Dog Days

Woof woof. Dog days are here again.

SanDisk, at $18±, is bumping along at 5 year lows. Personally, I feel the potential SanDisk story is intact- in other words the upside is still there. It may take a while, but the emerging solid state memory markets seem to be shaping up as gorillas and SanDisk seems to have the technical expertise and IP to make a run at the big time.

SanDisk could certainly come up short. Mr. Market seems to think the odds don’t favor the company – at least near term- and has priced SNDK accordingly.

Personally I worry a bit about focus within the company. Its all fine and well to throw dollars at potential sexy consumer niches, when the times are flush. Its another matter when belts need tightening.

Why SanDisk needed to buy MusicGremlin (10 June) is beyond me. Purportedly it was about digital content distribution technologies.

Maybe this one will work out where FanFare didn’t- but I’m not holding my breath. Let’s face it, SanDisk isn’t going to come up with a viable alternative to iTunes. The whole TrustedFlash-ecosystem vision is looking a bit dated.

That said, I like the SNDK strategy of leveraging its 220,000+ storefronts worldwide into higher-margin, flash-memory-intensive consumer products. I also like the strategy of targeting high-growth emerging markets with their 1 billion new consumers.

(Somewhat) Recently, Craig Ellis @ Citi, a longtime SNDK bull, threw in the towel and downgraded SNDK from Buy to Hold. Excerpts of his 19 June report follow, along with follow-on comments from 27 June, and 27 June UBS comments on Micron. These reports although a bit stale are, nonetheless, still relevant.

With NAND prices in free-fall, and worldwide macro economic concerns, Ellis’s move was understandable, as Shlomi Cohen put it. I find Shlomi’s comments regarding SanDisk’s expectations for massive demand from 2010 onwards particularly interesting, but that’s going to have to wait for a later post.

The big NAND headache today is that too much supply is chasing too little demand. This is hurting all players, SanDisk included. I find it remarkable that the Intel division that makes memory chips managed to lose $706 million last quarter on $300 million of sales.

Find it somewhat ironic that it appears that the winners in the commodity NAND business will likely be those that have special expertise in the fundamental technology.

SanDisk claims that all NAND is not created equal.

The next couple of years should be telling as to whether SanDisk’s x3 ABL-enabled NAND is ready for prime time. 43-nanometer x3 will be sampling by year end 2008 and in production early 2009. By year end 2009, SanDisk says that 50% of its bits are expected to be 43nm x3.

SanDisk says ABL- enabled x3 gives a cost benefit of 15% to 20% at the product level compared to the same technology on two-bit per cell.

Equally promising, ABL purportedly boosts regular MLC (x2) program throughput by 240% (16Gb chip), leveling the playing field with SLC. ABL purportedly also has an edge when it comes to energy efficiency.

ABL MLC (x2) sounds tailor-made for MLC SSDs.

Given the scheduled 2009 volume ramp of 43nm x3, would be surprised if SanDisk isn’t also confident of the prospects of ABL-enabled x3 SSDs for 2010 and beyond.

Besides commoditization and oversupply of NAND, SanDisk is also facing uncertainty over Samsung’s IP licensing renewal (August 2009 expiration). Personally think that Samsung is going to renew (at lower rates), but likely at the last minute.

I’m thinking that at some point a post devoted just to IP issues would be worth the effort. Other IP issues besides Samsung include controller etc litigation and strategic considerations including the Intel/Micron JV and an update on Solid State Storage Solutions LLC.

The (long term) extreme upside, and pot-of-gold, would seem to be 3D.  My sense is that the licensing agreement between SNDK and Toshiba hints that progress has been made. I have suspected for a while that SNDK has 3D switchable R/W working in the lab.

The Toshiba agreement would seem to clear the way for working out manufacturing issues/techniques. R&D for 3D OTP at 45nm is being carried out at TSMC. In parallel, additional 3D R/W activity is focussed on scaling the diodes down to 32nm/22nm.

From my perspective the Toshiba agreement probably focusses on 32nm/22nm. No doubt analysts will ask Eli about 3D tomorrow on the quarterly results call. I don’t expect him to say much. Stakes are too high.

The 3D payoff, if there is one, looks several years out. And the dog days drag on.

