16 October Analyst Digest

A change of plan here. No new post this week.

Analyst excerpts feel like the right way to coast into earnings. This post has been popular too. Suspect folks are coming back and re-reading sections. A lot here packed full of interesting angles/perspectives and worth contemplating.

Just added excerpts from BS’s and Lazard’s recent comments to those of Citi, CS, JPM & GS below. BS & Lazard seem to be mostly OK with things. GS is pretty negative, but there is a lot left out that other excerpts cover. Most notably the drag of 200 mm replacement requirements on 2008 NAND growth.

In any case, believe SanDisk will report a good Q3, maybe a great Q3. As JPM put it, the stars are aligned for this quarter. Q4 looks to be more of the same.

Q1/Q2 2008 seems to have a lot of folks worried about a replay of Q1/Q2 2007 where NAND oversupply ruined the party.

While there is cause for caution, 2008 looks to be shaping up very differently from 2007. First and foremost, the Koreans (Samsung & Hynix) look like they’re going to have one big 2008 200 mm headache that they didn’t have this year. 200 mm fabs are not going to be competitive for much longer.

According to Citi, Samsung has a 50% problem, while Hynix has an 80% problem. This is a big deal with no quick fix. As Citi puts it, these “extensive 200 mm replacement requirements” will create “a headwind to supply growth next year.”

Thanks to fab 3 and fab 4, SanDisk/Toshiba seems to be sitting pretty for 300mm wafers and to add icing on the cake, according to Credit Suisse, SanDisk/Toshiba is “about half a year ahead of the Koreans” in NAND technologies. Read half year ahead on geometry shrink/ yields.

Would not be shocked if SanDisk blows out earnings on Thursday and not much happens on share price. At some point though, if SanDisk continues to execute, it will have its day in the sun.

Analyst consensus for Q3 is $0.32 EPS on $931M in revenues. Given Q3 pricing strength in NAND and SanDisk’s product momentum, think there is a real chance SanDisk will exceed on the top line. Maybe approaching or topping $1B.

If so, the primary explanation is likely to be “strength in mobile.” Both cards and embedded were likely stellar. Believe mobile is now SanDisk’s strongest product segment, and will continue to be for a long, long time.

Also suspect that the rate of growth of memory in mobile, may be underestimated by many. It alone may soak up enough NAND to get to the SSD ramp.

While $1B is a nice round number with lots of zeros and would look most attractive sitting on the top line, the most impressive Q3 number might be gross margins. SanDisk took a real GM hit in Q1 and Q2. JPM thinks there is a chance that Q3 GMs will be more in line with historical averages.

If this comes to pass, SanDisk’s bottom line will pop. Personally think SanDisk will be in $0.4x+ territory and $0.5x+ is not out of the realm of the possible. GM expansion is a beautiful thing.


**** 16 October 2007 ****

Bear Stearns

SanDisk Corp


Gurinder Kaira
Dinesh Moorjani
Andrew Liersch

3Q07 Earnings Preview

• SanDisk is scheduled to report 3Q07 results on Thursday 10/18. We are looking for total revenues of $959M (+16% QoQ) and EPS of $0.38, with potential for upside to this estimate. Our product revenue estimate of $849M and product gross margin estimate of 25% are both above the guidance ranges of $750-825M and 20-24% respectively. Published consensus stands at total revenue of $931M (+13% QoQ) and EPS of $0.32, though we believe expectations are much higher.

• We have raised our 3Q estimates (our prior EPS estimate was $0.32) based on higher ASP and product gross margin assumptions. Our retail checks indicate that promotional activity was lower than expectations coming into the quarter. Based on our analysis, we estimate SanDisk’s retail ASP declined 12% QoQ in 3Q.

• For 4Q, we are looking for bit growth and ASP decline of 70% QoQ and -18% QoQ respectively, and we expect a significant improvement in product gross margin to 29%, as SanDisk should benefit significantly from its 56nm ramp. Our 4Q EPS estimate is $0.71, adjusted slightly from our previous estimate of $0.70.

