South China Vocational Education

The business is growing and profitable again once a one-off tax bill is removed. The discount exists because no profit has ever left the schools for shareholders: every dividend was IPO money handed back. We stay on the watchlist until the board pays a cash dividend with its first August-year results, due by 30 November 2026.

All revenue comes from two Guangdong colleges that the listed company controls by contract through a sponsor owned by the chairman and his sister. The larger college is chartered as a non-profit. For a minority holder the business is easy to understand and hard to collect from.

The economic engine is simple. Students pay a year's tuition and boarding before term starts, the schools teach through the year, and revenue is recognised over the school year. In FY2025 tuition was RMB712.7m and boarding RMB64.7m of RMB791.8m revenue. Revenue = full-time students × tuition and boarding per student, plus a small amount of training income. Profit is what is left after teachers, cooperative-education partners, campus depreciation and, since 2025, a 25% income tax. ¹

The Institute is a diploma-level higher-vocational college founded in May 2002 with campuses in Guangzhou and Qingyuan, over 46 majors and 19,993 students. It ranked 16th among private higher-vocational colleges in ShanghaiRanking's 2026 list (13th in 2025). The Technician College, founded in July 2005 in Guangzhou, teaches secondary-vocational and technician programmes to 16,883 students. Both are inside the Greater Bay Area, which the company presents as its recruitment advantage. ^1,2,27

The share's claim runs through four steps: school surplus, a service fee to the wholly owned WFOE, dividends up to the Cayman parent, and dividends to holders. The contracts entitle the WFOE to the schools' surplus only after costs, taxes and reserves required by law. The Institute's charter, approved by Guangdong's education department in 2022, describes it as a non-profit school whose assets no organisation or individual may divide during its existence. The company itself says both schools have not yet registered under the 2021 for-profit or non-profit rules, and Guangdong sets no deadline. ^1,2,19

The family controls every layer. Chairman He Huishan and his spouse hold 50.75% of the shares, chief executive He Huifen 18.52%, and Fangyuan Education 5.70%; together the controlling shareholders hold 75.0%. The same two siblings own the sponsor. On 28 October 2026 the shares traded HKD264,396; the day before, HKD14,170. A 25% free float worth HKD43.4m means any re-rating needs a company action, not just better results. ^2,20

Volume growth is the strongest part of the case: 11,079 more students since 2021/22 and fees rising every year. The mix has swung between the two schools, and the first half of 2026 brought the first fall in fee cash received since the 2021 listing. The 2026/27 intake is the number that decides the revenue path.

The June 2026 count of 37,191 restates the same school year rather than adding a new intake, so growth since 2021/22 is 43%–44% depending on the count used. The more telling point is the switch between schools. Each school's growth has come in spurts: the Technician College when the Institute held back for its upgrade plan, then the Institute once its new Qingyuan capacity arrived. Revenue is therefore less exposed to one school than the total suggests, but each swing changes the fee mix. ^1,2,7

This is the contradiction in the numbers: H1 revenue grew 6.9% while fee cash received fell 22.2%. The benign reading is timing. Full-year fee cash still rose 6.1% in FY2025, and the rise in current other receivables in H1 2026 fits fees collected later, or waived secondary-vocational fees that government pays the school afterwards. The adverse reading is that fewer continuing students paid the next year early. The new August year-end settles it, because the 31 August 2026 balance will hold the 2026/27 prepayments. ^1,2

Without a new national policy, the local tax authority decided the schools' academic income had been taxable since 2020. The bill was RMB165.2m, more than the whole market value, and most of it is now paid. The schools now pay tax like businesses while the larger one remains chartered as a non-profit.

The reassessment came from practice, not law. The FY2025 annual report says no new tax rule requires private schools to pay corporate income tax, but the local authority began reassessing the region's vocational schools and required academic income for 2020–24 to be treated as taxable. The schools booked RMB103.8m of back tax, RMB58.6m of surcharge on it and RMB2.7m of surcharge on other taxes. They also provided RMB28.9m of tax on FY2025 academic income. ¹

The cash has largely gone out. By 30 March 2026 the schools had paid RMB74.9m of back tax and RMB6.3m of surcharge, and in H1 2026 income tax paid was RMB122.7m against RMB1.7m a year earlier. Tax payable fell from RMB146.6m to RMB43.2m even after RMB19.3m of new tax for the half. Other expenses also doubled to RMB25.1m in H1 2026, which the company attributes to domestic taxes for this and prior years and their surcharges. ^1,2

The upgrade ambition sits on top of this. The company aims to turn the Institute into a vocational university within a decade. The IPO plan to buy about 400,200 sq m of land for it has been dropped; the last HK$50.1m of IPO money went to existing facilities in H1 2026, and all HK$446.0m is now spent. An upgrade would need more buildings, staff and approvals, so it is a capex risk before it is a fee opportunity. ^1,2

On normalised profit the shares trade at 1.4 times, against a 2.4 times median for four listed private-college peers. The discount is not mainly about the multiple: South China pays a 25% tax rate where three of the four peers pay less than 5%, and its margins are about half of theirs.

