PDF 原檔：報告_UBS_智慧型手機記憶體成本_20260922_original.pdf
圖片清單（已驗證 2026-09-24）
| 檔名 | size | 分類 | 親眼所見內容 |
|---|---|---|---|
報告_UBS_智慧型手機記憶體成本_20260922_002.png |
54.9KB | 真資料圖 | 主要手機品牌 YoY Sell-Through 成長率折線圖（Jan'25-Jul'26），Apple/Samsung 表現優於 Xiaomi/Oppo/Vivo/Transsion（後者多數已跌破 -20%） |
其餘 001/003/004（<40KB）為 DRAM/NAND 占手機 BOM 成本比重圖，依 trimmed 文字描述判斷為真資料圖但未逐張放大驗證。
原始內容
Global I/O Smartphones
Memory to continue to weigh in on Smartphones in 2027
Smartphone unit Sell-In -3% YoY in '27E; initial '2 E forecasts at +2% YoY
UBS 3 26 APAC Tech Tour takeaways (link) continue to point towards (1) further downside for Android ex-Samsung smartphone unit builds into 2027, and (2) relative resilience for Apple and Samsung. Memory procurement could turn out to be incrementally more challenging for smartphone OEMs in 2027 than in 2026. Memory makers continue to prioritise hyperscale and AI customers. Hence memory supply could be even more of a bottleneck next year. As such, we further de-risk our '27E smartphone unit Sell-In forecasts to 1.11bn (-3% YoY) from 1.16bn (+2% YoY) prior and roll out initial '2 E forecasts at 1.13bn (+2% YoY). We leave our '26E forecasts unchanged at 1.14bn (-10% YoY).
Memory costs continue to go higher
We now estimate memory to account for 35% of the bill of material (BOM) costs of flagship/premium smartphones and 56% of those of low-end smartphones by 4 26E (from 13%/29% in 4 25), based on our current base case scenario of the DRAM/NAND upcycle extending until at least 2 28E/4 27E. This is broadly consistent with Xiaomi's commentary that memory now accounts for 40% of average smartphone BOM costs and 50-60% of low-end BOM costs. We believe the current memory cycle will remain a key challenge for OEMs as (1) increasing costs continue to pressure margins, and (2) ASP hikes aiming to offset such pressures are likely to pose incremental downside risk to end demand. Over the long term, we believe cost control will be more manageable for Apple and, to a lesser extent, Samsung, given their premium product portfolio, while more detrimental for others.
Apple & Samsung outperformance likely to continue
Global smartphone unit Sell-Through (source: Counterpoint) is now down 8.4% YoY Jul '26 YTD. Within this, we note Apple (+8.9% YoY) and Samsung (-4.8% YoY; Sell-In up YoY implies some channel inventory build) continue to outperform their Chinese Android peers (Vivo/Xiaomi/Oppo/Transsion: -16.4%/-17.9%/-18.4%/-22.4% YoY). With premium models like to fare better in this downturn, we believe Apple and, to a lesser extent, Samsung will continue to outperform the broader market and gain incremental share. iPhone suppliers continue to expect '27E to be a strong year for Apple, with the iPhone Duo (c.10m units in first 12M) as well as the new launch schedule likely to drive YoY unit growth. Samsung continues to believe it will be able to grow unit and gain market share, though profitability will remain a concern.
Selective on smartphone plays; prefer memory/MLCC suppliers
Within the smartphone supply chain, we continue to prefer MediaTek and Samsung (both APAC Key Call Buy) as well as ASE, Hon Hai, JCET, Micron, Murata, SEMCO, SK Hynix and TSMC (all Buy). We remain cautious on LG Display, Maxscend, Xiaomi (all Neutral) and Largan (Sell). We have Neutral ratings on OEMs Apple and Transsion.
