Fujikura (5803) Analysis 2026: From AI Data Center Favorite to an Expectation-Confirmation Trade

Summary

Fujikura (5803) has been one of the most closely watched AI infrastructure stocks in the Japanese market in 2026.

The company is often classified as an electric wire or non-ferrous metals stock. But that is no longer the main reason the market cares. The current story is optical fiber, optical components, high-density cabling, and data center interconnects for AI data centers.

The fiscal year ended March 2026 was strong. Fujikura reported revenue of 1.1824 trillion yen, operating profit of 188.7 billion yen, ordinary profit of 199.5 billion yen, and profit attributable to owners of the parent of 157.2 billion yen. Operating profit rose 39.2% year on year, and net profit rose 72.5%.

The problem was not the past result. It was the next-year outlook.

For the fiscal year ending March 2027, management guided for revenue of 1.2430 trillion yen, operating profit of 211.0 billion yen, and net profit of 156.0 billion yen. Operating profit is expected to increase, but net profit is projected to decline slightly by 0.7%.

The market reacted harshly. After the May 14 earnings announcement, the stock plunged, including a limit-down move. It closed at 5,819 yen on May 15. By June 10, it had fallen further to 4,194 yen. Yahoo Finance historical data showed a PER of 44.51x and a PBR of 12.39x as of June 10. In morning trading on June 11, the stock was around 4,029 yen, still under pressure.

The essence of the move is not business deterioration.

The numbers are strong. The expectations embedded in the stock had been stronger.

Fujikura is not a dead theme stock. But the market is no longer simply buying the words AI data center. From here, the stock is in a confirmation phase: optical communications demand, margins in the information and communications business, raw-material constraints, North American hyperscaler capex, and the possibility of upward revisions at the next earnings release all matter more than the theme label itself.

This article updates the original May 16 piece using stock price, earnings, and external information available as of June 11, 2026. It is a framework for evaluating Fujikura, not a personalized buy or sell recommendation.

Bottom Line: A Strong Company, Not a Cheap Stock

If Fujikura is viewed simply as an electric wire stock, the valuation is difficult to explain.

Even at the June 10 close of 4,194 yen, the stock traded at a PER of 44.51x and a PBR of 12.39x. That is down meaningfully from roughly 61.8x PER and 17.2x PBR on May 15, but it is still not a normal valuation for a conventional manufacturer or non-ferrous metals company.

The market is valuing Fujikura as an AI infrastructure growth stock built around:

• Optical communications for AI data centers
• High-density optical fiber cables
• North American hyperscaler demand
• High-density cabling technologies such as SWR and WTC
• Rising profitability in the information and communications business

That is why the dividend yield is not the main point. Fujikura's company forecast for the fiscal year ending March 2027 is 38 yen per share after the stock split. Against the June 10 close of 4,194 yen, that implies a forecast dividend yield of about 0.91%.

38 yen / 4,194 yen = about 0.91%

The dividend is not the issue. The issue is that the stock is being driven by AI data center demand and growth expectations, not by income support.

Fiscal Year 2026 Earnings Review

Fujikura's fiscal year ended March 2026 was strong on the numbers.

The information and communications segment was the key driver. Demand for data centers, optical fiber cables, and optical-related products rose on the spread of generative AI, while profitability improved.

This part is genuinely strong.

Looking at revenue, profit, and then cash flow, Fujikura was not just a theme stock. Operating margin was around 16%, and operating cash flow was 132.9 billion yen.

Yet the market sold the stock anyway. The reason was not the historical earnings result. It was the gap between the next-year outlook and market expectations.

The Market Was Not Disappointed by Weak Earnings. It Was Disappointed by the Shape of Upside

The company forecast for the fiscal year ending March 2027 is as follows.

Operating profit is expected to rise, and ordinary profit is expected to reach another record high. Still, the stock sold off.

For a high-PER AI-related stock, simply delivering profit growth is not always enough. The market had already bought the possibility of much larger upside from AI data center demand. Before earnings, Fujikura had been treated as one of the representative AI infrastructure names in Japan, and expectations had become heavy.

Then the company presented a slight decline in net profit and flagged procurement risk for some materials, including hydrogen, as optical cable production ramps up rapidly.

