadeaway Mode" is even more stunning when we focus on the trio of NVDA, MU, and AMD.

Total Return - Past Three Years (to and including 3/13/2019)

The past three years have been nothing but heaven for the trio, even if we include the last few very rough months.

AMD (TR of +828%), NVDA (+431%) and MU (+241%) have left dust to all other benchmarks, whether diversified (SPY, QQQ) or focused (SOXX, SMH):

  • iShares PHLX Semiconductor ETF (SOXX) +114%
  • VanEck Vectors Semiconductor ETF (SMH) +101%
  • SPDR® S&P 500 ETF (SPY) +47%
  • Invesco QQQ Trust (QQQ) +71%

Total Return - FY 2018

Nevertheless, if we look at 2018 on a standalone basis, including its devastating fourth quarter, it's more of a mixed bag.

While AMD managed to finish the year as the best performing S&P 500 stock of the year, NVDA and MU will remember 2018 much less fondly.

Source: FactSet

Having said that, 2018 mustn't be looked at as one piece. Since the last quarter has been so different than the three quarters preceding it, here's a split view of 2018, until and from the decisive date (October 1st 2018):

  • Total Return from 1/1/2018 to 10/1/2018

The "Golden Era" during which AMD (+206%), NVDA (+50%) and, to a much lesser extent, MU (+10%) outperformed the semiconductors' leading ETFs (SOXX +10%, SMH +9%), as well as the main indices (QQQ +20%, SPY +11%) by a mile.

  • Total Return from 10/1/2018 to 3/13/2019

Then came 10/1/2018 and from that date on, things weren't the same for semis anymore. Ever since then, it's like a mirror-opposite image to what we experienced during the first three quarters of 2018. NVDA (-42%), AMD (-26%) and MU (-14%) are way behind the semiconductors leading ETFs (SOXX flat, SMH -1%), as well as the main indices (QQQ -5%, SPY -3%).

Clearly, we are witnessing a real and significant fadeaway. The question now is what the future holds for these semis, and throughout the rest of this article, we will try to find out.

Past Earnings and Other Financial Aspects

  • Market Cap

First of all, it's important to note that NVDA alone has a market size that is almost 50% greater than the market-cap sizes of MU and AMD, combined.

Having said that, it's noticeable that while the current trend is down for NVDA and MU, and up for AMD. Putting it differently, the gap (in size) is shrinking, and we expect it to shrink further, going further.

The amazing run of semis is illustrated fantastically through the below chart, showing the development of the market-caps of the trio over the past decade.

Ten years ago, all three companies were tiny operations that hardly anyone looked at. At the peak of their valuation (each one on a stand-alone), the three companies were worth over $300B (!), with NVDA accounting for 2/3 of it.

With the current combined market-cap standing at over 40% (or circa 60%, three months ago) below the combined peak levels, it might seem (to some of you) as if these companies lost their "mojo". We, however, wouldn't make such a generalization; at least, not yet...

  • Revenue (Quarterly)

Over the past two years, MU was on a league of its own here, more than doubling its average quarterly income from 2016 to 2018. However, MU golden time is now over, and its flagship DRAM and NAND flash products are expected to experience a continuous decline in the foreseeable future.

This is exactly why the analysts at Wells Fargo (WFC) just lowered their expectations out of Micron, citing DRAM price concerns. WFC analysts are now expecting revenue of only $22.58B (down from $29.9B; -25%!) and EPS of $5.26 (down from $5.87; -10%) in CY19.

Risks include highly volatile pricing for DRAM and NAND flash, the need for relatively high levels of capital investment, and large swings in Micron’s profitability that have occurred in the past and which we think are likely to continue in the future.

Nonetheless, WFC still rates MU as an "Outperform" with a $50 PT, as analysts at the bank still believe that the "long-term positive secular demand drivers remain intact."