Equities
Global
Technology
Nicolas Gaudois
Analyst
nicolas.gaudois@ubs.com +65-6495 5148
David Vogt
Analyst
david.vogt@ubs.com +1-212-713 4296
Randy Abrams
Analyst
randy.abrams@ubs.com +886-2-8722 7338
Timothy Arcuri
Analyst
+1-415-352 5676
Jimmy Yoon
Analyst
jimmy.yoon@ubs.com +65-6495 4617
Luke Yoo
Analyst
luke.yoo@ubs.com +82-2-3702 8132
Sunny Lin
Analyst sunny.lin@ubs.com +886-2-8722 7346
Shingo Hirata, CFA
Analyst
+81-3-5208 6224
Jimmy Yu
Analyst
S1460517080002 jimmy.yu@ubs.com +86-21-3866 8880
Kenji Yasui
Analyst
kenji.yasui@ubs.com +81-3-5208 6211
Joseph Koh
Analyst
joseph.koh@ubs.com +603-2781 1182
KEY CHARTS
Figure 1: Most/Least Preferred lists
| Company | Ticker | Analyst | Rating | Price | PT | '27E PE | '28E PE | '27E P/BV | '28E P/BV |
|---|---|---|---|---|---|---|---|---|---|
| Most preferred | |||||||||
| ASE | 3711.TW | Sunny Lin | Buy | NT$638 | NT$835 | 20.5 | 16.4 | 5.39 | 4.66 |
| Hon Hai Precision | 2317.TW | Randy Abrams | Buy | NT$250.5 | NT$340 | 14.7 | 12.8 | 1.87 | 1.70 |
| JCET Group | 600584.SS | Jimmy Yu | Buy | Rmb73 | Rmb135 | 31.8 | 23.2 | 3.76 | 3.30 |
| MediaTek | 2454.TW | Sunny Lin | Buy | NT$4710 | NT$6500 | 26.0 | 15.6 | 12.96 | 9.84 |
| Micron Technology | MU.O | Timothy Arcuri | Buy | US$1015.8 | US$1625 | 5.4 | 3.7 | 4.48 | 3.00 |
| Murata Mfg. | JP6981 | Shingo Hirata | Buy | ¥7763 | ¥13200 | 23.9 | 17.7 | 4.51 | 3.92 |
| Samsung Electronics | 005930.KS | Nicolas Gaudois | Buy | Won261000 | Won535000 | 3.6 | 3.1 | 1.52 | 1.16 |
| SEMCO | 009150.KS | Jimmy Yoon | Buy | Won1388000 | Won2700000 | 24.8 | 15.6 | 6.84 | 4.85 |
| SK Hynix | 000660.KS | Nicolas Gaudois | Buy | Won1857000 | Won3000000 | 3.0 | 2.4 | 2.27 | 1.62 |
| TSMC | 2330.TW | Sunny Lin | Buy | NT$2460 | NT$3650 | 16.5 | 12.4 | 6.08 | 4.44 |
| Least preferred | |||||||||
| Largan | 3008.TW | Randy Abrams | Sell | NT$6430 | NT$5000 | 30.0 | 25.7 | 3.96 | 3.64 |
| LG Display | 034220.KS | Jimmy Yoon | Neutral | Won8400 | Won11600 | 7.2 | 5.2 | 0.57 | 0.52 |
| Maxscend | 300782.SZ | Jimmy Yu | Neutral | Rmb73.74 | Rmb80.2 | NM | 146.9 | 3.40 | 3.32 |
| Xiaomi | 1810.HK | Jimmy Yu | Neutral | Rmb26.4 | Rmb30.4 | 19.6 | 13.5 | 1.97 | 1.73 |
Source: UBS estimates. Prices as of 18 September 2026.
Figure 2: DRAM/NAND as % of flagship smartphone BOM costs
Source: Counterpoint, Fomalhaut, UBS estimates
Figure 4: YoY smartphone unit Sell-Through growth by major OEM, Jan '25 - Jul '26
Source: Counterpoint
Figure 3: DRAM/NAND as % of low-end smartphone BOM costs
Source: Gartner, UBS estimates
Figure 5: YoY smartphone unit Sell-Through growth by major OEM, Jul '26 YTD
Source: Counterpoint
Figure 6: UBS smartphone unit Sell-In forecasts
| Smartphone volume | (m) | 2025 | 1Q26 | 2Q26 | 3Q26E | 4Q26E | 2026E | 1Q27E | 2Q27E | 3Q27E | 4Q27E | 2027E | 1Q28E | 2Q28E | 3Q28E | 4Q28E | 2028E |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Samsung | 240 | 60 65 | 66 | 60 | 250 | 67 64 | 66 | 61 | 258 | 67 | 65 | 67 | 62 | 261 | |||
| % YoY | 7% | -2% 12% | 8% | 0% | 4% | 12% | -1% | 0% | 2% | 3% | 1% | 2% | 2% | 2% | 1% | ||
| % of total | 19% | 20% 23% | 24% | 20% | 22% | 24% 24% | 25% | 21% | 23% | 24% | 24% 25% | 20% | 23% | ||||
| Apple | 241 | 60 56 | 55 | 90 | 261 | 63 59 | 58 | 94 | 273 | 66 | 61 60 | 96 | 283 | ||||
| % YoY | 9% | 17% 23% | 3% | 0% | 8% | 5% | 5% | 5% | 4% | 5% 4% | 4% 4% | 3% | 4% | ||||
| % of total | 19% | 20% 20% | 20% | 30% | 23% | 23% | 22% | 22% | 32% | 25% | 24% | 22% | 22% | 32% | 25% | ||
