Evidence library · Hong Kong · Luxury watch and jewellery retail
Emperor Watch & Jewellery HKEX: 0887
Investor snapshot
Business model
Emperor Watch & Jewellery buys finished European luxury watches as a dealer and designs its own gold jewellery, then sells both from sixty-nine shops in Hong Kong, the Chinese Mainland, Macau, Singapore and Malaysia.
Evidence now
Revenue for the six months to 30 June 2026 rose 5.0% to HK$2,934.0m and profit attributable to owners rose 58.0% to HK$310.4m on a gross margin of 33.0% against 30.1%, while inventories rose to HK$3,488.7m, or 321 days of cost of sales, against HK$1,564.7m of net cash and no bank borrowings.
Main risk
The central risk is that 3.11 percentage points of the first-half margin gain has no disclosed cause and no territory split is published at the half year, so the largest single driver of the reported profit cannot be checked against anything the Group has filed.
Next proof
The next test is the FY2026 results announcement expected in March 2027, which carries the full-year gross margin, the inventory write-down inside cost of sales and the first territory margin split since Other Asia Pacific revenue fell 75.2%.
Publication state: evidence library — no public letter rating, no valuation.
Information cut-off: 3 September 2026. The latest financial statement in scope is the interim results announcement of 20 August 2026, for the six months to 30 June 2026; the 2026 interim report had not been filed at the cut-off, and the last audited accounts are those for the year to 31 December 2025.
About the private research record
Also on file for this company, and not published here: a separate access-controlled working record holding the analyst judgments, the forecast and valuation work, the review artifacts and the full source register. It remains private, access-controlled and outside this public page.
On this page
Business anatomy · from inputs to customer value
Allocated watches and bought gold become cash only when a shopper walks in
The Group owns no watch brand and makes no watch. It funds the stock, holds it on prime retail streets, and earns the difference between what the shelf cost and what the shopper pays.
Follow the operating chain from demand or inputs to customer outcome and cash.
InputsDealer supply
Buy the stock
What happensEuropean brands allocate watches to the Group as a dealer; gold is bought for its own jewellery brand.
Capital at workAllocation is granted by the brand, never bought outright.
Route to marketPrime streets
Stock the shops
What happensSixty-nine shops in Hong Kong, the Mainland, Macau, Singapore and Malaysia hold the stock on prime streets.
Revenue triggerRent and salespeople are fixed; the shelf turns about once a year.
Customer outcomeOver the counter
Sell to the shopper
What happensResidents and Mainland tourists buy a watch at the brand price, or jewellery priced off gold plus workmanship.
Who paysGross margin is set by the metal cost and the discount given.
Cash conversionTill to shelf
Turn stock into cash
What happensTakings settle supplier and gold-loan balances, fund new stock and new stores, and leave a dividend.
Cash triggerCash returns only as the shelf sells; a write-down destroys it.
Original SMID Research comic. AI-assisted, analyst-directed monochrome artwork depicts the evidenced operating steps of Emperor Watch & Jewellery; deterministic captions and dated mix figures state the economics using issuer disclosures available to 2026-09-03. The drawings are representative—not issuer artwork, an exact product or site design, a statement of scale, or a company forecast.
Investor translation
What matters after the operating picture
Four questions connect the business model to cash and balance-sheet risk. This is a factual reading aid, not a valuation or recommendation.
Value lever
Revenue per shop and the gross margin on brand-allocated watches and own-brand gold jewellery.
Cash bottleneck
Inventory: about a year of cost of sales sits on the shelf before a shopper buys it.
Credit breakpoint
No bank borrowings; gold loans and pledged deposits are the only funded lines.
Next proof
FY2026 results in March 2027: full-year gross margin, the write-down and the store count.
Text version of this comic
Inputs · Buy the stock European brands allocate watches to the Group as a dealer; gold is bought for its own jewellery brand. Capital at work: Allocation is granted by the brand, never bought outright.
Route to market · Stock the shops Sixty-nine shops in Hong Kong, the Mainland, Macau, Singapore and Malaysia hold the stock on prime streets. Revenue trigger: Rent and salespeople are fixed; the shelf turns about once a year.
Customer outcome · Sell to the shopper Residents and Mainland tourists buy a watch at the brand price, or jewellery priced off gold plus workmanship. Who pays: Gross margin is set by the metal cost and the discount given.
Cash conversion · Turn stock into cash Takings settle supplier and gold-loan balances, fund new stock and new stores, and leave a dividend. Cash trigger: Cash returns only as the shelf sells; a write-down destroys it.
