Elliott Smith, CFA

Transcription

1 SECTOR UPDATE January 13, 2016 Tech, Semis, Components and Distributors The Industrialization of Semis Favor growth over safety Formalizing our Industrialization of Semis thesis. After writing on this topic for the last 3 years, we formalize our industrialization thesis today, showing detailed regression analyses: Our analysis shows that semi investors have historically paid up for safety characteristics; however, growth characteristics have been better at generating alpha. As semis have become a more mature, slower-growing industry (more like industrials), investors, companies, and boards of directors should shift their favor away from safety to growth. As long as debt capital markets cooperate, expect M&A to be a major part of growth (as is the case in industrials). Key ways to invest in this thesis are to buy NXPI, AVGO, MSCC, and ADI. Semi investors have historically "paid up" for safety. Semiconductor investors have historically "paid up" (with high P/E and EV/EBITDA multiples) for high gross margins, and low net leverage, characteristics we call "safety". Industrials investors have not consistently paid up for these metrics. But, safety characteristics have not generated alpha. Our analysis shows that across the three time periods we studied, investing in semi companies with safety characteristics has not generated alpha. Semi investors should instead "pay up" for growth. Across the three time periods we studied, the characteristics of semi companies that were most strongly correlated with total shareholder return were (1) earnings growth, (2) revenue growth, and (3) free cash flow growth. The same characteristics are the three most strongly correlated with shareholder return in industrials. This appears extremely supportive of continued M&A in semis (as long as debt capital markets cooperate), reflecting the same approach industrial companies have taken. Semi Boards of Directors should shift to compensate growth metrics. We reviewed dozens of 10- K filings to investigate how compensation metrics align with our "growth over safety" view. Semi BoDs compensate for high margins in 85% of cases, but compensate for EPS and cash flow in only 35% and 13% of cases, respectively. We advocate semi BoDs should more prominently target EPS and cash flow growth, like their industrial counterparts which use these metrics in 48% and 30% of cases. We believe the best ways to invest in the industrialization of semis is through companies that have embraced M&A to grow sales & EPS. The most prominent of these are NXPI, AVGO, MSCC, and ADI. Companies Impacted in This Note Ticker Price Rating Target ADI $51.65 Buy $71.00 APH $47.43 Neutral $55.00 ARW $48.32 Buy $65.00 ATML $7.68 Neutral $8.83 AVT $39.08 Buy $50.00 BDC $41.78 Buy $80.00 LLTC $39.51 Neutral $45.00 MCHP $41.85 Neutral $50.00 MXIM $33.15 Neutral $39.00 NXPI $77.78 Buy $ PLAB $12.20 Buy $15.00 ST $39.64 Neutral $50.00 TEL $58.46 Buy $74.00 TXN $51.53 Neutral $57.00 XLNX $43.36 Neutral $45.00 AVGO $ Buy $ MSCC $31.33 Buy $49.00 INTC $32.68 Neutral $35.00 CY $8.39 Neutral $9.00 William Stein, CFA william.stein@suntrust.com Joseph Meares, CFA joseph.meares@suntrust.com Elliott Smith, CFA elliott.smith@suntrust.com SEE PAGE 18 FOR REQUIRED DISCLOSURE INFORMATION Page 1 Tech, Semis, Components and Distributors Equity Research

2 The Industrialization of Semis Contents PM Summary: favor growth over value... 3 Introduction & A brief word on our methodology... 3 Semi investors have historically paid up for safety... 3 Safety characteristics have not generated alpha... 7 Semi investors should instead pay up for growth Best individual metrics Multivariate analysis Semi Boards of Directors should shift compensation metrics to favor growth over safety Best ways to invest in the industrialization of semis theme NXP Semiconductors (NXPI, $77.78, Buy) Avago Technologies (AVGO, $129.92, Buy) Microsemi Corporation (MSCC, $31.33, Buy) Analog Devices (ADI, $51.65, Buy) Page 2 of 20

