|Shares Out. (in M):||174||P/E||0||0|
|Market Cap (in $M):||12,154||P/FCF||0||0|
|Net Debt (in $M):||-2||EBIT||0||0|
Summary Recommendation: I recommend going long LRCX at ~$70 / share, assuming a 12-24 month holding period and a potential absolute return of 30-50%. LRCX sets up as a compelling long-term earnings growth story, with a track record and continued forward view of outgrowing its market while maintaining - and progressively expanding - its strong margin profile (in-line to slightly better than many of its competitors and comparable companies). In addition to the company’s personal success, LRCX is in midst of (accretively) acquiring KLAC, a related company (not a competitor) that provides an entirely complementary set of products and helps diversify LRCX’s customer exposure. As a combined entity, management (who holds a strong track record of execution) projects FY18 (ending June) revenue of $10B, with non-GAAP operating margin of 27%, which implies more than $8 / share in non-GAAP EPS, with a path to more than $9 if things break right.
However, due to short-term fears regarding cyclical oversupply in one of LRCX’s primary markets, shares are currently trading at less than 10x my pro-forma CY17 non-GAAP EPS estimate (see below). As a result, I see LRCX is an attractive long-term opportunity on an absolute basis, on a historical valuation basis, and vs. peers, with the latter two implying multiples closer to 13x.
Company Background / Description: The process of manufacturing semiconductors (e.g., computer processors, flash memory, LEDs, etc.) is tedious and highly technical, requiring a dedicated industry of specialized equipment providers for manufacturers. Lam Research (LRCX) is one of these semiconductor capital equipment makers, with a primary focus on front-end production (i.e., the process of actually creating the chip, as opposed to back-end testing or packaging). Lam’s particular strengths lie in:
Etching - chemical removal of material layers from the surface of the silicon wafer on which the chip is created. Etching is an essential step, with wafers typically undergoing multiple rounds and steps of etching before being ready for use / the next step in the process. LRCX is the etching market leader with high-50% market share.
Deposition - applying thin films of material onto the substrate (i.e. wafer) for insulation and conduction purposes; also a critical step in the manufacturing process. Lam’s deposition business is essentially from its acquisition of Novellus in 2012 and holds a solid market position of high-30% / low-40% market share
Currently, LRCX is specifically benefitting from numerous secular tailwinds related to the general trend toward smaller chips that can deliver greater power at higher levels of efficiency, as well as the particular demand for flash memory as a replacement for legacy disk-based technology. The complexity involved in manufacturing increasingly smaller-sized chips increases the importance and relevance of LRCX’s solutions. These trends have allowed LRCX to grow revenue significantly faster than the ~$33B wafer fab equipment (WFE) market over the last 3 years (22% vs. 6% annualized).
At Friday’s (2/19) closing price of $69.77, Lam maintains a $12.7B market cap, along with roughly $2B in net cash. For CY15, Lam generated revenue of $5.9B (+21%) and non-GAAP EPS of $6.29 (+35%), with gross, EBIT, net, and FCF margins of 46%, 22%, 18%, and 18%, respectively. Lam also pays an annual dividend of $1.20 a share, for a 1.7% yield.
Highest sales growth among direct competitors and comparable companies while maintaining comparable margin levels (slightly better or slightly worse vs. nearly everyone in the space, depending on the metric; in the aggregate, in-line with the field but with better growth).
Riding strong secular trends in the semiconductor industry - FinFET, multi-patterning, 3D NAND, and advanced packaging - all related to shrinking node sizes (i.e. smaller chips) and the demand for more powerful and energy-efficient chips. As node sizes get increasingly smaller, the manufacturing process becomes more difficult and complex, increasing the need for (and amount of) Lam solutions.
Additionally, the flash memory market in particular is in the early-ish stages of replacing disk-based storage as the de-facto standard memory media, due to clear performance and efficiency benefits, combined with falling flash costs. (LRCX has a particularly strong position among memory chip manufacturers)
Favorable capital return policies - currently in midst of a $1B capital return program, including $850M in buybacks. Over CY15, LRCX has returned $463M to shareholders (~44% of FCF) between dividends ($153M) and buybacks ($310M).
Pending acquisition of KLAC creates a more diversified business (= potential for multiple expansion), with no overlap (i.e. all complimentary revenue) and at higher margins. While management M&A and synergy projections should always be taken with a grain of salt, LRCX has a strong M&A track record, with the 2012 purchase of Novellus widely-viewed as a significant success.
Earnings multiples are cheap both on an absolute and historical basis compared to recent and expected earnings growth, as well as versus competitors and comparable companies, who are growing slower and with comparable margins.
3 customers (Samsung, Micron, and TSMC) make up >50% of revenue.
Memory comprised 66% of CY15 shipments, up from 56% in CY14 and 48% in CY13 (expected to be roughly 50% post-KLAC merger).