What Emperor Watch & Jewellery does
Emperor Watch & Jewellery is a shopkeeper. It buys finished European luxury watches as an authorised dealer and resells them at retail; it designs and sells gold and gem-set jewellery under its own ‘Emperor Jewellery’ brand; and it does both from its own shops in Hong Kong, the Chinese Mainland, Macau, Singapore and Malaysia. It owns no watch brand and makes no watch. Rolex, Patek Philippe, Tudor and Cartier are the four brands the issuer names. Almost everything else on this page follows from that: the scarce assets are the dealerships and the prime-street leases, and the capital is the stock.
Revenue splits by product as well as by place. In FY2025, the year to 31 December 2025, watch retail supplied HK$3,492.4m and watch commission income HK$36.2m — 61.2% of group revenue between them. Jewellery retail supplied HK$2,123.6m and a new wholesale line to Mainland franchisees HK$113.2m, 38.8% together. The jewellery half is where the Group moves one step upstream: it owns the brand and designs the product. It is also the thinner half, priced off the gold price plus a workmanship charge, and it is where the store network is being rebuilt.
At 30 June 2026 the Group operated 69 shops: 29 in Hong Kong, 26 in the Chinese Mainland, nine in Macau, three in Singapore and two in Malaysia. Sixteen of the 26 Mainland stores are jewellery stores and nine of those sixteen are franchised. The Mainland jewellery business sits inside a joint venture formed in April 2025 in which the Group holds 51% and an outside partner, Xiaoduocai, holds 49%. Hong Kong includes a self-owned five-storey retail complex at 4–8 Canton Road, Tsim Sha Tsui, whose upper floors the Group bought from its controlling shareholder’s property arm in 2025.
Operating disclosure gap. The filings do not give revenue by brand, same-store sales, store-level profitability, the split of Mainland revenue between watch and jewellery or between own and franchised stores, or the terms and renewal dates of the watch dealerships that produce three-fifths of revenue. Growth therefore cannot be decomposed into volume, price and mix from the public record.
The four-year record, and what is not comparable in it
Revenue has risen in each of the last three years, from HK$3,684.3m in FY2022 to HK$5,765.3m in FY2025, a 56.5% increase. FY2022 itself was a fall: revenue was HK$3,926.6m in FY2021. Profit attributable to owners more than doubled over the same three years, from HK$222.1m to HK$458.4m, and the store network shrank from 93 shops at the end of FY2023 to 64 at the end of FY2025. Revenue per average store rose from HK$59.8m in FY2024 to HK$79.0m in FY2025 on that shrinking base.
HK$m unless stated, years to 31 December
FY2022
FY2023
FY2024
FY2025
Revenue
3,684.3
4,823.2
5,230.3
5,765.3
Gross profit
1,177.3
1,450.3
1,480.9
1,779.5
Gross margin
31.95%
30.07%
28.31%
30.87%
Inventory write-down inside cost of sales
—
—
122.7
50.0
Other gains and losses
(21.3)
(16.4)
(25.2)
(85.1)
Profit before taxation
277.2
367.7
317.0
531.4
Profit for the year
222.1
299.2
256.7
431.3
Profit attributable to owners
222.1
299.2
256.7
458.4
Basic earnings per share
3.28c
4.41c
3.79c
6.34c
Dividends declared per share
0.62c
1.32c
1.10c
1.69c
Payout of basic earnings
18.9%
29.9%
29.0%
26.7%
Net cash from operating activities
605.6
403.6
772.4
1,096.7
Inventories at 31 December
2,703.4
3,060.3
3,003.4
3,078.2
Cash, time deposits and pledged deposits
664.4
619.6
949.8
1,805.8
Gold loans
—
—
—
207.7
Equity attributable to owners
4,845.1
5,046.7
5,214.1
5,617.4
Stores at 31 December
—
93
82
64
Sources: annual reports 2022 to 2025 and the annual results announcements for FY2023, FY2024 and FY2025, all filed on HKEXnews. Figures are as reported. Dividends declared per share are the interim and final dividends declared for that financial year added together; the payout line divides that total by basic earnings per share and is arithmetic on the two rows above it. The inventory write-down is disclosed inside cost of sales for FY2024 and FY2025 only; dashes mark figures the filings read for this page do not carry. Profit for the year and profit attributable to owners diverge from FY2025 onwards because non-controlling interests appear for the first time in that year.