The big shocker here is Nvidia, a company that has seen massive growth over the past five years, until the fourth quarter of 2018. On 1/28/2019, Nvidia caught the market off-guard, by cutting significantly its Q4/2018 guidance , due to weaker gaming and data-center sales.

Possibly learning from the Apple (AAPL) experience, Nvidia has prepared the market for the worst, and then - when the actual numbers came in (February 14th) - both companies pleased investors with more upbeat guidance for future years.

In the case of Nvidia, the company guided for flat-to-slightly-down revenues in FY 2020, still better than the market, more-bearish, consensus (following the 1/28/2019, earlier "preparation") expecting a 5% decline.

Q4 was an extraordinary, unusually turbulent, and disappointing quarter. Looking forward, we are confident in our strategies and growth drivers. - Jensen Huang, Nvidia's CEO

Earlier this week, the company announced the ~$6.9B acquisition of Mellanox (MLNX). Although, once the deal closes, Nvidia expects the purchase to be immediately accretive to its non-GAAP gross margin, EPS, and FCF, we don't think that this acquisition is going to move the needle by too much.

Looking at the most recent FY data:

  • Revenue: NVDA: $11.72B, MLNX: $1.09B (Implying an increase of 9.3%, assuming zero cannibalization)
  • Net Income: NVDA: $4.14B, MLNX: $134.26M (Implying an increase of 3.2%, assuming zero cannibalization)

So, even if we assume a "sum of all parts" (with zero cannibalization), based on last year's numbers (and we don't believe NVDA would be able to repeat those...), this deal is not a game-changer.

While Nvidia is likely to be the better buyer of Mellanox - as opposed to other interested parties such as Microsoft (MSFT), Intel, Xilinx (XLNX), or IBM (IBM) - this doesn't mean that Mellanox is in position to create miracles for Nvidia. A nice, profitable, addition? Sure. A game-changer? Certainly not!

AMD also came up with a downside guidance for Q1/2019. The company sees revenues of only $1.2B-$1.3B (vs. market consensus for $1.47B), citing graphics softness. (Large crypto-currency mining inventory needs to be cleared out.)

All in all, everybody is talking/playing down the numbers in 2019 vs. 2018.

  • EBITDA (Quarterly)

Micron benefited from phenomenal margins (more on that below) that has been translated into an unbelievable EBITDA. To put this in the right context: MU's most recent quarterly revenues were 3.6x and 5.6x bigger than those of NVDA and AMD, respectively. At the same time, MU's EBITDA is 12.4x and 68.5x greater than that of NVDA and AMD, respectively.

  • Operating Income (Quarterly)

The exact same picture (as the EBITDA) is drawn when it comes to the operating income. MU is (i.e., was) living on a (profitability) planet of its own.

Note the sharp drop of Nvidia's operating income, bringing it back to levels the company has seen only more than five years ago.

  • Net Income (Quarterly)

Looking at the below, and the previous charts, one can't wonder why AMD, a company that is growing its top line very nicely, and enjoying pretty hefty margins (more on that hereinafter) - is unable to translates its growth into an improved profitability.

When it comes to bottom line, AMD still has a long way to go.

  • Operating Margin (Quarterly)

We spoke about Micron's unreal operating margins - and there you have it. While Nvidia was getting close, touching the 40% mark not too long ago, MU has been making 40%+ (and at some point, even 50%!) over the past 18 months or so. That's nothing short of insane.

If you asked yourself why AMD is struggling in turning (revenue) growth to profit - here's your answer. Too high costs/expenses are eating into most of the growth, leaving too little for shareholders to take home.

  • Profit Margin (Quarterly)

Even more overwhelming than the operating margins are the profit margins. Micron is almost able to turn this on a 1:1 ratio. Every $1 of operating profit is losing very little along its way to become an almost $1 of net income.

If only MU would be able to continue operating like this forever, I assure you that its stock wouldn't be trading with multiples of 4-6x for so long...