The business is strong. But the forward picture was not as explosive as the stock price had implied.

That gap drove the PER compression from mid-May into June.

Stock Position as of June 11

On May 15, the short-term focus was whether the stock could hold the 5,700 yen zone after the earnings shock. By June 11, that framing had become stale.

Yahoo Finance historical data showed a June 10 close of 4,194 yen, volume of 54.54 million shares, PER of 44.51x, and PBR of 12.39x. At 11:30 a.m. on June 11, the stock was quoted around 4,029 yen, down 165 yen or 3.93%.

The 5,700 yen battle is no longer the relevant near-term reference point. After the earnings shock in May, the stock briefly rebounded to 5,163 yen on June 3, but was then pushed back into the low 4,000 yen range.

The current pattern is closer to this:

Earnings shock
↓
Short rebound
↓
Selling into the rebound
↓
Waiting for revaluation in the low 4,000 yen range

From here, the question is no longer whether 5,700 yen holds. It is whether selling pressure dries up in the low 4,000 yen range, and whether the stock can recover the 4,500 to 5,000 yen area.

Why Fujikura Was Revalued in the AI Trade

AI data centers do not run on GPUs alone.

Fujikura has been revalued because of the connectivity layer.

In generative AI, huge amounts of data move inside and outside GPU clusters. Compute power alone is not enough. If communication and interconnect capacity become bottlenecks, the efficiency of the entire data center suffers.

That makes optical fiber, optical components, and high-density cables a quiet but important bottleneck in AI data center buildouts.

This is why the market bought Fujikura. It is not a semiconductor company, but it supplies part of the physical infrastructure needed to run semiconductors at scale.

Why SWR and WTC Matter

SWR and WTC are often cited as Fujikura's differentiating technologies.

SWR stands for Spider Web Ribbon, and WTC stands for Wrapping Tube Cable. Fujikura offers ultra-high-fiber-count, small-diameter, high-density optical fiber cables using these technologies.

The company describes SWR/WTC optical fiber cable with 13,824 fibers in an outer diameter of less than 40 mm. For data center construction, space efficiency, shorter installation time, easier high-density deployment, and future expandability all matter. That makes this type of cable well aligned with AI data center demand.

This is a real strength.

But a real product strength and a justified stock price are not the same thing. The more the market values SWR/WTC, the more it will ask how much of that demand remains as profit.

Linkage to North American Hyperscaler Capex

Fujikura's AI infrastructure valuation is closely tied to capital spending by North American hyperscalers.

The demand drivers include Microsoft, Amazon, Google, Meta, Oracle, large cloud operators, and AI model companies.

That means Fujikura is effectively treated as a leveraged play on North American AI capex.

If AI data center investment in North America continues, optical communications and high-density cabling demand should remain firm. But if concerns emerge around AI investment ROI, data center project delays, power shortages, GPU procurement cycles, or capex discipline at cloud companies, Fujikura's valuation can come under pressure.

Fujikura is a Japanese company, but the real driver of the stock is global.

Valuation Has Compressed, But It Is Still High

On May 15, Fujikura was trading at roughly 61.8x PER and 17.2x PBR. By June 10, those had fallen to 44.51x PER and 12.39x PBR.

That is a large compression.

Even so, this is still not a conventional manufacturing valuation. The market is still pricing in AI data center demand beyond 2027, North American hyperscaler investment, optical cable supply constraints, pricing power, margin resilience, and the possibility of upward revisions.

For the stock to turn higher again, the story that AI data centers will grow is not enough.

The market likely needs evidence that:

• Margins in the information and communications business remain high
• Raw-material constraints are lighter than feared
• North American demand is not slowing
• The next earnings release leaves room for upward revisions
• Growth visibility is strong enough to support a still-high PER

When an expectations-led stock loses part of that expectation premium, good news can be met with a muted reaction. Fujikura is now in that confirmation phase.

Volume and Positioning: Large Investors Are Still Rotating Through the Name

One of Fujikura's defining features is the concentration of domestic and overseas capital in the stock.

The participant base likely includes domestic retail investors, overseas growth funds, AI theme funds, CTAs, short-term algorithms, short sellers, and earnings-event traders.

