Hiding In Plain Sight: Micron's Q4 Beat

Summary

• The Q3 earnings call was great news for Micron longs.
• Sanjay Mehrotra makes his first appearance leading the call and emphasizes his strategy.
• Some things left unsaid are important to consider.
• Ernie reminds us to take a hard look at bit output in Q4, and we like what we see.

Micron’s (NASDAQ:MU) recently concluded Q3 concall was a good one, packed with mostly promising news for investors. The fact that the market reacted the way it did is disappointing, but I’ll leave the reading of those particular tea leaves to other commentators. Let’s instead spend some time unpacking the news that is pertinent to the fundamentals of the long position.

Call me weird, but I personally find the analysis of these concall transcripts endlessly fascinating. My academic background, (eons ago) was in the arcane area of what used to be called “Sovietology.” There was so little information coming out of the evil empire that anyone seeking to understand what was going on needed to look for subtle clues in the information flow that we did get. These corporate concalls are very similar. Time is limited, and there is so much that Micron wants to keep hidden and so little that it really wants us to know. This leaves us in a position of reading between the lines, looking for clues that might come from patterns, phraseology, and intonation – looking for the tells that come from what was said and what was not said.

The headlines of the call were excellent. A beat on revenues – a quarterly record at $5.57B - and best-ever non-GAAP earnings of $1.62/share to go along with that, both numbers above Q2 guidance. So, that was good but the guidance was even better, with Q4 revenues projected to climb 6% to a range centering on $5.9B and operating income to jump 11% to a range that finds $2.3B as its center – all of this resulting in an EPS number of $1.73 to $1.87. Who could complain about that? Laissez les bon temps rouler.

And, there is every indication that the good times will continue. Bit growth is still strong, though it will moderate somewhat as the company aggressively pushes the pace of process upgrade in both DRAM and NAND. On the DRAM side of the house, the 1X upgrade continues apace as 1Y develops in the lab, and that situation has its analogue in NAND as the company rolls out Gen 2 3D NAND (the 64 layer process), with Gen 3 in development. 1X and Gen 2 will both see material output this FY and are on track to be a majority of bit output in FY 2018.

So, all is good on the production and development side of the house. When we pair this with the strong demand environment we are seeing green lights well into the future. Here’s how Mehrotra characterizes it:

“For calendar 2017, we expect DRAM industry bit supply growth of between 15% and 20%, slightly below our view of demand growth. For NAND, we expect 2017 industry supply growth in the high 30% to low 40% range, constraining what would otherwise be higher demand. We expect healthy industry demand to persist into 2018, supported by continued strong growth in both DRAM and NAND demand, reflecting broader trends in the data center and mobile markets, as well as increased adoption of SSDs across enterprise, cloud and client PCs.” - SA Transcripts, Micron Q3 earnings call

Mehrotra is mirroring comments made by his industry peers when it comes to demand staying strong well into 2018. One area where Micron is out of phase with its competitors is its consistently bullish view of industry production. Western Digital (WDC) is on record predicting low to mid 30s industry supply growth, and Samsung (OTCPK:SSNLF), the industry’s share leader with 40%+ of the NAND business, has stated it will only grow output by roughly 30% this year. So, why is Micron guiding so high for the industry? Could it be that it is so significantly outgrowing its competitors that it expects to make an outsized impact on industry numbers? We’re going to explore that supposition below.

For now, though, let’s talk about what was not said. Notably – and significantly - Sanjay did not mention 3D XPoint. (Strangely, the analysts on the call did not ask him directly about it.) In any case, his silence in this regard is not good news for the program. It may not be terrible news, but the failure to mention XP was not inadvertent. It may not mean anything substantive at this point, and everything may be fine with the product and the program, but we just don’t know. Commentators and analysts that I respect, including Arthur Hanson, have recently voiced concerns about the program. One thing we can be sure of – Sanjay Mehrotra will not be saddled with erroneous expectations about a program he had no responsibility for. If he is deeply concerned about XP, he would have used the concall to reposition and reset expectations, and he chose not to do that. So, no good news on XP and no really bad news – yet. At this point, the operative expectation remains unchanged. QuantX product will ship this year but not represent material revenue for the company. Hopefully, we’ll get more (and positive) news on this vital technology in Micron’s Q4.

On to some other substantive news from the concall. First off, Sanjay did an excellent job of communicating his strategic priorities, and we will be well served by focusing on them. He couldn’t have put it more succinctly. As he said - “execution and competitiveness are my primary focus.” Here’s the expanded message:

“Our execution and competitiveness are my primary focus, particularly accelerating the ramp of new technologies into volume production and introducing new products quickly, both of which are essential to delivering innovative solutions at lower costs and strengthening Micron’s business fundamentals. Micron has a tremendous portfolio of technologies and core capabilities. Our goal is to leverage these to provide high-value products and solutions that improve our revenue mix. We will target high-growth opportunities and seek out partnerships with leading companies in the ecosystem to position Micron for long-term success. - Micron Earnings Call Prepared Remarks

We would do well to keep in mind the nexus between “competitiveness” and “high value” in Sanjay’s formulation. How do we know Sanjay is serious about this? Consider his first two hires: Sumit Sandana as Chief Business Officer and Jeff VerHeul as SVP of Non-Volatile Engineering. Both report directly to the new CEO, and both are ex-SanDisk executives whom he knows well. These are his guys – known performers – who live and breathe high-value product solutions. In particular, we should remember that Sandana’s role at SanDisk incorporated the Enterprise Solutions portfolio – a position from which Sandana drove the Enterprise business from nothing to almost a billion dollars at the time of the acquisition. Sandana will have all of the BU’s reporting to him, and with the addition of the corporate Business Development team, he will have all the resources that he needs to integrate and improve how the company goes to market and what products it goes to market with.

