SMIC:Higher demand and EPS,but depreciation skews the story;Sell

What's changed

Historically, second-tier foundries have always run at lower utilization thanindustry leader TSMC, but SMIC has kept utilization above 97% since 2Q15and consistently above that of TSMC, likely due to a desire among itscustomers to support the Chinese government’s efforts to strengthen thedomestic foundry industry. We expect SMIC to maintain 97% averageutilization and surpass its target of a 20% revenue CAGR in 2016-2019. Wealso expect SMIC’s R&D grants to sustain and rise from US$34mn in 2015to US$65mn and US$95mn (19% and 20% of earnings) in 2016E and 2019Erespectively, referencing the industry’s development path in Taiwan.

Implications

We have raised our assumptions for utilization and government grants,which increases our 2016E-2018E EPS by 15%/146%/84%. However, wenote the company’s 7-year depreciation policy (vs. peers’ 5 years), whichwe believe leads to SMIC’s high margins and ROE relative to peers despiteits lower revenue/net PP&E ratio. Even using 7-year depreciation, our2017E-2018E depreciation estimates for SMIC are 37% above consensus.As a result, even as we turn more positive on SMIC’s demand and R&Dgrants outlook, we stay cautious on its returns and valuation. Additionally,to offset the differences in depreciation policies, we change our valuationmethodology from P/B-ROE to EV/GCI-CROCI/WACC, which is based ongross assets.

Valuation

Our 12-month target price rises to HK$8.40 from HK$7.40 (ADR to US$5.40from US$4.70) after adjusting for December’s reverse stock split andchanging our valuation basis from P/B to EV/DACF, which is derived fromEV/GCI-CROCI/WACC and factors in SMIC’s superior growth in 2016-2019E.Our new target price implies a NTM P/B of 1.19X (above UMC’s 0.61X andHua Hong’s 0.74X) and 24% downside potential; we maintain Sell.

Key risks

Higher-than-expected R&D grants and stock purchases by gov’t entities.