Suggested path: the verdict, the four-line chain block, the numbers, the playbook and the update log, about 8 minutes. The full argument is folded; tap it to open.建議讀法：先看結論，再看卡點四行、數字、操作框架和更新紀錄，約 8 分鐘。完整論證已收合，點開才會展開。
Company deep dive · personal research公司深度研究 · 個人研究
Himax Technologies (HIMX)
Himax sells no AI chip. It sells the chips that decide what a car screen shows and how it reads your touch, and cars are now over half of its revenue. In Q2 its gross margin (what is left of each $100 of sales after direct costs) rose from 30.4% to 33.1%, and both revenue and margin beat the company's own guide.Himax 賣的不是 AI 晶片，而是決定車用螢幕「顯示什麼、怎麼讀你的觸控」的驅動晶片，車用已占營收一半以上。第二季毛利率（每賣 100 元，扣掉直接成本剩下的錢）從 30.4% 升到 33.1%，營收和毛利率都比公司自己的財測好。
Verdict: a real automotive leader, but the price is 20% above the $12.4 target, and the research view is avoid.結論：真正的車用龍頭，但現價比目標價 12.4 美元高 20%，研究觀點是迴避。
Three cases: bear $4.4 (30% odds) the cycle turns down and there is no optics revenue; base $12.5 (50%) automotive grows double digits plus about $50M of optics revenue in 2027; bull $24.3 (20%) optics scales to about $150M and gross margin reaches 36%.
One line: the 08-06 Q2 report beat the guide, yet the price closed at $14.85 on 10-06, almost the same as the $14.92 this page showed on 06-30. This is the first time the page gives a target price, $12.4, so the view moves from "no strong edge" to avoid.
$14.85
Close 10-0610-06 收盤
$12.4
Target (16% below the price)目標價（比現價低 16%）
33.1%
Q2 gross margin (Q1: 30.4%)第二季毛利率（第一季 30.4%）
If you hold it: default to trimming in steps, not all at once. New money: not yet; buy only below $10.8 and after the Q3 report (11-05) confirms margin and the first optics shipments.
The automotive lead is real. Cars were "well over half" of Q2 revenue, and the company guides automotive driver-chip sales up by a "solid double-digit" percentage again in Q3 (company statement).
Margin is rising, but still thin. 33.1% means $33 left from each $100 of sales; after operating costs, about $11 is operating profit. The Q3 guide is about 34%.
The price has already paid. About 73× the last twelve months of profit (the P/E: how many times earnings per share the price is), about 37× this year's expected profit. This research's target of $12.4 is 16% below the price. The sell-side (analysts at brokers who write research for clients) average target, $30.2 (3 analysts), is more than double the price and needs profit to double.
Risks: about 74% of revenue comes from China-headquartered customers (2025 annual report); the core business moves with the cycle; co-packaged optics (CPO: moving data as light, built right next to the chip) and smart glasses ship only small volumes in 2026.
Sources: IBKR daily bars 10-06; company Q2 release and call 08-06; stockanalysis 10-07. Line-by-line sources are at the bottom.來源：IBKR 日線 10-06；公司第二季財報與法說 08-06；stockanalysis 10-07。逐項來源見最下方。
Chain block — where Himax sits, in four lines卡點四行 — Himax 在供應鏈上到底是什麼角色
One line: Himax is an important supplier of car-screen chips, not a choke point for the whole industry. Margin is recovering, and the price already pays for this recovery and part of the new themes.
1. 角色：不是 AI 晶片。是決定車用螢幕畫面和觸控的驅動晶片。收入跟著車廠和面板廠的出貨走，約 74% 來自中國系客戶，所以它押的是單一地區的客戶，不是整個 AI 擴建潮。
1. Role: not an AI chip. It is the chip that drives a car screen's picture and touch. Sales follow car and panel makers' shipments, and about 74% comes from China-headquartered customers, so it is a bet on one region's customers, not on the whole AI buildout.
2. Choke type: not a choke point. Car-grade qualification takes 2 to 5 years (the 06-30 version's claim, not re-checked). The "qualification choke" label also needs quarters-to-qualify and a repeat-order quarter; neither has a public source, so the label is not given. The company says it has "difficulty securing sufficient capacity": its foundries hold it up. (Unverified)
3. Rent: only part of it stays with Himax. In Q2, each $100 of sales left $33.1 after direct costs ($30.4 in Q1, $31.2 a year earlier), and about $10.8 of operating profit. Automotive share and gross margin rose together, so part of the extra money stayed with Himax, but far less than at the foundries upstream.
4. Already priced in: both. 73× (last twelve months) and 37× (this year's estimate) pay for profit recovering from last year's low, and also pay part of the way for optics going from small to large volume. Optics ships only small volumes in 2026, so that part has no revenue behind it yet.
Numbers — how the target price is built, and Q2 against the guide數字 — 目標價怎麼算，和第二季對照財測
The target price is $12.4, 16% below the $14.85 price. Book value is what is left for shareholders on the balance sheet after debts; P/B is how many times book value per share the price is.
This research expects end-2026 book value of $5.40 per ADS and a target of $12.4, based on end-2026 book value times 2.3 on P/B. Over ten years P/B ranged from 0.73 to 5.19, with a median of 1.73. It gets 2.3 because the base case's 2027 return on equity (ROE: what $100 of book value earns in a year) is about 11% against a ten-year median of 7.5%; scaled by that ratio the multiple would be 2.6, and we take about 88% of it because optics revenue has not shipped and book value still holds part of the 2021 peak. This multiple pays for profit growth, including about $50M of optics revenue in the base case; optics scaling into a real role change is only in the bull case.
FY2025A
FY2026E
FY2027E
Revenue營收
$832M
$922M
$1,046M
EPS / ADS每 ADS 盈餘
$0.25
$0.40
$0.62
P/E at the 10-06 close以 10-06 收盤計本益比
~59×
~37×
~24×
2026、2027 年是本研究推算（2027 為基準，含光通訊營收）；輸入見「更多數字」。
2026 and 2027 are this research's estimates (2027 is the base case, with optics revenue); inputs are under "More numbers".
