Clifford Modern Living Holdings Limited (3686.HK) - Deep Dive Research Report

Prepared 19 July 2026. All figures in Renminbi (RMB) unless a dividend is quoted, in which case Hong Kong cents (HK) as declared. Fiscal year ends 31 December; the company reports on a half-yearly cadence.

A note on reporting cadence and "concalls" before we begin

Clifford Modern Living is a small-capitalisation Hong Kong Main Board issuer in the consumer/property-services space. Companies of this size and type in Hong Kong do not hold quarterly earnings calls, and no transcript aggregator (or the company's own IR site) carries any recorded management call for this name. What the company does produce, on the standard HKEX half-yearly rhythm, is a results announcement plus a management discussion and analysis (MD&A) section twice a year. This report therefore substitutes the last six reporting periods for the "six concalls," and Sections 7 and 9 are built from the MD&A and results-announcement outlook statements attached to each. The six periods used throughout are:

1. FY2025 (year ended 31 Dec 2025) - results announced 27 March 2026; annual report late April 2026
2. H1 2025 (six months ended 30 June 2025) - announced 28 August 2025
3. FY2024 (year ended 31 Dec 2024) - announced 28 March 2025
4. H1 2024 (six months ended 30 June 2024) - announced August 2024
5. FY2023 (year ended 31 Dec 2023) - announced March 2024
6. H1 2023 (six months ended 30 June 2023) - announced August 2023

Recency check: with a 31 December year-end and a half-yearly cadence, the next release is H1 2026 (six months ended 30 June 2026), expected by the HKEX deadline of end-August 2026. As of today (19 July 2026) that period is not yet due, so FY2025 is correctly the most recent released period. It is genuinely the latest and is fully incorporated.

Section 1: What the company does

Clifford Modern Living is, at its core, the captive services company for a single, enormous residential estate - and, since 2020, an increasingly large precious-metals treasury vehicle bolted onto that services business. Both halves matter, and the second half now drives the reported profit.

Start with the estate. Clifford Estates (祈福新邨) in Panyu, Guangzhou, is the largest gated community in China: roughly 500 hectares, about 60,000 homes, and a resident population north of 200,000 drawn from dozens of countries. It has been under continuous development since 1991 by the Clifford Group (祈福集团), chaired by Peng Lengji (彭磷基). Peng's wife, Man Lai Hung (孟丽红, "Wendy Man"), is Chairman and CEO of the listed Clifford Modern Living and, together with her holding vehicle Elland Holdings Limited, controls roughly 72.89% of it. So the listed company is not an arm's-length third-party manager competing for mandates; it is the household-services operating arm of the family that built and still runs the estate.

What does that arm actually do? It runs the daily life of Clifford Estates and the surrounding Clifford developments. It manages the properties (security, cleaning, common-area upkeep, resident support). It runs the shops residents buy from - supermarkets, wet markets, convenience stores inside the estate. It teaches residents' children after school (language and interest classes). It wires up the buildings (security systems, IT and hardware/software integration, telecoms). It renovates and fits out flats. It brokers property sales and rentals within the community, places domestic help, does the laundry, and consults on the estate's catering outlets. If you live inside Clifford Estates, a meaningful share of the money you spend on services in a given month flows to this company.

The value proposition is captive-market density. A conventional property manager wins buildings one tender at a time and then tries to cross-sell. Clifford starts with a self-contained city of 200,000 people that its controlling family developed, and layers service after service onto that fixed, loyal base. The "moat," such as it is, is incumbency and physical presence inside a walled community - not technology.

The second half of the story is unusual for a property-services firm and is now the swing factor in results. Since 2020, management has deployed the group's substantial surplus cash not into bank deposits but into precious metals - initially unallocated silver bullion - as what it calls a treasury-management and capital-preservation strategy. When silver ran up hard through 2025, the mark-to-market gains on that bullion, not the shops or the property management, produced a record profit. FY2025 net profit of about RMB284.6 million was up roughly 192% year-on-year, even as revenue fell about 12.8% to RMB320.3 million - the net margin optically jumping to ~89% because the bulk of the profit was a fair-value gain sitting outside operating revenue. In January 2026 the company sold 680,000 ounces of silver for about RMB387.9 million and rotated part of the proceeds into gold and allocated silver.

