Global smartphones: UBS cuts 2027 smartphone sell-in outlook as memory costs become a larger OEM burden

Overview

UBS argues that a continuing memory upcycle will make procurement and cost management more difficult for smartphone OEMs in 2027, leading it to lower its global unit forecast. The firm expects premium-positioned Apple and Samsung to be more resilient than Chinese Android peers and remains selective across the supply chain.

Core views

UBS's APAC technology-tour takeaways point to further downside for Android smartphone builds excluding Samsung in 2027, alongside relative resilience at Apple and Samsung. The firm therefore reduced its 2027 global smartphone sell-in estimate to 1.11bn units, a 3% year-on-year decline, from its prior forecast of 1.16bn units and 2% growth. It initiated a 2028 estimate of 1.13bn units, up 2% year-on-year, while leaving its 2026 forecast unchanged at 1.14bn units, down 10% year-on-year. The central constraint is memory. UBS expects memory suppliers to continue prioritizing hyperscale and AI customers, potentially making memory supply a greater bottleneck for handset makers in 2027 than in 2026. Under UBS's base case for the DRAM upcycle to last at least through 2Q28E and the NAND upcycle through 4Q27E, DRAM and NAND are estimated to account for 35% of flagship/premium smartphone bill-of-material costs and 56% of low-end smartphone BOM costs by 4Q26E, versus 13% and 29%, respectively, in 4Q25. UBS notes Xiaomi's commentary that memory already represents about 40% of average smartphone BOM cost and 50-60% for low-end models. Higher component costs pressure margins, while OEM price increases intended to offset them could further weaken end demand. UBS sees the cost shock as more manageable for Apple and, to a lesser extent, Samsung because of their premium portfolios; it considers the effect more detrimental for other OEMs. Counterpoint data show global smartphone sell-through down 8.4% year-on-year in July 2026 year-to-date. Apple was up 8.9%, Samsung was down 4.8%—with positive sell-in suggesting some channel inventory build—while Vivo, Xiaomi, Oppo and Transsion fell 16.4%, 17.9%, 18.4% and 22.4%, respectively. UBS expects premium models to hold up better in the downturn, allowing Apple and Samsung to continue outperforming and gain incremental share. For Apple, suppliers expect 2027 to be strong, supported by the iPhone Duo, estimated at roughly 10m units in its first 12 months, and a new launch schedule that UBS believes can drive year-on-year unit growth. Samsung expects to grow units and gain share, although UBS flags profitability as a concern. UBS's forecasts show Apple units rising from 261m in 2026E to 273m in 2027E and 283m in 2028E, while Samsung rises from 250m to 258m and 261m, respectively. By contrast, UBS forecasts Xiaomi at 123m units in 2026E, 111m in 2027E and 119m in 2028E; OPPO at 99m, 88m and 88m. Within the supply chain, UBS prefers MediaTek and Samsung, both APAC Key Call Buys, as well as Buy-rated ASE, Hon Hai, JCET, Micron, Murata, SEMCO, SK Hynix and TSMC. The report remains cautious on Neutral-rated LG Display, Maxscend and Xiaomi, and on Sell-rated Largan. Apple and Transsion are Neutral-rated OEMs. The preference reflects UBS's view that memory and MLCC suppliers are better positioned than the broader smartphone complex amid elevated memory costs and pressured handset demand.

Analysis framework

UBS combines APAC technology-tour observations with smartphone sell-through data, company commentary, supplier expectations and its own shipment forecasts. It links the expected DRAM/NAND cycle to smartphone BOM inflation, then assesses the resulting margin and demand effects across premium and lower-end OEMs before identifying relatively preferred supply-chain beneficiaries.

Methodology notes

• Industry AnalysisSupply-demand framework

  Memory supply-demand analysis

  UBS assesses how memory makers prioritizing hyperscale and AI customers can restrict memory availability for handset OEMs and raise component costs.

• Industry AnalysisUpstream-Midstream-Downstream Transmission

  Memory-cost pass-through across the smartphone supply chain

  The report traces higher DRAM and NAND costs into OEM margins, handset pricing, end demand and relative performance among component suppliers.

• Valuation methodsP/E and PEG Valuation

  Forward P/E multiple analysis

  The preferred and least-preferred list presents 2027E and 2028E P/E multiples as valuation reference points for covered supply-chain stocks.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• MediaTek (2454.TW)

  Preferred smartphone supply-chain exposure

  Strengths

  APAC Key Call Buy

  Comparison

  Preferred over the least-preferred smartphone names

  Risks

  Cyclical handset demand and inventory-cycle exposure

• Samsung Electronics (005930.KS)

  Preferred memory supplier and relatively resilient smartphone OEM

  Strengths

  Buy-rated; UBS expects relative resilience and potential market-share gains

  Weaknesses

  Profitability remains a concern for the handset business

  Comparison

  Expected to outperform Chinese Android peers

  Risks

  Memory-cycle and smartphone-demand sensitivity

• ASE (3711.TW)