Three things in that table are not what they look like. FY2025’s two profit lines diverge for the first time. Profit for the year is HK$431.3m and profit attributable to owners is HK$458.4m, because non-controlling interests took a HK$27.1m loss — the outside 49% of the Mainland jewellery joint venture formed in April 2025. Before FY2025 there were no material non-controlling interests and the two lines were identical. FY2024 is a write-down year. HK$122.7m of inventory write-down sits inside that year’s cost of sales and 1.49 percentage points of the FY2025 margin improvement is simply its absence. And the FY2025 balance sheet moved for reasons that are not trading. Cash, time deposits and pledged deposits nearly doubled, but HK$210.5m of that is the outside partner’s subscription for its 49% of the joint venture and HK$79.3m is the net proceeds of a share placing; both are financing, not operating cash.
The first half of 2026: what the print actually says
The six months to 30 June 2026 are the most recent evidence on this page, and they are the reason the record above reads better at the profit line than at the revenue line.
HK$m unless stated, six months to 30 June
1H2025
1H2026
Change
Revenue
2,793.6
2,934.0
+5.0%
Gross profit
840.4
968.8
+15.3%
Gross margin
30.08%
33.02%
+2.94pp
Other gains and losses
(14.4)
24.5
+38.9
Profit before taxation
240.8
389.8
+61.9%
Profit for the period
194.3
317.7
+63.5%
Non-controlling interests
(2.2)
7.3
—
Profit attributable to owners
196.5
310.4
+58.0%
Basic earnings per share
2.73c
4.28c
+56.8%
Interim dividend per share
0.55c
0.90c
+63.6%
Stores at 30 June
73
69
−4
Source: interim results announcement of 20 August 2026 and the comparatives it carries. Percentage changes are arithmetic on the two columns. The 1H2025 comparative for basic earnings per share is the weighted average of 7,203,973,322 shares used in the 2026 announcement, not the 7,256,708,129 printed as the denominator in the 2025 announcement; see the corrections section below.
A quarter of the increase in profit before taxation is one line. Other gains and losses swung from a HK$14.4m loss to a HK$24.5m gain, a HK$38.9m swing against a HK$149.0m increase in profit before taxation. Of the HK$24.5m gain, HK$11.9m is a fair value gain on gold loans and HK$13.9m is a net exchange gain. Strip the whole line from both halves and profit before taxation grew 43.1% rather than 61.9%. Neither component is a trading result and neither is disclosed as recurring.
The issuer told the market a month early, and the market moved then. A positive profit alert filed on 21 July 2026 said net profit for the half would be not less than HK$310m; the reported figure was HK$317.7m, 2.5% above the stated floor. An alert filed on 27 January 2026 did the same thing for FY2025: a HK$420m floor against HK$431.3m reported, 2.7% above. Both alert figures, and both reported comparisons, are total profit for the period including non-controlling interests — that is the measure the alerts use, and it is not the measure the profit-attributable-to-owners line uses. The two reconcile through the non-controlling-interest line, which was a HK$7.3m profit in the half and a HK$27.1m loss in FY2025, so profit attributable to owners sits below total profit in the half and above it in FY2025.
Where the revenue comes from, and what stopped
Three first halves within 2.2% of each other, then a 75.2% fall. External revenue by reported territory, six months to 30 June, HK$m. Other Asia Pacific is the issuer’s own segment for Singapore and Malaysia. Source: interim results announcements 2023–2026.
The Group reports four territories. Three of them grew in the first half of 2026 and one of them almost disappeared.
Six months to 30 June, HK$m
External revenue 1H2025
External revenue 1H2026
Change
Segment profit 1H2025
Segment profit 1H2026
Change
Hong Kong
1,583.7
1,753.8
+10.7%
215.2
285.2
+32.5%
Chinese Mainland
723.0
873.3
+20.8%
83.8
147.3
+75.7%
Macau
163.0
203.8
+25.1%
31.5
64.7
+105.5%
Other Asia Pacific
309.7
76.7
−75.2%
52.6
5.5
−89.5%
Total
2,779.3
2,907.6
+4.6%
383.2
502.8
+31.2%
Source: segment note, interim results announcement of 20 August 2026. External segment revenue of HK$2,907.6m plus commission income of HK$26.5m reconciles exactly to group revenue of HK$2,934.0m. Segment profit is the issuer’s own segment result before unallocated corporate expenses. Percentage changes are arithmetic on the two columns.
Take Other Asia Pacific out of both halves and the three continuing territories grew segment profit 50.4% year on year, against the 31.2% the total shows. That is the arithmetic reason the reported group figures understate what the continuing business did, and it is also the reason the history is not a clean base.