The one line that captures the company today: a modestly shrinking, cash-generative captive services operator for China's largest gated community, sitting on top of a family-controlled balance sheet that has chosen to store its surplus in gold and silver rather than in the bank.

Section 2: Business segments

The group reports eight segments, though four carry almost all the operating revenue. Below, the four material segments get full treatment; the four smaller ones are grouped. FY2025 segment revenue is shown to convey scale (total revenue RMB320.3m).

Retail Services (~RMB125.0m, the largest revenue line)

What it does. This is the shopkeeping arm inside Clifford Estates: supermarkets, traditional wet markets, and convenience stores selling groceries, fresh food, and daily necessities to the resident population. It is a bricks-and-mortar grocery and daily-goods operation with a fixed, walk-in catchment.

Core capability. The capability is not merchandising genius; it is location control inside a closed community. The company operates the retail floorspace physically embedded in an estate its parent developed, with limited competing formats inside the gates. Fresh-food (wet market) operations in particular require local supply relationships and daily logistics that an outside entrant cannot easily replicate for a single estate.

Why it exists separately. Retail has completely different economics from property management - it is a thin-margin, high-turnover goods business exposed to food-price inflation and resident spending, not a recurring service fee. FY2025 retail revenue fell (from ~RMB139.4m), consistent with softer consumer spending and a maturing resident base.

Competitive position. Inside the estate it enjoys a near-captive position; the real competition is leakage to external hypermarkets (RT-Mart, Walmart/Sam's Club, local chains) and e-commerce/community group-buying that residents can reach outside the gates or online. It wins on convenience and proximity; it loses on price and range versus large-format rivals.

Role in the group. The top-line anchor and the most consumer-cyclical segment - a cash contributor whose revenue is currently drifting down with the broader mainland consumption slowdown.

Property Management Services (~RMB114.5m, the fastest-growing operating line)

What it does. General property management and resident-support services - security, cleaning, landscaping, common-area maintenance, and resident services across the residential and commercial floorspace of Clifford Estates and related Clifford developments.

Core capability. Managing a 500-hectare, 60,000-home estate as a single integrated operation is genuinely hard to replicate, but the barrier is ownership and incumbency rather than proprietary process. The company has been consolidating its grip here: it acquired the remaining 24.5% of Guangzhou Panyu Clifford Property Management (for about RMB21.8m), taking that management subsidiary to wholly-owned.

Why it exists separately. This is a recurring-fee, contract-based service business with predictable revenue - structurally different from the goods-based retail line. FY2025 property-management revenue grew (from ~RMB86.8m), the one operating segment moving up while others declined.

Competitive position. It effectively has no direct competitor inside its home estate. The comparison set is China's large listed property managers, but Clifford does not meaningfully bid against them for third-party mandates - its book is its own community.

Role in the group. The recurring, defensive core - the segment management points to as the stable base under the volatility of retail and precious metals.

Off-Campus Training Services (~RMB28.4m)

What it does. After-school education for residents' children - language-learning classes and a range of "interest" (non-academic) classes such as arts, music, and enrichment programmes, run inside the estate's education facilities.

Core capability. A captive pipeline of local families and existing physical teaching space inside a community that markets itself partly on schooling (Clifford Estates hosts bilingual schools).

Why it exists separately. Education is regulated distinctly from property and retail, and its economics turn on teacher headcount and class enrolment. This segment is the most exposed to policy: China's 2021 "double reduction" (双减) crackdown on for-profit academic tutoring reshaped the whole sector, pushing operators toward non-academic "interest" classes. FY2025 revenue fell (from ~RMB36.5m), consistent with both regulatory pressure and a softer demand environment.