• Looking ahead to 2008, while we expect 1Q08 to exhibit typical seasonality (we are estimating ASP decline of 22% QoQ for SanDisk), we believe NAND supply-demand dynamics for 2008 as a whole will be more favorable compared to 2007. We expect moderating industry-wide supply growth as well as upside from notebook demand to lead to a more benign pricing environment compared to 2006 and 2007. We currently forecast SanDisk’s ASP to decline by 51% YoY in 2008, versus 57% and 60% in 2006 and 2007, respectively.

• While 3Q and 4Q should be solid quarters, we believe that 2H07 earnings strength is already largely priced into the stock. On the other hand, we believe that 1Q seasonality issues as well as the notion of NAND spot pricing continuing to decline near-term are also getting priced in, but are priced in to a lesser extent.

While we are unenthused about the near-term outlook for the NAND market, in our view, investors should look beyond 1Q seasonality, as we believe the stock’s upside potential significantly exceeds downside risk — we expect upside to be driven by improved NAND industry fundamentals in 2008 as compared to 2007. We believe downside risk is to the mid-to-low $40s, assuming that some of early 2008 seasonality is yet to be priced in, while upside potential is to the high $60s. We are maintaining our Outperform rating and rolling over to a 2008 year-end price target of $68.


Lazard Capital Markets

SNDK 3Q earnings preview: Expect solid results; BUY

Earnings Preview


Expect a strong quarter. Sandisk will report its 3Q numbers on October 18 at 5:30pm EDT. Overall, we expect the company to post strong results on better than expected ASPs for 3Q, strong Micro SD unit growth, and higher than expected yields on its 56nm technology.

Raising estimates for 3Q. We are raising our 3Q revenue and pro forma EPS estimates from $910M and $0.31 to $947M and $0.34, above consensus of $930M and $0.32 respectively. Our ASP and bit growth assumptions for 3Q are –11% and 31% respectively.

ASP declines ahead, though 4Q outlook should be favorable. We believe the highlights of the call will be the focus on ASP declines in 3Q and the outlook for holiday season demand. We expect ASPs to decline 14% for 4Q on increasing capacity from both Samsung and Hynix after market share losses in 3Q. Our channel checks indicate that the mobile business is seeing a significant uptick since September. Our 4Q revenues and pro forma EPS estimates are $1,217M and $0.58 respectively.

Valuation remains compelling. Sandisk is currently trading at 21x our forward 2008 pro forma EPS. This is relatively low considering that we estimate earnings growth for next year at 63%. Our sum-of-the-parts based price target of $63 implies a 2008E P/E of 28x.

The biggest risk in the Sandisk story remains the overcapacity in NAND in the short term which could lead to steeper than expected ASP declines.

We reiterate our BUY rating on Sandisk as we believe 2008 will see a better supply-demand NAND balance as 200mm becomes obsolete and video and mobile kick in as significant NAND growth drivers. We believe the company is well positioned with its 56nm and the ramping of Fab 4 to more than offset the ASP drops in the industry next year.

**** 14 October 2007 ****

Goldman Sachs
October 14, 2007

James Covello

Americas: Technology: Semiconductors
SNDK should rally on 3Q report but Samsung ruined 2008 for NAND

Expect solid 3Q results and guidance for SanDisk, but Samsung’s upwardly revised bit growth forecasts make us less bullish on ‘08

We expect SanDisk to report solid 3Q207 results and offer solid 4Q2007 guidance, driven by strong ASPs. While 4Q2007 ASPs will be down from 3Q2007 levels, the decline will be off of a much higher 3Q2007 base, suggesting upside to 4Q2007 numbers. While we continue to like the stock heading into earnings, we are forced to take a less bullish view on 2008 given Samsung’s upwardly revised bit growth expectations. We have said all along that one of the key risks to our view on SanDisk heading into 2008 was either Samsung and/or Toshiba significantly increasing their bit supply growth forecasts. Unfortunately, this risk has now become a reality with Samsung increasing its 2008 bit supply growth forecast to +120% yoy (vs. our +100% yoy estimate), leading to potentially significant industry over capacity in 2008.