The peer group is the four listed private-college operators closest in model: Kepei is the other Guangdong group, Xinhua runs vocational and undergraduate schools in Anhui, Gench runs an applied university in Shanghai, and JH runs universities and vocational schools in several provinces. Gingko Education (01851) is excluded because it trades at about 30 times rolling profit, far outside the group. [S22–S25]

The peer comparison changes two inputs. First, the margin gap is structural: South China's 28.8% gross margin compares with 46%–54% for peers, so we do not assume it closes; our base margin is 29%. Second, the tax gap is a risk for the sector, not just for this company, because the reassessment came from local practice in Guangdong rather than from a national rule. ^1,23

Revenue per student rose 11% in four years while cost of sales per student rose 32%. Gross margin fell from 40.4% to 28.8%, recovered one point in FY2025, and fell again in the first half of 2026. Our base case keeps it at 29%.

Staff are not the main cause. Remuneration rose 35.0% from FY2023 to FY2025 while employees rose 40.8%, and staff cost fell from 39.6% to 37.6% of revenue. The FY2025 annual report points instead to depreciation and amortisation, student practice and teaching spending, and new property-management costs as the campuses grew. In H1 2026 remuneration rose 10.4% against 6.9% revenue growth, so staff are now adding to the pressure. ^1,2

The second half is where the margin goes. Derived second-half gross margins were 21.7%, 16.1% and 24.5% in FY2023–25, against 34%–42% in the first halves. The change to an August year-end will show a cleaner full school year from FY2027. For valuation we use the last twelve months to June 2026: revenue RMB816.5m and gross margin 28.7%. ^1,2

Operating cash has run above profit every year because students pay ahead. But capex took 63%–99% of it, and the five-year cash left after reinvestment was RMB81m. Dividends of RMB186m were more than twice that, and the gap was filled with IPO money and bank debt.

Profit quality is not the problem. Operating cash flow was 1.3 to 3.1 times profit in FY2021–24, because tuition arrives before it is earned. The problem is where the cash went: capex was 99% of operating cash in FY2024 and 70% in FY2025. FY2025 cash after reinvestment of RMB35.8m was also flattered by tax: only RMB1.8m of income tax was paid that year while RMB132.8m was booked. ^1,4

The H1 2026 outflow needs to be read with the calendar. Operating cash was negative in H1 every year since 2024 because spring is when prepaid fees are used up, not collected. Before tax it was −RMB126.4m against −RMB34.0m a year earlier, so even without the tax bill the half was RMB92m weaker, mostly the RMB45.6m fall in fee cash and RMB38.5m more prepayments and receivables. ^2,8

Bank borrowings rose from RMB181.2m to RMB421.9m in 18 months to pay for dormitories and tax. Net bank debt is RMB141.7m, half a year of underlying EBITDA, with no pledged assets or guarantees. Survival is not the question; the rising debt is another claim ahead of holders.

Other claims rank ahead of holders too. Lease liabilities were RMB131.6m at June 2026, including RMB58.0m of new building leases in FY2025. Tax payable was RMB43.2m and net current liabilities RMB211.1m, normal for a school that holds prepaid fees, but a reason not to count group cash as surplus. The capex pipeline, by contrast, has emptied: contracted capex fell from RMB157.4m to RMB15.6m in six months as the RMB227.4m Qingyuan dormitory programme reached its June 2026 completion date. [S1, S2, S15–S17]

The company paid RMB185.8m of dividends in 2022–25, about 42% of its profit, and every yuan was charged to the share premium raised at the 2021 IPO. The parent has never received a dividend from the schools. RMB207m of that IPO-funded capacity is left, worth 1.4 times the market value, and the board stopped paying in 2026.

The parent's own balance sheet makes the point. At 31 December 2025 it held RMB0.9m of cash and RMB219.7m due from subsidiaries, and reported RMB207.0m of distributable reserves. In 2025 the subsidiaries repaid RMB39.8m of that balance, which matches the RMB40.3m of dividends paid. The money travels back up as repayment of IPO capital, not as profit, and the deferred-tax note says the onshore subsidiaries have no plan to distribute their RMB1,055.2m of earnings in the foreseeable future. ¹

Governance has failed twice in four years. In 2022 the schools' sponsor lent RMB234.5m to a third party, Guangzhou Haigemeina, and RMB10.0m to the chairman and his company Health Valley without board approval or the announcements the Listing Rules require; RMB221.0m was outstanding at the end of 2022, 89% of the group's cash. The board ratified the loans in March 2023 and all were repaid. For FY2024, EY qualified its audit opinion because there was no evidence that two labour-dispatch vendors had delivered RMB12.5m of student-recruitment services, 29% of that year's selling expenses. EY then left over an audit-fee disagreement, and the new auditor, AOGB, qualified FY2025 on the comparative figures. [S1, S12–S14, S28, S29]

There is no evidence of value moving to the controllers through operations: related-party rental income was RMB8.9m in FY2025, 1.1% of revenue, and balances with related parties were RMB8.3m at June 2026. The concern is control. The family owns the sponsor, approved the 2022 loans at subsidiary level, and holds 75% with the other controlling shareholder, enough to approve major transactions by written consent, as it did for the 2024 dormitory contracts. ^1,2,16

This is a License Ownership Gap: the profit is real, but it belongs to schools the listed company does not own, one of them chartered as a non-profit. The discount is deserved as long as no profit reaches holders. It becomes excessive the day the board resumes cash dividends from the remaining RMB207m of reserves.