Competitive position and role. A small, policy-sensitive segment; useful for resident stickiness and cross-sell, not a profit engine.

Information Technology Services (~RMB17.9m)

What it does. Technology services for the estate and its buildings - related engineering, security systems, and hardware/software integration, plus telecommunication services.

Core capability. Systems-integration and security-installation know-how tied to the estate's infrastructure; effectively the in-house technical contractor for a community it already manages.

Why it exists separately / role. Different (project- and engineering-based) revenue profile; a smaller supporting line that deepens the group's control of the physical estate rather than a standalone growth bet.

The four smaller segments (grouped, together roughly RMB34m)

• Renovation and Fitting-Out Services - flat renovation and equipment installation for residents, a project-based line tied to unit turnover inside the estate.
• Property Agency Services - sales agency, rental agency, and post-rental services for units within the community; revenue tracks secondary-market transaction volumes in a soft mainland property market.
• Catering Services - consultancy and management for the estate's food-and-beverage outlets. Management has flagged restructuring of the catering division as a drag on group revenue, one of the reasons FY2025 top line fell.
• Others - employment placement (domestic help) and laundry services.

Segment summary

Overlay, not a segment: the precious-metals treasury (silver and now gold bullion) is an investment activity carried on the balance sheet, and its fair-value movements - not any operating segment - drove FY2025's record profit.

Section 3: Products and business detail

The "product" is the bundle of daily-life services delivered inside one physical location, so the catalogue is best understood as everything a Clifford Estates resident can buy from the company across a typical month:

• Groceries and fresh food from company-run supermarkets, wet markets, and convenience stores.
• A managed home: security patrol, gate control, cleaning, landscaping, and common-area maintenance for their building and the estate.
• After-school classes for their children (languages, arts, interest programmes).
• Building technology: CCTV and security systems, network and hardware/software integration, telecom connectivity.
• Home projects: renovation and fit-out when they move in or upgrade.
• Transactions: agency services to buy, sell, or rent a unit inside the estate, plus post-rental support.
• Household help: domestic-worker placement and laundry.
• Dining: F&B outlets the company consults on or manages.

Geography. The business is overwhelmingly concentrated in Panyu, Guangzhou, centred on Clifford Estates, with the parent Clifford Group's related developments in Huadu, Nanhai, Zhaoqing (all Guangdong) and Lushan (Jiangxi) providing an adjacency. There is no export dimension; this is a single-metro, single-community services operator. The listing is in Hong Kong; the functional and reporting currency is RMB.

The treasury "product line." The most distinctive operational fact is the precious-metals programme, which functions like a segment in its financial impact:

• Since 2020, surplus cash was placed in unallocated silver bullion as treasury management/capital preservation.
• Carrying value was about RMB141.3m at end-2024.
• Silver's 2025 rally produced large fair-value gains that flowed through the P&L and made FY2025 profit a record.
• In January 2026, the company sold 680,000 ounces of silver (400,000 oz on 9 Jan, 280,000 oz on 12 Jan) for aggregate net proceeds of about RMB387.9m, booking a further ~RMB43.8m gain expected in FY2026.
• Proceeds were redeployed: ~RMB85.7m to the FY2025 dividend, ~RMB180m into 2,730 ounces of gold plus 150,000 ounces of allocated silver, ~RMB50m to repay payables, and ~RMB72.2m to working capital.
• The board set a cap of RMB400m on precious-metals investment, with internal controls and controlling-shareholder written approval obtained.

The shift from unallocated to allocated metal (and adding physical gold) is a governance detail worth noting: allocated bullion is specifically segregated and owned outright, reducing counterparty risk versus an unallocated claim.

Milestones that shaped the business. Consolidation of the Panyu property-management subsidiary to 100% (buying out the 24.5% minority for ~RMB21.8m); initiation of the silver programme in 2020; the special dividend and dividend-policy reset in 2025; and the January 2026 silver monetisation and rotation into gold.