Ten-year P/B (month-end closes, 2016-01 to 2026-09)十年 P/B（月收盤，2016-01 至 2026-09）
P/B
Price at $5.40 book對應股價（淨值 5.40）
Low (2019-08)低點（2019-08）
0.73×
$3.9
25th percentile25 百分位
1.32×
$7.1
Median中位數
1.73×
$9.3
Used for the target目標價採用
2.3×
$12.4
75th percentile75 百分位
2.81×
$15.2
High (2017-11)高點（2017-11）
5.19×
$28.0
Now (10-06 close)現在（10-06 收盤）
2.88×
$14.85
現價的 P/B 高過十年裡約 77% 的月份（做法見「更多數字」）。
The current P/B is above about 77% of the months in ten years (method under "More numbers").
Why historical P/E is not the anchor. Earnings per ADS went from -$0.08 in 2019 to $2.50 in 2021 and back to $0.25 in 2025, so year-end P/E jumped from 4.6× to 65×. A range that wide measures nothing; it is only a cross-check.
Cross-check: base-case 2027 earnings of $0.62 per ADS times 20× P/E gives $12.5. The 20× follows the median forward P/E of four peers (2026-10-07: Novatek, Raydium, LX Semicon, Synaptics), 19.1×; Himax grows faster, so it gets a little more. Detail is under "More numbers".
Scenario (odds are judgment)情境（機率為判斷）
2027 EPS2027 年盈餘
P/E本益比
Price股價
vs now對現價
Bear 30%: cycle down空方 30%：景氣下行
$0.31
14×
$4.4
-71%
Base 50%: modest growth plus small optics基準 50%：溫和成長，加小量光通訊
The three fair values agree: target $12.4 (P/B), base case $12.5 (P/E), probability-weighted $12.4. They are within 1%, and the target of $12.4 is canonical because book value has already happened while earnings are an estimate. All three rest on one assumption, that earnings recover, so they are not independent evidence.
Research view: avoid. The target is 16% below the price, past the -10% edge. (Avoid means an expected 6-month return below -10%; range means -10% to +15%; buy means above +15%.) The edge price is $13.8: the view becomes range only after five straight closes below $13.8, or one close below $13.35, so it does not flip on small moves. Buy starts below $10.8.
More numbers (tap to open)更多數字（點開）
Price as of股價時間 2026-10-06 close (IBKR daily bar)收盤（IBKR 日線）Financials財報數據 Q2 2026 (reported 08-0608-06 公布) + FY2025
Earnings per ADS and book value come from each year's 20-F (2016 to 2020) and stockanalysis annual data (2021 to 2025); book value is equity attributable to owners divided by year-end ADS. ROE is the year's earnings divided by year-end book value. Monthly P/B uses IBKR month-end closes divided by book value interpolated month by month between year-ends. Book value jumped from 2.78 to 4.99 in 2021 because the company earned $437M that year.
The optics (CPO) assumption rests only on company statements. On the 08-06 call the company said it will not give an optics revenue figure ("too uncertain to quantify"); its other 2027 comments exclude optics; optics revenue in 2026 is "still limited" and will see "very significant growth" in 2027, with a "meaningful contribution" "even in the most conservative scenario", especially to profit; asked what "meaningful" means, the CEO said it "is not going to be low single-digit percentage". Because no amount is given, the base case takes the smallest defensible figure: about 5% of revenue, roughly $50M, at the company-wide gross margin (34.5%) and with no added operating expense. This is this research's assumption, not company guidance. Sensitivity: each extra $10M of optics revenue adds about 1.6 cents to 2027 earnings per ADS, about $0.3 of price at 20×. The bull case uses $150M, three times the base, also this research's assumption; the bear case has no optics.
Inputs: 2026 revenue of $922M is Q1 199.0, Q2 227.4, the Q3 guide midpoint (up 9% on the quarter) of 247.9, and Q4 assumed equal to Q3. Revenue excluding optics in 2027 grows 8% in the base case (bear -3%, bull +11%), this research's assumption; the company says full-year 2026 automotive driver sales grow by double digits with momentum into 2027, and non-driver products rise to about 30% of revenue (the company states this excludes optics), while large-display drivers fall in Q3. Q3 gross margin is guided to about 34%. Q2 operating expenses were $50.7M; Q3 adds the $11.8M annual bonus; Q4 returns to $50.7M. The operating-to-net ratio of 0.80 is the actual 0.79 in Q1 and 0.81 in Q2. ADS count is fixed at 174.43M. 2026 earnings of about $0.40 are the two reported quarters (4.6 and 11.4 cents), the Q3 guide midpoint of 9 cents, and about 15.4 cents for Q4 (operating profit $33.6M × 0.80), which has no bonus like Q3. The one-time gain from selling an equity-method investee (about $23M to $24M, pending approval) is excluded.
These four are display-driver or edge-sensing peers with very different growth and profitability, so only the median is used as a reference. FocalTech (3545) is loss-making and has no forward P/E; Ilitek (3598) could not be pulled and is not used.
Value數值
06-30 page06-30 頁面
Market cap市值
$2.59B
$2.6B
P/E on last 12 months (EPS ~$0.20)過去 12 個月本益比（盈餘約 $0.20）
Full-year 2025 revenue was $832.2M and earnings per ADS about $0.25, exactly the 2025 dividend. First-half 2026 revenue of $426.4M is still 0.8% below first-half 2025's $429.9M. Q2 revenue mix: small and medium driver chips 71.4%, large driver chips 8.4%, non-driver products 20.2%. Q3 guide: revenue up 7% to 11% on the quarter (about $243M to $252M), gross margin about 34%, and earnings of 8 to 10 cents per ADS after the annual employee bonus.
