PUBLIC INFRASTRUCTURE
RESEARCH NOTE #005

Making Waste Treatment More Efficient:
Takuma

Published Oct. 1, 2026 · About 9 min
OUR INVESTMENT THESIS

As Japan faces both population decline and aging waste-treatment facilities, Takuma says it has established technology that combines AI combustion control and remote support to enable safe, stable operation with two operators—fewer than before. If that technology can be deployed through plant renewal and long-term operating contracts, reducing labor needs and costs, the essential social function of waste treatment could translate into recurring revenue and shareholder value.

Value to societyReduce the burden of plant operationMaintain safe waste treatment and energy recovery with fewer people.
Value to shareholdersTurn construction into long-term revenueExtend earnings from plant renewal into roughly 20 years of operations and maintenance.
Critical conditionTurn technology and investment into profitAI must improve long-term contract economics, while growth investment creates the next source of earnings.
Company & stock information
Takuma Co., Ltd. 6013:TYO
HOW TO READ THIS RESEARCH
VERIFIED FACTS

Renewal demand is showing up in earnings, while efficiency technology has reached practical use

1
Domestic environmental and energy operations are the core of revenueIn FY2026/3, the Environmental & Energy (Domestic) segment accounted for 76.6% of consolidated revenue. The company's FY2027/3 forecast implies 80.1%. Supplementary results materials
2
Japan's waste-treatment facilities are agingAccording to Takuma, more than 70% of roughly 1,000 operating facilities in Japan have been in service for over 20 years. Takuma has a domestic record of roughly 380 plant orders, the largest in Japan. These figures suggest continued short- to medium-term demand for renewal and life extension. Life-extension business
3
AI control has moved beyond demonstration into operational technologyTakuma says systems including its AI-based combustion control system, ICS, can reduce manual plant operations by 99%. Combined with remote monitoring and operation, the company says safe and stable operation can be maintained with two operators—fewer than before. The public materials reviewed here do not disclose how many facilities are actually operating with two people. Digital technology
4
Takuma owns its own remote-support base and data platformSolution Lab is a 24-hour center that monitors plants nationwide and supports operations. POCSYS is Takuma's proprietary operations and maintenance support system, centrally managing and analyzing data from more than ten facilities and feeding lessons back into operations. Takuma says these tools can improve operating quality and lower lifecycle costs, but it has not disclosed actual cost savings. Solution Lab and POCSYS
5
Takuma wins contracts covering design, construction and about 20 years of operationFor the Ichinoseki-area project, a corporate consortium won both design and construction and operations from August 2030 through July 2050. Total contract value is ¥34.39 billion. The announcement does not disclose Takuma's standalone operating revenue or margins. Ichinoseki order
6
Order backlog is at a record high, and the domestic environmental business is growing in sales and profitFY2026/3 orders reached ¥333.0 billion and backlog ¥745.1 billion, both record highs. Backlog was 4.50 times annual revenue of ¥165.6 billion. In FY2027/3 Q1, domestic Environmental & Energy revenue increased ¥10.342 billion year on year and operating profit increased ¥2.112 billion. The profit contribution from AI alone is not disclosed. Q1 results
7
Shareholder returns exceed planned growth investmentFor FY2024–FY2026, Takuma planned ¥21.0 billion of growth investment versus ¥36.0 billion of shareholder returns. Growth investment consists of ¥5.0 billion in equipment, ¥10.0 billion in M&A and business investment, and ¥6.0 billion in R&D. Shareholder returns consist of ¥18.0 billion in buybacks and ¥18.0 billion in dividends. Takuma executed ¥13.8 billion of growth investment in FY2024–FY2025 and completed three acquisitions by April 2026. For CO₂ separation and capture, performance evaluation is scheduled to begin in FY2027 with the goal of delivering the first commercial unit by FY2030. A larger return budget does not prove underinvestment, but progress toward the FY2031/3 ordinary-profit target of at least ¥20.0 billion depends on whether growth investments create future earnings. Supplementary resultsResults briefingCO₂ demonstration plan

What these seven facts tell us

Current growth can largely be explained by aging-facility renewal and a record order backlog, and it is already appearing in revenue and profit. AI-enabled labor saving, lower operating costs and long-term operating contracts are potential mechanisms for turning this demand into recurring income.

However, the number of facilities using AI, actual operating-cost savings and margins on long-term operations are not disclosed. What can be verified today is strong demand and growth in the existing business. The idea that AI will further lift margins and enterprise value remains a hypothesis.