  Preferred smartphone supply-chain exposure

  Strengths

  Buy-rated

  Comparison

  Included in UBS's most-preferred list

  Risks

  Cyclical handset demand and inventory-cycle exposure

• Hon Hai Precision (2317.TW)

  Preferred smartphone supply-chain exposure

  Strengths

  Buy-rated

  Comparison

  Included in UBS's most-preferred list

  Risks

  Cyclical handset demand and inventory-cycle exposure

• JCET Group (600584.SH)

  Preferred smartphone supply-chain exposure

  Strengths

  Buy-rated

  Comparison

  Included in UBS's most-preferred list

  Risks

  Cyclical handset demand and inventory-cycle exposure

• Micron Technology (MU.US)

  Preferred memory-supplier exposure

  Strengths

  Buy-rated; aligned with UBS's preference for memory suppliers

  Comparison

  Preferred over pressured smartphone OEM exposure

  Risks

  Memory-cycle sensitivity

• Murata Manufacturing (JP6981)

  Preferred MLCC-supplier exposure

  Strengths

  Buy-rated; aligned with UBS's preference for MLCC suppliers

  Comparison

  Included in UBS's most-preferred list

  Risks

  Cyclical handset demand and competitive pressure

• Samsung Electro-Mechanics (009150.KS)

  Preferred MLCC-supplier exposure

  Strengths

  Buy-rated; aligned with UBS's preference for MLCC suppliers

  Comparison

  Included in UBS's most-preferred list

  Risks

  Cyclical handset demand and competitive pressure

• SK Hynix (000660.KS)

  Preferred memory-supplier exposure

  Strengths

  Buy-rated; aligned with UBS's preference for memory suppliers

  Comparison

  Included in UBS's most-preferred list

  Risks

  Memory-cycle sensitivity

• TSMC (2330.TW)

  Preferred smartphone supply-chain exposure

  Strengths

  Buy-rated

  Comparison

  Included in UBS's most-preferred list

  Risks

  Cyclical handset demand and inventory-cycle exposure

• LG Display (034220.KS)

  Least-preferred smartphone supply-chain exposure

  Weaknesses

  Neutral-rated

  Comparison

  Less preferred than UBS's Buy-rated component and memory names

  Risks

  Cyclical demand, inventory cycles and competitive pressure

• Maxscend (300782.SZ)

  Least-preferred smartphone supply-chain exposure

  Weaknesses

  Neutral-rated

  Comparison

  Less preferred than UBS's Buy-rated component and memory names

  Risks

  Cyclical demand, inventory cycles and competitive pressure

• Xiaomi (01810.HK)

  Cautious OEM exposure

  Weaknesses

  Neutral-rated; sell-through was down 17.9% YoY in July 2026 year-to-date

  Comparison

  Expected to lag Apple and Samsung amid memory-cost pressure

  Risks

  Margin pressure from memory costs and downside to end demand from price increases

• Largan (3008.TW)

  Least-preferred smartphone supply-chain exposure

  Weaknesses

  Sell-rated

  Comparison

  Included in UBS's least-preferred list

  Risks

  Cyclical demand, inventory cycles and competitive pressure

Key data

• 2027E global smartphone sell-in1.11bn units-3% YoY; cut from 1.16bn units and +2% YoY previously
• 2026E global smartphone sell-in1.14bn units-10% YoY; unchanged forecast
• 2028E global smartphone sell-in1.13bn unitsInitial forecast, +2% YoY
• Memory share of flagship smartphone BOM35%4Q26E, versus 13% in 4Q25
• Memory share of low-end smartphone BOM56%4Q26E, versus 29% in 4Q25
• Global smartphone sell-through-8.4% YoYJuly 2026 year-to-date, according to Counterpoint
• Apple sell-through growth+8.9% YoYJuly 2026 year-to-date
• Xiaomi sell-through growth-17.9% YoYJuly 2026 year-to-date

Impact & implications

UBS expects persistent memory inflation to reduce the industry’s 2027 unit outlook and create a sharper divide between premium OEMs with better cost absorption and lower-end Android vendors. Its preferred positioning is toward memory, MLCC and selected supply-chain suppliers rather than the broader handset market.

Risks

• The handset sector is cyclically sensitive to macroeconomic conditions.
• OEMs can lose market share rapidly because of poor product timing, design or distribution.
• Supply-chain companies face both macro-driven cyclicality and inventory cycles.
• High competition makes product innovation and R&D investment critical to success.

What to watch

• Memory procurement conditions as suppliers continue to prioritize hyperscale and AI customers.
• The duration of the DRAM and NAND upcycles assumed through at least 2Q28E and 4Q27E, respectively.
• Whether higher handset prices to offset memory costs further weaken end demand.
• Apple and Samsung unit growth, market-share performance and Samsung profitability.