Despite our more cautious intermediate-term view, our bottom line for investors today is that there is still a short-term trading opportunity in the stock around earnings, as much of the near-term risk is already priced in at $47, with the stock likely to rally coming out of the solid earnings report.

SanDisk will be reporting its CY3Q07 results on Thursday, October 18 after the market close. We continue to expect CY3Q07 results to be solid, driven by strong ASPs during the quarter. Our model currently assumes a 10% qoq ASP decline for 3Q2007. However, our weekly retail card survey suggests that pricing is likely to be flat to slightly up in 3Q2007. While the market remains concerned about ASP declines in 4Q22007 negatively impacting SanDisk’s fundamentals, 4Q2007 ASPs will be coming off of a much higher 3Q2007 base, suggesting upside potential to 4Q2007 numbers as well. Further, Samsung’s 3Q2007 results imply solid royalty revenues (~$130mn) for SanDisk in 4Q2007 (recall that royalty revenues are recognized on a one quarter lag), which is a significant driver of earnings at greater than 50% of operating income.

While we continue to like the stock heading into earnings, particularly in light of the recent sell off, we are forced to take a less bullish view on 2008 given Samsung’s upwardly revised bit growth expectations for 2007 and 2008. We have said all along that one of the key risks to our view on SanDisk heading into 2008 was either Samsung and/or Toshiba significantly increasing their bit supply growth expectations for 2008. Samsung indicated on its earnings call on 10/11/07 that it expects the company’s bit growth to be +140% yoy in 2007 (30% qoq in CY407) and slightly less than 120% yoy in 2008. This compares to our previous expectation for 2007 bit supply growth for Samsung of 130% yoy in 2007 and 100% yoy in 2008. While Samsung’s management believes that the NAND industry will only add 120% bit supply in 2008, our estimates suggest that bit supply growth in 2008 would be closer to 140% yoy if Samsung were to increase its own bit production by ~120% yoy. As a result, we believe that the industry may face significant excess supply in 2008 given Samsung’s increased bit growth expectations (please see Exhibit 3).

To the extent that the NAND industry faces excess capacity in 2008, we believe that NAND pricing could decline in excess of 50%-55% yoy (the typical annual rate of decline for NAND), which would lead to further margin contraction for SanDisk in 2008.

Longer-term, we continue to see very good prospects for the NAND industry and for SanDisk, particularly around solid state drives. However, we need to stay true to our supply driven view of the industry and are therefore talking a less bullish stance on the stock in the intermediate-term. Despite our more cautious intermediate-term view, our bottom line for investors today is that there is still a short-term trading opportunity in the stock around the earnings report.

We are leaving our 12-month price target for SanDisk unchanged for now at $65. Recall that our price target is based on a fair value for royalty sales of ~$43, $9 in net cash/share, and an 18X multiple on 2008 product EPS (ex interest income) of ~$0.70. The key risk to our view continues to be incremental NAND supply coming on-line.

On a separate note, Samsung’s capex commentary did not change our bearish view of semi equipment. While Samsung significantly increased its memory capex for 2007 (see Exhibit 4), over 70% of that capex has already been spent, which implies that there is no upside to SPE orders in the coming quarters from this announcement. Importantly, we believe that increased capex in 2007 implies more downside to capex in 2008 (as 2008 spending is pulled into 2007), with Samsung’s management acknowledging that 2008 capex could be down yoy in 2008. As a result, we would use the rally in SPE stocks last Friday to increase short positions as we continue to expect the CY3Q07 earnings season to be a negative catalyst for many of the front-end SPE stocks.

**** 12 October 2007 ****

Samsung and Sony Ericsson Implications
Specialty Semiconductors
Citigroup Global Markets Equity Research
12 October 2007

Craig A Ellis
Asiya Merchant, CFA

What’s New — Thurs, Sony Ericsson and Samsung’s results held implications for many Specialty Semi cos and NAND and Wireless industry dynamics.