Section 4: Customers

The customer is, almost tautologically, the resident of Clifford Estates - and by extension residents of the related Clifford developments. This is a business-to-consumer operation aimed at a captive population of roughly 200,000 people living behind the same gates.

Who buys and how the decision is made. For retail, off-campus training, laundry, and domestic-help placement, the buyer is the household - a routine, low-consideration, high-frequency decision driven by convenience and proximity. For property management, the "customer" relationship is structural: residents and owners pay management fees as a condition of living in the estate, so there is no repeat purchase decision in the ordinary sense - it is closer to a utility. For renovation, agency, and IT/security projects, the buyer is the individual owner or the estate's own facilities function, and the decision is project-specific.

Why they choose Clifford. Physical presence and default availability inside a walled community. When the supermarket, the training centre, and the management office are all inside your estate, the company captures spend by geography rather than by winning a competitive pitch. Trust in the "Clifford" brand - the same name on the estate, the schools, and the hospital - reinforces the default.

Switching costs. For property management, switching is effectively impossible for an individual - you cannot unilaterally change who manages a shared estate. For retail and services, switching is trivially easy in principle (drive outside the gates, or order online) but frictional in practice, which is exactly why in-gate convenience is the whole model. The risk is not a customer "leaving" so much as customers spending less inside the gates as external retail and e-commerce improve.

Concentration. The concentration risk is unusual: it is not a single named customer, it is a single community. The entire operating business is exposed to the health, demographics, and spending power of one estate. If Clifford Estates ages, loses residents, or sees household budgets tighten, every operating segment feels it at once - which is visible in the simultaneous FY2025 declines across retail and training.

Contract structure and revenue predictability. A mix: property management fees are the recurring, predictable base; retail is high-frequency but volume-sensitive spot revenue; training is enrolment-based and seasonal; renovation, agency, and IT are lumpier project revenue. The operating revenue is therefore reasonably sticky but slowly declining, while the reported profit has become dominated by the far less predictable mark-to-market on precious metals.

Section 5: Competitive landscape

Clifford Modern Living sits in an odd competitive spot because its home market is a private, enclosed community it effectively controls. The competition is therefore less "who wins the mandate" and more "where does resident spending leak."

Property management. Nationally, Chinese property management is a large, consolidating, listed industry. The relevant peers - as size references, not as head-to-head rivals - include the names below. Clifford does not bid against these for third-party estates; it manages its own. Where it is exposed is the comparison: these players have scaled asset-light, third-party expansion models, while Clifford is deliberately captive and single-community.

Retail. Here the competition is real and external. Residents can shop at large-format grocers and warehouse clubs outside the gates, and buy from e-commerce and community group-buying platforms. Clifford wins on proximity for daily top-ups and fresh food; it loses on price, assortment, and increasingly on the convenience of same-day delivery.

Off-campus training. Post-"double reduction," the sector fragmented and shrank; Clifford competes with local interest-class providers and online enrichment, with the whole category capped by regulation.

Barriers to entry. Extremely high inside its estate (you cannot build a competing gated community on the same land, and the property-management franchise is locked), and essentially nil outside it (Clifford has no structural advantage competing for retail or management business elsewhere, and it does not try). This is a moat that is deep but narrow - one estate wide.

Market-cap figures are rough magnitudes as of mid-2026, shown only as peer-size reference; they move constantly and should not be treated as precise. None is a direct competitor for Clifford's captive estate business.

The honest read: Clifford's competitive position is not "we out-compete the big managers." It is "we are the only operator inside a community we own the relationship to." That is durable but capped in size, and the operating side is slowly eroding at the edges to external retail. The genuine differentiator on the reported numbers right now is the balance sheet strategy, not any operating moat.

Section 6: Industry

Clifford straddles three industries, each with its own demand drivers.