3 PM Summary: favor growth over value Over the last three years we ve written about our industrialization of semis thesis. Today we formalize our thesis with regression analyses. In summary, our work suggests that as management teams, boards of directors, and investors come to terms with a semiconductor industry that grows only slightly faster than global GDP, Companies will transition focus away from safety (high gross profit margin and low leverage) to growth (sales, EPS, and cash flow per share), often by acquiring other companies, cutting costs, and using appropriate financial leverage. Boards of directors will likewise shift performance targets away from safety to growth metrics, and Ultimately we believe investors will begin to reward companies that deliver faster earnings growth (through acquisitions & cost cuts, or otherwise) with higher valuation metrics. Introduction & A brief word on our methodology In 2012 when we launched coverage of the semiconductor industry, feedback from investors indicated that the key metric in evaluating semiconductors is gross profit margin, that acquisitions don t create value for equity because they rarely enable gross margin expansion, and that financially levered semiconductor companies represent very risky investments. We disagreed with those views, and over time built the regression analyses to support our opinions. We collected 10 fundamental metrics for approximately 90 companies across the semiconductor and industrial sectors over 3 separate time periods (the trailing 1, 3, and 5 year periods each ending 12/31/2014). We ran simple and multiple regressions of the fundamental data relative to valuation metrics (P/E and EV/EBITDA) to determine the answers to three questions for semis and (separately) for industrials: (1) what fundamental metrics were most highly correlated with high valuation?, (2) did investments that favored those fundamental metrics generate alpha?, and (3) if not, what were the characteristics most correlated with alpha generation? Separately, we studied whether performance criteria that boards of directors established for management teams were aligned with metrics that generate investment returns. We found that semi investors have historically favored safety (high gross margin and low financial leverage). These characteristics are not associated with alpha generation. Outsized returns are most correlated with growth metrics, specifically EPS, cash flow, and sales growth. We also found that BoDs place too much emphasis on safety (specifically margins) and not enough on EPS or cash flow growth. Semi investors have historically paid up for safety The valuation component of our study had the most striking differences between semi and industrial companies, and is an important area regarding our industrialization theme. We based this study on a simple regression of historical metrics versus valuation multiples at the beginning of the study period, which we then verified through multivariate iterations. For semis, investors have historically "paid up" Page 3 of 20

4 EV/EBITDA Tech, Semis, Components and Distributors (with high P/E and EV/EBITDA multiples) for high gross margins and low net leverage. By contrast, in industrials, investors do not place a consistent premium on any particular fundamental metric, meaning no metric was significant across all 3 periods of the study. While different time periods have slight differences in which metrics have the highest correlations (and some time periods have much stronger correlations than others), the important trends were evident semi investors apply premium valuations to companies with historically high gross margins and low net leverage. Other metrics were correlated during different periods, but we were more concerned with what investors have repeatedly paid a premium for. In the following exhibit, we show an example of the high correlation between historical gross profit margin and EV/EBITDA valuation for semiconductor companies in the study. Exhibit 1: Semiconductor simple regression study, historical GPM vs. EV/EBITDA valuation Historical GPM vs. EV/EBITDA Source: Company Data, FactSet, STRH Analysis y = x R² = % 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Historical GPM In the following exhibit, we outline the simple regression conclusions for each time period. Page 4 of 20

5 Exhibit 2: Simple regression study conclusions for what investors have paid up for What do Investors Pay For? Semis Industrials Regression Simple Regression Simple Regression Simple Regression P/E Valuation Gross Profit Margin Level Gross Profit Margin Level Net Leverage Net Leverage Net Leverage Gross Profit Margin Level ROIC level Gross Profit Margin Level Gross Profit Margin Level EV/EBITDA Gross Profit Margin Level Operating Profit Margin Level ROIC Level Valuation ROIC Level Net Leverage No significant metrics ROIC Level Operating Profit Margin Level P/E Valuation Gross Leverage Net Leverage EV/EBITDA Valuation Gross Profit Margin Level Gross Profit Margin Level Net leverage Source: Company Data, FactSet, STRH Analysis Next, we show the filtered, multiple regression conclusions. These results take the simple regression output a few steps further, removing high levels of multicollinearity and false positives to try and find the true drivers of valuation. In the following exhibit, we show an example of our multiple regression analysis for the same semiconductor study comparing EV/EBITDA valuation with historical gross profit margin. The multiple regression analysis verifies our simple regression above, and we show the results below. Page 5 of 20