NAND: A Little Better Than Feared — SEC was up-beat on 3Q pricing (we think +30% qq) and benign on 4Q’s outlook (-15%; bears feared dn 20-25%), blaming recent declines on competitor’s yield-induced spot market dumping. On supply, 3Q production was below target (3% vs 10%; SLC mix and yield issues), the 4Q outlook was 30% and 51nm mix shifts only gingerly to 30% (from 15%), with 07E growth of 130% reiterated. While naturally concerned about the 29% 07 memory cap ex increase (to 6.19T Won), this may be explained by extensive 200mm replacement requirements (50% of Samsung NAND supply; 80% for Hynix), a headwind to supply growth next year confirmed by below-industry avg bit growth expectations of <120%.

SNDK Implications – Samsung’s NAND output and pricing broadly support our 07E royalty modeling and 3Q07 and 4Q07 EPS of $0.34 and $0.73, respectively. We expect SNDK to increase its C07 royalty guidance toward our $465M est (from $425M guide) at its 10/18 call, though note mgt is conservative in setting an outlook. On the retail business, 4Q07’s supply growth only moderately increases 4Q07 retail pricing risks, though US sell-through bears watching given consumer spending vulnerabilities now evident in US same store sales data. Overall we expect the St to view the benign 08 bit
growth outlook favorably, and expect this to help intermediate term stock psychology, recently damaged by declines in both NAND and DRAM pricing.

Key Handset Takeaways — SNE’s unit shipments were modestly disappointing (25.9M vs CIR’s 27.6, St’s 26.8M) though SEC’s 42.6M units (vs CIR 40M) and high end mix (Ultra Edition, other 3G models) suggests 14% (or greater) global unit growth remains a reasonable expectation for 2007.

Wireless Component and Analog Takeaways — Specialty Semi cos. levered to SNE handsets include SWKS (21%), SanDisk (11%), and RFMD (6%) while SEC handset leverage includes SWKS (~12% C07E revs, Helios mini transceivers), RFMD (<7% revs, power amplifiers/front end modules), Micrel (~5% revs, power mgt ICs), and SLAB (< 5% revs, FM tuners). While somewhat mixed overall, in aggregate we view handset developments favorably for wireless component makers. RFMD rated Buy (1S) and SWKS Hold (2S).



12 October 2007
Asia Pacific/Japan
Credit Suisse Equity Research
Semiconductor Equipment (Electrical Machinery/Semiconductor Devices) /
Industrial electronics sector

Research Analysts
Hideyuki Maekawa

Samsung 3Q results: miniaturization problems?

First impression: A conference call with Samsung Electronics did nothing to change our fundamental view of the company’s prospects. The increase in demand for DRAM and NAND did not exceed normal seasonality, and Samsung seems to have struggled to raise yields on miniaturized design rules despite an attempt to boost margins through the competitiveness of its DRAM and NAND technologies. The upward revision of the consolidated capex budget for 2007 to W7.2tn from W6.5tn (management cited a front- loading of some spending from the 2008 budget) was a surprise, given that orders of front-end equipment have virtually halted since April. We are doubtful because the company left its capacity and output forecasts unchanged: the explanation for this apparent inconsistency could be that miniaturization has increased the number of processes and depressed capacity and yields. Management avoided giving a clear answer about whether it would reduce capex in 2008 as a reaction to the increase in 2007, but we believe a 20% reduction is possible. In light of the figures, we maintain our OUTPERFORM rating on Toshiba (6502, TP ¥1,200), and continue to favor waiting for the right time for a contrarian investment in Tokyo Electron (8035, NEUTRAL, TP ¥7,900).

Investment implications for Toshiba: We forecast that NAND prices will decline about 15% QoQ in Oct-Dec as demand from mobile-phone and MP3 applications remains robust (we expect prices to remain stable through November and then fall 20% in December). Such conditions support our scenario of a large outperformance in OP from Toshiba’s NAND business, and suggest that 1H earnings will provide a near-term catalyst for a share price run-up. We believe the share price could soften from December, once NAND unit pricing begins to ease, but over the medium term, we favor accumulating on weakness, based on the company’s leadership in NAND technologies (it is about half a year ahead of the Koreans, which we believe gives it an advantage of more than 20%) and the solidity of orders in its nuclear power business.