**** Citi 19 June 2008 SanDisk Excerpts****

SanDisk Corp (SNDK)
Not Enough Help from the Macro, Downgrade To Hold From Buy

Craig A Ellis
Asiya Merchant, CFA

Hold/High Risk 2H
from Buy/High Risk
Price (19 Jun 08) US$23.44
Target price US$27.00
from US$35.00
Expected share price return 15.2%
Expected dividend yield 0.0%
Expected total return 15.2%
Market Cap US$5,268M

Quick Call – We are reducing our C08 and C09 revenue and EPS forecasts and our rating to Hold from Buy based on cautious findings in this week’s global NAND pricing and product end demand fieldwork. On un-changed valuation our target price declines to $27 from $35.

Not Enough Help From The Macro In 2H08 To Stick With Buy – Positive on the SNDK long trade this yr, we expected rising contract pricing by mid-June (big tier-1 OEM buys, tier-2/3 allocations, capacity shut-downs) to be a 2H08 SNDK royalty and product revenue/gross margin tailwind. While SNDK’s 2Q08 product pricing looks firm (Asia checks, US April/May sell-through) and its 43nm mfg transition solid, expected mid-08 under-supply is not emerging as we had expected given recent demand weakness, eroding the basis for our call.

Asia Trip/Global Field Checks – In the past week we find: 1) Only limited big tier-1 customer engagements/buy programs at multiple high-volume mfgrs, 2) cautious expectations on big European embedded handset programs for 2H08, 3) slowing May/June industry and SNDK handset card demand in China/Asia, and thus 4) muted tier-2/3 customer allocation risks into August. Checks on the 2HJune contract pricing are mixed – we think a negative reset is likely (- 5%) w/a sustained increase potentially as late as August, lasting only through September. In sum, a more disappointing outlook than just one month ago.

Estimate Reductions – Estimate changes reflect: 1) lower 2Q08 and 2H08 product bit growth (cards, drives, SSD), 2) more aggressive 3Q08 and 4Q08 contract pricing assumptions (impacting royalty), but as a partial offset 3) more benign 2Q08 retail product pricing. CIR 2008E and 2009E new/old to $1.23/$1.29 from $1.40/$1.73 and are now 15% and 30% below the Street.

Tactical Edge – Bulls will push “cheap” valuation and a potential 3Q08 contract up-tick but: 1) co-specific C2H08  bit growth, inventory, and GM risks are growing, 2) the stock’s seasonal peak is nearing, and 3) lt buyers are likely to sell strength ahead of late-C08’s pricing and stock risks (0.7 r-squared).

Risks To Our Call – It appears 20mm de-commissioning is occurring largely as expected (Hynix, Toshiba/SanDisk, Samsung), a supply-side positive. Should product end demand snap back strongly in 3Q08 and 4Q08, our new call could prove overly conservative.

Recent Demand Erosion Undermines Seasonal Trade – Downgrading to Hold From Buy

We are downgrading SNDK shares to Hold/High Risk from Buy/High Risk. Our C08 and C09 revenue/ EPS estimates are now 5%/15% and 10%/30% below the Street.  As background, our positive stance was heavily dependent upon rising late-2Q08 and 3Q08 NAND contract pricing (tier 1 OEM product refresh, 200mm fab shutdowns) and prospects for favorable, macro-driven 2008E estimate revisions given a lack of company-specific new product activity this year (next year SSD’s should emerge). We had expected such macro strength to catalyse a share rally through the seasonally strong June through October time period (Figure 3, 4), similar to C2007 (see 5/12/08 “Moving into the Seasonal Sweet Spot, Reiterate Buy”).

So what has changed? While checks support the firm SNDK product pricing and solid 43nm manufacturing transition we expected, recent card and drive end demand erosion in Asia, and still-modest embedded handset orders (from European companies) have gone against our call.  For the stock, contract pricing has mattered twice as much as EPS revisions (R-squared of 0.70 versus 0.35) so the loss of the macro catalyst has significant implications even if SNDK can keep gross margins propped up near term.

On revised estimates our target price declines to $27 (from $35) suggesting relatively balanced risk reward. Upside of 15% exists to our new target versus downside to $20 on mid-teens multiple on new C08E EPS estimates. While a painful move with the stock off 30% from recent highs, the body of evidence we see no longer argues for putting new money to work in the name and thus we move to the sidelines.

Top picks in our coverage include small cap ONNN (target $14) and large cap MRVL (target $22), each of which is early in enjoying a structural up-tick in EPS power which we think can take estimates higher in 2H08 and 2009.  Add to positions in these names on any weakness.