6 EV/EBITDA Tech, Semis, Components and Distributors Exhibit 3: The Semi multi-regression study proved that semi investors apply a premium valuation for high gross margins Regression Statistics Multiple R R Square Adjusted R Square Standard Error Observations 48 ANOVA df SS MS F Significance F Regression Residual Total Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept Historical GPM y = 18.55x R² = % 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Historical GPM Source: Company Data, FactSet Next, we summarize the multiple regression conclusions for each time period which incorporate the broader, simple regression study. Page 6 of 20

7 Exhibit 4: Multiple regression study conclusions for what investors have paid up for What do Investors Pay For? Semis Regression Multiple Regression Multiple Regression Multiple Regression P/E Valuation Gross Profit Margin Level Gross Profit Margin level Net Leverage Net Leverage Gross Profit Margin Level ROIC level Gross Profit Margin level Gross Profit Margin Level EV/EBITDA Gross Profit Margin Level Valuation Industrials P/E Valuation EV/EBITDA Valuation No significant metrics No significant metrics No significant metrics Gross Profit Margin Level Gross Profit Margin Level Net leverage Source: Company Data, FactSet, STRH Analysis And finally, considering both regression sets across the 3 time periods, we show our conclusions. In this final set of results, it was important to discover the most important trends across all of the time periods as whole, not intermittent sources of significance. We found that while semi investors are willing pay a premium for high gross profit margins and low net leverage, industrial investors are not willing to a pay a consistent premium for any single metric. Exhibit 5: Final study conclusions for what investors have paid up for What do Investors Pay For? Semis Industrials Regression P/E Valuation EV/EBITDA Valuation P/E Valuation EV/EBITDA Valuation Final Conclusion Gross Profit Margin Level Net Leverage Gross Profit Margin Level No significant metrics No significant metrics Source: Company Data, FactSet, STRH Analysis Safety characteristics have not generated alpha Across the three time periods we studied, investing in semi companies with high margins and low net leverage (buying "safety") has not generated alpha. In the following exhibits, we demonstrate the lack of correlation between both historical GPM levels and net leverage with future returns over each of the 3 time periods. While some of regression lines do slope in a positive direction, the largest R-squared among the data set is 0.08, implying a correlation of approximately 28% in the best case. Page 7 of 20

8 Total Return Total Return Tech, Semis, Components and Distributors Exhibit 6: semi regression comparing historical GPM and net leverage to the company s future total return 200.0% 150.0% 100.0% Historical GPM vs Future Return y = x R² = % 0.0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% -50.0% % % Historical GPM Historical Net Leverage vs Future Return 200.0% 150.0% 100.0% 50.0% y = x R² = % % % % Historical Net Leverage Source: Company Data, FactSet Page 8 of 20

9 Total Return Total Return Tech, Semis, Components and Distributors Exhibit 7: semi regression comparing historical GPM and net leverage to the company s future total return 150.0% 100.0% Historical GPM vs Future Return y = x R² = % 0.0% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0% -50.0% % Historical GPM Historical Net Leverage vs Future Return 250.0% 200.0% 150.0% 100.0% 50.0% y = x R² = % % % Historical Net Leverage Source: Company Data, FactSet Page 9 of 20

10 Total Return Total Return Tech, Semis, Components and Distributors Exhibit 8: semi regression comparing historical GPM and net leverage to the company s future total return 120.0% 100.0% 80.0% 60.0% 40.0% 20.0% -40.0% -60.0% -80.0% % Historical GPM vs Future Return y = x R² = % -20.0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Historical GPM Source: Company Data, FactSet Historical Net Leverage vs Future Return 120.0% 100.0% -40.0% -60.0% -80.0% Historical Net Leverage y = x R² = % % % 60.0% 40.0% 20.0% Semi investors should instead pay up for growth Now that we know applying premium valuations for safe plays does not typically work, the next logical question becomes: What should I pay for? Across the three time periods we studied, the coincident metrics most strongly correlated with total shareholder return were (1) earnings growth, (2) revenue growth, and (3) free cash flow growth. Investors should note: the same characteristics are also the three most strongly correlated with shareholder return among industrials. Best individual metrics We used the same format to evaluate this section a diverse set of simple regressions, coordinated with a final multiple regression study. In the following exhibits as an example, we show the simple regression analysis of EPS growth and total return for semiconductor companies in the study, followed by a multiple regression analysis that verifies the result. Page 10 of 20

