Debt capacity remains beyond what the superficial Net Debt/EBITDA ratio suggests.

First and foremost, our borrowing covenants do not impose a uniform restriction which is “Net Debt/EBITDA = 3.0x.” We believe that temporarily exceeding the 3.0x threshold to execute high-quality M&A opportunities is well within an acceptable range, given the subsequent cash-generating capacity it would provide.

Furthermore, at the time of an M&A execution, while the “numerator (Net Debt)” increases, the EBITDA of the target company is added to the “denominator (EBITDA).” Consequently, the structure ensures that the Net Debt/EBITDA ratio does not increase as significantly as it might appear on the surface.

Let us provide a detailed explanation below.

• M&A simulation (until reaching a 3.0x leverage ratio)

The mechanical calculation for the case where Net Debt/EBITDA reaches 3.0x is as follows:

[The case where Net Debt/EBITDA reaches 3.0x]

※Assuming full debt financing and zero growth for the target company.

As shown, in the case of a reasonably priced M&A deal with an EV/EBITDA of 5.0x, even an additional investment of approximately 25.0 billion yen would still keep the Net Debt/EBITDA within the 3.0x range.

We have been in an aggressive investment phase and have maintained a negative free cash flow (FCF) until now; however, following a change in our capital allocation policy, we will now begin the full-scale reduction of interest-bearing debt (deleveraging) through operating cash flow.

If no new borrowing is undertaken, the “numerator (Net Debt)” will decrease through cash generation from existing businesses, meaning the Net Debt/EBITDA buffer will automatically expand over time.

Furthermore, stock-deal M&A is a means of pursuing M&A while preserving debt capacity. During the period of share price decline between December 2025 and March 2026, we acquired treasury stock at a low cost to be used as consideration for future M&A.

We continue to believe that allocating cash flow toward growth investments rather than shareholder returns is the best way to reward our shareholders, who have entrusted us with their capital based on our strategy of positioning M&A as a core pillar of growth.

Previously, we categorized dividends as a “pure return of profits” and did not consider implementing them from the perspective of capital efficiency.

However, through dialogue with the capital markets, we have come to recognize that the presence or absence of dividends is a factor that exerts a material impact – beyond theoretical equity value – on the investment eligibility for institutional investors and the decision-making of individual investors.

Therefore, we have determined that implementing dividend payments within a scope that does not impede growth investment holds strategic significance from the perspective of expanding our investor base. Accordingly, we have decided to commence dividend payments starting from the fiscal year ending on January 31, 2027, with a policy to maintain “disciplined dividend increases” every year in line with the improvement of our cash-flow generation capacity.

As we remain in a phase characterized by significant growth opportunities, the absolute level of dividends will be limited; our policy of prioritizing growth investment remains unchanged.

1. Amusement arcade business

Driven by the large scale of GiGO’s store network and sales volume, the deployment of “GiGO-exclusive prizes” has been increasing year after year.

This is because IP rights holders, attracted by GiGO’s expanding scale resulting from roll-up M&A and annual new store openings, have begun proposing “GiGO-exclusive prize” projects to us. As a result, we rolled out over 200 types of GiGO-exclusive prizes in the fiscal year ended on January 31, 2026, which has driven same-store sales growth. For the fiscal year ended on January 31, 2026, the annual same-store growth rate remained around 100 to 110% year-on-year throughout the period.

• Karaoke business

We have welcomed KAJI Corporation, the industry leader among karaoke equipment dealers, and ONTSU, the second-ranked player, into our group through M&A.

The establishment of ENNE Co., Ltd., born from the reorganization of these two companies, has created an overwhelmingly dominant industry-leading distribution network. As a result, our bargaining power has increased dramatically.

Furthermore, we have established a unique secondary distribution business that leverages our scale by taking in used karaoke equipment from Karaoke BanBan – the third-largest player in the karaoke box industry – and selling it to the nationwide night market and other segments through ENNE’s industry-leading wholesale network.

At Karaoke BanBan, the same-store growth rate has also maintained a stable level, averaging 101% annually.

• Foreign currency exchange machine business

By promoting DX, such as the utilization of AI, revenue has grown dramatically, achieving year-on-year increases in every month since joining the group.

In addition to being acquired at an appropriate valuation of 4.7x EV/EBITDA, the foreign currency exchange machine business is characterized by low maintenance CAPEX. By doubling earnings through operational efficiencies driven by DX, it has been transformed into a high-margin business delivering high ROIC.

As one of our carefully selected organic growth investments based on high IRR/ROIC, we envision aggressive deployment centered on major domestic store chains in the fiscal year ending on January 31, 2027.

We are working diligently on these improvements.