Chinese residential property services. Demand is driven by the installed base of managed floorspace and by property managers' ability to cross-sell "value-added" community services. The sector grew rapidly through the 2010s as developers spun off management arms, then de-rated sharply from 2021 as the mainland property developers behind many of them (Evergrande, Country Garden and others) hit liquidity crises, dragging sentiment across the whole listed group. Clifford is insulated from developer default risk in one specific way: its estate is built and long-established, and its parent is a private family group, not a leveraged national developer racing to deliver pre-sold units. But it shares the sector's exposure to soft secondary-market transactions (hurting its agency line) and to residents' willingness to pay for services.

Chinese consumer retail (grocery). Demand tracks household income and confidence in the Pearl River Delta. The mainland's post-pandemic consumption recovery has been uneven, and grocery is being reshaped by e-commerce, community group-buying, and warehouse clubs. This is the structural headwind behind Clifford's declining retail revenue.

Off-campus education. Reshaped by the 2021 "double reduction" policy, which banned for-profit academic tutoring in core school subjects and pushed the market toward non-academic interest classes and stricter licensing. The industry is smaller and more regulated than it was five years ago; growth is capped by policy, and demand is demographically pressured by falling birth rates in China's large cities.

The precious-metals overlay. Not an industry the company operates in, but the one that now moves its earnings. Silver and gold prices are set globally by macro forces - real interest rates, the dollar, industrial demand (for silver), and safe-haven flows. This exposes a Guangzhou services company's reported profit to the world bullion market, a genuinely unusual industry-level dependency for a firm in this sector.

Regulation and cyclicality. Property-management fees face local price scrutiny; education is tightly regulated; retail is competitive but lightly regulated. Operating cyclicality is moderate and consumer-driven. The reported cyclicality, however, is now dominated by commodity-price cyclicality via the bullion book - which can swing profit far more violently than any of the operating segments.

Section 7: Growth triggers

Drawn from the last six results announcements and their MD&A/outlook sections. As noted, there are no earnings calls; each trigger is cited to the corresponding results release.

• Precious-metals monetisation and rotation into gold and allocated silver (FY2025 results, 27 March 2026; discloseable-transaction announcements Feb-March 2026). The January 2026 sale of 680,000 oz of silver for ~RMB387.9m locks in a further ~RMB43.8m gain expected in FY2026, and RMB180m is being redeployed into 2,730 oz gold plus 150,000 oz allocated silver under a RMB400m cap.

Management frames the shift as "a defensive, prudent capital-management approach," moving surplus cash from deposits into precious metals as "better value stores during economic downturns."

• Full ownership of the Panyu property-management subsidiary (prior-period announcement; carried into FY2025). Buying out the remaining 24.5% of Guangzhou Panyu Clifford Property Management for ~RMB21.8m consolidates 100% of the estate's management economics, and property-management revenue grew in FY2025 against declines elsewhere.

• Property-management as the growth line within operations (FY2025 results, 27 March 2026; H1 2025, 28 August 2025). Property-management revenue rose to ~RMB114.5m (from ~RMB86.8m), the one operating segment management points to as expanding.

• Catering-division restructuring (FY2024 results, 28 March 2025; reiterated FY2025). Management attributed part of the revenue decline to restructuring the catering business - framed as a deliberate reset rather than lost demand, implying a cleaner base going forward.

• Balance-sheet firepower (repeated across FY2024, H1 2025, FY2025). Management has consistently emphasised a net-cash position with no borrowings, positioning the group to fund dividends, the metals programme, and opportunistic deployment without leverage.

"The Group maintained a net cash position and had no borrowings as of year-end," with a stated emphasis on "strong liquidity, strategic investments, and diversified growth."

• Revised dividend framework enabling flexible capital return (FY2024 results, May 2025 policy change; FY2025 final dividend, March 2026). The reset to a "no less than 25% of audited profit" final-dividend floor, paired with the doubled FY2025 final dividend and a prior special dividend, signals capacity for larger returns when investment gains are realised.