11 Total Return Tech, Semis, Components and Distributors Exhibit 9: Semiconductor simple regression study, EPS growth vs. total return EPS Growth vs. Total Return 300% 250% y = x R² = % 150% 100% 50% 0% -150% -100% -50% 0% 50% 100% 150% 200% 250% -50% -100% EPS Growth Source: Company Data, FactSet Page 11 of 20

12 Total Return Tech, Semis, Components and Distributors Exhibit 10: The Semi multi-regression study proved that EPS growth was the best coincident indicator of Total Return Regression Statistics Multiple R R Square Adjusted R Square Standard Error Observations 32 ANOVA df SS MS F Significance F Regression Residual Total Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept EPS Growth Correlation EPS Growth Rev Growth FCF Growth EPS Growth 1.00 Rev Growth FCF Growth % EPS Growth vs. Total Return 250% 200% 150% y = x R² = % 50% 0% -100% -50% 0% 50% 100% 150% 200% 250% -50% -100% EPS Growth Source: Company Data, FactSet In the following exhibit, we show the most significant individual metrics to explain total shareholder return for both Semi and Industrial investors across each of the three time periods. These are the final Page 12 of 20

13 results for the best coincident indicator, based on both regression sets. When looking for the single most important metric, EPS growth is hands down the largest driver of incremental shareholder value. Exhibit 11: Summary table showing the single best coincident indicator for total shareholder return over each period Semis Final Conclusion EPS Growth EPS Growth EPS Growth EPS Growth Industrials Final Conclusion EPS Growth EPS Growth FCF Growth EPS Growth Source: Company Data, FactSet, and STRH Analysis Multivariate analysis Aside from uncovering the best individual metrics, we also contemplated a traditional multivariate analysis that attempted to uncover the best combination of variables to describe total return. While most models ended up best described by only a single, independent variable (shown in the above section), two portions of the study were better explained through a combination of variables. The industrial model was best described by a combination of revenue growth and free cash flow growth, and with correlation between the two independent variables at only 39%. The semi model was also best described by a combination of revenue growth and free cash flow growth. The multicollinearity was slightly more prevalent here with a correlation of 46%. The multiple regression analysis results are shown in the 2 exhibits below. Page 13 of 20

14 Exhibit 12: The industrial multi-regression study proved that Rev growth and FCF growth together were the best coincident indicators of Total Return Regression Statistics Multiple R R Square Adjusted R Square Standard Error Observations 32 ANOVA df SS MS F Significance F Regression E-07 Residual Total Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept Rev Growth FCF Growth Correlation EPS Growth Rev Growth FCF Growth EPS Growth 1.00 Rev Growth FCF Growth Source: Company Data, FactSet, STRH Analysis Page 14 of 20

15 Exhibit 13: The semi multi-regression study proved that Rev growth and FCF growth together were the best coincident indicators of Total Return Regression Statistics Multiple R R Square Adjusted R Square Standard Error Observations 32 ANOVA df SS MS F Significance F Regression E-05 Residual Total Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept Rev Growth FCF Growth Correlation Rev Growth EBITDA Growth EPS Growth FCF Growth Rev Growth 1.00 EBITDA Growth EPS Growth FCF Growth Source: Company Data, FactSet, STRH Analysis Semi Boards of Directors should shift compensation metrics to favor growth over safety Armed with a view that, for both semis and industrials, growth in cash flow, earnings, and sales are more correlated with alpha generation than safety characteristics like high gross margin and low leverage, we investigated how well aligned management teams compensation metrics are with characteristics that drive alpha. We analyzed dozens of 10-K filings to first determine compensation metrics for both semi and industrial management teams. We then compared these results with the metrics that are most highly correlated with shareholder returns (growth in revenue, EPS, and FCF). Our analysis, summarized in the following exhibit, suggests BoDs, especially for semiconductor companies, could do a much better job of aligning performance metrics with alpha generating characteristics. Page 15 of 20