To double from $14.85 the price must reach $29.70, almost the sell-side average target of $30.2. At the bull case's 28×, that needs earnings of $1.06, 22% above the bull case's $0.87. So doubling needs facts rosier than this research's bull case. The downside is cushioned by $298.7M of cash, the dividend and automotive resilience, but there is still room below the target.
Playbook操作框架
If not holding未持有
Do not buy yet. The price is 20% above the target; wait for either the price or the evidence to change.先不買。現價比目標價高 20%，要等價格或證據其中之一改變。
If holding已持有
Default to trimming in steps, not selling everything at once. The bull case still has a 20% chance worth $24.3, so keep part; after five straight closes below $13.8 the view returns to range and trimming stops.預設分批減碼，不必一次出清。多方情境仍有 20% 的機率、價值 $24.3，所以留一部分；股價連續 5 個交易日收在 $13.8 以下，觀點回到區間，就停止減碼。
Entry zone進場區
Below $10.8, and with Q3 gross margin at 34% and the first CPO shipments booked as revenue. Price alone or evidence alone is not enough.$10.8 以下，而且第三季毛利率達 34%、CPO 首批出貨已認列營收，兩個條件同時成立。只有價格或只有證據都不夠。
Why not bolder為何不更積極
No short: the bull case of $24.3 is 64% above the price, the stock moved between $6.85 and $25.09 in 12 months, and founders hold about 24%, so the risk of a squeeze is high.不建議放空：多方情境 $24.3 比現價高 64%，過去 12 個月股價在 $6.85 到 $25.09 之間跳動，創辦人持股約 24%，空方風險大。
What breaks the view失效條件
Gross margin at 36% or more for two quarters, with CPO revenue: 2027 earnings go up to $0.87 and the view becomes range.連續兩季毛利率達 36% 以上，並有 CPO 營收：2027 年盈餘上修到 0.87 美元，觀點改為區間。
Gross margin below 30% for two quarters, or the company says optics revenue slips beyond 2027: the base case falls to the bear case and the target is cut.毛利率連兩季低於 30%，或公司說光通訊營收延到 2027 年以後：基準情境降為空方，目標價下修。
The company or a customer confirms Himax as the sole supplier of a key CPO part with repeat orders: the choke type changes to qualification choke.公司或客戶公開確認 Himax 是 CPO 關鍵零件的唯一供應商且已重複下單：卡點類型改為認證卡點。
Catalysts — threshold, then reaction催化劑 — 門檻，然後反應
These thresholds were added on 2026-10-07; the old version had no pre-set thresholds.這些門檻是 2026-10-07 為這一頁補上的；舊版沒有事先設定的門檻。
When時間
Event事件
Threshold, then reaction門檻，然後反應
Status狀態
2026-11-05
Q3 results call (company notice of 10-05; evening in Taiwan)第三季財報電話會議（公司 10-05 公告，台灣時間當晚）
Revenue at or above $243M and gross margin at 34%: base case holds, target unchanged. Gross margin below 33%: base earnings cut, target recomputed. Gross margin at 35% or more with CPO revenue booked: raise the bull odds and recompute the target.營收達 $243M 以上且毛利率達 34%：基準情境成立，目標價不變。毛利率低於 33%：基準盈餘下修，目標價重算。毛利率達 35% 以上且 CPO 開始認列營收：多方情境機率上調，重算目標價。
scheduled排程中
2026 H2下半年
First small CPO shipments (Gen 1)CPO 第一代小量出貨
The call confirms shipments and revenue, and the 2027 wording stays "significantly higher than 2026": the base case's $50M holds and the bull odds go up. Only "validation" with no shipments: the base case's optics revenue goes to zero and the target is recomputed.財報電話會議確認已出貨並認列營收，且 2027 年說法維持「明顯高於 2026」：基準的 5,000 萬美元成立，多方機率上調。只說「驗證中」、沒有出貨：基準的光通訊營收歸零，目標價重算。
watching觀察
2026 fall秋季
WiseEye smart glasses launchWiseEye 智慧眼鏡上市
Non-driver products stay above 20% of revenue: direction is right. Below 18%: cut base 2027 revenue.非驅動產品占營收維持 20% 以上：方向對。低於 18%：2027 年基準營收下修。
watching觀察
Scored checkpoint (tap to open)已兌現的檢查點（點開）
When時間
Event事件
Result結果
2026-08-06
Q2 results (against the guide this page quoted on 06-30)第二季財報（對照 06-30 頁面引用的財測）
Hit: revenue of $227.4M above the top of the guide (about $225M), gross margin of 33.1% above about 32%, earnings of 11.4 cents above 8.6 to 10.3 cents. No threshold had been set then, so it is scored against the guide only.達標：營收 $227.4M 高於財測上緣約 $225M，毛利率 33.1% 高於約 32%，盈餘 11.4 美分高於 8.6 到 10.3 美分。當時沒有事先設定門檻，所以只對照財測。
Update log — material news since publish更新紀錄 — 發布後的重要消息
Dated additions, newest first. The core thesis above stays as-is unless an entry says otherwise.
逐則標日期、最新在上。除非條目另外註明，上方核心論點維持不變。
✓ 2026-10-07Target price moves from $10.8 to $12.4: the base case now includes about $50M of optics revenue, and the view stays avoid目標價由 10.8 美元調到 12.4 美元：基準情境加進約 5,000 萬美元光通訊營收，觀點仍是迴避
One line: the $10.8 target from earlier the same day gave optics no value in the base case; the company says optics will see "very significant growth" in 2027, so the base case now adds about $50M and the target rises to $12.4. The $14.85 price is still 20% above it, and the view stays avoid.
Earnings moved: base-case 2027 earnings per ADS rise from $0.54 to $0.62 (about $50M of optics revenue added, this research's assumption, based on the company's 08-06 call); the bull case rises from $0.73 to $0.87 ($150M of optics); the bear case is unchanged at $0.31 with no optics.