What makes the fall worth a section rather than a footnote is what came before it. The same segment produced HK$302.98m, HK$307.64m and HK$309.67m of external revenue in three consecutive first halves — a 2.2% spread across three years — on eight or nine stores, at full-year segment margins of 17.4% to 19.4% in FY2022 to FY2024 against Hong Kong’s 9.3% to 10.9% over the same three years. It was the Group’s highest-margin territory. Singapore went from eight stores at 30 June 2025 to three a year later; Malaysia went from one to two.
Hold every comparison to FY2024, the last full year before the fall, so that none of it is a mixed-period effect. Other Asia Pacific earned HK$70.9m of external revenue per average store that year: 2.0 times the Chinese Mainland’s HK$34.6m and 1.5 times Macau’s HK$48.9m, though below Hong Kong’s HK$88.0m. It earned it on HK$11.9m of non-current assets against Hong Kong’s HK$1,599.1m — HK$54 of revenue for every dollar of fixed asset, against Hong Kong’s HK$1.83 — and carried FY2024 right-of-use depreciation of HK$10.8m against Hong Kong’s HK$203.4m. It is the asset intensity, not the revenue per store, that does not read like shop-floor retail.
A network rationalised, then rebuilt in a different place. Stores by territory at each reported disclosure date. The Mainland count fell by 24 shops over two years and then added six in the six months to 30 June 2026, of which the issuer attributes eight openings to jewellery. Source: ‘Number of stores’ disclosure in each results announcement, 2023–2026.
Evidence boundary. Four annual reports, three interim reports and the corporate-action filings retrieved for this page contain no sentence explaining what the Other Asia Pacific business was, why it fell, whether the fall is an exit, a loss of supply or a loss of a customer channel, or whether any part of it moved elsewhere in the Group. The remaining rows of the 196-row 2022–2026 announcement tape were enumerated by title and classified rather than opened, so what is recorded here is an absence across what was read, not across the whole tape. Five ESG reports on that tape carry territory headcount and were not retrieved.
Margin, and the write-down inside it
The reported margin and the margin before the write-down move in different directions. Territory margins are recomputed from the segment note’s disclosed external sales, commission income and cost of sales; the four territory gross profits sum to group gross profit plus the group write-down in both years, which is the arithmetic check that the decomposition is right. Source: segment note and cost-of-sales note, annual reports 2024 and 2025.
FY2025’s reported gross margin rose 2.55 percentage points to 30.87%. The inventory write-down inside cost of sales fell from HK$122.7m to HK$50.0m over the same period. Before write-downs the margin rose 1.07 percentage points, from 30.66% to 31.73%. The difference — 1.48 of the 2.55 points, or 58% of the headline improvement — is the absence of a FY2024 charge rather than anything the shops did.
The segment note discloses cost of sales by territory, which lets the same split be done one level down. Against external sales plus commission income, FY2025 margins before write-downs were Hong Kong 31.70%, Macau 43.39%, the Chinese Mainland 30.32% and Other Asia Pacific 28.54%, against FY2024’s 29.37%, 33.12%, 33.24% and 29.87%.
The Mainland margin fell 2.92 points before write-downs while its reported margin rose 2.19 points, because the Mainland write-down fell from HK$87.0m to HK$21.5m. Gold jewellery is a thinner business than watches, and the Mainland is where it is being scaled. The direction of that recomputed series is the opposite of the direction the reported series shows, and it is not visible anywhere in the headline numbers.
The first half of 2026 is a different case. The margin rose 2.94 points while the write-down moved the other way — HK$5.1m against nil — so the underlying gain is 3.11 points and is not a write-down effect. It is also not geographic mix: removing Other Asia Pacific from both halves leaves 2.86 of the 2.94 points intact. The issuer attributes the improvement to ‘the remarkable performance of the Group’s overall sales’ and says nothing further. Territory margins are not disclosed at the half year, so the largest single movement in the half-year accounts is also the least explained.
Evidence boundary. There is no half-year territory margin split, no disclosure of the write-down by territory at the half year, and no explanation of the 3.11 points of underlying margin gain. One external observation is on the record and belongs beside it rather than inside it: the COMEX front-month gold future settled at US$4,519.6 an ounce on 3 September 2026, against a calendar-2025 average of US$3,447.3 over 252 sessions, a premium of 31.1%. That is context, not a cause: the sign is not obvious either, since a gold-denominated liability marked at a higher gold price should produce a loss, yet the half booked a HK$11.9m gold-loan fair value gain while the gold-loan balance fell from HK$207.7m to HK$137.1m.