OUR THESIS

Combine efficiency and long-term operations to align social need with corporate profit

1Municipal waste-treatment facilities age while the workforce available to operate them shrinks.
2Renewal and major upgrades incorporate AI combustion control and remote support.
3Fewer operators can maintain safe, stable treatment and local energy recovery.
4Construction orders lead into operating and maintenance contracts lasting roughly 20 years.
5Accumulated data supports continuous improvement and repeat orders, reducing earnings volatility.
6Profits are reinvested into AI, acquisitions, CO₂ capture and other future growth areas.

The core of the thesis is a cycle in which AI improves profitability in today's business and those profits are reinvested into future growth. Even if efficiency and long-term operations improve current earnings, the thesis is incomplete if growth investment fails to create the next source of profit.

RISKS & DISCONFIRMING EVIDENCE

Social necessity does not guarantee profit growth

WHERE THE THESIS STANDS

Demand and technology are visible; the incremental profit from AI and future growth is not

ConfirmedDemand is converting into revenue and profit

We can verify renewal demand, record backlog, revenue and profit growth in the domestic environmental business, AI labor-saving technology and long-term operating models. Capital allocation currently favors shareholder returns over growth investment.

Investment weaknessThe incremental effect of AI and future growth is unclear

The number of AI installations, margin improvement and effect on repeat orders are undisclosed. Earnings from acquisitions and CO₂ capture also remain to be proven.

Next checkMargins and the next capital-allocation plan

We will watch domestic environmental margins, recurring-revenue growth and the balance between growth investment and shareholder returns in the next plan.

WHAT WE WILL VERIFY NEXT

Watch whether backlog becomes profit and recurring revenue

Every quarterTrack order backlog, domestic Environmental & Energy revenue and margin, and projects where construction profitability deteriorates.
At each full-year resultReview the sales mix of operations, maintenance and after-sales services, the number of AI-enabled facilities, and whether lower manual operation has translated into staffing and profit improvements.
FY2027/3Check actual growth-investment spending and outcomes, earnings contribution from acquired companies, and the next capital-allocation policy after buybacks.
Through 2030Look beyond the ¥20.0 billion ordinary-profit target and test growth in recurring revenue, earnings from acquisitions and commercialization of CO₂ capture technology.
VALUATION AT PUBLICATION

Earnings growth is visible, but the shares are not deeply discounted

Share price¥3,005
Forward P/E13.98x
Trailing P/B1.94x
FY2026/3 ROE12.4%
Forecast dividend¥108
Dividend yield3.59%

Our view at publication: renewal demand, record backlog and the dividend make Takuma a credible long-term candidate. We do not, however, view the current price as deeply undervalued; we want evidence that AI and growth investment create the next layer of profit.

A forward P/E of 13.98x is not extreme relative to current earnings expectations. Aging-facility renewal and record backlog are already converting into revenue and profit, so the multiple looks reasonable if current earnings persist. However, annual profits can swing with the timing of large projects, and backlog alone does not guarantee future earnings.

A P/B of 1.94x means the market values Takuma at nearly twice book equity, so it is not cheap on asset value alone. Trailing ROE of 12.4%, however, shows that the company is generating meaningful profit from equity, providing a reason for valuation above book. To sustain that premium, the company needs to show healthy margins in the domestic environmental business, more recurring revenue from long-term operations and maintenance, and measurable economics from AI-enabled labor saving.

The forecast dividend is ¥108 per share, implying a 3.59% yield. Takuma targets the higher of 50% of profit or 4% of equity as its distribution benchmark, making shareholder-return policy clear. But over FY2024–FY2026, shareholder returns exceed planned growth investment. The return policy is attractive, while long-term growth after 2030 depends on whether acquisitions, R&D and CO₂ capture investment create new earnings.

Overall, current earnings and dividends are attractive, but a 1.94x P/B already reflects expectations for returns above book value. In upcoming results we will focus on domestic environmental margins and recurring revenue; in the next plan, we will focus on the results of growth investment and the balance with shareholder returns. Market data

Metrics are based on the Sep. 28, 2026 closing price. This is a subjective assessment based on public information and the thesis in this article.

Sources

Worth sharing?Send this research to someone
This article is a research note prepared from public information and a human–AI discussion. The article body is not rewritten after publication. It does not recommend any security. Investment decisions remain the reader's responsibility.