16 Exhibit 14: The percentage of companies that reward their management team based on each of the criteria listed Metric Semis Industrials Drives Returns? Margin / cost 87% 61% No Sales/Bookings growth 78% 57% Yes EPS-related 35% 48% Yes Cash flow 13% 30% Yes Returns (fundamental) 9% 48% No Returns (stock) 13% 30% Yes Source: Company Data, FactSet, STRH Analysis In particular, we highlight that the most commonly used compensation metric for Semiconductor companies is margins, despite our analysis showing that high margins isn t an alpha-generating characteristic. Meanwhile, the big alpha generating characteristics EPS and Cash Flow are used by semiconductor boards in only 35% and 13% of the time, respectively. Simply put, to better align management s and shareholders interests, semiconductor boards of directors should shift the metrics on which they measure management from safety (specifically margins) to growth (specifically EPS and cash flow). Best ways to invest in the industrialization of semis theme While M&A has been an ongoing phenomenon of semiconductor investing for the last three years, we anticipate it will continue, so long as debt capital markets cooperate. Companies in our coverage that we believe are most aligned with our industrialization thesis include: NXP Semiconductors (NXPI, $77.78, Buy) We believe NXPI is a leader relative to our industrialization of semiconductors theme. Even before the company acquired Freescale, it targeted outsized EPS growth through a combination of superior sales growth (with leading positions in ID and automotive) and shareholder friendly capital allocation (share buybacks). Freescale gives NXP a significant step-up in EPS and a $500 million per year cost savings EPS accelerator. Ultimately we believe NXP s earnings growth will be approximately 17% per year. Our PT is $103 based on 14x our 2016 EPS. Avago Technologies (AVGO, $129.92, Buy) We see AVGO as a clear innovator relative to our industrialization of semiconductors theme. The company s RF filtering expertise affords it a stable growth engine that it supplements with significant acquisitions. With BRCM we believe AVGO will have long-term organic earnings growth of 13% that bumps to 17% with future potential M&A. Our PT is $180 based on 14x our 2017 EPS. Page 16 of 20

17 Microsemi Corporation (MSCC, $31.33, Buy) Microsemi s intermittent acquisitions fuel a mostly inorganic growth profile. In the near-term we anticipate approximately 40% of company sales are poised to accelerate in Following its pending acquisition of PMC-Sierra (PMCS, $11.63, NR) we believe MSCC will have long-term organic earnings growth of 14% that bumps to 19% with future potential M&A. Our PT is $49 based on 12x our 2017 EPS. Analog Devices (ADI, $51.65, Buy) ADI is a more recent fit into our industrialization of semis theme. We believe management is slowly transitioning the company s under-levered balance sheet to one that uses leverage to fuel growth. This is most recently evidenced by the company s raise of $875m net incremental debt that it will use to fund either an acquisition or stock repurchase. Our PT is $71 based on 20x our C16 EPS. Page 17 of 20

18 Companies Mentioned in This Note Analog Devices, Inc. (ADI, $51.65, Buy) Amphenol Corp (APH, $47.43, Neutral) Arrow Electronics, Inc. (ARW, $48.32, Buy) Atmel Corp. (ATML, $7.68, Neutral) Avnet, Inc. (AVT, $39.08, Buy) Belden Inc. (BDC, $41.78, Buy) Linear Technology Corp. (LLTC, $39.51, Neutral) Microchip Technology Inc (MCHP, $41.85, Neutral) Maxim Integrated (MXIM, $33.15, Neutral) NXP Semiconductors N.V. (NXPI, $77.78, Buy) Photronics, Inc. (PLAB, $12.20, Buy) Sensata Technologies Holding NV (ST, $39.64, Neutral) TE Connectivity, Ltd. (TEL, $58.46, Buy) Texas Instruments Inc. (TXN, $51.53, Neutral) Xilinx, Inc. (XLNX, $43.36, Neutral) Avago (AVGO, $129.92, Buy) Microsemi (MSCC, $31.33, Buy) Cypress (CY, $8.39, Neutral) Intel (INTC, $32.68, Neutral) PMC-Sierra (PMCS, $11.63, NR) Analyst Certification I, William Stein, hereby certify that the views expressed in this research report accurately reflect my personal views about the subject company(ies) and its (their) securities. I also certify that I have not been, am not, and will not be receiving direct or indirect compensation in exchange for expressing the specific recommendation(s) in this report. Required Disclosures Analyst compensation is based upon stock price performance, quality of analysis, communication skills, and the overall revenue and profitability of the firm, including investment banking revenue. As a matter of policy and practice, the firm prohibits the offering of favorable research, a specific research rating or a specific target price as consideration or inducement for the receipt of business or compensation. In addition, associated persons preparing research reports are prohibited from owning securities in the subject companies. Charts indicating changes in ratings can be found in recent notes and/or reports at our website or by contacting SunTrust Robinson Humphrey. Please see our disclosures page for more complete information at Page 18 of 20