In that role, VerHeul will be his partner for NVM technology while Brian Shirley retains the DRAM and “emerging memory” technology portfolios. This move, taking NVM off Shirley’s plate, was long overdue. The focus that NAND gets is now equal to DRAM, and this is very good news for Micron investors. Micron is still a DRAM company, but the day is coming – certainly by the end of the decade – when NAND/NVM products outweigh DRAM products in revenue and profit contribution. And yes, in case you were wondering, XP is in VerHeul’s portfolio of NVM product.

The best news of the call was the came when Ernie signaled the Q4 beat that is baked into quarter when he discussed bit output for DRAM and NAND in FY’17. Here’s the tell. First DRAM.

DRAM non-GAAP gross margins for the third quarter increased 10 percentage points sequentially to 54%, driven by a 6% cost per bit reduction and better product mix. As a reminder, we noted last quarter that second half fiscal year 2017 DRAM bit output would be about 10% higher than first half fiscal year 2017. – SA Transcripts, Micron Q3 earnings call

Now NAND.

Trade NAND revenue increased 21% compared to the prior year quarter, reflecting a 17% increase in bit shipments and a 3% increase in ASPs. Non-GAAP gross margin was 41%, up 10 percentage points, driven by a 12% cost per bit reduction and better product mix. As a reminder, we noted last quarter that second half fiscal year 2017 bit growth would be about 30% above first half fiscal year 2017. […] Consistent with DRAM, we considered this bit growth pattern when we provided our 2-year bit growth CAGRs earlier in the year. -SA Transcripts, Micron Q3 earnings call

Hmmm. Do you notice the pattern here? Do I need to remind you that we are being reminded to take heed of its two-year guidance supplied in the analysts conference of 2017 and its intersection with its guidance from Q2’s concall? Something about this second half bit growth pattern is obviously very important to Ernie. Maybe we should take a closer look, no?

So, let’s work the math and see what a 30% half over half increase means for Q4 NAND production. Here are the numbers from this year.

So, we start with a nominal output of 1,000 at the end of Q3 FY ’16, and let’s work the reported output growth numbers quarter by quarter. We need to solve for a number that will provide 30% growth of bit output in the second half versus the first half, and what we come up with is 21%. Let’s plug that into their NVM revenue last quarter, add Micron’s projected ASP increase of 3% and voila, we get a very nice bump in NAND earnings in Q4.

Note some assumptions. I have bumped trade NAND gross margin by 1% to 42%, and further assumed a 3% GM for the Intel (INTC) business, and 35% GM for its NOR business. The end result of following Ernie’s reminder is an NVM business that adds up to $2.48B in Q4 revenue generating almost $1B in gross margin.

If we run the same exercise on the DRAM side, we get to a 12% bump in bit output in Q4. That in turn delivers the following on the revenue and profit ledger (assumptions as noted on bit cost).

(Note my assumption of a bit cost reduction in line with Q3 that I believe is reasonable given Micron’s rapid ramp on 20nm and 1X product. This assumption leads in turn to a 2.5% GM increase from Q3’s 54% figure.)

Put all this in the mix and Ernie’s reminder gets you get the following:

Will it be that good? Is Ernie hedging that much on his guidance? More closing thoughts on that below, but for now, let’s reflect on that earlier question regarding why Micron is consistently higher on its industry forecast for NAND growth than its compatriots. Take another look at Micron’s year.

110%. As in one one zero and change. Wrap in Q4 of 2016’s 16% bit increase and it gets even more impressive. Put another way, Micron will be producing more than 1.5 times more bits in Q4 of 2017 than it did in Q3 of FY 2016. And, there is more to come in FY ’18, though the pace of growth will moderate.

Let’s close. There is a lot more to discuss, but we’ll hold it for another day. Perhaps a parting sentence or two regarding Ernie’s guidance. On balance, it’s reasonable. If we could have more certainty about pricing over the next two months, we could quibble, but we can’t. One example – change the 3% ASP increase for DRAM in Q4 to a 3% decrease, and the EPS number goes to $1.93. The NVM number isn’t quite so sensitive since it’s a much smaller business, but that same 6% swing would bring the number down 5 cents. So, all in all, his $2 is a sporting proposition - one that he is likely to beat with room to spare. Thanks for the reminder, Ernie.

Analyst’s Disclosure: I am/we are long MU, WDC. 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.