Bit growth is said to have been 3% in Jul-Sep (target as of July: 10-15%) and is forecast at 36-39% for October-December. The shortfall for Jul-Sep is primarily attributed to a higher proportion of SLC (single-level cell) production and persistently low 50-nm process yields. Management currently targets annual bit growth at 140% for 2007 (target as of July: slightly above 130%) and slightly below 120% for 2008. Unit prices rose 15% in Jul-Sep and management projected them to fall 15% in Oct-Dec.

Semiconductor capex

Management raised its 2007 semiconductor capex budget to W7.2tn from W6.5tn, commenting that this reflects an acceleration of capacity additions at Line 15 (a DRAM line) and Austin (NAND line), together with upgrades to each line. Management declined to comment on whether capex will be higher or lower in 2008. In Jan-Sep, Samsung Electronics spent 73% (W4.99tn) of the amount budgeted for semiconductors overall in 2007 (parent level) and 72% (W4.48tn) of that for memory operations.

**** 3 October ****

SanDisk Corp (SNDK)

Citigroup Global Markets Equity Research
3 October 2007

Buy/High Risk 1H
Price (03 Oct 07) US$53.36
Target price US$71.00
Expected share price return 33.1%
Expected dividend yield 0.0%
Expected total return 33.1%
Market Cap US$12,158M

Craig A Ellis
Asiya Merchant, CFA

Contract Pricing a Negative, New Products a Partial Offset

What’s New – NAND news flow has been mixed this week, with a negative contract pricing undertone. Specifically, early-week 2H Sept contract pricing declined to close the gap with spot, a change affirmed in MU’s results commentary. That said, new high-density NAND prds emerged, and we believe supply headwinds for 2008 remain under-estimated. Overall, developments remain within the parameters of our above-consensus 2H07 and 2008 estimates in our view. For the stock, we acknowledge that the window for the classic seasonal trade is narrowing, though given expectations for relatively benign 1H08 seasonality and with downside risk to the high-$40s and upside potential to $71, we continue to think reward/risk is favorable. Reiterate Buy.

MU More Negative Than Positive For NAND — Aug qtr NAND bits grew 50%+ qq (vs 50% guide) while pricing jumped ~30%, confirming strength in June- Aug; Nov qtr prices could decline 25%, reflecting the pull-back in contract this month. For F08, MU expects 40% qq bit growth, or 300% for the yr (vs 230% CIR ests), a negative; NAND capex to decline ~30%. IMFT is <10% industry supply though Korean mfgrs, with ~65% NAND supply, face significant 200mm capacity de-commission headwinds in 2008 (50%/80% of SEC/Hynix output to sub-20% by YE08), a key contributor to our positive view on 2008 supply.

NAND Contract Price Updates and Declines – Earlier this week, 2H Sept NAND contract prices fell 11% on a wtd avg basis or 400-600 bps greater than we expected. Data is consistent with MU comments. That said, pricing was still up 34% in 3Q07, above levels used in our SNDK royalty models and 4Q07 pricing is tracking at down 9% (if flat from here). Overall we remain comfortable with our -15% royalty modeling for 4Q07, though we have lost some conservatism. Alternatively, we are encouraged by new product activity (Zune’s 4G/8G flash MP3 players, new music phones at Verizon) and note last week Toshiba indicated order fulfillment at 70%, tighter than 75% four weeks ago.


Resilient Retail Pricing during 3Q – Our analysis of card+drive retail price
trends affirm 3Q retail pricing was resilient overall, underpinning CIR modeling
for 2H07 gross margin acceleration. Checks indicate pricing is more benign in
cards, more severe in MP3 products, though the latter likely due to channel
drain ahead of pre-holiday new product transitions. We remain comfortable
with 3Q07 and 4Q07 price/MB estimates of -13% and -15%, respectively which
support EPS of $0.34 and $0.73, 6% and 36% above Consensus respectively.

