AIDMA HOLDINGS INC

1. Company Overview

AIDMA HOLDINGS INC. is a Japanese small- and medium-sized enterprise services company. Its core business is outsourced sales execution and sales-process design, supported by its Sales Crowd software and database; it also operates remote-work and employment-support services. It is a single reportable “work innovation” segment, not a diversified holding company despite the group structure.

Data freshness matters here. The latest clean official annual base is FY2025, ended 31 August 2025 and filed 28 November 2025. More recent data is partial, unaudited, or estimated: the latest official update is the nine months to 31 May 2026, released on 10 July 2026, together with revised company guidance for FY2026. FY2026 had not ended as of the supplied market date of 7 August 2026. The FY2025 annual results are audited; the FY2026 nine-month results are unaudited quarterly data; FY2026 forecasts are company guidance.

FY2025 audited revenue was ¥13.270 billion, operating income ¥3.103 billion, and net income ¥1.955 billion. The company then reported FY2026 nine-month revenue of ¥10.853 billion, up 12.6% year on year, but operating income of ¥1.850 billion, down 19.8%, and net income of ¥996 million, down 27.3%.

At the supplied market-data date of 7 August 2026, the share price was ¥912 and equity market capitalization was ¥13.555 billion. The supplied FY2025 balance-sheet-derived net cash figure is ¥5.923 billion, equal to 43.7% of market capitalization. This is a substantial financial buffer, but it is not synonymous with distributable surplus: the group is acquisitive and operates with customer receivables and contract liabilities. The supplied FY2025 book value was ¥7.208 billion. No defensible liquidation value is available because necessary inputs for inventories and real estate were unavailable; “not available” must not be read as zero.

2. How the Company Makes Money

The economic engine has two principal service lines. In FY2025, sales support generated ¥8.796 billion of revenue, or about 66% of the total. AIDMA designs sales campaigns for business customers, builds target lists, conducts multi-channel outreach and test marketing, and supplies inside-sales support. It monetizes consulting, implementation, outsourced operating work, and its Sales Crowd sales-support tools. Human-resource support generated ¥3.924 billion, or about 30% of revenue, principally through the Mamaworks remote-work employment platform and related operating-process consulting. Other services generated ¥549 million.

The operating model is asset-light in the conventional physical-capital sense. FY2025 property, plant and equipment was only ¥158 million against ¥13.270 billion of revenue. Its real inputs are sales know-how, customer data, workflow design, software, brand, employee management, and a flexible network of remote workers. The company reported more than 15,500 cumulative assisted companies, a database covering approximately 2.1 million registered companies, more than 650,000 Mamaworks members, and approximately 3,500 contracted crowd workers as of August 2025. These figures are company disclosures, not independently audited measures of competitive advantage.

There is some useful operating leverage: FY2025 revenue grew 25.0% while operating income grew 25.7%, retaining an operating margin of 23.4%. However, the model is not capital-light in the broader owner-earnings sense if acquisitions are required to sustain growth. FY2025 included ¥535 million of cash outflow for acquired subsidiaries, and recorded goodwill reached ¥1.203 billion. Acquisitions can enlarge the offering, but also create integration risk and recurring amortization; they are not free growth.

The durable elements are accumulated campaign data, operating playbooks, existing customer relationships, the worker network, and the ability to combine software with managed service. The fragile elements are equally important: outsourced sales is a crowded category, the underlying corporate data are substantially replicable, customers can test alternative providers, and much of the value proposition is operational execution rather than an irreplaceable platform. This looks like a capable service franchise with evidence of historical high returns, not yet a proven high-return compounder protected by a wide moat.

3. Why the Stock Fell

The immediate, verified catalyst was the 10 July 2026 FY2026 guidance revision. Management cut its full-year revenue forecast from ¥17.0 billion to ¥14.0 billion, operating income from ¥4.0 billion to ¥2.2 billion, and net income from ¥2.3 billion to ¥1.2 billion. Those are reductions of 17.6%, 45.0%, and 47.8%, respectively, from the prior company forecast. The dividend forecast of ¥30 per share was maintained, and the company simultaneously authorized a buyback of up to ¥900 million or 450,000 shares. The buyback did not offset the signal from the earnings reset.