The balance sheet: about a year of stock, and no bank borrowings
The stock is the business, and it is funded by the balance sheet rather than by a bank. The three bars are components of one balance sheet and must not be netted by eye; the pledged portion of the deposits is not freely available. Source: balance sheets in the annual reports 2022–2025 and the interim results announcement of 20 August 2026.
At 30 June 2026 the Group held HK$1,505.0m of cash and cash equivalents, HK$68.0m of time deposits with an original maturity over three months and HK$128.9m of pledged bank deposits, against HK$137.1m of gold loans and no bank borrowings. Net cash on the issuer’s own definition is HK$1,564.7m, which reproduces the HK$1,565m stated in the management discussion to within HK$1m. Undrawn banking facilities were HK$219m, against HK$150m at 31 December 2025.
At the balance-sheet date, HK$m
31 Dec 2025
30 Jun 2026
What it is
Inventories
3,078.2
3,488.7
Watches and jewellery on the shelf and in the warehouse
Cash and cash equivalents
1,292.1
1,505.0
—
Time deposits over three months
318.2
68.0
Outside cash and cash equivalents on the balance sheet
Pledged bank deposits
195.5
128.9
Security for banking facilities; not freely available
Gold loans
207.7
137.1
Twelve-month fixed-rate RMB loans whose fair value gain or loss runs through profit or loss
Bank borrowings
—
—
None reported at either date
Lease liabilities, current and non-current
405.7
410.4
Shop leases capitalised under HKFRS 16
Trade receivables
23.1
22.1
Against HK$2,934.0m of half-year revenue
Equity attributable to owners
5,617.4
5,876.3
—
Net cash, issuer definition
1,598.0
1,564.7
Cash plus time and pledged deposits less gold loans
Sources: balance sheet and notes, annual report 2025 and interim results announcement of 20 August 2026. Net cash on the issuer’s definition is recomputed from the four balance-sheet lines above it and reproduces the figure stated in the management discussion to within HK$1m in both periods. Lease liabilities are the current and non-current lines added together.
Per share, on the 7,256,708,129 shares in issue at 30 June 2026, net cash is HK$0.2156, inventory is HK$0.4808 and book value attributable to owners is HK$0.8098. Those three are components of one balance sheet and must not be added together. They are given separately because each answers a different question, and none of them is a statement about what the shares are worth.
Inventory rose HK$410.5m in the half against HK$140.4m of revenue growth, taking closing inventory from 282 to 321 days of cost of sales. The Group opened one Hong Kong multi-brand store and eight Mainland jewellery stores in the half, and a multi-storey Rolex boutique on its own Canton Road building shortly after it, so an investment reading is well supported by what the filings say. It is not proved by them. The write-down inside first-half cost of sales is disclosed — HK$5.1m against nil a year earlier — and notes 8 and 9 carry ageing analyses of trade receivables and trade payables, but nothing in the announcement breaks the inventory balance down by age or the write-down down by territory, and no post-balance-sheet sell-through is given.
Evidence boundary. The interim results announcement carries the condensed statements, the segment note and the significant-accounting notes, but no cash-flow statement, so first-half operating cash flow cannot be read from it. The 2026 interim report, which historically follows in the second week of September and adds the share-option and directors’-interests notes, had not been filed at this page’s cut-off.
Control, related parties and where the capital went
A discretionary trust vehicle associated with the Yeung family holds 4,393,970,000 shares, 60.55% of the company, down from 64.813% before a January 2025 placing. Five other Hong Kong-listed companies sit under the same discretionary trusts, and at least two of them transact with this one.
The recurring cost of those dealings is small, and saying so plainly matters more than implying otherwise. FY2025 related-party charges were about HK$26m in total — HK$18.0m of information-system and administrative service charges, plus lease interest, variable lease payments and advertising — against HK$531.4m of profit before taxation, roughly 5%. Lease liabilities owed to related companies were HK$45.6m of the HK$405.7m total at 31 December 2025.
The sequencing is the observation. On 9 January 2025 the Company placed 477,250,000 new shares at HK$0.167 through Emperor Securities Limited, a related company, raising HK$79.24m net from independent third parties, stated to be for retail-network expansion and working capital. On 28 February 2025 it agreed to buy the upper floors of 4–8 Canton Road from a wholly-owned subsidiary of Emperor International Holdings for about HK$79.8m, completing in August 2025. The two amounts match to within HK$0.6m and are seven weeks apart. The purchase went to independent shareholders with an independent valuation of HK$80.3m and was approved; the Group held about HK$950m of cash and deposits at the end of FY2024 and did not need the placing to fund it. Both readings survive the evidence, and the page records the sequence rather than choosing between them. What is not in doubt is who bore the dilution: the controlling trust did not subscribe, and its stake fell 4.26 percentage points.