Note the character of this list: the genuinely new trigger is a treasury/commodity decision, not an operating expansion. There are no new plants, no new customer wins, no new-market entries - because the operating business is a fixed, single-community franchise. The growth optionality here is financial, not industrial.

Section 8: Key risks

• Earnings are now driven by bullion prices, not the business. FY2025's record profit was a fair-value gain on silver; the ~89% net margin is an accounting artefact of a mark-to-market gain sitting on top of shrinking operating revenue. Mechanism: if gold and silver fall, the same fair-value line reverses and reported profit can collapse - or turn to a loss - with the operating segments unable to offset it. This is a high-probability volatility risk (metals will move) and a moderate-to-high-probability drag risk if the 2025 rally unwinds. It is the single most important thing to understand about the reported numbers.

The company itself states the metals are held "to preserve capital and hedge against market uncertainties" - an acknowledgement that this is a macro bet, not operating income.

• Single-community concentration. The entire operating business depends on one estate's population and spending. Mechanism: demographic ageing, out-migration, or a Pearl River Delta consumption slump hits retail, training, agency, and catering simultaneously - as the across-the-board FY2025 operating declines already show. High-probability, moderate-magnitude drag.

• Controlling-shareholder dominance and related-party structure. The Peng/Man family controls ~72.89% and also controls the estate, the schools, and the broader Clifford Group. Mechanism: minority holders have little say over capital allocation - including the decision to route surplus cash into precious metals - and related-party dealings between the listed company and the family's other Clifford entities require trust in governance. The metals programme was approved by written controlling-shareholder consent in lieu of an independent shareholder vote, a governance shortcut that concentrates the decision. Low-probability catastrophic, but a persistent structural overhang.

• Structural erosion of operating segments. Retail is leaking to external hypermarkets and e-commerce; off-campus training is capped by "double reduction" and falling birth rates; property agency depends on a soft secondary market. Mechanism: slow, compounding revenue decline that the property-management line only partly offsets. High-probability, moderate drag - visible in the 12.8% FY2025 revenue fall.

• Dividend reliability tied to volatile gains. Management cut the minimum payout floor from 50% to 25% in 2025 "for greater flexibility... amid a challenging operating environment." Mechanism: with operating profit soft, large dividends (the doubled FY2025 final, the prior special) are effectively funded by realised metals gains; if metals reverse, the capacity to sustain elevated payouts shrinks. Moderate probability, moderate impact.

• Liquidity and float. As a tightly-held small cap (~72.89% controlled), the free float is thin, which amplifies price moves and limits the reliability of the market price as a signal. Structural, ever-present.

Section 9: Walk the talk

Six reporting periods, assessed for management credibility. Concall dates used: FY2025 (27 Mar 2026), H1 2025 (28 Aug 2025), FY2024 (28 Mar 2025), H1 2024 (Aug 2024), FY2023 (Mar 2024), H1 2023 (Aug 2023). The most recent (FY2025, 27 Mar 2026) is within ~90 days-adjacent to today's window as the latest released period; H1 2026 is not yet due.

The arc that emerges is of a management team that has been consistent and, in the specific case of the treasury strategy, vindicated - but whose consistency is easier to judge on capital allocation than on operating growth, because they have never really promised operating growth.

Start at the early periods. Across FY2023 and into FY2024, management's story was steady, incremental: operating revenue edging up (FY2024 revenue +6.5%, net profit +13.6% to RMB97.5m), a net-cash balance sheet, and an ongoing treasury programme in silver initiated back in 2020. There was no grand operating-growth guidance to fail - the framing was always defensive and liquidity-first. That matters for credibility: a team that never over-promises operating growth cannot be caught having missed it.