The multiple moved too, and that is the direct cause of this change: P/B goes from 2.0× to 2.3×. Book value of $5.40 did not move, so the whole gap between $10.8 and $12.4 comes from the multiple. It moves because base-case 2027 ROE rises from about 9.9% to about 11.3%; scaled by the ten-year median ROE of 7.5% the multiple would be 2.6×, and we take about 88% of it, the same ratio as last time's 2.0× against 2.3×.
Result: target $12.4 (before: $10.8), 16% below the price; probability-weighted $12.4 and the base-case P/E method $12.5 are within 1%. The three cases now sit on the first screen.
The optics assumption rests only on company statements: no amount given ("too uncertain to quantify"), the other 2027 comments exclude optics, a meaningful contribution "even in the most conservative scenario", and "meaningful" is not "low single-digit percentage". About $50M is roughly 5% of revenue, the smallest figure those statements support. Gross margin is the company-wide 34.5% with no added operating expense.
Sensitivity: each extra $10M of optics revenue adds about 1.6 cents to 2027 earnings per ADS, about $0.3 of price at 20×.
New edge prices: the view becomes range only after five straight closes below $13.8, or one close below $13.35; buy starts below $10.8. This keeps the view from flipping on small moves.
The playbook is rewritten to match: entry zone is below $10.8 plus evidence, the bull case is worth $24.3, and trimming stops at $13.8.
Verdict: a real automotive leader, but the price is 38% above the $10.8 target, and the research view is avoid.
One line: the 08-06 Q2 report beat the guide, yet the price closed at $14.85 on 10-06, almost the same as the $14.92 this page showed on 06-30. This is the first time the page gives a target price, $10.8, so the view moves from "no strong edge" to avoid.
If you hold it: default to trimming in steps, not all at once. New money: not yet; buy only below $9.4 and after the Q3 report (11-05) confirms margin and the first optics shipments.
This research expects end-2026 book value of $5.40 per ADS and a target of $10.8, based on end-2026 book value times 2.0 on P/B. Over ten years P/B ranged from 0.73 to 5.19, with a median of 1.73. It gets 2.0 because return on equity (ROE: what $100 of book value earns in a year) is recovering from about 5% to about 10%, against a ten-year median of 7.5%. This multiple pays for profit growth. The price pays for the optics role change; the target does not.
The three fair values agree: target $10.8 (P/B), base case $10.9 (P/E), probability-weighted $10.8. They are within 1%, and the target of $10.8 is canonical because book value has already happened while earnings are an estimate. All three rest on one assumption, that earnings recover, so they are not independent evidence.
Research view: avoid. The target is 27% below the price, past the -10% edge. (Avoid means an expected 6-month return below -10%; range means -10% to +15%; buy means above +15%.) Below $12.0 the view becomes range; below $9.4 it becomes buy.
Cross-check: base-case 2027 earnings of $0.54 per ADS times 20× P/E gives $10.9. The 20× follows peers' forward P/E (2026-10-07): Novatek 15.3×, Raydium 13.9×, LX Semicon 22.8×, Synaptics 23.1×, median 19.1×; Himax grows faster, so it gets a little more.
Default to trimming in steps, not selling everything at once. The bull case still has a 20% chance worth $20.5, so keep part; if the price falls below $12.0 the view returns to range and trimming stops.
One line: using ten years of price-to-book and peers' forward P/E, this gives a target of $10.8, and the $14.85 price is 38% above it. It is the page's first target price, so there is no earlier target to compare with; the view becomes avoid because the computed return of -27% is past the -10% edge.
Method: end-2026 book value of $5.40 per ADS times 2.0 on P/B gives $10.8. The ten-year P/B median is 1.73, with a range of 0.73 to 5.19 (IBKR month-end closes and 20-F book values). Cross-check: base 2027 earnings of $0.54 times 20× P/E gives $10.9; the probability-weighted value is $10.8.
This is the first target, so there is no earlier target and no factor that moved it. It rests on a profit-growth argument: ROE recovering from about 5% to about 10%, so the multiple is 2.0×. The optics role change is paid for by the price, not by the target.
The view, the playbook and the catalyst thresholds are rewritten around the new target. The 2026 earnings estimate moved from $0.34 to $0.40, because the annual bonus lands only in Q3 and Q4 has none.
The ten-year P/B uses 129 month-end closes from 2016. The low is 0.73× (2019), the high 5.19× (2017), and the current 2.88× is above about 77% of the months.
Historical P/E is not the anchor: earnings per ADS went from -$0.08 in 2019 to $2.50 in 2021 to $0.25 in 2025, and year-end P/E from 4.6× to 65×.
Peer forward P/E (stockanalysis, 10-06 and 10-07): Novatek 15.3×, Raydium 13.9×, LX Semicon 22.8×, Synaptics 23.1×. FocalTech is loss-making and the Ilitek page could not be pulled, so neither is used.
Band edges: below $12.0 the view becomes range; below $9.4 it becomes buy. The price is far from the edge, so no buffer note is needed.
This entry replaces the lines "target price not applicable" and "view unchanged" in the earlier entry of the same day; that entry is kept under earlier entries.
Verdict: a real automotive leader, but the price has already paid for most of the good news, so there is no obvious bargain.
One line: the 08-06 Q2 report beat the guide and the Q3 guide went up, yet the price closed at $14.85 on 10-06, almost the same as the $14.92 this page showed on 06-30. Every number is refreshed and the view is unchanged.
If you hold it: hold; no add because of price. New money: not yet; wait for the Q3 report (11-05) to confirm margin and the first optics shipments. This page has no target price and no entry price.
Why this page has no target price. Only 3 analysts cover it and their earnings estimates could not be found, so there is no reliable denominator, and without one a target price is a guess. The sell-side average target is $30.2; to hold it up at the base case's 40 to 55×, earnings per ADS would have to reach $0.55 to $0.76, above this research's base of $0.30 to $0.50 and 2.7 to 3.8 times the $0.20 of the last twelve months.