• The circumstances surrounding the discovery of the operational errors in the North American business

In North America, our business model involves acquiring stores en masse through M&A and either replacing or supplementing non-character plush prizes in claw machines with Japanese IP prizes that are popular in the region. Immediately following this transition, store sales surged significantly -approximately tripling – and the turnover rate for prizes improved.

However, upon investigating a group of stores where sales began to decline some time after the transition, it was discovered that these locations had been left for extended periods with insufficient prize replenishment.

The failure to keep up with prize replenishment was not due to a lack of management on the ground, but rather a “judgment error” during the PMI (Post-Merger Integration) process. In North America, we expanded our scale to approximately 12,000 locations by acquiring NEN in 2024, followed by Player One, BARBERIO, VENUplus and Newo Enterprises in 2025. During the integration of these companies, we standardized all operations based on those of Player One, as it was the largest entity and had the most established operational framework.

The issue was that Player One was a company primarily focused on “manned arcade locations (approximately 110 stores).” Under that same operational model, instructions were issued to NEN – which primarily consists of “unmanned mini-locations (approximately 10,000 stores)” – to “deposit collected cash at the bank every single day.” As a result, NEN’s approximately 200 field merchandisers were overwhelmed by the high-hurdle task of visiting four to five locations daily while also making bank runs, despite the fact that distances between stores in North America are unlike Japan and require one to two hours of driving. Consequently, the frequency of patrols for the primary purpose of prize replenishment decreased by 40%.

• Countermeasures

We noticed the decline in sales around November 2025, and after conducting an on-site investigation, we identified “operational chaos caused by daily bank deposits” as the root cause and decided to abolish this rule. As a result, the frequency of patrols has been on a recovery trend since February.

• The “Kiddleton Force” AI business app and the Timeline for operational normalization

To ensure that our field merchandisers can perform replenishment tasks efficiently, we have introduced our internally developed AI business application, “Kiddleton Force.” This app improves the operational efficiency of field merchandisers by utilizing AI to calculate optimal replenishment routes and perform image processing to monitor prize stock levels.

The application was developed based on systems and content that have already proven successful in our domestic operations.

One is the “AI for calculating cash replenishment routes” used in the foreign currency exchange machine business. In this business, we replaced the manual, human-led process of planning cash replenishment routes with an internally developed AI tool. Specifically, we have the AI perform calculations based on various variables, such as;

• which exchange machines should be replenished (taking into account factors such as daily usage frequency and currency levels)
  • which areas should be prioritized
  • what is the most efficient route to take

As a result, the number of replenishments per person increased by 20%. Due to this, revenue from the foreign currency exchange machine business has grown year-on-year in every month since joining GENDA, and a comparison of EBITDA for the 11 months before and after joining the group shows a 1.5-fold increase in profit.

The second is “GiGO NAVI,” an app used by staff at our domestic “GiGO” amusement arcades. This contributes to the standardization of tasks that previously relied on manuals, Excel, and human experience or intuition – such as checking per-machine revenue and prize inventory levels (which formerly required returning to the back office to check), as well as integrating operation manuals and digitizing inventory auditing.

Based on these tools, which have already delivered proven results within the company, we developed “Kiddleton Force,” the business application for our North American operations.

In the vast territory of the U.S., “Kiddleton Force” is more than just a route navigation tool; the AI calculates the optimal route by considering every variable – such as prize sales trends, cash accumulation, travel time between locations, and contract terms with installation sites – and displays “where and how to travel today” on the app.

The AI learns the differences in sales speeds for each individual prize and provides specific instructions on “which prize should be placed in which machine,” without relying on the intuition of on-site staff.

Furthermore, by uploading photos after prize replenishment to the app, the status of restocking at that specific location can be confirmed at a glance. Additionally, we have built a system where sales data from each game machine – which previously had to be reported manually to headquarters – is now automatically analyzed by AI and converted into a database simply by taking a photo and uploading it to the app.

Although the systems are now in place, it will take some time before every field merchandiser at every location can fully execute this cycle. For this reason, while we remain in a recovery phase during the first half of this fiscal year, we expect these improvement measures to take full effect starting from the second half, leading to a full-scale recovery in earnings.

• New store openings

Additionally, against the backdrop of our expanding network of locations in North America, bulk installation contracts with major nationwide chains are gaining momentum.

As previously reported, regarding Walmart, we plan to progressively withdraw from the small spaces currently housing only one or two mini claw machines. Instead, we intend to open new locations in tenant sections that are five to ten times larger in floor area.

In addition to AMC, one of the world’s largest movie theater chains, we are also making progress with store openings in other major chains. We have signed a contract with AMC to open locations in 170 sites, with installations scheduled to be completed by April. With Applebee’s, a restaurant chain operating over 1,500 locations across the U.S., we have initially signed an agreement to open in 85 stores.