| Huawei | 48 | 14 16 | 14 | 13 | 56 | 14 | 16 | 14 | 13 | 57 | 14 | 16 | 14 | 14 | 58 | ||
| % YoY | 0% | 3% 24% | 27% | 12% | 16% | 4% | 3% | 0% | 0% | 2% | 0% | 0% 0% | 8% | 2% | |||
| % of total | 4% | 5% 6% | 5% | 4% | 5% | 5% | 6% | 5% | 4% | 5% | 5% | 6% | 5% | 5% | 5% | ||
| OPPO* | 102 | 31 29 | 21 | 18 | 99 | 22 | 22 | 22 | 22 | 88 | 22 | 22 | 22 | 22 | 88 | ||
| % YoY | -3% | 31% 20% | -24% | -33% | -3% | -28% | -24% | 5% | 22% | -11% | 0% | 0% 0% | 0% | 0% | |||
| % of total | 8% | 10% 10% | 8% | 6% | 9% | 8% | 8% | 8% | 7% | 8% | 8% | 8% | 8% | 7% | 8% | ||
| Vivo | 106 | 21 21 | 20 | 21 | 83 | 19 | 19 | 19 | 19 | 76 | 19 | 19 | 19 | 19 | 76 | ||
| % YoY | 6% | -7% -22% | -31% | -22% | -21% | -10% | -10% | -5% | -10% | -9% | 0% | 0% | 0% | 0% | 0% | ||
| % of total | 8% | 7% 8% | 7% | 7% | 7% | 7% | 7% | 7% | 6% | 7% | 7% | 7% | 7% | 6% | 7% | ||
| Xiaomi | 165 | 34 31 | 29 | 29 | 123 | 29 | 28 | 27 | 27 | 111 | 31 | 30 | 29 | 29 | 119 | ||
| % YoY | -2% | -19% -26% | -32% | -24% | -26% | -14% | -10% | -8% | -7% | -10% | 7% | 7% | 7% | 10% | 8% | ||
| % of total | 13% | 12% 11% | 11% | 10% | 11% | 11% | 10% | 10% | 9% | 10% | 11% | 11% | 11% | 10% | 11% | ||
| 96 | 19 20 | 18 | 17 | 74 | 19 | 20 | 19 | 18 | 75 | 19 | 21 | 20 | 21 | 81 | |||
| Transsion | % YoY | -9% | -10% -16% | -37% | -25% | -23% | 0% | -3% | 1% | 6% | 1% | 0% | 8% | 8% | 18% | 8% | |
| % of total | 7% | 6% 7% | 5% | 6% | 7% | 7% | 6% | 6% | 6% | 6% | 7% | 7% | 6% | 7% | |||
| 6% | 13 | 13 | |||||||||||||||
| Lenovo | % YoY | 55 3% | 13 -8% | 13 | 13 | 13 | 51 | 12 | 12 | 14 | 13 3% | 50 | 12 | 12 6% | 0% | 51 2% | |
| 6% | -8% | -15% | -7% | -10% | -6% | 3% | -3% | 5% | -2% | 4% | 5% | ||||||
| % of total | 4% 69 | 4% 5% 19 | 5% | 4% | 5% | 4% | 4% 14 | 5% 14 | 4% 14 | 5% | 4% | 5% | 5% | 14 | 56 | ||
| Honor | % of total | 5% | 12% | 14 | 14 | 14 | 61 | 14 | 56 | 14 | 14 | 14 | |||||
| % YoY | 8% | -13% | -20% | -24% | -12% | -26% | 0% | 0% | 0% | -8% | 0% | 0% | 0% | 0% | 0% | ||
| 6% 5% | 5% | 5% | 5% | 5% | 5% | 5% | 5% | 5% | 5% | 5% | 5% | 5% | 5% | ||||
| Other | % YoY | 143 | 24 12 | 23 | 22 | 80 | 15 | 15 | 16 | 16 | 62 | 13 | 13 | 14 | 14 | 54 | |
| % of total | 8% 11% | -42% | -65% | -39% -32% 7% | -44% | -38% | 26% | -29% | -24% | -22% | -13% | -13% | -13% | -16% | -14% 5% | ||
| 1,265 | 8% 294 276 | 4% | 8% 273 | 295 | 7% 1,139 | 5% | 6% 268 | 6% 268 | 6% | 6% 1,105 | 5% 277 | 5% 274 | 5% 272 | 5% 304 | 1,127 | ||
| Total | % YoY | -13% | 273 | 296 | |||||||||||||
| 3% | -4% | -7% | -2% | 1% | 2% | 2% | |||||||||||
| -7% | -15% | -10% | -3% | 0% | -3% | 3% | 2% |
Source: Gartner, Company data, UBS estimates. *Note: Oppo units include OnePlus but exclude Realme.
Valuation Method and Risk Statement
The handset industry, as part of consumer electronics, is cyclically sensitive to the economy and so upside and downside risks relate to macro drivers. Other downside risks include the rapid loss of market share due to poor product timing, design or distribution. The supply chain companies into the handset industry also operate in a highly cyclical market both from macro concerns and inventory cycles within the supply chain. These companies also operate in a highly competitive environment with product innovation and R&D investments critical to success. Valuation methods across the sector include multiple analysis.