The price has already paid. About 73× the last twelve months of profit (the P/E: how many times earnings per share the price is), about 44× this year's expected profit. The sell-side (analysts at brokers who write research for clients) average target, $30.2 (3 analysts), is more than double the price and needs profit to double.
Do not buy yet. The price has paid for the recovery and optics has no revenue yet; re-assess after the Q3 report confirms the margin and the first optics shipments.
Hold; no add because of price. The Q2 report overturned no old argument, so it gives no reason to sell either.
Not applicable. This page has no target price, so it does not invent a level; entry depends on the evidence in the table below, not on price.
4. Already priced in: both. 73× (last twelve months) and 44× (this year's estimate) pay for profit recovering from last year's low, and also pay part of the way for optics going from small to large volume. Optics ships only small volumes in 2026, so that part has no revenue behind it yet.
The $14.85 price sits in the lower half of the base range. The odds and ranges were set on 06-30 and are unchanged; they are judgment, not measured figures.
FY2026E revenue is the sell-side consensus (3 analysts, stockanalysis 08-07). FY2026E earnings are this research's own arithmetic: two reported quarters (4.6 plus 11.4 cents), plus the Q3 guide midpoint of 9 cents, with Q4 assumed equal to Q3, about $0.34 in total. (This research's estimate)
✓ 2026-10-07Q2 beat the guide and the Q3 guide went up, but the price barely moved, so the view stays: a real leader at a fair price第二季比財測好、第三季財測也上調，股價卻幾乎沒動，所以觀點不變：真的龍頭，價格合理
One line: every number is refreshed to the 10-06 close, with the Q2 report and new news added. The view, odds and ranges did not change, and this page still has no target price.
Numbers refreshed to the 10-06 close: price $14.92 to $14.85, last-twelve-month P/E 79× to 73×, market cap about $2.6B to $2.59B. The sell-side average target rose from $23.70 to $30.2 (3 analysts).
New news: Q2 revenue $227.4M, gross margin 33.1%, earnings 11.4 cents, all above the guide (company release, 08-06). The Q3 guide is revenue up 7% to 11% and gross margin about 34%. Both CPO generations are in engineering production ramps, with 2027 shipments "significantly higher than 2026"; WiseEye smart glasses launch this fall.
The view is unchanged, so the target price is not applicable and no factor moved it. The page had no playbook, no catalyst thresholds and no update log before; all three are added now, and the layout was reorganised for phones.
Rest of this entry (tap to open)這一則的其餘內容（點開）
新增卡點四行：角色（不是 AI 晶片）、卡點類型（不是卡點）、利潤守不守得住（只留下一部分）、已經反映（獲利回升和一部分新題材都已付錢）。這一頁在 10-07 之前沒有這一區。
Added the four-line chain block: role (not an AI chip), choke type (not a choke point), rent (only part of it stays), already priced in (both the profit recovery and part of the new themes are paid for). The page had none of this before 10-07.
The old page had one paragraph of qualitative "roughly fair" conclusion. It is now split into the verdict, the chain block, the numbers and the playbook, with the full argument and the glossary folded.
The old page said "about 74% of revenue from China-headquartered customers", a figure from the 2025 annual report that was not re-pulled; the founders' roughly 24% stake was not re-pulled either, and stockanalysis shows 29.05% for insiders in total, on a different basis.
The TradingView pull saved only the call transcripts and the news list, so there is no sell-side earnings estimate; the sell-side target comes from stockanalysis on 10-07. The company also says the sale of an equity-method investee should close in Q4 for a one-time pre-tax gain of about $23M to $24M (pending regulatory approval). It is not core business and is not in any estimate here.
The close is the last IBKR daily bar (10-06). The $14.92 shown on 06-30 was the live price then; the daily close that day was $15.34, so the +82% year-to-date figure was based on the live price.
Himax is the clear leader in automotive display ICs (~40% DDIC, >50% TDDI share, company-stated) with credible optionality in AI sensing, optics and AR. But after +82% YTD to $14.92 (✓ IBKR · 06-30), it trades at ~79× trailing earnings on thin ~30% gross margins — the consensus growth story is largely in the price, and the average sell-side target (~$23.70) embeds multi-year EPS growth as the base case.
~80% of revenue is display driver ICs — cyclical, with inventory swings and ongoing ASP pressure. Q1'26 revenue was actually down 2% QoQ; the growth is a guided H2 reacceleration, not yet in the trailing numbers.
~79× trailing / ~41× EV-EBITDA on ~30% margins prices a clean recovery. Any growth or margin miss (foundry cost, demand) compresses the multiple hard. live · 06-30
What it takes to double from $14.92 (~$5.2B market cap): sustained well-above-base EPS growth held at a high multiple, or a re-rating on proven new-TAM scale (CPO/AR/WiseEye shipping in volume). That exceeds near-term guidance and current targets — it's closer to optimistic long-term pipeline realisation than to management's stated outlook. Downside is cushioned by ~$206M cash, the dividend, and auto resilience, but exposed to a cyclical de-rating. ◆
06-30 header line: TTM P/E = live price ÷ TTM EPS ~$0.19 per ADS. 52-week range $6.85 → $25.09 — ~+118% off the low, now ~41% below the high. Dividend $0.252/ADS for 2025 (~1.7% yield), paid 2026-07-10. ADS = American Depositary Share; 1 ADS = 2 ordinary shares.
Content-per-car is structural and Himax leads, with hundreds of design wins and only ~30% in mass production (prior commentary). H2'26 ramps (LTDI, OLED-auto, Tcon) are partly visible in guidance — double-digit QoQ growth guided for auto driver IC + Tcon.