  • PE Ratio (Trailing Twelve Months)

Speaking of which, Micron TTM P/E is probably the most attractive Wall Street has ever seen from a company of that size, over such a (relatively) long period. The problem of the company/stock is that everybody knows that this "gold rush" is going to come to an end, and basically, the stock reflected this ever since the rush has started.

AMD, on the other hand, is trading at multiples that can't be justified from a pure financial perspective. Of course, this is the mirror image of MU, as investors expect AMD top and (especially) bottom lines to improve over time, thus taking the P/E down.

NVDA is (or was supposed to be) the "adult in the room", delivering more reliable/consistent results. Came the last quarter and reshuffle the cards. We expect NVDA's TTM P/E to cross the 30-mark soon, on its way to 40!

Future Earnings and Other Financial Aspects

  • PE Ratio (Forward)

If you think to yourself, MU is a bargain, trading with a forward P/E of only 5.2x - think again. Recall that the TTM P/E is 3.2x, so we're talking about a 63% rise in the multiple. That could be a big ouch for shareholders, unless profitability won't go down as harsh and as quickly as some analysts expect.

In light of the disappointing earnings by NVDA, it's no wonder to see its forward P/E almost meeting that of AMD in the 30s. It won't be the first, nor the last time this duo might meet on this chart (going forward). Nonetheless, we believe that over time the red line (of AMD) will be lower than that of NVDA (blue line).

And who knows? In 2-3 years, we might have a gathering of all 3 companies, trading at (more or less) the same multiple. That would be something that we haven't seen for a long-long time.

  • Revenue Estimates*

*For Current Fiscal Year, Next Fiscal Year, and 2 Fiscal Years Ahead

Here are the current market revenue estimates for the foreseeable future:

Now, let's present the above data a bit differently:

Source: Author, based on Y-Charts data (as of 3/13/2019)

What can we learn out of this table when it comes to future revenue growth?

  1. MU is out of the growth game completely.
  2. Although NVDA and AMD are expected to grow about the same, on average, AMD growth is expected to be more smooth, steady and consistent, while for NVDA, this is more a future thing, following the recent bumps (and assuming the company overcome those).
  3. Past growth is indeed no indication for future growth. The latter is going to be much softer and slower for these companies.
  • EPS Estimates*

*For Current Fiscal Year, Next Fiscal Year, and 2 Fiscal Years Ahead

Here are the current market EPS estimates for the foreseeable future:

Now, let's present the above data a bit differently:

Source: Author, based on YCharts data (as of 3/13/2019)

What can we learn out of this table when it comes to future revenue growth?

  1. AMD is in a league of its own. However, it's important to bear in mind that AMD is coming off a very low point, thus it's easy for it to outperform. Furthermore, expectations from AMD to finally start making serious money (at the bottom line) are long due. Will this time be different (and AMD finally move closer to making $1/year)? Only time will tell.
  2. MU is also in a league of its own, but about 3 levels below that of AMD... The market is basically seeing MU's 2018 EPS getting halved in ~3 years. If so (i.e. $5.75 EPS) and based on a stock price of ~$40, the multiple will be about 7x, more than doubling the current TTM P/E!
  3. Currently, the market believes that Q4/2018 and 2019 would only be a soft dent and a short-term delay in Nvidia's long-term massive growth story. Just like few other analysts, we strongly doubt that NVDA would be able to grow at an >30% pace in such a short notice. It seems unrealistic because such a growth rate has been achieved only during the crypto-boom, which even the company itself acknowledged is gone.