The market reaction was immediate. The stock closed at ¥1,255 on 10 July before the market had time to absorb the post-close announcement, then closed at ¥1,013 on 13 July, a one-day decline of 19.3%. The supplied 7 August price of ¥912 was 27.3% below the 10 July close. The observed sell-off therefore maps directly to a severe cut in expected earnings, rather than to a vague change in risk appetite.

The underlying business event was more serious than ordinary quarterly cost inflation. Management said that the mix of orders had shifted toward sole proprietors and very small owner-managed businesses; longer lead times before service started and weaker initial or ongoing collections reduced the conversion of booked orders into recognized revenue. It introduced a February rule requiring full prepayment for sole-proprietor orders and a May rule that only projects beginning service within 20 business days count as orders. These actions intentionally reduced order intake but were intended to improve cash collection and revenue quality.

The nine-month accounting evidence corroborates real pressure. The allowance for doubtful accounts rose from ¥327 million at FY2025 year-end to ¥987 million at 31 May 2026. Goodwill amortization also rose from ¥160 million to ¥312 million year on year for the nine-month period. The third quarter alone produced approximately ¥3.605 billion of revenue and ¥330 million of operating income, an operating margin of roughly 9.2%, versus about 24.4% in the comparable prior-year quarter. This is why the market treated the episode as a potential impairment of the profit model, not simply a timing issue.

4. What the Market Is Assuming

The market is no longer capitalizing FY2025’s ¥1.955 billion of audited net income as a stable run rate. After assigning full value to the supplied ¥5.923 billion net cash balance, the market value attributed to the operating business is approximately ¥7.633 billion. At a 10% required owner-earnings yield, that corresponds to only about ¥763 million of sustainable annual owner earnings. This is an analytical inference, not a reported market forecast, and it assumes all net cash is excess.

One-time or cyclical factors: the abrupt increase in the doubtful-debt allowance, delayed service starts, and amortization from recent acquisitions can depress reported FY2026 profit without permanently reducing customer demand. The FY2026 nine-month gross profit rose 13.9%, from ¥6.667 billion to ¥7.593 billion, while revenue rose 12.6%. The gross-profit engine therefore did not collapse. The loss of operating leverage instead occurred below gross profit, where selling and administrative expense rose 31.7%. This supports the view that part of the damage is repairable, but it does not prove that it is temporary.

Medium-term headwinds: the revised FY2026 guidance implies a weak final quarter. With nine-month operating income of ¥1.850 billion and full-year guidance of ¥2.2 billion, the implied fourth-quarter operating income is only about ¥350 million on implied revenue of ¥3.147 billion, or an 11.1% operating margin. The guidance therefore does not assume an immediate return to the FY2025 margin. It also explicitly warns that the stricter order rules will reduce backlog and affect the following year. The evidence gap is that AIDMA has not disclosed a sufficiently detailed cohort table showing order volume, collections, cancellation rates, service-start conversion, and contribution margin by customer type over multiple periods.

Long-term threats: the market appears to be allowing for the possibility that the historical growth model was partly fueled by weak underwriting of smaller customers, rather than by purely economic customer demand. It is also discounting the risk that generative AI and larger sales-tech vendors lower the value of basic lead generation, scripting, outreach, and reporting. The available financial record supports strong historic expansion—FY2021 to FY2025 revenue rose from ¥3.716 billion to ¥13.270 billion and operating income from ¥830 million to ¥3.103 billion—but it does not isolate organic growth, acquisition contribution, customer retention, or returns on acquired capital. That evidence gap matters more now than it did during the growth phase.

5. Temporary or Structural?

Credit underwriting and revenue conversion: structural but survivable. The damaged economic mechanism is not merely sales volume; it is the conversion of orders into cash-paying service revenue. If the company must reject or prepay-screen a meaningful part of its former customer base, addressable demand, backlog, and growth all fall. This is realistically repairable within three years only if the revised customer segmentation, prepayment terms, service packages, and collection processes produce comparable contribution margins from lower-risk customers. It is not repaired by reporting more orders under looser definitions.