The non-driver bucket (Tcon, WiseEye, optics) is higher-margin and growing off a small base; operating leverage + mix could push margins above the recent ~30%. Q1/Q2 guidance beats and auto resilience are the proven part.
Q1 2026 segment split ✓ Q1'26 release. FY2025 was ~80% driver / ~20% non-driver. Automotive runs through SMDDIC + Tcon and has been a major contributor (management cited ~half of revenue in recent quarters).
At launch on 06-30 the price was $14.92 (IBKR live, +82% year to date), about 79× last-twelve-month profit, with a sell-side average target of about $23.70. The conclusion was the rare "no strong edge".
Full argument (tap to open): how it works, business, bull, bear, variant view, competition完整論證（點開）：原理、本業、多方、空方、變異觀點、競爭
The argument in full論點全文
Himax 是車用顯示 IC 的明確龍頭（公司宣稱 DDIC 約 40%、TDDI 超過 50% 市占），在 AI 感測、光學與 AR 也有可信的選擇權價值。以 10-06 收盤 $14.85 計，本益比約 73 倍（過去 12 個月）、毛利率約 33%，市場的成長故事大致已反映在股價；賣方平均目標價（約 $30.2，3 位分析師）是把「未來幾年獲利成長」當成基準情境在算。
Himax is the clear leader in automotive display ICs (~40% DDIC, over 50% TDDI share, company-stated) with credible optionality in AI sensing, optics and AR. At the 10-06 close of $14.85 it trades at about 73× trailing earnings on ~33% gross margin, so the consensus growth story is largely in the price, and the sell-side average target (about $30.2, 3 analysts) takes multi-year earnings growth as the base case.
The honest variant view: there isn't a strong one. Unusually, this analysis converges with consensus rather than fighting it — the auto leadership and near-term guidance support are real, but the business is fundamentally cyclical, margins are thin, and ~74% of revenue rides on China-headquartered customers. The edge isn't in the idea; it's in whether the H2'26 automotive ramps and early new-market shipments beat an already-optimistic bar.
How it works原理
A display is millions of sub-pixels that each need an exact voltage to show the right brightness and colour. Himax makes the mixed-signal chips that turn a video signal into those precise per-pixel voltages — and increasingly, chips that let a device sense as well as show. Here's the signal chain; the glossary is folded at the bottom.
The display signal chain. Himax is the mixed-signal specialist panel makers rely on because precise high-voltage analog design and timing is hard — pull the chip and the panel shows dead pixels, failed touch or non-compliant automotive safety.顯示訊號鏈。面板廠依賴 Himax 這種混合訊號專家，因為高壓類比設計與精確時序很難 — 把晶片抽掉，面板就會出現壞點、觸控失靈，或車規安全不合格。
WiseEye AI 感測
Sense presence/gesture without waking the main chip.不喚醒主晶片就能偵測有人/手勢。
WLO / CPO 光互連
Light, not copper, between chips in AI servers.AI 伺服器內，晶片間用光、不用銅。
LCoS AR 光引擎
A tiny light engine for AR glasses.AR 眼鏡的微型光引擎。
The three "new-TAM" optionality bets, illustrated — each is real and funded, but small and early today.三個「新市場（TAM）」選擇權的示意 — 都是真的、有投入，但目前都還小、還早。
The business — where the money comes from本業 — 錢從哪裡來
Revenue is overwhelmingly one chip per panel/device, booked once on shipment (no subscription). About 70% is small/medium driver ICs (much of it automotive + smartphone/tablet); the higher-value, higher-margin growth is the "non-driver" bucket (Tcon, WiseEye, optics).
~74% of FY2025 revenue came from China-headquartered customers (BOE and affiliates, etc.) ✓ 20-F. Great when China panel demand and policy are friendly; a real single-region dependency if either turns — and Chinese IC suppliers (ESWIN, Chipone) are pushing into the same sockets.
Content-per-car is structural and Himax leads, with hundreds of design wins and only ~30% in mass production (prior commentary). Q2 automotive driver sales rose by double digits on the quarter, and management guides a "solid double-digit" rise again in Q3 and double-digit growth for the full year, with several large-display LTDI projects entering mass production.
The non-driver bucket (Tcon, WiseEye, optics) is higher-margin and growing off a small base; management expects it to approach 30% of revenue in 2027, from about 20% now. Gross margin already rose from 30.4% to 33.1% in one quarter, and the Q3 guide is about 34%. The Q1 and Q2 beats and auto resilience are the proven part.
WiseEye (edge-AI sensing), WLO/CPO (optical interconnect for AI servers) and LCoS (AR glasses) are real, funded efforts with named partners. None is material yet: CPO's first generation entered an engineering production ramp in Q3 with small shipments in 2H26 and 2027 guided "significantly higher", and a leading brand launches WiseEye smart glasses this fall. Each is still a call option on a large new market if it converts to volume.
WiseEye（邊緣 AI 感測）、WLO/CPO（AI 伺服器光互連）、LCoS（AR 眼鏡）都是有名有實、有合作夥伴的投入。目前都還不大：CPO 第一代第三季進入工程量產爬坡、2026 下半年小量出貨、2027 年公司預期「明顯更高」，某領導品牌今秋推出 WiseEye 智慧眼鏡。每一個仍是「若能放量、就打開一塊大新市場」的買權。
Owner-operators + dividend老闆經營＋配息
Founders (Wu brothers) own ~24% ✓ Bloomberg — high skin-in-the-game. Conservative capital allocation, annual dividend when profitable ($0.252/ADS for 2025), no empire-building.
~80% of revenue is display driver ICs — cyclical, with inventory swings and ongoing ASP pressure. Q2'26 revenue rebounded 14.2% on the quarter, but first-half 2026 revenue ($426.4M) is still 0.8% below first-half 2025 ($429.9M); the growth is a guided H2 re-acceleration, not yet in the trailing numbers.