19 STRH Ratings System for Equity Securities 3 designations based on total returns* within a 12-month period** Buy total return 15% (10% for low-beta securities)*** Reduce total return negative 10% (5% for low Beta securities) Neutral total return is within the bounds above NR NOT RATED, STRH does not provide equity research coverage CS Coverage Suspended *Total return (price appreciation + dividends) **Price targets are within a 12-month period, unless otherwise noted ***Low Beta defined as securities with an average Beta of 0.8 or less, using Bloomberg s 5-year average Beta Legend for Rating and Price Target History Charts: D = drop coverage I = initiate coverage T = transfer coverage SunTrust Robinson Humphrey ratings distribution (as of 01/13/2016): Coverage Universe Investment Banking Clients Past 12 Months Rating Count Percent Rating Count Percent Buy % Buy % Neutral % Neutral % Sell/Reduce % Sell/Reduce % Page 19 of 20

20 Other Disclosures Information contained herein has been derived from sources believed to be reliable but is not guaranteed as to accuracy and does not purport to be a complete analysis of the security, company or industry involved. This report is not to be construed as an offer to sell or a solicitation of an offer to buy any security. SunTrust Robinson Humphrey, Inc. and/or its officers or employees may have positions in any securities, options, rights or warrants. The firm and/or associated persons may sell to or buy from customers on a principal basis. Investors may be prohibited in certain states from purchasing some over-the-counter securities mentioned herein. Opinions expressed are subject to change without notice. The information herein is for persons residing in the United States only and is not intended for any person in any other jurisdiction. SunTrust Robinson Humphrey, Inc. s research is provided to and intended for use by Institutional Accounts as defined in FINRA Rule 4512(c). The term Institutional Account" shall mean the account of: (1) a bank, savings and loan association, insurance company or registered investment company; (2) an investment adviser registered either with the SEC under Section 203 of the Investment Advisers Act or with a state securities commission (or any agency or office performing like functions); or (3) any other person (whether a natural person, corporation, partnership, trust or otherwise) with total assets of at least $50 million. SunTrust Robinson Humphrey, Inc. is a registered broker-dealer and a member of FINRA and SIPC. It is a service mark of SunTrust Banks, Inc. SunTrust Robinson Humphrey, Inc. is owned by SunTrust Banks, Inc. ("SunTrust") and affiliated with SunTrust Investment Services, Inc. Despite this affiliation, securities recommended, offered, sold by, or held at SunTrust Robinson Humphrey, Inc. and at SunTrust Investment Services, Inc. (i) are not insured by the Federal Deposit Insurance Corporation; (ii) are not deposits or other obligations of any insured depository institution (including SunTrust Bank); and (iii) are subject to investment risks, including the possible loss of the principal amount invested. SunTrust Bank may have a lending relationship with companies mentioned herein. SunTrust Robinson Humphrey, Inc All rights reserved. Reproduction or quotation in whole or part without permission is forbidden. ADDITIONAL INFORMATION IS AVAILABLE at our website, or by writing to: SunTrust Robinson Humphrey, Research Department, 3333 Peachtree Road N.E., Atlanta, GA Page 20 of 20