The treasury bet is where management said something specific and it came true. Having committed surplus cash to silver since 2020 as capital preservation, they held the position, and when silver ran in 2025 the payoff was exactly the "store of value in uncertain times" thesis they had described. FY2025 delivered a record RMB284.6m profit on that gain. Then, crucially, they acted on their own logic rather than getting greedy: in January 2026 they monetised 680,000 oz at a large gain and rotated into gold and allocated silver, tightening governance (segregated metal) and capping the programme at RMB400m. That is a promise-and-execution loop that closed cleanly.

Guided approach: surplus cash deployed into precious metals "to preserve capital and hedge against market uncertainties." Outcome: silver appreciation drove a ~192% profit jump in FY2025; position partly realised in Jan 2026 for ~RMB387.9m and rotated into gold and allocated silver.

Where the record is more mixed is on capital-return messaging. Management cut the minimum dividend-payout floor from 50% to 25% in May 2025, citing a "challenging operating environment" - a downgrade of a prior commitment.

"Greater flexibility in managing cash flow amid a challenging operating environment."

Yet in the same breath they paid a special dividend of HK11.2 cents (~HK$113.8m) in March 2025 and then doubled the FY2025 final dividend to HK9.50 cents (from HK4.50). So the floor came down while actual cash returned went up. A charitable read: they lowered the binding promise to preserve optionality, then over-delivered when metals gains gave them room. A skeptical read: the payout policy is now discretionary and tied to volatile investment gains rather than durable operating cash flow. Both reads are defensible; what is clear is that the policy was softened even as the cheque got bigger.

On the operating side, the "catering restructuring" flagged in FY2024 and repeated in FY2025 is the clearest example of a soft promise that translated into a real revenue drag - management framed it as a deliberate reset, and revenue did fall, so the framing was honest even if the outcome was a decline.

Assessment. This is a management team that does what it says on capital and balance-sheet strategy, and has been honest (if unexciting) about operating softness rather than papering over it. The credibility caveat is not dishonesty; it is that the wins have come from a commodity bet and the payout policy has become discretionary, so "walking the talk" here means trusting a controlling family's treasury judgment more than trusting an operating growth plan. On its own terms, the team has been reliable.

Section 10: Shareholder friendliness index

Dividends. The last three fiscal years show a company that pays, and has recently paid generously - but off an increasingly investment-gain-funded base. For FY2024 the final dividend was HK4.50 cents per share, and separately a special dividend of HK11.2 cents (about HK$113.8m) was paid on 31 March 2025. For FY2025 the final dividend was doubled to HK9.50 cents. Prior years were smaller (broadly on the order of ~HK2 cents for FY2022 and ~HK5 cents paid in 2023 for the FY2022 period, per dividend-history aggregators). The notable policy event: in May 2025 management cut the minimum-payout floor from 50% to 25% of audited consolidated profit, citing cash-flow flexibility in a tough operating environment - so the binding commitment weakened even as actual cash paid rose, because the FY2025 payout was underwritten by realised silver gains rather than by operating earnings. Payout ratio is not especially revealing here given the distortion from fair-value gains; the more honest framing is that recent dividends are partly a distribution of investment profits.

Buybacks and dilution. At the June 2026 AGM the board proposed the standard general repurchase mandate (the routine HKEX authorisation to buy back up to 10% of issued shares) alongside the dividend. This is an authorisation, not evidence of execution. On the two required sources: (1) MoatMap's database records zero buybacks in the trailing ~90 days (since 20 April 2026); (2) an external search of results announcements and news for the prior three years surfaced no evidence of any actual on-market share repurchases - only the customary annual mandate. Share count has been stable at roughly 1.0 billion shares (implied by the ~HK$113.8m special dividend at HK11.2 cents ≈ ~1,016m shares), with no meaningful buyback-driven reduction and no signs of material option dilution. So over three years the count is essentially flat: neither retiring nor issuing shares in size.

Verdict: Returns Capital - via dividends, not buybacks - but with an asterisk: the generosity is real yet increasingly funded by volatile precious-metals gains rather than durable operating cash, and the minimum-payout promise was recently softened.