~73× trailing / ~47× EV-EBITDA on ~33% margins prices a clean recovery. Any growth or margin miss (foundry cost, demand) compresses the multiple hard. (10-06)
~74% China-customer revenue is a demand and policy single-point-of-failure. Domestic Chinese IC suppliers, backed by policy and panel-maker relationships, are qualifying into core sockets — the most likely way share erodes.
約 74% 營收來自中國客戶，是需求與政策的單點風險。受政策與面板廠關係加持的中國本土 IC 廠，正切進核心料號 — 是市占被侵蝕最可能的途徑。
New TAMs are still early + all-or-nothing新市場仍早期、成敗全有或全無
WiseEye / CPO / AR carry the usual semiconductor execution and qualification risk and are small today. The valuation already grants them partial credit; slipped ramps or lost design wins de-rate the optionality.
A variant-perception lens looks for where consensus is wrong. Here it mostly isn't. Consensus reads Himax as an auto-display leader with new-product optionality, fairly but fully valued — and the evidence supports that. The two views converge; that is the finding.
Most theses on this page argue with the market. This one doesn't — and saying so plainly is more useful than manufacturing a contrarian angle that the evidence won't carry. The honest output is "fairly valued, no strong edge."這頁多數論點都在跟市場對作。這一檔沒有 — 而老實說出來，比硬擠一個證據撐不住的反向角度更有用。誠實的結論就是「合理估值、沒有明顯超額」。
Competitive position競爭地位
Himax's moat is real but share-specific — strongest in automotive, where 2–5-year design-in + safety qualification create high switching costs. Weaker in fast-moving consumer segments, where it meets scale players and rising Chinese suppliers.
The moat is real but share-specific — long where switching costs are high (autos), thin where they're low (consumer). Longer bar = harder to displace Himax.護城河是真的、但「特定市占」型 — 轉換成本高的地方（車用）厚，低的地方（消費）薄。長條越長＝越難取代 Himax。
The key share-erosion risk in China sockets中國料號市占被侵蝕的關鍵風險
Jargon decoder (tap to open)術語解碼器（點開）
Jargon decoder術語解碼器
The core products核心產品
DDIC顯示驅動 IC
Display Driver IC
Display Driver IC — turns digital video into the exact analog voltage each sub-pixel needs. Source drivers do the digital-to-analog per column; gate drivers scan rows.
顯示驅動 IC — 把數位影像轉成每個次畫素需要的精確類比電壓。源極驅動做每行的數位轉類比，閘極驅動逐列掃描。
Why it matters:為何重要： Himax's bread and butter — ~80% of revenue. Himax 的主食 — 約 80% 營收。
TDDI觸控顯示整合
Touch and Display Driver Integration
Touch and Display Driver Integration — folds capacitive touch sensing into the DDIC, so one chip both drives the screen and reads your finger. Fewer layers, thinner, cheaper.
Timing Controller — takes video from the host, processes it (scaling, colour, local-dimming zones for mini-LED), and feeds precisely-timed data to the DDICs. The "conductor" for big/high-res panels.
Why it matters:為何重要： Higher value; local-dimming Tcon is a Himax strength. 價值較高；區域調光 Tcon 是 Himax 強項。
Local dimming區域調光
mini-LED backlight zoning
Splitting a mini-LED backlight into zones the Tcon dims independently — deeper blacks, better HDR contrast, lower power. The Tcon decides each zone's brightness.
Why it matters:為何重要： Premium-panel content growth, esp. autos/TVs. 高階面板的含量成長，尤其車用/電視。
The "new TAM" optionality新市場（TAM）選擇權
WiseEye超低功耗 AI 影像感測
Himax's ultra-low-power always-on AI sensing
An always-on, ultra-low-power (sub-1mW) image sensor + a tiny edge-AI processor (Arm core + NPU) that runs small models on-device — detects presence/gesture/motion without waking a power-hungry SoC or sending video to the cloud.
常時開啟、超低功耗（<1mW）的影像感測器 ＋ 一顆小型邊緣 AI 處理器（Arm 核心＋NPU），在裝置端跑小模型 — 偵測有人/手勢/動作，不必喚醒耗電的主晶片，也不必把影像送上雲。
Why it matters:為何重要： Edge-AI sensing TAM (notebooks, IoT) — early. 邊緣 AI 感測新市場（筆電、IoT）— 早期。
WLO / CPO晶圓級光學／共同封裝光學
Wafer-Level Optics → Co-Packaged Optics
Wafer-Level Optics — making micro-lenses/diffractive elements with chip lithography. A path into Co-Packaged Optics (CPO): optical links that replace copper inside AI servers for more bandwidth at lower power.
晶圓級光學 — 用晶片微影製作微透鏡/繞射元件。是切入「共同封裝光學（CPO）」的路徑：用光連結取代 AI 伺服器內的銅線，換更高頻寬、更低功耗。
Why it matters:為何重要： AI-datacenter optionality (e.g. FOCI partnership). AI 資料中心題材（如與 FOCI 合作）。
LCoS矽基液晶微顯示
Liquid Crystal on Silicon
Liquid Crystal on Silicon — a reflective microdisplay (liquid crystal on a silicon backplane) that's tiny, bright and high-resolution. The light engine for AR glasses, pico-projectors and HUDs.
矽基液晶 — 反射式微顯示（液晶做在矽背板上），體積小、亮度高、解析度高。是 AR 眼鏡、微型投影、抬頭顯示（HUD）的光引擎。
Why it matters:為何重要： AR / smart-glasses optionality — execution-dependent. AR／智慧眼鏡題材 — 取決於執行。
Fabless / OSAT無晶圓廠／委外封測
Fabless design + Outsourced Semiconductor Assembly & Test
Himax designs the chip IP and outsources manufacturing: wafers from foundries (TSMC/UMC + some China), then bumping/assembly/test at OSATs (ASE, ChipMOS…). Low capital, but it captures less margin than the foundries upstream.