Competitive commoditization and AI: structural but survivable. AI can automate portions of outbound calling, content generation, lead prioritization, meeting notes, and sales coaching—the same tasks embedded in parts of AIDMA’s offering. The economic mechanism at risk is pricing power and labor-arbitrage margin. AIDMA’s own AI products may defend costs and service quality, but that is not proof of a moat because competitors have access to many of the same foundation models. The risk is survivable if AIDMA remains superior at campaign data, customer-specific workflow integration, compliance, quality control, and conversion accountability. It is not truly structural if AI primarily lowers delivery cost industry-wide while AIDMA retains customers and shares in the productivity gain.

Acquisition and goodwill risk: structural but survivable. The mechanism is capital allocation. A service company that buys revenue at inadequate returns can report growth while reducing per-share owner earnings through cash consumption, amortization, and later impairment. The FY2026 increase in goodwill amortization makes this risk visible. It is reversible within three years only through disciplined acquisition pacing, transparent post-acquisition returns, and evidence that acquired businesses generate cash above their cost of capital. Until then, the cash balance deserves a haircut in downside valuation.

Founder and execution concentration: structural but survivable. The company identifies founder-president Yohei Miura as important to management and strategy. The operational model depends on recruiting, sales process design, and integration discipline, all areas in which key-person risk is material. A deep second layer of operating leadership would take years to demonstrate, not quarters.

Time-as-a-moat test: a well-funded competitor could build sales software and hire remote workers within two years. Within five years it could accumulate a comparable customer database, workflows, and case history. Over ten years, the only defensible barrier would be a sustained feedback loop: better campaign data creates better client conversion, which attracts more clients and workers, which creates better data. AIDMA has components of that loop, but the FY2026 collection failure shows that scale alone does not prove its quality. The moat is therefore narrow to moderate, not wide.

The overall diagnosis is mixed: a genuine short-term earnings and credit-quality problem, with structural questions about customer selection and differentiation. It is not evidence that the company’s core demand has disappeared; nine-month revenue still grew. But it is also not safe to classify as a simple temporary earnings dip until collections and cohort economics normalize under the tighter rules.

6. Is the Market Wrong? By How Much?

Moat-and-mispricing score: 5/10. The balance sheet, historical profitability, and current valuation create a potentially favorable asymmetry. Against that, the company has not yet proven that it can preserve revenue conversion and margins after ceasing to underwrite riskier customers. The current price is attractive only if normalized earnings settle materially above roughly ¥0.8 billion and the net cash remains genuinely available to shareholders.

The supplied FY2025 owner-earnings estimate is ¥1.615 billion, with a stated owner-earnings yield of 11.92% at the supplied market capitalization. This is a model estimate, not audited earnings, and it reflects FY2025 conditions before the collection problem became evident. It should not be carried forward unadjusted. For valuation, the bridge below uses my normalized distributable-earnings proxy rather than asserting a precise owner-earnings calculation: sustaining capex, acquisition needs, stock-based compensation expense, and future working-capital requirements are not separately disclosed sufficiently to support a cleaner owner-earnings reconstruction.

All figures are my intrinsic-value estimates in Japanese yen, not price targets. Operating value equals normalized earnings divided by the required equity yield; intrinsic equity value equals operating value plus credited net cash. Per-share values use 14.863 million shares outstanding excluding treasury shares from the latest quarterly disclosure. This does not fully incorporate all potential stock-option dilution: the FY2025 annual report disclosed options equivalent to 2.7% of issued shares, and the cash proceeds or exercise prices are not available here. Stock-based compensation expense is also not separately available, so it cannot be verified as immaterial.

The base case implies that the market is underpricing the equity by about ¥4.37 billion, or ¥294 per share. Put differently, if all supplied net cash is credited, the current operating-business value supports only about ¥0.76 billion of normalized annual earnings at a 10% yield—37% below the ¥1.2 billion FY2026 company forecast and 57% below the supplied FY2025 model owner-earnings estimate. That gap is the variant perception, but it is conditional. If the cash is needed for acquisitions, credit losses, or working capital, or if normalized earnings settle near ¥0.7 billion, the apparent discount largely disappears.

7. Key Facts, Estimates, and Judgments