The Mainland joint venture is the other place the capital went. In April 2025 the Group put its Mainland jewellery business into companies owned 51% by it and 49% by Xiaoduocai, an independent third party, and received HK$210.5m for the 49% interest, booked in financing activities and directly in equity with no gain in profit or loss. The PRC vehicle, Emperor Jewellery (Jiangsu), produced HK$161.7m of FY2025 revenue against HK$222.6m of expenses and a total comprehensive loss of HK$55.8m. Two features of the agreement are easy to miss: the first RMB540.0m of accumulated joint-venture losses is borne one-third by Xiaoduocai and two-thirds by the Group, against a 51/49 equity split, so in the loss-making phase the Group carries more than its ownership share; and the FY2025 accounts allocate HK$26.1m, or 46.7%, of the Jiangsu comprehensive loss to the 49% interest, which is neither one-third nor 49%. The filings do not reconcile the two.
In December 2025 the Yeung-family consignment arrangement was renewed with the annual cap raised from HK$35m to HK$50m, on the stated expectation of significantly increased consignment activity from the family. Historical consignment value over the eleven months to November 2025 was HK$30.5m.
Evidence boundary. The filings do not disclose what sits inside the joint venture’s HK$219.3m of non-current liabilities, do not break the non-controlling-interest line out by vehicle at the half year, and do not explain the 46.7% loss allocation. Whether the non-controlling-interest line turning from a HK$27.1m loss in FY2025 to a HK$7.3m profit in the half reflects Jiangsu or another vehicle is not decidable from the public record until the FY2026 annual report.
The market record, and what a price move can be traced to
The shares closed at HK$0.295 on 3 September 2026, the last close before this page’s cut-off. Market capitalisation at that price was HK$2,141m. With the controlling trust at 60.55%, the estimated free float is 39.45%, or HK$844.5m. Median turnover over the trailing 20 sessions was HK$750,225 a day, and over the trailing 250 sessions HK$1,593,893 a day; three sessions in the last 250 traded no shares at all. On the 20-session measure the freely traded portion of this company changes hands slowly enough that the price is set at the margin by small amounts of money.
Each detected move over the three years to 3 September 2026 was tested against the Hang Seng Index and against a four-stock control set of Hong Kong-listed jewellery and watch retailers — Chow Sang Sang (0116), King Fook (0280), Chow Tai Fook (1929) and Luk Fook (0590) — and then matched against the enumerated HKEX announcement tape for the window. Thirty-six sessions or weeks qualified as moves. Two line up with an issuer filing.
Session
Move
Filing in the window
28 January 2026
+36.8%
The session after the Company said FY2025 net profit would be not less than HK$420m against HK$257m in FY2024
22 July 2026
+11.8%
The session after the Company said 1H2026 net profit would be not less than HK$310m against HK$194m a year earlier
Session returns are closing-price moves against the previous close. Both filings are positive profit alerts under Listing Rule 13.09(2)(a). The results announcements themselves moved the shares −6.0% and −1.7% on the following sessions.
Five more of the 36 moves track the index or the peer set, including a 14.6% fall on 7 April 2025 in a market-wide sell-off in which the Hang Seng fell 13.2% the same session. The remaining twenty-nine have no issuer announcement in their window.
Sixteen of those twenty-nine are one episode. Between late April and mid-June 2025 the shares rose from about HK$0.20 to a high of HK$0.64 on 16 June 2025, then gave it all back: two July sessions fell 17.1% and 20.8%, and by the cut-off the price was HK$0.295. Across those sixteen qualifying sessions the median return in excess of the benchmark was about 15 percentage points, volume reached as much as 124 times the trailing median, and on the largest single session the four-peer median return was +6.1% against the Group’s +43.8%. No issuer announcement precedes any of them, and no unusual-price-movement announcement appears anywhere on the tape for the window.
The mapping is deliberately mechanical: it reads a price series against two controls and an announcement archive. It does not examine broker notes, block trades, substantial-shareholder timing, trade press or index reviews. An unexplained move here means no issuer disclosure was found in the window, not that nothing happened. What the record does establish is that this register reprices by tens of per cent without news, which is a fact about holding the shares and is separate from the business.
What the filings still do not answer
What caused the 3.11 percentage points of underlying gross-margin gain in the six months to 30 June 2026. This is the largest single movement in the half-year accounts and the issuer attributes it only to overall sales performance.
