2400 — XD Inc.

Deep-value: XD Inc. (2400.HK) — September 14, 2026

Issuer: XD Inc. (心动公司), Cayman Islands, listed on HKEX on 12.12.2019, code 2400 Reference price: HKD … (14.09.2026; 52-week low 37.66; 52-week high 92.45) Market cap: HKD 17,986 mil. = CNY 15,384 mil. (HKD/CNY rate 0.8553) Reporting currency: CNY. Trading currency: HKD. They are not the same — every yield reported against market cap in this document is explicitly converted. Primary sources: HKEXnews — 2025 annual report (29.04.2026), FY2025 annual results announcement (27.03.2026), H1 2026 interim results announcement (28.08.2026), FY2024 annual results announcement (28.03.2025), H1 2025 interim announcement (29.08.2025), the two buyback programmes (05.01.2026 and 10.06.2026), the August 2026 monthly return (04.09.2026). Market data: yfinance. Market context: research brief of 14.09.2026.

Data pack correction. The pack generated initially today (data-pack-2400-20260914.md, 1,980 characters) was EMPTY: the script had been given the raw ticker 2400, which yfinance cannot resolve, and it silently wrote “(data unavailable)” to all four tables. I regenerated it with the correct Yahoo symbol (data-pack-2400.HK-20260914.md, 8,325 characters) and rebuilt every figure from the HKEX filings regardless, which are the primary source. This is exactly the known pattern for numeric HKEX tickers.

Brief corrections. (1) The brief left unreconciled the “RMB 15.35 亿 vs 16.57 亿” discrepancy in FY2025 profit. Reconciliation from the primary announcement: RMB 1,656.5 mil. = group profit, RMB 1,535.3 mil. = profit attributable to shareholders of the company; the RMB 121.2 mil. difference is the minority interest share. Both figures are correct, they are different things. (2) The brief cited the HKD 0.40 final dividend as a current fact. It is stale. The board decided on 27.03.2026 NOT to recommend a final dividend for FY2025 (FY2024: HKD 0.40) and on 28.08.2026 not to declare an interim dividend. The tracker still carries div_yield=0.01032 — and that yield no longer exists. (3) The brief made no mention at all of the two HKD 400 mil. buyback programmes from 2026, which are by far the most important capital-allocation event of the last two years.

Executive summary

XD Inc. does two things: it develops and publishes games (65.9% of FY2025 revenue) and it operates TapTap, mainland China’s gaming community and distribution platform (34.1%). The second business is the one that matters for valuation: TapTap produced FY2025 revenue of RMB 1,967.7 mil. with an 86.2% segment gross margin, i.e. 39.9% of group gross profit from one-third of revenue. Its model is unusual and is the source of the moat: TapTap takes no cut at all from developers’ sales (“zero revenue share”), monetizes exclusively through performance advertising, and over ten years has thereby built up the catalogue and community that the rest of China’s app stores charge 30… for.

The current numbers are good, but the direction no longer is. FY2025 was an exceptional year — revenue RMB 5,763.7 mil. (…), attributable profit RMB 1,535.3 mil. (…), gross margin 73.8%, operating cash flow RMB 1,712.6 mil., zero financial debt. But broken down by half-year, the picture splits in two: H1 2025 was … revenue and … profit; H2 2025 was … revenue (2,681.8 vs 2,791.5 mil.); H1 2026 recovered to … revenue but … attributable profit. TapTap, the margin engine, went from … in FY2025 to … in H1 2026. The market responded brutally: from HKD 91.80 in January 2026 to … today, … off the peak, right at the 52-week low.

Three things make the case interesting at this price. First: the balance sheet. At 30.06.2026 the company held RMB 3,897.4 mil. in cash and short-term investments, zero borrowings (the only liability is RMB 112.8 mil. in IFRS 16 lease obligations), for net cash of RMB 3,784.7 mil. — 24.6% of market cap, or HKD 9.14 per share. Second: cash conversion. TTM operating cash flow of RMB … mil. exceeds TTM group net profit of RMB 1,570.2 mil. (ratio 1.14); there is no accruals problem here. Third, and most important: on 05.01.2026 and 10.06.2026 the company launched two automatic HKD 400 mil. buyback programmes, executed HKD 493.2 mil. in H1 2026 for 7,882,600 shares that were CANCELLED, cancelled a further 5,427,600 in August, and has explicitly replaced the dividend with buybacks. The chairman’s letter in the 2025 annual report states outright: “Going forward, we plan to replace traditional cash dividends with ongoing share repurchases and cancellations — a more tax-efficient approach for all shareholders.” Issued shares fell from 494,889,562 (31.12.2025) to 483,880,364 (31.08.2026), … in eight months.

The counter-case is equally concrete. H1 2026 profit fell not because the business deteriorated, but because the effective tax rate jumped from 3.6% (FY2025) to 16.2% — pre-tax profit was nearly flat (…), post-tax …. The 3.6% rate was an artefact of carried-forward tax losses from 2022–2023; they’ve been exhausted. Any multiple calculated on FY2025 profit embeds a tax rate that no longer exists. In addition, FY2025’s record 73.8% gross margin is largely a recognition-mix artefact: game revenue recognized on a NET basis (excluding distribution costs) rose from 4.3% to 13.5% of total, then fell back to 5.0% in H1 2026 — and the margin dropped to 70.3%.

Estimated value. I triangulated five models (detail in the valuation chapter): bear FCFE DCF …, Greenwald EPV …, adjusted EV/EBITDA 9x …, base FCFE DCF …, bull DCF …. Monte Carlo with 20,000 scenarios over the same DCF gives a median intrinsic value of HKD … (median MOS …), with a …% probability of undervaluation and a P10–P90 range from … to …. Declared normalized owner earnings: RMB … mil. (9.10% yield on market cap, 12.07% on enterprise value). Minority-shareholder access to the net cash is modeled with λ = 0.35 — justified below.

Verdict: MONITOR, with small tentative sizing acceptable. The business is genuinely high quality (ROE 42.0%, zero debt, ~100% conversion, PwC auditor with a clean opinion, frugal management that cancels shares), but a median margin of safety of ~10% doesn’t pay for a hit-driven model, with a tax rate mid-normalization and the flagship asset (TapTap) in revenue contraction for the first time. This is not a value trap — it’s a good business at close to a fair price, where a quarter of the value sits in a cash pile that the minority shareholder only partially reaches. Convergence with the tracker’s GBL score of 58.3% (“MONITOR”).