Section 11: Insider activities

Per the MoatMap cross-market disclosure database (market: HK), which is the canonical source here because the HKEX Disclosure of Interests portal is API-gated and returns blocked stubs to web search:

Insider transactions - last 12 months: none recorded. MoatMap shows 0 insider transactions for 3686.HK over the trailing twelve months, and 0 buybacks in its ~90-day window. There were no HKEX Disclosure of Interests filings (Forms 3A/3B for directors/CEO or substantial-shareholder change notices) captured in this period.

Context on the ownership backdrop. The absence of trading is itself the signal here. Chairman and CEO Man Lai Hung (孟丽红), together with Elland Holdings Limited, holds roughly 72.89% of the company. A controlling family at that ownership level, with no disclosed director dealings in the window, has neither trimmed nor added to its stake on the open market - consistent with a long-term, control-oriented holding rather than a trading posture. There is no cluster of insider buying to flag as bullish, and, equally, no insider selling to raise concern.

Net assessment. With no reportable insider transactions in the last 12 months, the read is neutral. There is neither the conviction signal of open-market insider buying nor the caution flag of insider selling. The controlling shareholders' stake appears static and dominant. For a tightly-held small cap, a quiet insider record of this kind is unremarkable and carries no directional information.

Section 12: Scenarios

Bull case. The precious-metals bet keeps working. Gold and allocated silver hold or extend their gains, and the RMB400m treasury book throws off further realised gains that management, having shown discipline by monetising silver near strength in January 2026, harvests prudently and returns to shareholders through the enlarged, flexible dividend framework. Underneath, the operating core stabilises: property-management revenue, now 100%-owned at the Panyu subsidiary, keeps grinding higher; the catering restructuring is complete so the operating revenue base stops shrinking; and the Pearl River Delta consumer recovers enough to steady retail. The result is a small, net-cash, family-run services company that has quietly become a high-quality dividend payer backed by a hard-asset balance sheet - re-rated by income investors who like the combination of gold exposure, no debt, and cash returns.

Base case. The most likely path is bifurcated. The operating business continues its slow drift: retail and training keep declining modestly with mainland consumption and demographics, property management provides a stabilising counterweight, and total operating revenue is broadly flat-to-down. The reported profit, meanwhile, swings with bullion prices - some years spectacular, some years mediocre - because that is now the dominant earnings driver. Management keeps the balance sheet unlevered, pays a dividend sized to whatever gains are realised, and runs the estate competently. Investors get a captive services annuity wrapped around a commodity trading position, with the dividend as the main tangible return. Nothing breaks; nothing dramatically compounds.

Bear case. Gold and silver reverse hard. The same fair-value line that manufactured FY2025's record profit runs in reverse, and because the operating segments are shrinking and can't offset it, reported profit collapses or turns negative in a down year for metals. The reduced 25% payout floor and the discretionary, gain-funded nature of recent dividends mean distributions get cut just as the share price is weak. Simultaneously, the single-community concentration bites: an ageing estate and a soft Delta economy pressure retail, agency, and catering together, while "double reduction" and falling birth rates cap training. Minorities, holding ~27% against a 72.89% controlling family that makes treasury decisions by written consent, have no lever to pull. The company remains solvent and cash-rich, but the equity behaves like a leveraged, illiquid bet on precious metals with a slowly-eroding services business attached - and the market prices it accordingly.

Sources: Yahoo Finance 3686.HK, MarketScreener company profile, irasia IR page, Minichart FY2025 annual report summary, Minichart silver bullion disposal (Feb 2026), Minichart precious-metals treasury investment (Mar 2026), Globe and Mail - dividend payout cut, Clifford Group (祈福集团), HKEXnews FY2024 results PDF, HKEXnews FY2025 results PDF, MoatMap cross-market disclosure database (HK insider/buyback feed, as of 2026-07-19).