What the Other Asia Pacific business was, why it fell 75.2%, and whether the fall is an exit, a loss of supply or a loss of a customer channel. No filing read for this page contains a sentence about it.
The terms, tenor and renewal dates of the watch dealerships that produce 61.2% of revenue. No dealership agreement appears in any row of the 196-row 2022–2026 announcement tape, and none is required to be filed.
The split of Mainland revenue between watch and jewellery, and between own and franchised stores.
What sits inside the Jiangsu joint venture’s HK$219.3m of non-current liabilities, and how the 46.7% allocation of its comprehensive loss to a 49% interest reconciles with the one-third loss-sharing tranche in the joint-venture agreement.
Why the recomputed Macau gross margin before write-downs rose 10.3 percentage points in FY2025, on the Group’s smallest territory.
Territory gross margins at the half year, an ageing of the HK$3,488.7m inventory balance, and any write-down by territory.
First-half operating cash flow, which the interim results announcement does not carry.
What Luck Treasure Limited, named in the group structure, does.
The next public tests
The 2026 interim report, historically filed in the second week of September. It adds the share-option note and the directors’-interests update to the August announcement. It does not add a half-year cash-flow statement or a territory margin split.
A profit alert, if one comes. The Group has pre-announced a profit floor before each of its last two results and beaten it by a low single-digit margin both times, roughly four weeks before the print. On this name the alert has been the price event, not the result.
The FY2026 results announcement, expected March 2027. The full-year gross margin against FY2025’s 30.87%, the inventory write-down inside cost of sales, the store count, and the first territory split covering the period in which Other Asia Pacific fell.
The FY2026 annual report, expected April 2027. The territory cost-of-sales disclosure that makes the margin decomposition on this page possible, the related-party note, the joint-venture disclosures and the non-controlling-interest breakdown.
The Mainland store count. Sixteen of the 26 Mainland stores are jewellery stores and nine are franchised; whether the rebuild continues, and at what margin, is the open question the FY2026 accounts answer first.
What you can watch yourself
Every other test on this page waits for the company to file. These do not. Each row is a series you can look up yourself, free, today — with the level this reading was built against, the levels that would put it in question, and what the series cannot tell you.
Search HKEXnews for stock code 887 and look for an announcement titled Positive Profit Alert or Profit Warning in the weeks before a results date. Emperor Watch & Jewellery Limited, on HKEXnews
Last recorded
310 HK$ million: the net profit floor the alert states, for the period it covers, 2026-07-21
What the reading assumes
194 HK$ million: the net profit floor the alert states, for the period it covers (the prior-period net profit the alert itself compares against, 1H2025)
Watch / alert
194 and 0 HK$ million: the net profit floor the alert states, for the period it covers, on a move below — currently at or better than the level the reading assumed
How often to look
monthly (the series prints event)
What it points to. The Group has pre-announced a profit floor before each of its last two results and beaten it by a low single-digit margin both times, so an alert is a usable point estimate about four weeks before the print and both moved the share price the next session.
Direction only — this pack does not carry a coefficient from this series to reported earnings.
What it cannot tell you. Two observations are not a calibration. The alert states a floor, not a figure, and the Group is under no obligation to issue one at all: it issued none before the FY2023 or FY2024 results. Silence is not a signal, and this row cannot tell a reader anything in a year without an alert.
Settled by The results announcement the alert precedes, due 2027-03-31. Lead time: about four weeks before the results announcement.
Open the Yahoo Finance page for the gold futures contract GC=F and read the last price at the top of the page. CME Group, retrieved through the Yahoo Finance chart API
Last recorded
4,520 US dollars per troy ounce, 2026-09-03
What the reading assumes
3,447 US dollars per troy ounce (calendar-year average of the same daily series, over the 252 sessions stored in market/external-series.json, 2025)
Watch / alert
3,700 and 3,447 US dollars per troy ounce, on a move below — currently at or better than the level the reading assumed
How often to look
weekly (the series prints daily)
What it points to. Gold sets both the selling price and the replacement cost of jewellery, which was 38.8 per cent of FY2025 revenue with gold products 72.4 per cent of that. A rising gold price lifts revenue per gram and produces a holding gain on metal already bought; it simultaneously raises the fair value of the RMB gold-loan liability, which is disclosed inside other gains or losses. The sign is defensible; no coefficient is.
Direction only — this pack does not carry a coefficient from this series to reported earnings.