How Do We Play This Out

Let's start with Advanced Micro Devices:

With the above chart in mind, here's how we play/ed AMD, at the Wheel of FORTUNE, over the past 16 months:

No. Date Action
1

December 12th, 2017

BUY AMD @ $9.90 (initial position)

2

February 5th, 2018

BUY AMD @ $11.11 (adding to position)

3 July 31st, 2018

SELL (to open) AMD 01/17/2020 20 (covered) CALL for $5

4

September 4th, 2018

SELL AMD @ $27.78 (closing long and opening a short)

5a

5b

October 25th, 2018

October 30th, 2018

BUY AMD @ $17.91 (closing the short position)

BUY (to close) AMD 01/17/2020 20 CALL @ $3.50

6 March 12th, 2019 SELL (to open) AMD 04/26/2019 26 CALL @ $1.05

Note: Once we sold the long position on 9/4/18 (No. 4), the $20 CALL that we sold on 7/31/18 (No. 3) had switched from "covered" to "naked". This naked CALL sale got closed on 10/30/18 (No. 5b)

Here's what we wrote to our subscribers regarding the most recent trade (No. 6) that was executed earlier this week:

This is a sale of a naked (i.e. without owning the stock) CALL option, thus the risk is obviously high!

As you already know, our conviction when it comes to semiconductors is quite weak. As a reminder:

  • Last week, we pulled our "top pick" designation from Micron.
  • Three weeks ago, we sold a covered call on Nvidia (Trapping Value is even shorting it)
  • On 09/04/2018, we didn't only sell out of our long Advanced Micro Devices position, but we also sold the shares short (i.e., reversed our position) at $27.28

To make a long story short, we don't expect this segment to come up with great news that will lift prices much higher from here, even under the pinkish scenario.

Furthermore, if AMD was overvalued (in our view) at $27.28 over six months ago, you can imagine that today, we are even less enthusiastic on the stock at that price.

The option we are selling will make us short AMD only if the stock move above $26 and get assigned to us. Even then, the net price (from which we will be short) is $27.05 = strike ($26) + premium ($1.05)

At the moment, AMD is estimated to report earnings on 4/24/2019, two days before the option expires. Of course, any delay in reporting (to a date later than 4/26/2019) would make this option less risky than it currently is (under the assumption that AMD will report earnings two days prior to the expiry date). The reason for that is simple: (extra) volatility.

Since a big chunk of an option is attributed to the current and expected volatility of the stock, a removal of the post-earnings move threat will greatly diminish the value of the premium here.

In any case, here are the possible scenarios out of selling this naked call option:

Keep in mind that selling a naked call means that the upside risk is (mathematically) unlimited, as the stock can rise to infinity (and beyond). Therefore, the margin requirement here is high, and this is exactly why we rather play this trade with a relatively short-term expiry date (only 6.5 weeks away); otherwise, selling the AMD 07/19/2019 26.00 CALL for $2 (almost twice as much) would make a lot of sense, as 7/19/2019 is likely to fall before AMD reports earnings for Q2/2019. Putting it differently, in both cases we are likely to face only one ER till expiry date. If you don't mind locking the amount of margin that this trade requires - feel free to execute the July 19th expiry instead of the April 26th expiry.

As for Nvidia:

With the above chart in mind, here's how we play/ed NVDA, at the Wheel of FORTUNE, over the past 17 months:

No. Date Action
1

November 13th, 2017

SELL (to close) NVDA @ $217

2

December 12th, 2017

BUY (to open) NVDA @ $184.50 (only 1/3 position)

3 June 4th, 2018

SELL (to open) NVDA 01/18/19 280 (covered) CALL @ $22.10

4 February 20th, 2019 SELL NVDA 01/17/20 160 CALL @ $26 (covered and naked!)

Note: The last trade on 2/20/19 (No. 4) was executed with a "double dose": i) covered sale (risk rating: 1; the safest) against our long (1/3 of a full) position; ii) naked sale (risk rating: 5; the riskiest), which means that the risk is much greater.

Here's what we wrote to our subscribers regarding the most recent trade (No. 4) that was executed three weeks ago:

We held NVDA before the service was launched, and we sold it at $217. Then, when the stock pulled back in November-December 2017, we opened a smaller (about 1/3 of what we sold) position rode with it all the way to circa $300 (October 2nd 2018)

In mid-2018, we already felt the stock is too pricey so we sold a $280 CALL but this option never been assigned, as the price dropped like a rock and at the expiry date, the stock price already more than halved (!)