We derive our $27 target price using selected forward 12-month price-to earnings and enterprise value-to-sales applied to C09 estimates. Our target multiples are set on a 15-30% discount to the 3 yr historical medians. We believe selection of the discount on the three-year time period is warranted as it includes periods of supply imbalances (over/under), driven by increasing supply competition, end demand driver growth phases (handsets, USBs), and inventory correction representing a good cross section of NAND industry phases going forward. Our greater 30% discount on EV/S reflects visible recession risks on consumer demand, lower royalty revenues, and multi-year sector multiples compression. For our P/E, we use a 17x target multiple applied to our calendar 09 EPS estimate to yield a target price of approximately $21. Our 17x multiple is derived from a 15% discount on the company’s three-year historical range of P/E multiples that ranges from 15.4x to 30.4x. On an EV/S basis, a 1.7x multiple applied to our 2009 revenue per share estimate and net cash gives us a target price of approximately $34. We derive our 1.7x EV/S multiple from a 30% discount on the 1.2x to 3.7x 3-yr historical range. Taking the average target price derived from the two methodologies and rounding it, we arrive at our target price of $27.

**** Citi 27 June 2008 SanDisk Excerpts ****

SanDisk Corp (SNDK)
Three Fresh Data Points Support A Cautious Stance

Craig A Ellis
Asiya Merchant, CFA

Hold/High Risk 2H
Price (26 Jun 08) US$20.14
Target price US$27.00
Expected share price return 34.1%
Expected dividend yield 0.0%
Expected total return 34.1%
Market Cap US$4,526M

What’s New – Last week we reduced our stance on SNDK shares to “Hold” from “Buy”. New data points this week support this more guarded stance.

MU Cautious on NAND Balance — In framing NAND fundamentals on its F3Q08 call, MU repeatedly pointed to market oversupply, a view consistent with our Asia findings last week (SNDK handset memory card and USB drive channel inventory two weeks above normal). Elsewhere, MU guided to ~17% F4Q08 bit growth, or ~300% for the yr (vs 360% CIR ests), a supply positive. That said, ramping 43nm 16Gb product and the late year introduction of a 64Gb 3xnm product suggests upside potential (as in C07).

2H June NAND Contract Price Move 3x Worse Than Expected – Early Wednesday’s 15% drop (weighted average) was 3x greater than we expected and suggests demand weakness is more than a regional issue and casts June’s US bit-growth in questionable light. While holiday builds are ahead in mid-3Q, flat pricing through quarter’s end would imply a quarterly drop of 15% to 20%. Recall SNDK stock has a 0.7 r-squared relationship to contract pricing.

And 3Q08 Order Cuts Have Emerged – Checks mid-week point to moderate 3Q08 order cuts from a large OEM to a large memory manufacturer. While unclear if due to demand weakness or excessive original orders, implications for 3Q08 spot pricing are negative as wafers are likely in production.

Stock Strategy – Negative EPS revision risks appears heightened heading into SNDK’s 2Q08 results on royalty and product bit growth parameters (we model C08 and C09 EPS of $1.23/$1.29 vs Street of $1.45/$1.88). For the stock, at 52-week lows and suffering severe YTD underperformance (-40%) vs SOXX (- 9%) bottom-fishing looks tempting, though we discourage such action until fundamentals can stabilize and then point toward a sustainable price improvement. Risk to this cautious view is: a) a sharp snap back in 3Q and 4Q product end demand, and/or, b) broader NAND supply hiccups on manufacturing transitions (not observed presently). See SNDK note 06/19/2008 “Not Enough Help from Macro. Downgrade to Hold” for additional detail.

**** UBS 27 June 2008 Micron Excerpts

UBS Investment Research
Micron Technology Inc.

Global Equity Research Americas Semiconductors
12-month rating Buy
12m price target US$9.00
Price US$6.99

27 June 2008

Trading data
52-wk range US$13.98-5.46
Market cap. US$5.40bn
Shares o/s 773m (COM)
Free float 99%
Avg. daily volume (‘000) 5,947
Avg. daily value (US$m) 43.9

Balance sheet data 08/08E
Shareholders’ equity US$6.87bn
P/BV (UBS) 0.8x
Net Cash (debt) (US$2.31bn)

In Line F3Q Results, Persistent NAND
Weakness. Trim estimates and PT.