What it cannot tell you. Gold sets the price and cost of the jewellery half only. It says nothing about the watch half, which was 61.2 per cent of FY2025 revenue and is priced by the brands, so a gold rally alongside falling Hong Kong footfall can leave group revenue flat. The level here is a US dollar futures settlement while the Group buys in renminbi and sells in Hong Kong and renminbi, so read the gap directionally, never to the dollar.
Settled by The FY2026 annual report gross margin, the note 4 inventory write-down and the other gains or losses note, due 2027-04-30. Lead time: one inventory turn, roughly six to twelve months at 286-321 days of stock.
Search HKEXnews for stock code 887, open the latest results announcement, and find the 'Number of stores' table in the Presence in Prime Retail Locations section. Emperor Watch & Jewellery Limited, on HKEXnews
Last recorded
69 stores, 2026-06-30
What the reading assumes
93 stores (the same disclosure at 31 December 2023, the peak of the network, FY2023)
Watch / alert
64 and 60 stores, on a move below — currently at or better than the level the reading assumed
How often to look
quarterly (the series prints event)
What it points to. Store count is one of the two terms of the revenue identity in every territory driver: revenue equals average stores times revenue per average store. It is also the fixed cost base, so a change moves both sides of the model.
What it cannot tell you. A store count says nothing about store size, location quality or productivity, and it counts nine franchised Mainland jewellery stores that produce wholesale revenue rather than retail revenue. Two networks of 69 stores can earn very different money, which is exactly what happened here between 2023 and 2026.
Settled by The next results announcement's store table, due 2027-03-31. Lead time: none - it is the observation.
run_off — nothing public to watch
Nothing free and public tracks the Other Asia Pacific run-off between filings. Searched: the Group's own disclosure, which names only Singapore and Malaysia and gives no country split of revenue; Singapore Department of Statistics retail sales, whose watches-and-jewellery category covers the whole country and cannot isolate five stores; and the Singapore ACRA register, which does hold the subsidiary's own accounts but is a paid, per-document service filed annually with a long lag, so it fails both the free and the between-filings tests. The honest position is that this engine has no reader-checkable signal, and the register says so rather than substituting a proxy that would not actually move with it.
Levels last recorded as at the information cutoff, 2026-09-03. A series moving past a level is a reason to re-read the case, not a recommendation. The same series across every company covered: what you can watch.
Basis. Figures are as reported by the issuer unless the arithmetic is described as recomputed or derived. The 1H2026 figures are unaudited and unreviewed. Territory gross margins before write-downs, revenue per average store, revenue per dollar of non-current assets, the payout ratios, the continuing-territory growth rate and the free-float estimate are computations from stated figures and are labelled where they appear. Twenty-eight primary documents were retrieved by direct PDF fetch at URLs taken verbatim from the enumerated HKEXnews title-search tape for 1 January 2022 to 3 September 2026; the exchange’s own total for that window is 196 rows and 196 were returned, so the enumeration is complete by the exchange’s count, and the rows that were not retrieved were classified by title. Price data is a three-year daily closing series to 3 September 2026, with the Hang Seng Index and four Hong Kong-listed peers as controls. This page publishes no forecast, no valuation and no view on the share price.
Where the filings differ from each other. One conflict is live and is resolved here on the arithmetic. The 2025 interim results announcement prints 7,256,708,129 as the earnings-per-share denominator for the six months to 30 June 2025, which does not reproduce the 2.73 HK cents of basic earnings per share printed on the face of the same document. The 2026 interim announcement gives the same comparative as a weighted average of 7,203,973,322 shares, which does. This page uses the weighted average. Separately, two announcements on the tape are second issues of the same subject on consecutive days — the interim dividend form of 20 and 21 August 2026, reissued as a clarification of the record date, and the FY2025 final dividend form of 25 and 26 March 2026. Neither is marked a replacement and the figures they carry are identical; this page takes its dividend figures from the results announcements in both cases.
Corrections log
No post-publication corrections recorded as of 4 September 2026. Corrections are welcome through the contact page and will be dated and described here.
Reader questions & corrections
Ask about this company, challenge a source or suggest a factual correction. Only selected questions that SMID Research has reviewed and answered are published.
No reader questions have been published yet.
Submit a private question or correction
Submissions are private by default. Your email is encrypted and never published. Please do not send confidential or inside information, allegations about individuals, promotions, or requests for personalised investment advice.
Behind the lock The private vault holds the author-only working record for this company: the written view, the forecast and valuation work, the model, the claims ledger and the independent-review artifacts. It is not published and not available for sharing, and it is not investment advice.
Private boundary. The author-only record is not approved for public distribution. This public page deliberately omits any rating, valuation, scenario output, forecast and portfolio action.