Hong Kong Stock Exchange (HKEX) · Healthcare
Gushengtang Holdings 2273
MonitorScore band: 40–50
The thesis, in one sentence
The largest private traditional-Chinese-medicine chain, with high returns on capital and a founder buying back stock, but both headline numbers are artifacts: most of the reported profit jump is a convertible-bond revaluation, and real shareholder cash flow does not cover the committed dividend.
Written for this site in plain English, without figures. The arithmetic is in the full report.
The thesis is from the deep report of 2026-09-03; the verdict badge reflects the latest scoring of 2026-09-06.
Key risks
- Committed dividend exceeds the cash flow the business generates
- Almost all revenue is consolidated through contractual arrangements, not ownership
- Book value is mostly never-impaired goodwill
What would change the verdict
- The multiple rises further on profit that is largely bond revaluation
- Leverage rises to fund the distribution gap
- Share count rises on full conversion of the bond
- The liquidity reserve thins out, forcing a dividend cut or a halt to acquisitions
Read it yourself
Step five of the method is the reader's: open the latest annual or quarterly report and read it before acting on anything here. A score is a summary, and a summary is not understanding.
This exchange has no stable per-company address, so the link opens the disclosure portal — search there for the ticker.
The deep report was written against the filings available on 2026-09-03; anything published since is not in it.
Price, one year
DCF valuation range
range of the DCF modelsbase casetoday's price
Valuation range, shape only. Figures in the full version.
Monte Carlo outcome shape
scenarios below today's pricescenarios above today's pricetoday's price
Distribution of simulated outcomes around today's price. Bar heights only; the scale and the percentiles are in the full version.
Where the money goes
Last four reported quarters, 2022-12 → 2025-12, in CNY. Filings data as gathered on 2026-09-03. Figures rounded to three significant digits.
Chapter one: Executive summary (1 page: thesis, estimated value, verdict)
Gushengtang Holdings (2273.HK) — deep-value analysis
Date: September 3, 2026 · Price: … · Market cap: … (RMB 4.947 bn) Shares outstanding: 211,791,651 (HKEX Monthly Return, 08/31/2026) · Reporting currency: RMB · Trading currency: HKD · FX rate used: … = 0.8568 CNY (yfinance, 09/03/2026)
Primary sources: the issuer's HKEX filings — Interim Results 08/28/2026, Annual Results 03/31/2026, Annual Report 2025 (04/29/2026), Interim Results 08/29/2025, Annual Results 03/31/2025, convertible bond announcements 01/26/02/12/2026, Dividend Plan 04/21/2026, Monthly Return 09/03/2026, acquisition announcements 03/03 / 06/23 / 07/05 / 08/27/2026. Supplemented with the internal data pack data-pack-2273.HK-20260903.md and yfinance. The company does NOT file with the SEC (HKEX-only listing) — there is no EDGAR file and the downloader was not run.
Executive summary (1 page: thesis, estimated value, verdict)
Gushengtang is China's largest private traditional Chinese medicine (TCM) chain: 93 medical institutions in Mainland China and 26 in Singapore as of 06/30/2026, 3.16 million patient visits in H1 2026, RMB 3.25 bn of revenue in 2025. The model is simple and, at the unit level, very good: it leases space, attracts renowned physicians (a mix of in-house and collaborating doctors), and sells consultations plus decoctions from its own herbal supply chain, at a stable ~30… gross margin. Return on invested tangible capital is exceptional — FY2025 ROIC ex-goodwill ≈ 74% (NOPAT 383.3 mn / invested capital 518.7 mn, my own calculation from the audited balance sheet) — a sign that a mature clinic pays for itself quickly.
The thesis, in three sentences. The market seems to be buying "net profit … in H1 2026" and an 8.4% dividend yield; both are, to a large extent, artifacts. Of the RMB 219.2 mn profit in H1 2026, RMB 54.6 mn is a non-cash accounting gain from the fair-value revaluation of the company's own convertible bond (it fell because the stock fell) — and if you strip out just the gains on financial instruments, the half-year's core pre-tax profit is … versus last year, not …. In parallel, the dividend plan announced on 04/21/2026 (a minimum of … mn/year for 2026–2028) exceeds free cash flow to the shareholder, which in FY2025 was RMB 344.1 mn = … mn; the gap is funded from the US… mn convertible bond issuance subscribed by Boyu Capital.
The FCF bridge, briefly. CFO − capex gives RMB 494.0 mn, a figure external screeners also publish (stockanalysis.com: FCF 509.9 mn, 2.10 CNY/share). But all the clinics are in leased premises, and under IFRS 16 rent disappears from CFO and reappears as "lease principal" in financing: RMB 123.7 mn in 2025. Adding interest paid of 26.2 mn, FCFE = 344.1 mn RMB, 43.5% below simple FCF. This is the base used in the DCF.
Estimated value. Five triangulated models give a wide range, … – … per share, with a median of … The Monte Carlo (20,000 scenarios, OE 344 mn, g1 7%, r …%, gt 2%) gives a median intrinsic value of … and a median MOS of ** …**, with an undervaluation probability of …%. Broker consensus is radically different: 17 analysts, average target … (min … max …), "strong buy" rating — the divergence is explained in the valuation chapter and is, essentially, the difference between valuing "reported EBITDA" and valuing the cash left after rent, interest and tax.
Verdict: MONITOR, no position at … It isn't a company to avoid — the balance sheet's accounting quality is honest, receivables are falling, inventory is only raw material, accruals are negative (a good sign), and management repurchases aggressively and contrarian. But the current price offers no margin of safety on the real flow, and the "8% yield …/year buybacks" thesis rests on a distribution that, in 2026, exceeds free cash flow generated. The entry point that would change the conclusion: below …–22, where the median MOS turns positive even on the base assumptions. The signal to watch: the cash flow in the full interim report (expected publication in September 2026) and the FY2026 final dividend in March 2027 — if it's funded from operations, not from the convertible bond balance, the thesis shifts in the company's favor.
Figures were removed from this excerpt. The full report opens free for five tickers a week — see what's open now.
Full report contents
- Executive summary (1 page: thesis, estimated value, verdict)
- 🔒 The business and its moat (how it makes money, competitive advantage, durability) (Available in the full report)
- 🔒 Management and capital allocation (track record, buybacks/dividends/acquisitions, skin in the game) (Available in the full report)
- 🔒 What changed in the last 4 quarters (balance sheet position by position from the data pack, margins, cash conversion — explains EVERY large swing) (Available in the full report)
- 🔒 Balance sheet analysis — Quality of Earnings (Thornton O'Glove method) (Available in the full report)
- 🔒 The CEO profile — Outsider traits (William Thorndike method) (Available in the full report)
- 🔒 Accounting red flags (accruals, dilution, one-offs, accounting policy changes) (Available in the full report)
- 🔒 Triangulated valuation (conservative DCF with explicit assumptions + earnings power value + 5-year historical multiples + Monte Carlo from step 5; a range, not a point) (Available in the full report)
- 🔒 Pre-mortem (why the thesis might be wrong — 3 concrete scenarios) (Available in the full report)
- 🔒 Verdict compared with the tracker's GBL score (convergence/divergence and why) (Available in the full report)
Evaluation history
| Date | Verdict |
|---|---|
| 2026-08-23 | Monitor |
| 2026-09-03 | Monitor |
| 2026-09-06 | Monitor |
Want the rest of this report?
Subscribe to get one full deep report a week by email, the day before it opens on the site.