Telecom added ¥53B of operating profit in FY2025; the other four segments combined added ¥13B.

• The engine. AFL's optical fiber and connectivity segment grew revenue 52% to ¥451B and posted a 20.4% operating margin, up from 13.2%, as hyperscalers raced to qualify ultra-high-density cable for AI campuses. Telecom is now 46% of revenue but 68% of group operating profit.
• The conglomerate around it. Auto wire harnesses (3.3% margin), HDD and electronics components (12%), Japanese power cable (8%), and legacy Tokyo real estate are commodity-priced businesses — none earning anything close to Telecom's incremental margin.
• The question. Whether 20.4% is a forward-loaded floor — earned on the old plant footprint before Sakura Works (Feb 2025), the ¥300B US capacity tripling, and the SDM4 four-core MSA contribute — or a cycle peak that reverts as hyperscalers digest 2024–26 capex.

Margins doubled, the dividend doubled, and the multiple now embeds the next two years.

Operating profit nearly doubled in one year (¥69B → ¥136B) on +22% revenue, and net income jumped from ¥51B to ¥91B — gross margin moved 21% to 27%, SG&A held, and AFL fiber did the rest. Balance sheet is a fortress (net cash ¥77B, 49% equity ratio, ¥87B FCF), capex still only 3.1% of sales. P/B is 27× on a group whose ROE was 14% pre-AI; the price is paying for FY2027 earnings power, not what the FY2025 income statement booked.

May 14 FY2026 results are the binary trigger both bull and bear name.

• The number to watch. Telecom segment operating margin. Bull thesis (¥9,800 PT) needs ≥17% with a credible new mid-term plan and an explicit ROIC hurdle attached to the ¥300B (~$1.9B) US capacity tripling. Bear thesis (¥3,800 PT) needs <17% or single-digit Telecom revenue growth.
• The setup is constructive but stretched. 9-month FY2026 net profit is already ¥111.9B — 75% of the full-year ¥150B guide; dividend doubled to ¥215; a 5-for-1 split landed April 1; the ¥300B capex was approved in March; Reuters and Morgan Stanley MUFG validated the thesis between October and December 2025.
• The tape says size carefully. 30-day realised volatility 86% (stressed regime), 50-day average volume halved as price tripled, and price sits 104% above the 200-day. A close above ¥6,800 is the watch level for an add; a close below the 50-day at ¥4,860 breaks the uptrend.

Aggregate sell-side targets at ¥5,485 — 17% below spot — look like coverage lag, not analytical conviction.

• Consensus is mechanically stale. Most published targets sit on partial-year TTM income that pre-dates the FY2026 ¥150B guide raise, the doubled dividend, the 5-for-1 split, and the ¥300B capex. Forward P/E on management's own guide is in the high-40s, not 73×. A clean May 14 print plus a normal 30–60-day revision window is the mechanical path that would close most of the gap.
• The bear's FY2020 analogue is misapplied. The 0.5% margin and ¥38B loss were a five-driver shock — carrier capex pause, COVID, restructuring, impairments, and a 30% equity ratio. None of those drivers replicate today: hyperscaler customers (not carriers), 49% equity ratio (FY2025-end; 56.5% at 9M FY2026), net cash, 51% Americas mix. The right downside floor is Telecom margin 12–15% and group NI of ¥110–130B — not a reprint of FY2020.
• The 1.19B → 7B authorized-share lift is being priced as dilution risk. Japanese practice tilts toward maintaining pre-emptive headroom that mostly stays unused. With ¥185B cash, ¥87B FCF, and no announced M&A target, conditional issuance probability is plausibly under 10%. The June 2026 AGM language likely settles it.