What the price already assumes

At HKD0.130 the market value is RMB148.5m: 1.4 times normalised profit, 0.09 times book value and 1.56 times underlying EBITDA including debt and leases. Against our base case, the price pays for 24.9% of free cash being distributed, about RMB21.9m a year from FY2027, compared with RMB40.3m paid in 2025. The market is not doubting the schools; it is doubting the payout. ^1,2

Three fears explain the discount. First, cash cannot leave the schools: no service fee has been disclosed, the parent has made a loss every year, and the larger school's charter forbids dividing its assets. Second, tax and capex keep absorbing cash: RMB165.2m of back tax, a permanent 25% rate, and a five-year capex average of RMB211.9m. Third, governance and liquidity: two control failures, two qualified opinions and a free float worth HKD43.4m.

Cheap or value trap?

The deserved part of the discount is large. A holder of the share owns a claim on service fees that have never been disclosed, from schools that now pay full tax and, in one case, are chartered as non-profit. Valuing that claim at group profit, as a peer multiple does, would credit RMB107m a year that has never been paid. We therefore name the pattern the License Ownership Gap: revenue from licensed schools the listed company does not own, with value contingent on a legal or contractual transfer. The rule that comes with it holds here: the thesis is contingent on the transfer route working.

The excessive part is the capital route. The board does not need to change any school's status to pay holders: it paid RMB185.8m from share premium in four years and has RMB207.0m left, 1.4 times the market value. Paid at our base rate, that reserve alone is worth HKD0.100 a share, 77% of the price. What could permanently block value is a decision to spend the reserve on the university upgrade, new campuses or acquisitions, all of which management lists as strategy.

The skeptical-investor checks add four flags: audit risk (qualified opinions and an auditor change within two years), contract-control risk (100% of revenue in contract-controlled schools), shareholder alignment (a P/E below 5 with no dividend now), and high regulatory risk (the tax reassessment). With those flags the framework caps the classification at Watchlist.

Why is it cheap: temporary or permanent?

The cause that decides value is temporary and has a date: the board stopped dividends in 2026, and it can restart them from the parent's RMB207.0m of reserves with the 8-month results. The permanent part is legal, not cyclical: the larger school's charter makes its surplus non-distributable, which is why we value only cash paid out. The tax reassessment is regulation, temporary by default; the company has already paid most of it and can live with the 25% rate, so we keep that rate in every case and treat any reversal as upside we do not count.

We value only the cash we expect holders to receive. Base value is HKD0.155 against HKD0.130; the bear case is HKD0.009 and the bull case HKD0.489. The 20% upside to base does not compensate for a 93% downside until a dividend proves the route.

Risk-reward at HKD0.130: 0.21 to 1. The bear case loses 93.1% for a 19.6% gain to base, so neither entry gate passes. The probability-weighted return is 9.6% a year over three years, about half the 20% the house requires; the full panel is on the decision page.

Why this method

A multiple of group profit would value cash that holders have never received. We forecast free cash for the first three August years and apply a payout ratio; retained cash gets no value. We credit no opening excess cash, because group cash is matched by prepaid fees, tax payable and debt. Terminal value pays the FY2029 free cash out at the same ratio for ever with no growth.

The bear case does not need anything to go wrong in the schools. It needs only that the board keeps the money: another building programme or acquisition funded by the remaining reserves and debt, and no dividend. The first test is dated: the 8-month results due by 30 November 2026.

Three other risks are quantified in our cases. A second tax reassessment, for example on the other taxes behind the RMB2.7m surcharge, would hit cash before the schools gain any distribution right. A fall in 2026/27 students would hit revenue at a 29% gross margin, so each 1% of revenue is worth about RMB2.4m of gross profit. And the governance record means a new related-party loan would end the case, which is why it is a kill trigger rather than a haircut.

• Upgrade to Deep-Value Recovery if: a cash dividend is declared with the results for the 8 months to 31 August 2026 (due by 30 November 2026), the audit opinion on those accounts is clean, and capex for the six months to 28 February 2027 is RMB70m or less with no new campus or land commitment above RMB150m.
• Downgrade to Value Trap if: the 8-month accounts show a further under-provision of prior-year income tax; or a new land, campus or acquisition commitment above RMB150m is made with no dividend declared; or the audit opinion on the 8-month accounts is qualified on current-period figures.
• Kill trigger: any new loan or financial assistance to a related party or an undisclosed third party, a going-concern doubt, or a default on bank borrowings.
• Price and position-size discipline: with a free float worth about HKD43m and daily turnover often below HKD300,000, any position must be small enough to exit over weeks. At HKD0.130 the reward to risk is 0.2 to 1; we would need the bear case lifted by a paid dividend, or a price near HKD0.10 (the value of the reserves alone), before the balance becomes acceptable.