NPC INCORPORATED

Company Overview

NPC Incorporated is a small Japanese industrial equipment maker whose economics are tied mainly to solar manufacturing equipment, especially for a dominant U.S. customer, with smaller businesses in solar panel recycling equipment, factory-automation machines, spare parts, and related services. This is not a broad industrial franchise. It is a niche, project-driven machinery company with meaningful engineering know-how, a strong balance sheet, and very uneven earnings.

Data freshness matters here. The latest clean official annual base is FY2025, ended 2025-08-31 and filed on 2025-11-28. More recent data is partial, delayed, unaudited, or estimated: the latest official company update is the FY2026 Q3 earnings disclosure dated 2026-07-10, which is unaudited and heavily affected by project timing. Current market data in your supplied authoritative snapshot is as of 2026-07-24: share price ¥581 and market capitalization ¥12.56 billion.

On hard official annual data, FY2025 revenue was ¥9.27 billion, operating income ¥1.92 billion, and net income ¥1.33 billion. On your supplied deterministic block, FY2025 owner earnings were estimated at ¥3.55 billion, but that is a model estimate, not an audited fact. The supplied authoritative snapshot marks net cash/debt as unavailable. The FY2025 audited filing does show cash of ¥6.42 billion and very light liabilities, but I am not restating a precise net-cash figure as a hard fact because your preferred deterministic snapshot explicitly leaves it as unavailable.

The core question is straightforward: is FY2026 a temporary drought caused by order timing and customer caution, or is it proof that NPC’s earning power is too dependent on one customer to deserve much premium over book value?

How the Company Makes Money

NPC makes money by designing and delivering custom production equipment. In FY2025, official product-category sales were dominated by solar-cell manufacturing equipment at ¥7.16 billion, followed by parts at ¥0.97 billion, FA equipment at ¥0.77 billion, solar-panel recycling equipment at ¥0.23 billion, and environment-related services at ¥0.14 billion. The company is effectively a single operating segment.

The economic engine has two layers. The first is large-ticket custom equipment, where revenue comes in lumps and often depends on customer acceptance timing. The second is follow-on parts, upgrades, and services tied to the installed base. The second layer is usually the better business: smaller ticket, higher margin, and less dependent on greenfield capex. The problem is scale. That steadier layer is still too small to stabilize the whole company when the large projects slow.

Capital intensity is lower than the word “manufacturer” suggests. This is not a giant plant-and-equipment business. Fixed assets were modest relative to the balance sheet, and FY2025 operating cash flow was strong. The real intensity is working capital, not maintenance capex. Inventory, work in progress, receivables, and advance payments can swing sharply as large projects move between design, shipment, and acceptance. That makes reported earnings and cash flow lumpy even when the franchise is intact.

There are some real but narrow sources of durability. NPC has customer-specific process knowledge, qualification history, an installed base that feeds parts demand, and proprietary recycling methods. Those matter. But they do not amount to a wide moat. This is not a high-return compounder with long reinvestment runway. It is a specialized engineering supplier that can look outstanding in good years and very ordinary in weak capex years.

Why the Stock Fell

The observable catalysts are clear and dated. On 2025-04-10, management cut FY2025 guidance. The stated reasons were not accounting noise. The company said its primary U.S. customer reduced upgrade projects, prioritized factory operations, and held back some orders because of uncertainty around the new U.S. administration’s policies. The revised FY2025 revenue forecast fell from ¥10.93 billion to ¥9.29 billion, and net income guidance fell from ¥1.55 billion to ¥1.20 billion. FY2025 actual results later came in slightly better than that cut guidance, but the damage to confidence was already done.

The second leg down was FY2026. The official company updates show a brutal year-on-year collapse in reported results: Q1 FY2026 revenue down 76.8%, H1 down 64.0%, and Q3 YTD down 53.9%. By Q3 YTD, operating income was down 94.1% and net income had slipped to a small loss. Those are not market rumors. They are company disclosures.

The dominant narrative is that FY2024 and FY2025 were peak years inflated by one customer’s capex cycle, and FY2026 is the hangover. My inference is that the market stopped paying for the “perovskite and recycling optionality” story and started valuing NPC as what it currently is: a small, lumpy, highly concentrated equipment supplier whose near-term results depend on whether a few large projects are recognized on time.

What the Market Is Assuming

The quantitative check across years supports a mixed conclusion. Revenue has been volatile rather than compounding. The supplied deterministic block shows an 18-year revenue CAGR of essentially 0% and net-income CAGR of only 2.76%. At the same time, the balance sheet is strong and backlog has not disappeared. So the market is not pricing bankruptcy. It is pricing unstable earning power.

Temporary or Structural?

My diagnosis is mostly TIME, with one important ESSENCE risk inside it. The current earnings collapse looks mainly temporary: delayed revenue recognition, a low-capex year at the dominant customer, and the usual violence of a project business. The balance sheet, backlog, and rising work in progress all argue against permanent impairment today. But the essence risk is customer concentration. That is not a short-term accounting issue. It is a structural weakness in the business model.

Time-as-a-moat test: in two years, a well-funded competitor would struggle to displace NPC quickly because customer-specific qualification, process familiarity, and installed-base relationships matter. In five years, that protection weakens materially; a larger automation company could reproduce much of the capability if the customer wanted a second source. In ten years, there is no convincing evidence of a deep moat. Recycling equipment may prove more defensible because of proprietary methods and glass-recycling ecosystem links, but it is still too small to redefine the whole company today.

Is the Market Wrong? By How Much?

Worth-dead and worth-alive point in different directions. On your deterministic worth-dead snapshot, NCAV is ¥7.57 billion, or 60.3% of the current market cap. That gives real asset support, but this is not a classic net-net. On your deterministic worth-alive snapshot, book value is ¥10.84 billion, price-to-book is 1.16x, and model owner earnings are ¥3.55 billion, implying a startling 28.3% owner-earnings yield. If that FY2025 owner-earnings figure were even close to durable, the stock would be obviously cheap.

I do not treat that owner-earnings figure as a decision-grade run rate. FY2026 official YTD results already falsify that. The right bridge is to normalized earnings, not peak-looking FY2025 owner earnings. I therefore use conservative normalized earnings estimates and capitalize them at demanding equity yields. I apply no net cash adjustment in the bridge because your authoritative deterministic block marks net cash/debt as unavailable. That is conservative relative to the audited FY2025 cash position.

Moat-and-mispricing score: 5/10. The balance sheet and backlog reduce downside. The moat is shallow. The market is probably too pessimistic on near-term optics, but not obviously wrong about the structural concentration discount.

Per-share values above use the implied share count in the supplied market-cap snapshot so that the equity value and current market price stay on the same basis. The latest official FY2026 Q3 treasury-adjusted share count is lower because of the buyback, which would modestly increase per-share intrinsic value if used instead.

So is the market wrong? Only modestly, if at all, in the base case. The stock is not pricing NPC as a durable compounder, and it should not. It is pricing a balance-sheet-backed but fragile project business. The possible error is that the market may be underestimating how much of FY2026 is timing rather than destruction. But until diversification is visible in actual sales, not slides, the concentration discount is deserved.

Key Facts, Estimates, and Judgments