Results in line as margins improve modestly, NAND pricing remains weak

MU F3Q sales of $1.50b (+10% q/q) were slightly above UBS ests while EPS loss of -30c was a penny below. Despite continued solid cost reductions, GM only improved to 3% (UBSe 5%) as NAND ASP eroded further and new products ramp. We believe NAND ASP pressure is unlikely to abate near term and thus lower our margin and EPS estimates. We reduce PT to $9 from $10 but maintain Buy rating.

Slowing near term NAND bit growth a concern; Balance Sheet is healthy

We view new NAND bit growth outlook of mid-teen % q/q vs prior 30-40% as a concern due to reduced scale relative to larger peers that continue to invest in capex and a chronic market oversupply that is unlikely to lead to a meaningful ASP recovery. Balance sheet remains healthy with $1.5b in cash, F08 capex target at the upper end of the prior $2.5-3.0b range, and est operating cash flow of $1.4b.

Revising estimates down slightly on lower NAND bits, more ASP declines

Our F4Q sales est of $1.61b is approx unchanged but EPS goes to -18c from -16c on lower GM assumptions. F08 sales est increases 1% to $6.00b with EPS of – $1.24, down from -$1.19. F09E rev/EPS decline slightly to $6.71b/$0.15 from $6.75b/$0.16 on lower NAND bit growth in C2008 and lower GM.

Valuation: Reducing Price Target to $9 from $10; Maintain Buy Rating

Our new DCF-based PT of $9 is equiv to 1.0x P/BV vs historic range of 0.7-2.0x. We assume a WACC of 7.8% and terminal growth of 2% in our DCF.

F3Q08 Results Review

Micron reported May 2008 quarter sales of $1.498b (+10% q/q, +16% y/y) and EPS loss of -$0.30 that was slightly above our sales forecast for $1.447b, but a penny below our -$0.29 estimate. Despite continued solid execution from a manufacturing and cost reduction standpoint, gross margin only improved to 3.2%, below our 5.0% estimate due to further price declines in NAND flash while commodity DRAM pricing actually increased in the quarter. With no inventory write downs in the quarter, the 3.2% result was a 640bps improvement from F2Q. Operating expenses of $277m were below our $284m estimate despite further spending on new product qualifications. Micron continues to view current commodity/PC DRAM market conditions as well balanced with inventories lean. Specialty DRAM demand remains healthy from the handset and enterprise markets but increased pricing pressure from peers that have increased supply is offsetting commodity DRAM strength. NAND flash remains in oversupply with further ASP erosion likely to continue.

Gross Margins

Product gross margins improved 640bps sequentially as DRAM/NAND cost reductions of 15%/25% helped to largely offset the blended price declines of – 5%/-20%. After three consecutive quarters of inventory write downs, Micron did not mark down the value of any products in F3Q. However, if NAND pricing continues to deteriorate at aggressive rates as NAND cost reductions have slowed, further write downs in flash products could occur again.

Capital Expenditures

Given still healthy DRAM demand and slowing NAND bit growth, Micron noted that capex could be at the upper end of its original $2.5-3.0b capex guidance. This is somewhat of a change from management commentary in recent quarters for approximately $2.5b. We believe this variability in spending is subject to opportunistic purchases of equipment, goals of its manufacturing partners, and a response to slowing bit growth and cost reductions in both DRAM and NAND businesses.

Business Review

Micron reported slightly better than expected F3Q08 sales while lower than expected gross margins were impacted by ASP declines in specialty DRAM and NAND. Bit growth guidance was lower than expected, especially for NAND, as supply growth and yield outperformance in recent quarters is leading to slower supply expansion and cost reductions in the near term.

Micron F4Q08 Guidance


– Bits: Up mid to high single digit %
ASP: Approx flat if flat from today
Cost Reduction: Mid single digit %


– Bits: Up mid double digit %
– ASP: Down -20% if flat from today
– Cost Reduction: Mid double digit %

R&D + SG&A Expenses: Flat to down slightly from F3Q

Outlook and Estimate Changes

For the August 2008, quarter, we leave revenues essentially unchanged at $1.607b that reflects lower DRAM and NAND bit growth, but offset by better DRAM ASP assumptions and Lexar retail NAND sales. GAAP EPS of -$0.18 is 2c below our prior forecast on a lower GM estimate of 9%, down from 11%. This compares to Consensus estimates for $1.618b and -$0.15. F2009E rev/EPS decline slightly to $6.710b/$0.15 from $6.745b/$0.16 on lower NAND bit growth in 2008 and lower GM.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: