Opportunity to Build Positions on the Dip in SEMCO and TDK vs. Elevated Expectations in Murata and Taiyo Yuden We prefer SEMCO and TDK... In this report, we give a quick update on stocks that look most compelling to us after the recent sell-off in MLCC stocks over the last month, and our views on the stocks in Korea and Japan. Over the past month, the highest-beta MLCC stocks in Taiwan and Taiyo Yuden have been the worst performers, followed to a lesser extent by Murata and SEMCO. TDK (least exposed) fared best. Of these, SEMCO and TDK look oversold to us. In our minds, risk-reward in these two stocks has become more compelling after the sell-off, and we think investors should focus on them. For Japanese MLCC stocks, we prefer TDK because we expect not only auto MLCC earnings expansion but also continued growth in rechargeable battery earnings. ...over Murata and Taiyo Yuden: Murata and Taiyo Yuden will sustain competitive advantages in high end MLCC technology and productivity over non-Japanese competitors. However, market expectations for their earnings in F3/20 and thereafter look too high to us. Some in the market are assuming that price hikes across all MLCC products and resulting top-line growth will feed entirely into the OP line. Our view is that most of the price hikes will be for high-end and automotive (which also includes low-end) products, and that initial pricing on high-value-added (small, high- capacitance) products for mobile devices (smartphones, etc.) was quite high. As such, prices, despite being cut more gradually than usual because of tight overall supply/demand for MLCC, will have to come down. In addition, we believe that Murata and Taiyo Yuden will need to incur considerable capex to respond to growing demand for automotive MLCC products. Therefore, production-related costs will also climb. 1608-sized MLCCs, the primary automotive type, have cases that are 19x larger than 0603-sized MLCCs (the main size used in smartphones): The primary size of automotive MLCCs, 1608, has a case that is 19x larger than 0603-sized MLCCs (the main size for smartphones). This means more limited volumes being produced for these automotive types on the same production equipment. High capacitance/high-voltage products used in automobile power trains primarily have larger sizes of at least 3216. These require higher voltage of at least 50V, and are prone to electrical shorting when MLCC layers get thinner (because of movement/impact). Temperature control also degrades at smaller MLCC sizes, which makes it difficult to miniaturize. We think these problems are the reason for more larger-sized MLCCs. MLCCs historically have trended toward smaller sizes and larger capacities. With sizes shrinking and capacity increasing, volumes produced from the same equipment have risen and prices have climbed. As the number of MLCCs built into each smartphone has risen, so2too has the need for smaller and higher-capacity MLCCs, and this has helped earnings grow for Murata, the pioneer of MLCCs in this regard. We believe the growing number of MLCCs built into each car/vehicle will allow MLCC demand to expand, and fuel further earnings growth for the limited number of players, which includes Murata and TDK, capable of supplying a stable source of highly-reliable automotive products. That said, boosting automotive MLCC production volumes requires significant capital expenditure, and with marginal profit ratios for automotive MLCCs not as high as for high-end smartphone MLCC products, we see risk that near-term earnings may fall short of inflated market expectations. Japan – curbing excessive expectations of price hikes: Market expectations for MLCC price hikes have corrected from July peak levels, and shares of Murata and Taiyo Yuden have retreated 16% and 29%, respectively (as of September 13's close) vs. the TOPIX remaining relatively unchanged and 7-9% pullbacks for Japanese component stocks in our coverage. Murata is negotiating price hikes for all MLCC products for all customers, aiming to do so by end-December, and implement increases from Jan 2019. We think hikes are feasible for low-end and automotive (low-end products also included) products for two reasons: Meanwhile, given the high prices for small, high-capacitance high-value-added products in high-end smartphones, we expect price declines to be more muted than past trends. The net impact is still positive for companies' earnings. For details, refer to Shoji Sato's September 14 report (Electronic Components: July Japan Capacitor Output (METI): Healthy Demand for Ceramic Capacitors Continues). Exhibit 1:MLCC: Size comparison 45323216201216081005060304020201 Length4.5mm3.2mm2.0mm1.6mm1.0mm0.6mm0.4mm0.2mm Width3.2mm1.6mm1.25mm0.8mm0.5mm0.3mm0.2mm0.1mm Thickness2.0mm0.85mm0.6mm0.8mm0.5mm0.3mm0.2mm0.1mm Volume28.8mm34.35mm31.5mm31.024mm30.25mm30.054mm30.016mm30.003mm3 Volume ratio533.380.627.819.04.61.00.300.04 Source: Company data, Morgan Stanley Research Exhibit 2:Number of MLCCs per vehicle ConventionalLow-endMid-classHigh-endUltra high-end Pure ICEISSMicro HEVEV, ECV (units: pcs) Power Train450-600600-800800-1,0001,900-2,3002,700-3,100 Safety Comfort Infortainment Others Mild, Strong HEV, PHEV 1,000-1,400 500-800 400-700 500 Source: Company data, Morgan Stanley Research 1. Hikes so far for Murata's low-end products (20%+ of its MLCC sales) have been very limited, yet low-end market prices have risen substantially. 2. Automotive MLCCs, where demand is expanding, have larger sizes than smartphone MLCCs, with a complex inspection process and heavier production burden.3 After the reset – is now the time to build positions We think opportunities in MLCC stocks have arisen partly because, in each case, noise around speculative low-end spot markets has caused disproportionate stock underperformance despite robust 2H earnings fundamentals. However, we do not think these news flows have as many longer-term fundamental repercussions (as we detail below) as short-term sentiment would indicate. We hold the following views: Exhibit 3:Increases in FY1 Earnings Estimates over the Past Month – Remain Positive, Led by High-Beta MLCC Companies 6% 5% 3%3% 1%1% 0% 0% 1% 2% 3% 4% 5% 6% 7% SamwhaTaiyo Yuden YageoSEMCOTDKMurataWalsin 1MFY1 Earnings RevisionAvg Source: Thomson Reuters Exhibit 4:MTD Stock Performance – Midrange/Low-end and High- Beta MLCC Stocks Underperformed Most -27%-26% -21% -12%-12%-10% -8% -35% -25% -15% -5% 5% YageoWalsinTaiyo Yuden MurataTDKSEMCOSamwha MTD Share Price Perf.Avg Source: Thomson Reuters, Price as of 13th Sept, 2018 Customers will not give up LTA (long-term agreement) orders in 2019. Content growth and new TAM creation (auto, industrial, 5G base stations, data centers) will not diminish. Direct customer demand has been steady vs. news of some distributor pushback on low-end markets. Exhibit 5:Share Price Performance of MLCC Stocks (YTD) – Recent Weakness In All Spectrum 80100 120140 160180 200220 240 SEMCOMurataTDK Taiyo YudenYageo Source: Bloomberg Exhibit 6:Earnings Estimate Revision Breadth (FY2) for MLCC Stocks Continues Higher -50% 0% 50% 100% -100% -50% 0% 50% 100% Avg Earning Revision Breadth (3MMA) Avg Share Perf YoY (RHS) Source: Thomson Reuters; Note: MLCC Stocks are SEMCO, Murata, TDK, Taiyo Yuden4。。。。。。