129 month-end closes 2016-01 to 2026-09 (IBKR monthly bars) ÷ book value interpolated between year-ends: low 0.73 (2019-08), p25 1.32, median 1.73, p75 2.81, high 5.19 (2017-11); now 2.88 = 14.85 ÷ 5.16 (stockanalysis Jun-26 book value/ADS), above 76.7% of months
Year-end P/E (completed years)年底本益比（完整年度）
✗→ computed
Dec close ÷ year EPS, years with EPS above $0.10: 4.6× (2022) to 65× (2017), median 20.9×; 2018 (EPS $0.05) and 2019 (loss) excluded
End-2026 book value per ADS2026 年底每 ADS 淨值
◆ judgment
$5.40 = Jun-26 $5.157 (stockanalysis common equity $899.51M ÷ 174.43M; the 2026 dividend is already deducted in Q2, as in 2025) + H2 EPS 24.4¢; no pending-gain, no SBC
Target price目標價
◆ judgment
$5.40 × 2.3× P/B = $12.42; -16.4% vs $14.85; band edges: $13.80 (-10%), $10.80 (+15%); buffer: five closes below $13.80 or one close below $13.35 (over 3 points past the edge). 2.3× = 88% of the ROE-scaled 2.6× (1.73 × 11.3% ÷ 7.5%); the 06-30-to-earlier-today 2.0× was 88% of 2.28×
base $50M (about 5% of revenue), bull $150M, bear $0. Company statements, Q2 2026 call (2026-08-06, saved at sources/himx/transcripts/FQ2-2026-earnings-call-transcript.txt): will not quantify (too uncertain); other 2027 comments exclude optics; 2026 revenue still limited; very significant growth in 2027; meaningful contribution even in the most conservative scenario, especially to the bottom line; meaningful is not low single-digit percentage. Sensitivity: $10M optics revenue = 1.6c EPS = about $0.3 at 20×. Third-party analyst figures not used
Peer forward P/E同業預估本益比
✓ stockanalysis
Novatek 3034 15.31× (2026-10-07), Raydium 3592 13.90× (10-06 data), LX Semicon 108320 22.84× (10-07), Synaptics 23.06× (10-06); median 19.07×; FocalTech 3545 loss-making; Ilitek 3598 page not found
P/E and weights本益比與機率加權
◆ judgment
bear 14× (Raydium low) · base 20× (peer median 19.1 plus growth) · bull 28× (about 20% over the top peer); 30/50/20 carried from 06-30; prices $4.37 / $12.46 / $24.30; weighted $12.40
Financials & guidance財報與財測
Q2 2026 results第二季業績
✓ release + call
rev $227.4M (+14.2% QoQ, +5.9% YoY) · GM 33.1% · op income $24.6M (10.8%) · net $19.9M · EPS 11.4¢/ADS · 2026-08-06
SMDDIC $162.3M (71.4%) · LDDIC $19.2M (8.4%) · non-driver $45.9M (20.2%) · automotive "remained the largest revenue contributor ... representing well over 50% of total revenues" (release text as published by QuiverQuant; the call transcript, saved under sources/himx/transcripts, does not repeat the percentage)
Q3 guide第三季財測
✓ call
rev +7–11% QoQ · GM ~34% · EPS 8–10¢/ADS after annual bonus; auto driver "solid double-digit"; full-year auto double-digit; non-driver ~30% in 2027 (Q2 call transcript 2026-08-06, saved at sources/himx/transcripts/FQ2-2026-earnings-call-transcript.txt)
Cash / inventory現金／庫存
✓ release
cash and other financial assets $298.7M · inventory $151.5M (2026-06-30)
$949.4M (+14.1%) · stockanalysis, as of 2026-08-07
Business & moat本業與護城河
China customer revenue中國客戶營收占比
⚠ 20-F, not re-pulled
~74% of FY2025 (carried from the 06-30 build)
Automotive share車用市占
⚠ co.
"40% share in DDIC, well over half in TDDI" · company-stated (Q1 release 2026-05-07)
CPO statusCPO 進度
⚠ co.
Gen 1 (1.6T/3.2T) and Gen 2 (6.4T) in engineering production ramps from Q3; small shipments 2H26; 2027 "significantly higher" · Q2 call 2026-08-06; FOCI partnership. Gen 2 anchor customer not named
WiseEye smart glassesWiseEye 智慧眼鏡
⚠ co.
a leading brand launches WiseEye-powered smart glasses this fall (Q2 call 2026-08-06); brand not named
Equity-method investee sale出售權益法投資
⚠ co., pending
pre-tax gain of about $23M-$24M expected on closing, expected Q4 2026, subject to customary conditions and regulatory approval (Q2 call); not in any estimate on this page
Capacity constraint晶圓產能受限
⚠ co.
"higher manufacturing and procurement costs, extended lead times, difficulty securing sufficient capacity" (Q2 call)
Capital & ownership資本與持股
Dividend配息
✓ company
$0.252/ADS for 2025, $44M, paid 2026-07-10 (Q1 release; Q2 call)
Founder ownership創辦人持股
⚠ not re-pulled
~24% (Bloomberg, carried from 06-30); stockanalysis shows 29.05% for all insiders (2026-10-07), a different basis
Thursday 2026-11-05, 8:00 am EST (GlobeNewswire notice published 2026-10-05, seen in the TradingView news list)
Judgment (not data)判斷（非資料）
Odds機率
◆ judgment
30/50/20 · set 2026-06-30, unchanged; the 06-30 EPS and multiple ranges were replaced on 2026-10-07 by the dated scenario table above
Chain block卡點四行
◆ judgment
role, choke type (not a choke point), rent, already priced in · 2026-10-07
Not obtained未取得
⚠ dropped
TradingView quote, consensus and financial history (the pull saved only the news list, the document list and two call transcripts); forward P/E shown by stockanalysis (17.27) not used, it does not reconcile with $0.40 EPS