Company description

SanDisk Corporation (SNDK) was founded in 1988 as a flash memory storage company and has evolved to become a leading NAND flash memory cards/USB drives and embedded memory supplier worldwide. SanDisk pioneered numerous flash technologies, and currently holds over 279 patents, from which the firm generated royalty revenue of $331 million in 2006 (9.2% of sales). The company has vertically integrated captive NAND manufacturing through its FlashVision joint venture with Japan’s Toshiba (50/50). Vertical integration has been a gross margin and earnings positive for SanDisk over the last three years, helping it garner a richer product gross margin than many other memory card/USB competitors. Memory products are sold to both OEMs (i.e., digital camera, PC, cell phone handset, PDA, etc) and consumers through over 190,000 retail outlets worldwide. By channel, OEM sales were 33% in 4Q06, versus 67% for retail.

Investment strategy

We rate the share of SNDK Buy/High Risk (1H). Firming NAND fundamentals creates a favorable backdrop ahead of the seasonally advantaged 3Q. Long term product catalysts (mobile cards, video products and PC drives), a memory cost advantage (x3/x4 technology) and potential royalty upside (Hynix) are positives. At the macro level, we still expect 2H07 undersupply. Muted bit growth targets from competitor suppliers (Samsung) + SanDisk’s own 2H07 shrink/volume cost benefits improve SNDK’s retail pricing and margin prospects for 2Q07 through 4Q07. Risks include product launch delays and seasonal 1H softness. Tactically, we encourage investors to leg into shares ahead of powerful 3Q07 seasonality and positive Street EPS estimate revisions.


We derive our $71 target price using selected forward 12-month price-to earnings, enterprise value-to-sales, and price-to-book value (P/B) multiples applied to C08 estimates. Our target multiples are at a 34% premium to historical medians. We believe selection of a four-year time period is reasonable as it includes periods of supply imbalances (over/under), driven by increasing supply competition, end demand driver growth phases (nano, handsets, USBs), and inventory correction representing a good cross section of NAND industry phases going forward. While our revised multiples are still below the 25 percentile range (of prior 4 year period multiples), we believe 34% premium to median multiples is warranted to reflect the potential for incremental higher margin royalty revenues and pricing leverage given the undersupply NAND fundamental backdrop we now see ahead (in prior undersupply periods in 2H03 and 2H05, F12 M P/E multiples expanded by ~38%). In summary, our multiples are now in line with median trading multiples observed in prior undersupply periods.

For our P/E, we use a 27x target multiple applied to our calendar 08 EPS estimate to yield a target price of approximately $82. Our 27x multiple is derived from the company’s four-year historical range of P/E multiples that ranges from 15.1x to 35x. On an EV/S basis, a 2.59x multiple applied to our 2008 revenue per share estimate and net cash gives us a target price of approximately $58. We derive our 2.59x EV/S multiple from a 0.7x to 4.2x historical range. On a P/B basis, our 2.82x multiple gives us a target price of $72. We derive our 2.82x P/B multiple from a 0.8x to 4.2x historical range. Taking the average target price derived from the three methodologies and rounding it, we arrive at our target price of $71.


We rate SanDisk High Risk, primarily due to the stock’s own volatility, semiconductor industry volatility and NAND pricing volatility. Additional risks include oversupply concerns from new market entrants, in turn spurring intense price competition, especially during seasonally-slow periods; margin pressures if ASP declines exceed cost savings; delays in ramp-up at internal fabs and lower manufacturing yields of internally-sourced wafers, requiring SanDisk to rely more on less cost-effective external sources; abrupt demand slowdowns or lower-than-estimated demand, especially in emerging product categories (handsets, MP3s) which could result in inventory write down; maturing of traditional markets (such as digital camera cards); fab capex funding requirements potentially requiring the company to source funds from the capital markets and any issues on challenges with business integration of the recent msystems acquired business. Additionally, SanDisk has contingent indemnification obligations for certain liabilities Toshiba incurs (operating leases, indemnification) arising from its fab joint venture with SanDisk. None of these obligations are reflected as liabilities on the balance sheet. Downside risk to our target price exists if the company’s pricing deteriorates throughout the remainder of 2007 on loose NAND supply/demand fundamentals or on unfavorable royalty revenue negotiations with Samsung in 2009. If the impact on the company from any of these factors proves to be greater than we
anticipate, it may prevent the stock from achieving our target price.