So, we end up with the premium we received, but also with the stock...

Our revised price target for the stock is now $100-180; yes, there's no typo in here. We won't be surprised with any move (short-term) up or down. However, as you can clearly see, the downside ($60) is 3x the upside potential ($20), so clearly, we lack conviction to the upside, and we fear of the downside more. Therefore, this is a very easy trade for us, as it bears zero risk for us.

If the option gets assigned - we would receive a total (net price) of $186 (= strike of $160 + premium of $26), a bit higher that our PT and even higher than the price we paid ($184.50).

If the option doesn't get assigned - we are getting almost half the downside risk, upfront. It might be needed to ease the pain, if and when...

Since we don't see much chance for the stock to trade significantly higher, we are actually trading this in both proposed ways/dual hats:

On one hand, we sell covered calls that fully protect our long position. Should the option get assigned, we will book those against the sale price, as we usually do with assigned options (P&L goes fully to the underlying stock).

On the other hand, we do the same trade as a naked sale (obviously for a much smaller allocation), which is very risky, but we keep it small.

So, on the monthly report, you'll see two trades, one with risk rating 1 and one with a risk rating 5. Don't get confused by this dual presentation.

Here are the possible scenarios of this trade:

Last but not least - Micron

Truth is that we named MU one of our "top picks for 2019*", among the information technology sector, when the stock was trading at extremely distressed levels (along with the entire market) in late December 2018.

We also sold MU 03/15/2019 31.00 PUT option on 01/03/2019 for premium of $2.75. This option is expiring today (March 15th 2018) worthless. You can see the annualized return on this trade, per the below image (stating the possible scenarios at the time of sale).

However, at the end of February, we actually "dethroned" MU from its "top pick" designation after a quick run of circa 50% from trough to peak in a matter of just two months!

Chart Data by YCharts

With today's expiration of the option we've sold, we now have no direct and/or indirect position related to MU.

Outlooks And Price Targets

We already touched upon each company's own guidance for the next quarter/year, so there's no point repeating ourselves. Wall Street, anyhow, is taking these outlooks into consideration when assessing and estimating future earnings.

Below, you can see the current EPS estimates for the next two years. Pay close attention to the sharp down revisions that EPS estimates for both NVDA and MU have gone through over the last three months.

Wall Street has reduced its expectations by 20%-30% across the board, and this might only be the beginning, not necessarily the end.

Source: Author, based on Wall Street Journal Data (as of 3/13/2019)

At the same time, the EPS estimates for AMD remain very stable and haven't changed a bit. This means one of two things: either AMD is a very predictable company (which it isn't...), or a much more promising operation (it is!) compared to the other two semis.

As for what Wall Street sees for the three companies when it comes to PTs, it's all over the place, ranging from euphoria to depression.

As always, we ignore the extreme forecasts (i.e. "highest" and "lowest") and stick to Main Street, namely referring to the average and median estimates, as those tend to be more reliable and a better reflection of the overall sentiment. Having said that, when it comes to analysts' expectation our main motto is: Respect and suspect; more of the latter than of the former...

Source: Author, based on Wall Street Journal Data (as of 3/13/2019)

Looking at the above table, here are few quick-immediate observations:

1. MU has the highest potential to the upside (based on the average, median, and highest PTs) as well as the lowest downside potential (based on the lowest PTs).

2. In spite of Wall Street counting on AMD to perform more solidly than its peers, there's not much meat left on this bone, per the average and median PTs. Putting it differently, AMD is fairly valued as is (another justification to our recent sale of a naked call here).

3. Although NVDA seems to benefits from a bigger upside potential, compared to AMD, you can see that even the highest PT is not that upbeat. Nobody thinks that the stock is going to meet its 2018 highs anytime soon.

Bottom Line

At the very minimum, 2019 is going to be a tough year for NVDA and MU, compared to 2018. While for MU it's going to be tougher longer than this, with a lot of luck, NVDA might come up strong again in 2020. Having said that, the expected decline (in the company's earnings) in 2019 is so deep that we wouldn't buy the dip that early in the year. We expect more pain before a meaningful gain.

Note that our friend and partner, Trapping Value ("TV"), estimate that "NVDA will hit (brace for this) just $2.28 in earnings per share", based on this calculation:

Source: Nvidia: Game Over

While we are not as bearish as TV is, we certainly see the stock trading closer to $100 than to $200 inside 2019, just as it did last year.

Our strategy here, as outlined above, means that should NVDA trades above $160 towards/at expiry date - we would sell our long position and open a short position at a net price of $186. We certainly don't believe that the stock is worth more than that, to say the least.

Overall, our view for NVDA is bearish, and, as such, our current rating for the stock is "a conditional sell" = hold for now, but sell ~$180-185.

MU is the one we find it most difficult to make a call on, as we wouldn't be surprised to see it above $50/share, or below $30. Both cases have good arguments, and it's all about how severe the "injuries" will be, now that its main activities are over the peak. The key question is: Will it be a gradual slide along the mountain, or more like jumping off a cliff?

Our strategy when it comes to MU is to wait. We fear that the upside here is very limited, in light of the market expectations that it's going to look more like falling off a cliff rather a gradual adjustment to a shrinking demand for the company's products. Therefore, at the current juncture, we would rather step to the sidelines and watch. Should the stock trades back to the $20s, we might take charge again.

Overall, our view, for MU is bearish too, but we don't view the stock as being as vulnerable as NVDA is. As such, our current rating for the stock is "neutral", but we would be buyers at much lower levels, due to the excessive risk (and limited upside) that we identify here.

The most interesting, and promising, name among this trio is AMD, but even here, most of the growth potential is expected for periods beyond 2019. For us, as you will be able to see below, AMD has always been a valuation story. As long as the company is making peanuts on the bottom line, we deem the stock attractive below $16, and overvalued above $25.

Only once we see AMD making $1/year in EPS, would we start thinking of the stock as possibly worth moving into the $30s to stay there.

Overall our view for AMD is quite bullish for the long-run, just as it has been over the past 16 months. The main problem we see for AMD is its inability (thus far) to translate revenue growth and better margins to net income. AMD isn't a start-up anymore and when such a mature company earns only $0.32/year, there's no way we can justify valuations that represent multiples of 81-107x, when the stock traded at prices of ~$26-$34.

As such, our current rating for the stock is "neutral", as we wait for AMD to "show us the money". There's also the risk that its peer-group (NVDA, MU and possibly even the 3 musketeers: AVGO, QCOM, NXPI) might drag AMD's stock price down along with them.

Don't forget what we always say and strongly believe in: Macro Trumps Micro. Even the best, most resilient, stock will find it hard, if not impossible, to withstand a stormy, unhappy, market.

Over the long-run, it's macro - global, general, market trends - that rules (and leads) the world, much more than micro - company's specific attributes/news.

You light the skies, up above me A star, so bright, you blind me, yeah Don't close your eyes Don't fade away, don't fade away, oh

Source: "Rule the World" - Take That

All in all, it's safe to say that semiconductors have ruled the world in recent years, up to the end of Q3/2018.

Don't get me wrong: Semis aren't going anywhere, and it's not like we expect this boiling-hot segment to fade away completely. That is true for the trio in this article, as much as it true for the trio in my previous article.

Nevertheless, just as the majestic performance of "Take That", from the London 2012 Olympics Games, Closing Ceremony implies: Semis, you've lit the skies, up above me, Many stars, so bright, you've blinded me, yeah, Don't close your (production) lines, Don't fade away, don't fade away, oh

Sooner or later, you can rule the world (again)!

Disclosure: I am/we are long NVDA. 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: Short NVDA 01/17/2020 160.00 CALL
Short AMD 04/26/2019 26.00 CALL