The price action looks less like a defensive TOPIX constituent and more like a NASDAQ-style AI momentum stock.

Volume remains heavy. May 15 saw more than 128.9 million shares traded. June 3 saw 82.34 million shares, June 9 saw 60.69 million shares, and June 10 saw 54.54 million shares. That does not look like the market has lost interest. It looks more like the shareholder base is still being reshuffled.

In the short term, the question is whether the stock can stabilize in the low 4,000 yen range on heavy volume. If it breaks below 4,000 yen with volume, the PER compression could have another leg.

Upside and Downside Reference Levels From Here

As of June 11, the old May framework centered on the 5,700 yen support zone is no longer useful. It is more realistic to focus on the low 4,000 yen range.

Upside Resistance

Downside Support

These are not trading instructions. In a high-momentum stock, news flow, market tone, NASDAQ moves, and AI-related positioning can move the price quickly. If these levels are used at all, they should be read together with volume and the thickness of intraday selling into rebounds.

Risks

The biggest risk for Fujikura is the loss of AI investment expectations.

Unlike a bank stock or high-dividend stock, Fujikura does not have a high dividend yield to anchor the downside.

When expectations fall, the stock can adjust quickly. The move from May into June is a clear example.

Scenarios for the Second Half of 2026

Main Scenario

The main scenario is that Fujikura absorbs the post-earnings expectation adjustment and is revalued again as a core AI infrastructure stock.

The conditions would be:

• Selling pressure calms in the low 4,000 yen range
• The stock recovers the 4,500 yen area and absorbs rebound selling
• Margins in the information and communications business remain high
• North American AI investment continues
• Raw-material constraints do not materially damage earnings
• The next earnings release revives expectations for upward revisions

If these conditions line up, Fujikura could again be seen as one of the leading candidates in Japan's AI infrastructure trade.

Risk Scenario

The risk scenario is that high-PER compression continues and the stock moves from a high-level momentum range into a mid-level box, or even a lower range.

That could happen if:

• The stock breaks clearly below 4,000 yen
• Rebound selling remains heavy and the 4,500 yen area cannot be recovered
• NASDAQ or AI-related stocks correct
• Concerns emerge about North American AI capex
• Raw-material constraints pressure margins more than expected

Even in that case, it would not necessarily mean AI data center demand is over. But the stock may be treated as a name to wait on until the next upward revision becomes visible.

Overall Assessment

Fujikura remains one of the representative AI infrastructure names in Japan.

The strengths are clear:

• AI data center demand
• High-density optical communications
• SWR/WTC technology
• North American growth exposure
• Rising profitability in the information and communications business
• A theme that foreign investors can understand

But the stock had priced in very high expectations. The selloff after the May 14 earnings release, the rebound selling in early June, and the June 10 close of 4,194 yen all show that the market is no longer satisfied with the theme alone. It wants visible upside to earnings.

Fujikura is not an income stock. It is a high-momentum AI infrastructure stock where expectation management is the key variable.

So the question from here is not simply whether AI is still strong.

The question is whether information and communications margins hold up, North American capex continues, raw-material constraints stay manageable, volume confirms a bottom, and the next earnings release can beat market expectations again.

Final View

As of June 11, 2026, Fujikura is not a dead AI theme stock.

But the earlier phase, where investors were willing to pay a high PER simply because Fujikura was seen as a prime AI data center stock, has paused. The market is now watching whether supply-demand stabilizes in the low 4,000 yen range, whether the stock can recover the 4,500 to 5,000 yen area, and whether the next earnings release can restore expectations for upward revisions.

The business is strong. The theme is still alive.

The problem is how much of that had already been priced in.

Fujikura is best viewed not as a finished AI stock, but as an AI infrastructure growth stock adjusting the gap between strong business growth and expectations that had become too high.

Sources

• Fujikura | FY2026 Financial Results
• Fujikura | SWR/WTC Optical Fiber Cable
• Fujikura | Optical Fiber Cable Products
• Fujikura | Data Transmission Research and Development
• Yahoo! Finance Japan | Fujikura (5803) Stock Price and Indicators
• Yahoo! Finance Japan | Fujikura (5803) Historical Data