**** 27 September ****

North American Equity Research
27 September 2007

Sandisk Corp

Toshiba Chips In; Reiterate Overweight Rating

26 September 2007

Paul Coster, CFAAC
Mark Strouse

Toshiba’s commentary over the last two days reinforces our view that the NAND industry will be supply constrained through 2H07, benefiting SanDisk. If channel inventory is low exiting 4Q, we should be well positioned for a strong 1Q08. Reiterate Overweight rating.

• Supply constraint; part 1. Toshiba’s device group CEO stated at JPMorgan’s Asia Pacific Equity Conference that Toshiba would be supply constrained through 2H07, as reported by JPMorgan’s Industrial Electronics Analyst, Yoshiharu Izumi (09/25). Importantly, the Fab 4 ramp is on schedule, and Toshiba is confirming that super MLC (3
bit/cell) is on schedule for early 2008. Super MLC could be a new source of licensing for SanDisk.

• Supply constraint; part 2. Separately, EE Times yesterday reported comments by the CEO of Toshiba’s semiconductor business; Toshiba is currently meeting only 70% of customer demand. Available production is being directed at the OEM end market.

• What’s good for Toshiba is good for SanDisk. SNDK should be positioned to direct its NAND Flash supply from the JV to highest margin retail product/markets, and the priority Toshiba gives to the OEM market means that SNDK should also garner market share gain in retail where it chooses (Europe and Asia Pacific). Finally, we believe the priority given to OEM should lead to thin retail inventory at the end of 4Q, leading to strong channel demand in 1Q08. We expect the handset aftermarket to be a strong driver of demand in 1Q08.

• Reiterate Overweight. SNDK is currently trading at 20.8 times our FY08E PF EPS of $2.58, a 7.7% discount to the mean of our coverage. We believe consensus’ 07-08 estimates are too low and there’s potential upside to our above-consensus EPS and Revenue forecasts in 2H07.

Earnings Snapshot

In our view, the stars are aligned for SNDK this quarter. Though we are forecasting $0.36 PF EPS on $940mm of sales this quarter, there is potential for gross margins to move toward last year’s levels (PF gross margins were 32.7% in 3Q06). A re-run of last year’s margin story would yield $0.57 of PF EPS, all other things being equal. That looks a stretch, but we believe there’s upside to our above consensus forecast (street PF EPS stands at $0.31).


Valuation and Rating Analysis

We are Overweight SNDK. SNDK is trading at 20.8 times our FY08 PF EPS forecast of $2.58, a 7.7% discount to the mean of our coverage, which looks unjustified at present, given a projected 61% surge in EPS in 2008, and a 2-year (07- 09) CAGR of ~30%. We also believe consensus estimates are too low, for both FY07 and FY08. We favor owning SNDK during the course of these anticipated revisions and in anticipation of potential margin or volume upside surprises in the context of 2H07 NAND industry supply constraints.

On a next-year P/E basis, SNDK is trading at a significant discount to near-peers Micron and Apple, however, neither makes for a convincing comparable, in our view.

Risks to Our Rating

We could become less constructive on SNDK prospects for any of the following reasons:

• New or existing entrants ramp production, leading to a persistent over-supply of NAND product, leading to pressure on ASPs and margins

• SNDK loses key licensing revenue associated with its patent portfolio, specifically relating to MLC and Samsung.

• SNDK fails to transition to smaller geometries quickly and efficiently, leading to
a lagging cost structure.

• SNDK or other participants in the NAND industry lead retail and OEM prices down to unprofitable levels for sustained periods in an effort to win market share, or shake out weak players.