Semiconductor capex cyclicality - in general, while semiconductor growth is fairly consistent, the industry is prone to cyclical swings - fabs / IDMs are typically more fearful of inability to meet customer demand than being oversupplied in the short-term. As a result, fabs / IDMs can often overshoot estimates of demand (erring on the side of caution), causing a temporary / cyclical reduction in production (and thus equipment capex) as utilization moves back up.
EUV - [extreme ultraviolet lithography;] a new technology being developed by fellow semi capital equipment maker ASML, which would potentially diminish the need for traditional etch products, thus hurting LRCX’s long-term etch TAM
Potential issues related to the KLAC acquisition
Possible dilutive impact if KLAC underperforms, given the stock portion of the acquisition consideration and the incremental interest expense from the debt portion
Potential negative impact on current shareholder capital return policies (dividend and buyback program).
Lower growth profile of KLAC could limit multiple expansion
Recent Events / Why Does This Opportunity Exist?
Throughout a chunk of 2015, fears surrounding oversupply in the DRAM market began to increase, due to 1) softness in PC and handset markets; and 2) a pause in capital spending as memory manufacturers build out capacity for smaller node sizes. Fears hit their peak in late September, when LRCX, among other semi cap equipment stocks, declined ~20% (from ~$75 down to ~$62), amid a wave of negative memory market news. Eventually, shares recovered following strong CQ3 earnings and the announcement that LRCX would acquire KLAC - a move universally lauded by the market.
On the back of a volatile market to start 2016 amid macro concerns, sell-side comments following management meetings at CES in early January revealed that Samsung was likely to cut capex plans for 2016, bringing semi capex fears back to the forefront. This was followed by Samsung’s Q4 earnings call where, without giving specific guidance numbers, Samsung qualitatively confirmed these plans to the investment community, with expectations of reduced capex particularly in memory, amid DRAM weakness (expected PC and handset softness)
In mid-January, Intel, the world’s largest semiconductor manufacturer (~15% market share), announced a cutback to 2016 capex on its Q4 earnings call, which while still up $2.2B y/y (from $7.3B in 2015), was 1) unexpected; and 2) lower than the $10-11B spent annually from 2011-14.
Why Does The Opportunity Exist? Ultimately, despite a 1) a steady string of strong quarterly results; 2) numerous long-term secular tailwinds; 3) a universally-lauded move to acquire KLAC (which is unlikely to hit any antitrust hurdles, due to the complementary nature of their businesses, as opposed to the AMAT-Tokyo Electron deal that fell through); and 4) a solid / decent pro-forma balance sheet post-deal-close (i.e. debt load isn’t excessive relative to company size), fears surrounding a pull-back in capex from memory players has dragged down shares to the high-$60s / low-$70s (<10x pro-forma non-GAAP EPS for CY17), after previously reaching above $80 following the KLAC deal announcement.
LRCX successfully completes the KLAC merger and is able to achieve, at the very least, cost synergy targets over the next 18 months.
2016 plays out as management has projected since late 2015 (reiterated in early 2016), with strong 3D NAND spending in 1H16 more than offsetting substantial declines in DRAM spending as the industry rationalizes to supply / pricing, before a bounce-back in DRAM and both logic and foundry segments in the second half of the calendar year.
Management’s WFE estimates of a flattish 2016 remain in effect, eventually reaching ~$35B for FY18 (vs. the current CY16 estimate of ~$33B, or up ~6% over the following 18 months).
LRCX de-leverages down to promised levels (2.5x EBITDA) in the promised time frame (~18 months), reinstitutes share buybacks (as promised), and maintains the dividend.
Greater diversification of revenue base (both by customer and end-market), combined with a return in DRAM spending (as well as continued 3D NAND spending), expands LRCX’s multiple by 2-3 turns.
Based on my CY17 non-GAAP EPS estimate of ~$7.75 (to which there is upside; see below), and a 12x multiple (also with further upside), shares rise to ~$90-95 by the end of 2016 / beginning of 2017 (~30% return).
Based on the potential for $10B in revenue, reduced interest expense from deleveraging, and a return of the share buyback program, LRCX achieves non-GAAP EPS of $9 or more for CY18, which, combined with a 13x multiple would yield a price of $115-120 by the end of CY17 (~65% 2-year return).
KLAC merger fails to yield any synergies.
Response: certainly possible, though my CY17 estimate assumes no revenue synergies, only cost synergies from the perspective of reaching margin targets. In general, cost synergies are easier to achieve and mgmt has a strong track record of doing so, based on the Novellus acquisition
DRAM and logic spending cuts are more dramatic than expected and WFE declines in 2016, hurting LRCX in particular; revenue for the year is flat to slightly up.
Response: also certainly possible, but the Samsung capex cuts were noted prior to LRCX’s FQ2 call and management didn’t change its outlook. In fact, mgmt has explicitly noted that its projections assume a dramatic reduction in DRAM spending, with the bulk of 1H16 gains coming from 3D NAND. While a poor macro environment could make everything worse, based on both explicit statements and track record, it is clear that LRCX management is already factoring in the declines feared by the market. Further, current share prices are specifically already pricing-in this risk, providing significant upside potential to numbers were DRAM pricing to solidify sooner-than-expected. In terms of logic, like DRAM, the reduced outlook is temporary, as secular trends support longer-term growth. Unlike DRAM, LRCX has much smaller exposure to logic than other WFE players do, as memory comprises ~⅔ of LRCX shipments.
Better revenue diversification is offset by slower-growing KLAC revenue, and in addition to a one-sided sell-side opinion, multiple expansion is muted / non-existent.
Response: While the one-sided sell-side is not ideal, based on recent trading activity, it appears most of the multiple-compression is related to memory spending and LRCX’s memory concentration. KLAC, while slower-growing, is expected to roll out a new set of products, which should increase growth, at least in the short-term, with further upside from any revenue synergies. Further, based on my estimates, non-GAAP EPS should grow >10%, in CY17 over CY16, as well as another 15+% in CY18 over CY17 if things progress mostly on-track, providing solid earnings growth even absent any meaningful multiple expansion.
Slower-than-expected integration with KLAC pushes out timeline of de-leveraging and return of buyback program
Response: Hard to see this happening, given the strong cash flow generation of both companies (particularly KLAC) even if no synergies are reached. With mgmt’s stated focus on deleveraging as soon as possible, substantially all available FCF is likely to go toward debt reduction, which should be able to reduce the full increase in debt in as little as 18-24 months. Even without a full debt load reduction, even a partial reduction will lower interest expense, creating a virtuous FCF cycle. Management has pledged to resume share buybacks once debt is appropriately lowered.
Consensus revenue projections call for another flattish year, with revenue around $9-9.25B; Non-GAAP operating margins fall short of the 27% target and no debt is paid down nor shares repurchased, yielding non-GAAP EPS of ~$6.60-6.75; With the company’s current 10x multiple, shares stay flat / decline modestly in the high-60s / low-$70s
Response: Given that combined CY15 revenue without any synergies would yield a $8.7B revenue company, it is hard to see a similar level in 2 years from now, given all that’s going on in the WFE space. While $10B may seem ambitious, cross $9B is certainly not. Plus, even if the combined companies fall short of 27% operating margin, some level of margin expansion is all but assured given the substantially stronger margins at KLAC. With LRCX currently expecting 22% operating margins for 2016 on a standalone basis, mid-20s margin on a combined basis is highly likely. As a result, it’s difficult to see EPS staying below $7 / share. Additionally, assuming that memory isn’t secularly challenged (which it is most certainly not), there should be some multiple expansion, even if uninspiring, that would get the stock to, at least, above $80 as DRAM prices stabilize.
Concerns / Risks to Recommendation
Near-universal love from the sell-side, with 18 sell-side analysts and 14 buy recommendations (of the remaining 4, 2 are neutral and 2 are unrated).
Greater-than-expected cuts to DRAM spending from Samsung, others
Muted sales growth limits FCF expansion, limiting debt paydown and prolonging impact of increased interest expense
Continued declines in DRAM pricing
Macro issues hurting end-market demand for 3D NAND and general semi cap spending
Closing of KLAC deal
Resumption of buyback
Stabilization of DRAM pricing
Return of DRAM and logic spending in 2H
Samsung reverses course of DRAM spending plans due to other players (MU, SK Hynix) catching up
Realization of revenue synergies
Substantial uptick in PC and handset markets, pushing DRAM pricing up (as opposed to stable) and increasing DRAM spend
Estimates and Valuation
CY17 - Combined revenue for the 2 companies for 2015 is ~$8.7B. Both companies expect to grow over the coming year, as well as in CY17. Management’s $10B assumption for FY18 implies a 6% CAGR from today for the combined entity. Haircutting that to 4.5% for conservatism through CY17 reaches revenue of $9.5B, which basically removes the entire revenue synergy assumption of $600M by 2020. Assuming successful attainment of 27% operating margins, roughly $200M in interest expense, and 18% tax rate (based on mgmt guidance), CY17 non-GAAP EPS will reach $7.75. Obviously, there are revenue, interest, and share repurchase levers that can all be pulled to increase that number substantially
Interest expense estimate - ~$150M in incremental interest from the deal ($3.9B at 4%, as per mgmt guidance), on top of ~$100M in interest from existing debt, reduced by $50M from debt reductions of ~$600M / year
CY18 - Assuming $10B in revenue for CY18 (which is low relative to the $10B estimate for FY18, which ends 6 months earlier), 27% operating margin, $125M in interest expense (assumes most of new debt load reduced), 18% taxes, and a share count reduction of 20M, yields non-GAAP EPS of ~$9.15
Valuation - comps and historical valuation has generally been 13x forward year P/E. Being conservative on that number for CY17 gets to 12x on EPS of $7.75 = $93; For CY18, assuming $9.15 in EPS and the extra turn to 13x is achieved = $119
Closing of KLAC deal
Resumption of buyback
Stabilization of DRAM pricing
Return of DRAM and logic spending in 2H