• The cost of NAND supply fails to fall fast enough to substitute for magnetic disk storage, or to enable the introduction new applications

• Seasonal supply and demand fluctuations are not properly factored into analysts’
models and investors expectations.

• New, disruptive solid state storage technologies emerge that are significantly
more price-competitive than NAND.

• SNDK over-commits to capital investment in NAND capacity.


3 Responses to 16 October Analyst Digest

  1. Ron H says:

    Thanks Savo
    Can you comment on the D. Amir Lazard’s report (date,forecast, rec) ?

  2. savolainen says:

    You’re welcome, Ron.

    Here are the relevant portions of the recent Lazard SNDK Memory Pricing Update. My tentative plan is to do another Analyst Digest at some point following earnings. If that happens, will try to include some of D. Amir’s post – earnings SNDK commentary.

    **** 17 September ****

    Lazard Capital Markets
    Memory Pricing Update [excerpts]
    Daniel Amir
    17 September, 2007


    NAND pricing taking a turn. In the past few weeks we have seen a decline in NAND flash pricing, with ASPs for the 8Gb down 12% since September 1st. The decline is not surprising considering the strength in NAND pricing during the summer due to the Apple build, capacity issues at Samsung and speculative buying by memory brokers in Asia in anticipation of strong demand from the Apple channel and back to school.

    NAND likely to remain under pressure in near term. Overall, we anticipate that NAND will remain under pressure in the near term, as Tier 1 NAND vendors like Samsung and Toshiba are releasing more NAND parts into the market. We expect NAND pricing to decline by 30%-40% by 1Q08, with the majority of the declines coming in December post holiday.

    Mobile phone with card slots continues to roll. Currently, the area of most significant strength is the increasing penetration of handsets with card slots. We anticipate that by year-end, close to 40% of all handsets shipped will come with a card slot. This continues to be the top driver, in our view, for NAND flash cards. We believe this trend is especially positive for Sandisk, which is well positioned in the mobile space with its MicroSD cards for both retail and OEM.

    Apple and retail demand are the wild cards. We could be wrong in our thesis if we see 1) much stronger-than-anticipated demand from the Apple channel for the iphone or the new ipod video, resulting in higher NAND pull-ins than previously expected; 2) stronger-than-expected retail sell-through in the U.S. or larger-than-expected build for the Chinese New Year, which starts fairly early (February 7th).

    Takeaways on SanDisk

    Maintain BUY rating, $63 price target. Despite the NAND flash price decline we recommend SanDisk shares for the following reasons: 1) We think the company is well positioned given its cost edge versus its competition, which gives it the flexibility to ride out the near-term NAND flash price declines; we expect further product GM improvement; 2) SanDisk is well positioned in terms of pricing and products to capture additional market share in the seasonally stronger 2H. 3) We expect the company’s strong royalty and licensing growth to contribute ~$1.30 in EPS in 2008. 4) Current Street estimates for 2008 could prove conservative, as SanDisk is poised for gross margin expansion with its new Fab 4 facility and accelerating new markets of mobile phones, video, and Solid State Drive (SSD).

    Valuation and risk. SanDisk is currently trading at 22x our 2008 pro forma EPS estimate. Historically, in peak cycles, which we believe we could see in the next 12 months, SanDisk has traded at close to 30x estimated forward EPS (e.g.,in 2H05). Given our forecast for 67% earnings growth next year, we think the shares deserve a multiple closer to the historical level. Our sum-of-the-parts analysis of the IP and product revenues multiples compared with peers suggests a $63 per share valuation. Risk: NAND prices decline faster than manufacturing costs, potentially pressuring GM and earnings.

  3. afkkl8_99 says:

    Hi Savo

    Long time fan of yours, like to read and learn a lot….but after all I am here and believe the others including yourself as an investor who wants to see his capital rising, so far disappointed with SNDK.

    Regardless of the coming quarters, where do you see SNDK in 2010 ? Will it survive as an independent company? will remain a 10 B +/- company or will make it and join the 100B league ? I know its hard question, but appreciate your vision and ability to make an intelligent prediction



Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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: